Divergence and Convergence, Educating with Integrity: Proceedings of the 7th Annual JALT Pan-SIG Conference.
May. 10 - 11, 2008. Kyoto, Japan: Doshisha University Shinmachi Campus. (pp. 21 - 26)

Ethnographic research: Children's socializing processes in a Japanese pre-kindergarten class

by Wakako Kobayashi (Kunitachi College of Music)

The purpose of this study is to investigate the socializing processes of six Japanese pre-kindergarten kids. Three main theoretical frameworks were adopted: (1) data as exemplar, (2) the notion of learning in a community of practice, and (3) a concern for the validity of longitudinal research. Participants included one main teacher, a sub- teacher, six three-year-old kids and their mothers. The data was collected in the fall of 2007 by observing multiple interactions over a 120 minute time frame and interviewing the sub-teacher and two mothers. This data was analyzed on the basis of general descriptive analysis, disconfirmatory analysis, and in vivo coding of an emergent list. This study concludes by discussing prejudice concerning gender education and cultural schemata.

Keywords: ethnography, socialization, interview, observation, Japanese pre-kindergartens
Japanese Abstract

[ p. 21 ]

Erickson (1986, p. 192) states that the origin of the word ethnography comes from the Greek term “ethnoi” that means “others”. So the general aim of ethnographic research is to promote understanding of “others”. Erickson summarizes others as part of the invisibility of daily life and encourages ethnographers to ask, “What is happening, especially in social action that takes places in this particular setting?”(1986, p. 192). The second thing that Erickson advises ethnographers to do is to develop the need for specific understanding through documentation of concrete details of practice. “What do these actions mean to the actors involves in them, at the moment the action took place? To attempt to answer these questions with general responses is seldom useful. Interactive fieldwork research can illuminate such questions in adequately only if it is detailed and in depth. The third aim of ethnographic research concerns the local meanings of experiences for the people involved in them. Different classrooms, schools and communities may seem ostensibly identical, but they give rise to distinctly new information and local meanings. With these factors in mind, I believe that there are three steps towards understanding others as Erickson urged.
This paper outlines three requisite steps towards understanding others. First, we must realize a certain kind of difference or contrast like Herodotos (The History, Book Ⅱ) did when writing a travel narrative describing the ancient Egyptians, contrasting them with the Greeks from a colonialistic and obviously ethnocentric perspective.
Second, we need certain kinds of systematized knowledge. Without such systematized knowledge, various types of gaps would exist between the observer's reality or subjectivity and the native point of view (Malinowski, 1922, p. 43) underlining local people's immediate actions. That systematized knowledge is said to represent the local's objectivity, although many anthropologists have attacked Malinowski's ethnography in his collected reports about the Trobriander's explicit and implicit cultural knowledge as being too subjective, sensitive, and unscientific. To reduce such a possibility, we seek for a meaning of the particular, non-systematic, local knowledge of the local actors through thick descriptions (Geerts, 1973, p. 29), with bracketing of prejudgments and preconceptions.
". . . in order to conduct situated qualitative research, it is both uncommon and dangerous to begin from and associate with a well-defined theory."

A third requisite step towards understanding others in the Ericksian sense of the work is to adopt emic and etic perspectives dynamically. Through this interactive progressive process between dichotomized cultures, we can fill in the gaps between our culture and the culture we are observing. This fulfills the indispensable, humble and reflexive process of situated qualitative research. This adaptive process, which requires progressive focusing, may involve a gradual shift from a concern with describing social events and process towards developing and testing explanations or theories. (Hammersley & Atkinson, 1983, p. 131) Therefore, in order to conduct situated qualitative research, it is both uncommon and dangerous to begin from and associate with a well-defined theory. Instead, coding the data recurrently, researchers must tolerate uncertainty and ambiguity in their interpretations, avoiding any rush to determine a conclusion, since constantly new theories could be emerging.

Theoretical Frameworks

Three theoretical frameworks upon which this paper is based will be briefly be outlined.

Framework #1: All data as exemplar – the notion of ecological validity to understand complexity

At the heart of ethnography, a case study (or textual analysis), must first of all fulfill Kuhn's (1970, p. 79) claims about the notion of exemplar. Mishler (1990, p. 416) also insists the validity and trustworthiness of a particular study as the basis for tacit understanding of actual situated practices in a field. A working, socially constructed knowledge could be valid because the emergent consensus includes local social values and social consensus. Therefore, first, we think that no general, abstract rules can be provided for assessing the complex features in particular ethnographic studies. Second, no formal or standard procedure can be applied to comparing different types of validity. This particularlizability is a matter of meaning, understanding and interpretation in situated qualitative research. Concerning this, we should not rely too much on over-interpretation. The significance is that the weight should be in the account itself, which mirrors the situated reality because the biggest job for an ethnographer is to present a particular description.
Moreover, Hammersley and Atkinson (1983, p. 126) discuss that ethnographers have been reluctant to admit casual models. This is partly because it clearly stems the positivists' concept of causality and mainly because assessing the validity of casual relationship is recognized at extreme difficulty. Nevertheless, theories implying causal relationship are common in some ethnographic works just like the positivists emphasize on experiments. (Mishler, 1990, p. 416).

[ p. 22 ]

Framework #2: Learning in “communities of practice” – socialization by participating in social practice

A recent emphasis on “communities of practice” (Lave & Wenger, 1991, p. 41, Wenger, 1998, p. 172) in educational studies exemplifies the significance of the idea that all learning is a process of dynamic, adaptive and meaningful participation in social activity. Often this is with the guidance of more experienced social members, like teachers, parents, and peers. In the process, the whole person, activity and world are treated as mutually constitutive. By participation, Lave and Wenger (1991, p. 41) and Wenger (1998, p. 172) assert that education is at first legitimately peripheral but increases gradually in engagement and complexity.
This approach would trigger an important rethinking and reformulation of more conventional conception of learning, which normally and always implies the tyranny of the assumption that learning is a static reception of the factual knowledge or information from the teacher. In this research, I collected the data from a pre-kindergarten near Yokohama city known as the Airin kindergarten, which adopts a learner-centered, non-teacher centered approach as its main educational principle, treating all kids as whole human beings. Therefore, this should influence the kids' developmental lifelong procedure of ever changing identity construction to some extent.

Framework #3: Malinowski's perspective of “panopticon"

Ethnographers seek a profound understanding of particular contexts which consist of specific and concrete social actions by participating in and observing the situated cultures of others. Differing from the older concept of “armchair anthropology”, after conducting a two-year fieldwork research in the Trobrian islands, Malinowski (1967, p. 82) insisted on the significance of participating in a local society in order to acknowledge both implicit and explicit aspects of every day life in a particular cultural scene, adopting inductive and emergent approaches dynamically and simultaneously.
This panoptic position as a participating observer could be discussed in terms of Foucault's concept of automatization of power (Dreyfus & Rabinow, 1983, p. 184). That is the objectifying of the subject, which implies the observer's existence should influence the local people's social action. That is why relatively long-term observation or participation in a particular social scene is a necessary condition to decrease the effect of the observer in data collection. Nevertheless, it goes without saying that maintaining emic and etic perspectives simultaneously, in good balance, are relatively difficult for any researcher.


Three aspects of the research setting will be briefly explained: the reasons for selecting the given venue, background information about the setting, and a few characteristics of the actual class.

Reasons for choosing the venue

This first ethnographic study explores how Japanese 4 year–old kids in a tot class (a play class) which is attached to a private kindergarten are partly socialized in the concept of guided participation (Rogoff, 1992, p. 146). According to Rogoff, any social relationship inherent to families and communities should be included in learning activities, such as involving children and siblings, teachers and classmates in these cases. Why was this particular social scene chosen? One reason was to gain a holistic understanding of this legitimate social activity, where the mothers hope their children will play regularly in a private tot class.

Background information about the setting

First, since the Airin kindergarten is based on Christian (Baptist-Protestant) educational principles, the kids' autonomy or heuristics are regarded as significant. Consequently, no school uniforms are provided, and there are strict regulations and punishments in the overall educational process. Second, the Tsubomi (lit. “bud”) class is an informal play class, within the formal three- year kindergarten education program. This class is 90 minutes long and held once a week. Therefore, we imagine that many mothers are making a preliminary inspection to obtain more information about the kindergarten itself, besides how their kids play in the class. For instance, they can acquire information about class fees, overall atmosphere, lunch quality, bus service, recreation facilities, excursion activities, as well as whether their kids like the place.

[ p. 23 ]

Classroom characteristics

In the tot class, some noteworthy conflict and resistance to the social scene could be picked up. It is natural that not all kids could smoothly join the class, where they are physically separated from their mothers for 90 minutes, regardless of the fact that many mothers could stay in the back of the classroom if they wished. Likewise, we observed how two teachers – a main teacher who had the power to give instructions and a sub-teacher whose role was a helper – performed differently, representing uneven power relations in class.
Furthermore, the class itself was a good place not only for kids but also for mothers, who wanted to exchange various kinds of information by chatting while relaxing.

The Process of Analysis

The methods by which the data was analyzed will now be described. After mentioning the general descriptive analysis procedure, we will outline how disconfirmatory analysis, critical generating, and testing analysis were employed. Finally we will turn our attention to coding.

Procedure #1: General descriptive analysis

Spladley (1980, p. 56) specifies three features of all social situations which provide a possible grand tour description. These consist of (1) Space, (2) Actor, and (3) Activity. Each was identified this in the Airin Tubomi class.
In addition to these three main features, six additional factors to formulate the prime grand tour questions and to make observations were identified. The additional features are (4) Objects, (5) Act-1 Act-2, (6) Events, (7) Time, (8) Goals, and (9) Feeling. Although such elementary lists are a bit crude and rest on arbitrary classifications, I believe they indicate a range of relevant and reasonable features of context in the two observations (Observation 1 and Observation 2). Furthermore, in most senses, these nine dimensions can serve as a guide for the participant observer, which represents the researcher's positionality. Clear difference was provided that the sub-teacher had a chance to chat with some mothers about childcare, lunch, and leisure places in my second observation, which seemed to be a desirable activity for the sub- teacher, who didn't talk with mothers at all in the first observation. Despite the possibility that it happened purely by chance, I suppose that my existence as a participant observer, continuously taking notes in the classroom, dressed in a formal blue suit, made the participants feel awkward and unnatural to some extent.
The contrast between OA1 and OA2 (Observation Account 1 and Observation Account 2) indicates the reason why a long term observation is so to speak, sine qua non (necessary) in any ethnographic study. When we try to understand the other people's situated culture, it is a non-measurable aspect in the real life, by adopting the tool of the fieldwork, and observation. Whether she/ he is an expert or a novice, the observer will be in the panopticon-like position as Malinowski described. That is, we participate overtly and covertly in a particular social scene for an extended period of time, watching what happens and listening to what is said.
Next, considering the aspect of Time, every social interaction includes a temporal dimension – not only a host of events that occur before and after each interaction but also a temporal framework in terms of which the people involved locate them and focus on this dimension, naturally new observations or interviews could emerge. (Spradley, 1980, p. 56; Hammersley & Atkinson, 1983, p. 128) This tot class started in April and my short-term ethnographic study was in late September. Consequently, all social members, teachers, kids and mothers who participated in the class were familiar with each other to some extent.

Procedure #2: Disconfirmatory analysis and critical generating & testing analysis

As discussed earlier, disconfirmatory analysis is suitable for analyzing discrepant cases because the notion of ecological validity to understand complexity is represented in the data as an exemplar (Mishler, 1990, p. 415). However, as expected, unfortunately, the data corpus was so small that I could not find enough discrepancy. Hence, I must adopt a critical generating and testing approach, though it is completely deductive. Still I would like to try as an experiment to seek a more simultaneously inductive and emergent approach.
Figure 1 illustrates the key linkage between analyses and assertions (Erickson, 1986, p. 148). The merit of the Tubomi class as a general assertion is hypothesized. This is derived from the coding, and dividing it into three categories of assertion as described in a Figure 1.

[ p. 24 ]

Figure 1
Figure 1. The merits of Tsubomi class based on Erickson, 1986, p. 148, Figure 5.1

As the evidence above indicates, generating and testing analysis, which is scientific, deductive and governed by methodological rules, is not necessarily sufficient to offer ecological validity to a social discourse. This encourages us to view all types of research as form of life (Wittgenstein, 1953, p. 38) and to focus on ongoing activities rather than static properties of instrument and scores. Therefore, as Erickson (1986, p. 148) states, it should be noted that this kind of analysis often requires progressive problem solving, ongoing focus, emergent open-ended approaches in interactive ways during data collection and analysis. Furthermore, in judging the validity of researcher's analyses, it is important to keep in mind the best case for validity is to account for both frequent and rare events. Even though the researchers prove something in a causal sense, I believe that there should be discrepant case which is useful in illuminating local distinctive subtleties, though this is limited. The non-longitudinal ethnographic study like this didn't provide a sufficiently ample data corpus.
Therefore, we assume that the deliberate search for the disconfirming evidence is essential for data analysis. Moreover, to understand the complexity in a real social situation, the deliberate framing of an assertion must be tested against the data corpus ( Erickson, 1986, p. 148)

Procedure #3: Coding

Three aspects of coding will be briefly considered: the use of categories, the use of in vivo lists, and etic emerging codes.
  1. Categories:

    As for categories, many features were found from the two observation accounts and interview transcriptions. Seven items were categorized. Especially in terms of number 7, Linguistic Feature (polite expression), the sub-teacher who is younger and working part-time used mitigated expressions to indicate politeness such as “would you know – gozonjidesuka?” to senior mothers. This reflects the power relationship between the sub-teacher and mothers.

  2. In Vivo List (Emic, Situated Vocabulary, Folk taxonomy):

    One in vivo term worth noting is “asobu chikara”. This term closely relates to one of the educational principles of the Airin kindergarten. One of the principles is to regard each kids' autonomy, or heuristic as a significant factor and to develop their personalities through a learner-centered approach. I assume this term is also applied to the active, adaptive dynamic participation in the complex social practice. 

  3. Emerging codes (Etic, researcher's category)

[ p. 25 ]

Reflection and Conclusion

In ethnographic study, the analysis of data is not a distinct stage of the research since it begins in the pre-field work phase, in the formulation and classification of the data, and continues even through to the process of writing a study, which is absolutely an interactive, dialectical process between data collection and data analysis.
However, many ethnographers face the practical constraints on achieving this close interaction because any fieldwork, which includes the observation and interviews is a very demanding activity. Consequently, engaging in situated data analysis alongside data collection, or keeping emic and etic perspective simultaneously is often very difficult, especially for novice fieldworkers like me. Nevertheless, some level of reflexivity should be maintained, even if the reality is so complex that it reflects the “unstructured data” (Hammersley & Atkinson, 1983, p. 230). What this means is that the data are not always structured in terms of a finite set of analysis categories in the way that most positivistic research data are.
"[When] engaging in situated data analysis . . . some level of reflexivity should be maintained, even if the reality is so complex that it reflects the 'unstructured data'."

Rather, they take an open-ended style in analyzing the observational accounts and interview transcriptions because local situated data could represent a bigger whole social context as an exemplar to show the ecological validity. Therefore, ethnographic data collection and analysis is a process of capturing the holistic understanding in a specific social discourse. If we put it differently, the partial micro perspective connects to the whole, macro understanding because every cultural and social phenomena does not exists in a sole or decontexualized unit but has some significant functions in the culture or society as a whole.
Hence, the accounts themselves can be of great value. They might provide us with knowledge of culture hitherto unknown, therefore shake our predetermined and systematized assumptions about human beings and their social lives. Therein lies the interest of much ethnographic work.


Atkinson, D. (2003). Language socialization and sys-socialization in a south Indian college. In R. Bayley & S. R. Schecter (Eds.) Language socialization in bilingual and multilingual societies (pp. 147-162) Clavedon, UK. Multilingual Matters.

Drefus, H. & Rabinow, P. (1983). Power and truth. In H. L. Dreyfus & P. Rabinow. Michel Foucault Beyond structuralism and hermeneutics, 2nd ed. (pp. 184 - 204). Chicago: University of Chicago Press.

Erickson, F. (1986) Qualitative methods in research on teaching. In M. C. Wittrock & M. C. Wittrock (Eds.), Handbook of Research on Teaching, (Macmillan Research on Education Handbook Series) 3rd ed. (pp. 119 - 161). New York: Macmillan.

Geertzs, C. (1973). Thick description: Toward an interpretive theory of culture. In The interpretation of culture (pp. 3 - 30). New York: Basic Books.

Hammersley, M. & Atkinson, P. (1995). Ethnography: Principles in practice, 2nd ed. London: Routledge.

Herodotos. The History. (Book Ⅱ). Retrieved August 28, 2008 from http:// classics.mit.edu/Herodotus/history.2.ii.html

Kuhn,T. (1962/1970). The structure of scientific revolution, 2nd ed. (enlarged). Chicago: Chicago University Press.

Miles, M. B. & Huberman, A. M. (1994). Early steps in analysis. In M. B. Miles & A. M. Huberman. Qualitative data analysis: An expanded sourcebook, 2nd ed. (pp. 50 - 89). Thousand Oaks, Ca: Sage.

Malinowsky, B. (1967). A diary in the strict sense of the term. London: Athlone Press.

Misher, E. (1990). Validation in inquiry-guided research: The role of exemplars in narrative studies. Harvard Educational Review, 60, 415-442.

Lave,J. & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. Cambridge: Cambridge University Press.

Spladley, J. P. (1980). Participant observation (pp. 53 - 84). Forth Worth, TX: Harcourt Brace Jovanovich.

Wittgenstein, L. (1953). Philosophical investigations 3rd ed. Oxford: Blackwell.

Wenger, E. (1998). Communities of practice: Learning meaning and identity. Cambridge: Cambridge University Press.

2008 Pan SIG-Proceedings: Topic Index Author Index Page Index Title Index Main Index
Complete Pan SIG-Proceedings: Topic Index Author Index Page Index Title Index Main Index

Last Main Next

[ p. 26 ]