The Historical Methodological Foundation of Phenomenography

Abstract

Phenomenography is a research tradition originating from research on students’ studying and learning. The research methods of phenomenography have been discussed and criticized by many authors. There is a need for a presentation of the historical methodological foundation of phenomenography. The phenomenographic research methods are best understood as placed within the general research methodology and research approach of contextual analysis. The following description is directed to all interested in phenomenography and its methodological foundation. It is also directed to all interested in research methodology, especially in the fields of empirical educational and social science research. After a brief introduction, the common background of contextual analysis and phenomenography, and the origin and development of contextual analysis and phenomenography, are briefly outlined. Then, the methodological reflections and orientations in phenomenography, and the foundation of contextual analysis and phenomenography are presented. The article ends with a conclusion.

Introduction

The research tradition of phenomenography has been developed in relation to a wide range of philosophical and theory of science schools of thought, and in relation to a broad range of theories within the fields of cognition, learning, teaching, and language. In this article, this development is briefly outlined, with focus on research methodology. A general research methodology and research approach called contextual analysis has been developed in relation to the empirical research called phenomenography. Contextual analysis is a research methodology and a research approach, and not a specific research method. The reason for focusing on methodology and approach and not on method is the conviction that method should not come first. Research methods should be based on ontological, epistemological, and methodological assumptions specified in relation to investigated research objects. The expression, research object, is here used to mean that which is investigated, even if it is or includes subjects.

Generally described standard methods are seen as relevant to the extent that the research objects, and the knowledge of them, can be assumed to have shared characteristics, which justify the use of the same methods. The motivating of methods should be done in connection to, and as dependent on, development of knowledge within specific fields of research. This makes methodology and a systematic, critical, and creative thinking about methods, important. In the phenomenographic research, there has been a development and use of research methods argued for in relation to the investigated research objects and alternative possible methods. This development and argumentation is made based on the empirical work, and it can, to a large extent, be said to be in line with contextual analysis as a research methodology.

Before continuing, a clarification will be given concerning the use of some words and concepts in this presentation and in contextual analysis and phenomenography. It concerns the words object, phenomenon, case, and meaning. The word object is not used based on an objectivistic assumption, meaning that knowledge should be based on objectively given parts of the world. Knowledge is understood as relational, as existing in a relation between the knower and the part of the knower’s world the knowledge is about. The word object is used in the meaning of object of research, or object of knowledge, only meaning that which is investigated or studied, and that which is known. The objects investigated in contextual analysis and experienced by students are mostly called phenomenon and/or case.

Both phenomenon and case are seen as existing within the relation between knower and world. Phenomenon is seen as more of a conceptual unit, and case as more of an empirical unit, even if both are both conceptually and empirically based. Each case is a unique part of the world and a phenomenon if conceptualized as an organized whole, but a phenomenon usually corresponds to several cases (of the same phenomenon). Thus, case is used about the object of research and not as in classical case studies about an individual, a situation, an organization, or other unit that the information or data are linked to, but which is not in itself the object of research. How relations between descriptions of phenomena and cases are constituted must be clarified in each investigation. The word and concept meaning is throughout used about the meaning something is experienced to have. The meaning referred to is more delimited and situational than is often the case in using the word meaning in a language-based way, where the linguistic unit is considered to have a general meaning.

The Background of Contextual Analysis and Phenomenography

Both contextual analysis and phenomenography have their background in research carried out more than half a decade (contextual analysis) and more than a decade (phenomenography) before those terms were introduced. The present author introduced the term contextual analysis publicly in 1976 (Svensson, 1976) and Ference Marton introduced the term phenomenography publicly in 1981 (Marton, 1981). What these terms referred to were general ideas and programs based on already carried out research. The present author and Ference Marton were doing research during the late 1960s before together starting the research forming the basis for presenting contextual analysis and phenomenography.

The present author, from the midst of the 1960s at Gothenburg University, Sweden, developed an inventory to measure vocational interests, using semantic differential scales. I became critical to the neo-behavioristic meaning theory behind the semantic differential scales and the limitations of the dimension analytic methods used in investigating vocational interest. The work was finally reported in a report with the title, translated into English, “Language as the basis for compatibility. Problems of validity in a case of ‘measurement’ of interests” (Svensson, 1973). Before the work on vocational interests was finished, I carried out an investigation in a project led by Ference Marton. The project was an investigation of goals, teaching, and tests within education in English as foreign language. The project was carried out at, and in cooperation with, the Department of English, University of Gothenburg. Within the project, I carried out an analysis of study activity and study success (Svensson, 1970). I used statistical calculations on variable data, including test results, to clarify correlations between a range of activities foregoing test results and test results. The results from the analyses were interesting and revealing, but I also became concerned about the limitations of the psychometric and statistical methods.

Ference Marton, who was the leader of the project mentioned, simultaneously carried out a comprehensive work about tests and evaluation in higher education (Marton, 1967). He also carried out an experimental investigation on Structural Dynamics of Learning (Marton, 1970). The experiment concerned learning and recalling of a list of names of 48 famous persons. One main focus in the analysis was the relation between structure and recall performance. One conclusion was

The relationship between the degree of structure on one hand and performance on the other hand changed from decreasing negative to increasing positive. The higher degree of structure in the later phase of the experiment on the part of the better performing subjects was achieved by means of greater effort to build up a structure in the earlier phase of the experiment. (p. 8)

At the end of his book, Marton writes, “This highly artificial situation was seen as an instance of conflict between environmental complexity and human limitations. At the same time the course of learning provided a demonstration of man’s capacity for dealing with this conflict.” (p. 88). He ends with suggesting that the results should be used in further research on learning in relation to everyday problems of human life. The continuation of this research interest close at hand was to investigate studying and learning in higher education.

The Origin and Development of Contextual Analysis and Phenomenography

The present author and Ference Marton started as joint project leaders a research project in 1970 at University of Gothenburg, Sweden, concerning university students’ studying and learning in some different academic disciplines (financed by the Office of the Chancellor of the Swedish Universities). A main question focused in the project was how to observe and describe students’ studying, learning, and understanding of subject matter. The project was intended to shed light on three questions. These questions were asked in a devised order so that the answer to a preceding question would be contributory prerequisite if an answer was to be given to the next question. The questions were the following:

1. How can knowledge be observed and described in a profitable way? (By knowledge is meant what is learned).

2. How can study skill and learning be observed and described at a functional level?

3. How can study skill and learning be influenced?

The project both aimed at developing methods to observe and describe students studying, learning, and understanding, and making descriptions of those phenomena. Thus, methodology was a main question from the start of the research later called phenomenography. The focus on development of methodology and methods is expressed through the open questions about how to observe and describe knowledge, study skill, and learning. In the project, we developed a way to describe students’ understanding of subject matter in terms of qualitatively different conceptions of messages and phenomena. This methodological development formed the basis for the development of a research orientation that later was called phenomenography (Marton, 1981, 2015; Marton & Booth, 1997; Svensson, 1997). Development of research methods was an aim in the project and the way of doing the data analyses was new. In the most comprehensive report from the project (started in 1970), the approach to analyzing empirical data on students’ studying and learning was called contextual analysis (Svensson, 1976). In the concluding chapter about methodology, the following citation was part of what was said about contextual analysis. What is referred to as “a generally delimited relation,” in the second section of the citation, is the relation between study and learning activity on one hand and learning outcome on the other hand:

The reason for not using generally defined categorizations of data (variables) is that the use of such categorizations can seldom be defended because of the context dependency of the phenomena. A contextual analysis within the area of research dealt with here must first of all consider the individual as the most immediate context as regards interpretations of specific data. Thus, the analysis must give descriptions of the relations between specific data within individuals.

The individual and situational context is the starting point for the analysis. The aim is to describe a generally delimited relation. However, neither the full concrete meaning of the relation nor all relevant aspects and categorizations can be assumed to be known. The reason for this discussed earlier is that the meaning of the specific data is dependent on the context.

The “same” specific data may have totally different meanings in different contexts. The “same” amount of time, study technique and even concrete form of strategy and approach mean very different things depending on the amount and type of learning material. The same is also true about the relation between the aspects of study activity. The meaning of a concrete manifestation of one aspect of study activity will vary depending on the context of the activity.

A contextual analysis, then, must not only mean an aggregation of specific data with generally given interpretations, but a delimitation of specific data related to each other as aspects of the same phenomena. (Svensson, 1976, p. 187)

Our first theoretical methodological discussions both locally and internationally within educational research were with researchers representing the use of measurement and statistical methods. This discussion was very intense during the 1970s. For instance, we presented our research in the form of a symposium at the 1975 American Educational Research Association (AERA) meeting in Washington, D.C., where Richard C. Anderson, University of Illinois, and Richard E. Snow, Stanford University, were discussants. They appreciated our presented empirical studies and results. However, they raised questions about our approach and methods mainly in terms of suggesting that they were well chosen and carried out but had to be seen as preliminary investigations (pilot studies) that ought to be followed by definitions, measurement, and statistical analyses. We, on the contrary, explained that our approach was an alternative to what they suggested.

From this first research project, a research group developed, new projects were started and carried out, and research cooperations were developed. New projects concerning university students’ understanding within new fields and educational programs were carried out. Based on the research on university students’ studying and learning, a rather close cooperation was established with professor Noel Entwistle, then at University of Lancaster, his doctoral students, and his network of young researchers. Our common research had a focus on learning in the context of studying and teaching in higher education. The cooperation with Noel Entwistle resulted in a series of articles in British Journal of Educational Psychology, of which he was an editor (Marton & Säljö, 1976a, 1976b, Svensson, 1977). The cooperation with our British colleagues, among other things, later resulted in a book published in 1984 (Marton et al., 1997, 2nd ed.) with the title, The Experience of Learning. Implications for Teaching and Studying in Higher Education. In addition, Ference Marton and the present author published an article in Higher Education in 1979, with the title “Conceptions of Research in Student Learning” (Marton & Svensson, 1979).

At the 1980 AERA meeting in Boston, we presented our research in a symposium and our discussant, professor Elliot G. Mishler from Harvard university, was a representative of the ethnographic research tradition. He suggested that we, through the analytic character of our research, were too close to mainstream quantitative research and not enough ethnographically descriptive even if we were descriptive. On the contrary, we saw our approach as an alternative to a general ethnographic approach, an alternative more based on the understanding of the objects of research investigated.

The developed descriptions of qualitatively different conceptions were found to be very powerful in a lot of contexts also outside formal education. We carried out a research program, led by Ference Marton, with the title “Adult’s Conceptions of their Surrounding World” in the years 1976 to 1981 (financed by Bank of Sweden Tercentenary Foundation). Within this project, the focus was on describing qualitatively different conceptions within several different fields in contexts outside formal education. People’s conceptions of phenomena in the surrounding world are research objects in many disciplines, for instance, in developmental psychology, history of ideas, anthropology, sociology, and linguistics. It seemed to be motivated to consider the study of conceptions of the surrounding world as a special interdisciplinary knowledge area and to give this research a name. This research, together with the research on qualitatively different conceptions in cases of studying and learning, is the background to Marton’s (1981) suggestion of the term phenomenography.

After the term phenomenography was introduced in 1981, it was used to refer to both previous and coming research that had descriptions of conceptions and ways of experiencing as a central and basic part. The research within education on studying, learning, and teaching with this orientation expanded and was thus called phenomenography.

The further development of phenomenography not only consisted in a spread concerning subject matter, educational level, and a geographical spread but also in both a deepening of the dealing with theoretical issues and a diversification of the research. The diversification included a variation in theoretical and methodological orientation for inspiration. The inspiration from different academic works was part of the deepening of the theoretical interpretation and development of the empirical research. One deepening and broadening development has been especially extensive and significant when it comes to the theoretical basis and has been named Variation theory (Marton, 2015; Marton & Booth, 1997). The deepening part of this development is the description of variation as fundamental to conceptualization and learning. The broadening part is that this basis has been especially used in relation to teaching for learning compared with the previous main focus on studying and learning. Another deepening development has concerned the function of language use in developing and expressing conceptions (Anderberg et al., 2008; Svensson et al., 2009).

Methodological Reflections and Orientations

Many researchers in educational research have adopted or been inspired by the research orientation and research approach of phenomenography. As the research is not based in some theoretical and/or philosophical school, it is free to connect to such schools on the basis of researchers’ choices and understanding of the objects/phenomena investigated. This has also been made in phenomenographic research. Some work has been more related to cognitive research and others to sociocultural research. Some work has been more ethnographic and some more phenomenological. All work has, in some sense, been hermeneutic but not related to some specific hermeneutic tradition.

The phenomenographic research approach was, from the beginning, an answer to the limitations of a positivistic measurement and quantitative approach to investigate students’ studying and development of knowledge. This meant that an inductive development of scientific knowledge based on measurement and correlating of variables was found to be less fruitful. At the time, Popper’s deductive approach was much referred to and used (Popper, 1959, 1963). We did not find this approach suitable but stressed the need for an explorative approach different from working with definitions and falsifications. In addition, we found phenomenological and hermeneutic methods to be too deductive in depending on predefined concepts and transformations of empirical data according to philosophical concepts and theoretical language.

We think that theory is best tested and developed by theory meeting interpretations of verbal data based on the language of the individuals investigated, and not based on beforehand-defined theoretical concepts used in interpretations of specific data. However, contextual analysis was not then presented and argued for as a general methodology. Therefore, contextual analysis was then not referred to as a general methodology in later different phenomenographic investigations. However, the research was argued for and defended against critique in a way in line with contextual analysis.

An example of the arguing for phenomenographic contextual analysis is a contribution, by the present author, to a phenomenologically oriented conference (Ashworth et al., 1986). The contribution had the title, “Three Approaches to Descriptive Research” (Svensson, 1986). Contextual analysis in the first phenomenographic project resulting in categories of description of qualitatively different conceptions of the message of a text (Svensson, 1976) was compared with phenomenological reduction of a person’s report of an experienced case of learning by Giorgi (1978) and a linguistically based analysis of students’ telling about their reactions and the circumstances in which the reactions arise by Bliss, Ogborn, and Grice (Bliss et al., 1979). The comparison was commented on in the following way:

The specific differences in methods have to be understood in relation to the general differences in methodologies. The general differences between the approaches are related to that they represent different “turns” in the search for a basis for and for support in answering questions about methods. The three “turns” will be labelled “the scientific turn” (S), “the philosophical turn” (P) and “the linguistic turn” (L). (Svensson, 1986, p. 24)

The first way (S) used by the present author is contextual analysis. The second way (P) used by Giorgi is phenomenological reduction. The third way (L) used by Bliss, Ogborn, and Grice is construction of language networks. We here have three different ways of dealing with verbal data. The present author is convinced that contextual analysis is the generally more fruitful way.

Assumptions made in phenomenography as research orientation, focused on and restricted to description of qualitatively different conceptions, not including assumptions when this orientation is used to describe other more complex phenomena of which conceptions form part, have been discussed by the present author in previous publications. One of those publications (Svensson, 1997) is an article with the title “Theoretical Foundations of Phenomenography.” The conclusion given in this article is clarifying in the present context. It gives a summary of assumptions behind the phenomenographic empirical research, when it comes to describing conceptions, as understood by the present author:

Conclusion

Phenomenography has its roots in the general scientific tradition, not in philosophy or some specific school of thought. It represents a reaction against, and an alternative to, the then dominant tradition of positivistic, behaviouristic and quantitative research. It makes its own ontological, epistemological and methodological assumptions with inspiration from and similarities to several older and concomitant traditions, without agreeing entirely with any of those. The most fundamental assumptions are that:

1. knowledge has a relational and holistic nature;

2. conceptions are the central form of knowledge;

3. scientific knowledge about conceptions (and generally) is not true but uncertain and more and less fruitful;

4. descriptions are fundamental to scientific knowledge about conceptions (and generally);

5. scientific knowledge about conceptions is based on exploration of delimitations and holistic meanings of objects as conceptualised;

6. scientific knowledge about conceptions (and generally) is based on differentiation, abstraction, reduction and comparison of meaning. (p. 171)

The emphasis on holistic nature and holistic meaning says that conceptions and knowledge here are assumed to concern parts of the world as wholes. When this research orientation is used as an approach to investigate activities and processes of understanding, of learning, of communicating, of teaching, of culture, and other phenomena, further assumptions are made in relation to those phenomena in the forming and using of research methods. Such assumptions do not strictly follow from phenomenography but are based on the theoretical understanding of those more extensive and complex phenomena, including conceptions as phenomena. However, the assumptions stated in the six points above is also, in contextual analysis and phenomenography, made for these other more complex phenomena.

Phenomenography, as an educational research orientation, is about investigating people’s personal conceptions of phenomena in studying, learning, and teaching. In an article published in 2013, contextual analysis is argued for as a general methodology and approach exemplified in relation to research on teaching (Svensson & Doumas, 2013). The book, Contextual Analysis. A Research Methodology and Research Approach (Svensson, 2021) gives a comprehensive description of contextual analysis as a general research methodology and research approach. In the book, contextual analysis is described and argued for in relation to a broad field of scientific work. It is exemplified and discussed in relation to investigating physical motion, learning, teaching, and culture as research objects and fields of research. Contextual analysis is a general research methodology and research approach and not specific to phenomenography. However, it is in relation to contextual analysis that phenomenography is best understood and developed when it comes to research methods.

The Foundation of Contextual Analysis and Phenomenography

Against the background of the dominant tradition of using psychometric and statistical methods in educational research, we were very interested in and inspired by work within the philosophy of language, psycholinguistics, and sociolinguistics as an alternative basis for research methods. We related to empirical work within psycholinguistics, for instance, the work of Rommetveit (1968, 1972, 1974). We found that the use of language in creating scientific knowledge is a fundamental question also including the use of mathematics and statistics. We were inspired by the results of the logical positivist early work, especially Carnap’s (1936, 1937) work, and the difference between the early and the later Wittgenstein, between “Tractatus” (Wittgenstein, 1971) and “Philosophical Investigations” (Wittgenstein, 1974). Another source of inspiration was Polanyi’s (1969) writing about tacit knowledge. One difference of special importance we found was the difference between definition of meaning compared with exploring and delimiting meaning as a main difference in research approach.

When it comes to theoretical methodological orientations and schools the late 1960s and the 1970s involved an intense discussion of schools of thought, to a great extent discussed in terms of paradigms, inspired by Kuhn’s (1962) book The Structure of Scientific Revolutions and following publications (Kuhn, 1970, 1974). At Gothenburg university, there was generally, but especially within theory of science, a discussion of schools of thought and paradigms. Radnitzky (1970) presented an extensive description of fundamental differences between research traditions. He did not use the concept of paradigm but discussed differences between schools of metascience. He made a difference between two metascientific schools, the Anglo-Saxon and the Continental school. The distinction comes close to what is often described as a difference between a positivistic/analytic tradition and a dialectic/phenomenological/hermeneutic tradition. Radnitzky conducted his work in close cooperation with Törnebohm, professor of theory of science at Gothenburg university. Törnebohm (1975, 1977), in a later more logically based analysis, makes a similar grouping of paradigms in systemic and contextual paradigms. Although contextual analysis, as developed and described by the present author (Svensson, 2021), is more based in epistemology compared with Törnebohm’s basing of descriptions of paradigms in ontology, it has its background in those paradigmatic discussions.

There are three main characteristics of contextual analysis and phenomenography, in that development of knowledge is seen as case-based, analytic, and contextual. In a deep holistic approach, the understanding of objects is case-based holistic, and both analytic and contextual. Contextual analysis is case-based through delimiting research objects as cases, parts of the world, investigated to their parts, especially main parts and their interrelations. All information and data are treated as being about the specific case and its relation to a wider context. When it comes to students’ development of knowledge, it is in phenomenography described as case-based in terms of how they are delimiting a part of the world as presented in a text message or in direct observation as a whole, and how they understand this whole as constituted through main parts and their interrelations. In many cases, the description reveals a lack of such deep holistic understanding and knowledge. Thus, contextual analysis corresponds to what is seen as an ideal development of knowledge also for students and people in general.

In contextual analysis and phenomenography, the distinction between internal and external relation is used to clarify a difference in ways to give meanings to information units (data units) and parts of phenomena, and relate such units and parts, either in a discerning/delimiting way (internal relating) or a predefining/combining way (external relating). The relation of cases of phenomena to their contexts are also considered to be internal. It seems fruitful to differ between work with internal relations and work with external relations. All quantitative methods not only mean work with external relations, but also presuppose a preceding work with internal relations, if one accepts the thesis that a first determination of meaning always is made in and is dependent on a context.

The distinction between internal and external relations was introduced by Bradley (1908) and discussed by, among others, Moore (1922). Their discussions were mainly ontological discussions about the nature of relations. The use of the distinction between internal and external relations here does not focus on the nature of relations in the ontological sense, and no assumptions about the nature of relations in this ontological sense have to be made to use contextual analysis or in phenomenographic research. It has been common to consider relations described in natural sciences and in understanding natural science phenomena to be external. Hesse (1980) summarizes a post-empiricist account of natural sciences in five main points. One point is that “In natural science the law-like relations asserted of experience are internal, because what we count as facts are constituted by what the theory says about their interrelations with one another” (p. 172). This is an epistemological understanding of natural sciences pointing to the character of internal relations of the knowledge developed.

Analysis of internal relations is a matter of interpretation. In methodological discussions, it has been common to see analysis and interpretation as opposite methodological characteristics. This is partly due to the fact that analysis is given the meaning of partitioning into separate units that are considered to have generalized meanings. Due to this meaning given to analysis, transforming traditions (ethnographic, phenomenological, and hermeneutic traditions) have been critical to analysis. (Phenomenological reduction is here seen as one form of transformation.) Analysis equal to clarification of internal relations within cases, as argued for here, is a meaning of analysis clearly compatible with the understanding in those traditions of cases/phenomena, when it comes to relations between parts, and parts and wholes, as internal and not external. These transforming traditions, for different reasons, do not emphasize analysis.

It is an important difference between research approaches, that some privilege analysis, whereas others privilege transforming of data. There is a problem with using the terms analysis and transformation to refer to the difference between two categories of methods. Those concepts are used with different meanings and do not exclude each other. A book about research methods by Wolcott (1994), for instance, has the main title Transforming Qualitative Data and the subtitle Description, Analysis, and Interpretation. In Wolcott’s presentation, analysis is discussed as one form of transformation. Transformation is seen as superordinate and description, analysis, and interpretation as forms of transformation. In a corresponding way, analysis is seen as superordinate in analytic approaches, which include more or less of transformation. The background to this variation in emphasis on analysis and transformation in the treatment of data is that the development of knowledge in all the cases means both analysis and transformation of data.

Our concern is with knowledge about parts of the world assumed to be organized wholes. The research needs a more clearly analytic orientation toward phenomena as wholes. An emphasis on starting from what is studied or investigated, rather than from information and data, also means a demand for analytic qualities of the study and research methods. In principle analysis, as understood here, means a discerning of constituent parts by use of internal relations. A criterion in evaluation of both students’ development of knowledge and research ought to be how clear it is what is studied/investigated, and how clearly units included form parts of what is studied/investigated. Treatment of information and data to a greater extent should be a treatment in relation to the studied/investigated phenomenon, as a ground for relating information/data to each other and less an ordering of information/data in a compiling or transforming way.

Conclusion

Phenomenography is a human science tradition of empirical research developed over the past 50 years. It has been developed in relation to a broad range of theories within the fields of cognition, learning, teaching, and language and in relation to a wide range of philosophical and theory of science schools of thought. Different researchers have been inspired by different theoretical and methodological traditions, and related phenomenographic empirical research to those. This is seen as natural and fruitful as long as it is done in a clear relation to what is investigated, in line with that the research is empirically based. Within phenomenography, there has been a theoretical development in relation to the understanding of the objects of research.

In the research on university students’ studying and development of knowledge, we developed a focus on describing qualitatively different conceptions of and ways of experiencing subject matter as a central part. Descriptions of qualitatively different conceptions and ways of experiencing were then seen as a special field of research given the name phenomenography, including description of qualitatively different conceptions of and ways of experiencing subject matter within educational contexts. The research was further developed in theorizing about the basis for conceptualization and ways of experiencing in Variation theory, which was also used especially in research on teaching. Another further development concerned the function of language use in developing and expressing conceptions.

Discussion and development of research methods has been a part of the development of phenomenography throughout the years. The question of methods was central in the original research, later called phenomenography, then with a focus not only on how to observe and describe students’ understanding of subject matter, but also how to do this as part of investigating students’ studying and learning. The methodology and approach was called contextual analysis and argued for in relation to quantitative and qualitative alternative methodological approaches and methods. Different specific investigations and writings within the tradition varies in focus on methodology. In some cases, there is a clear inspiration from ethnographic, phenomenological, and/or hermeneutic thinking. Still, the phenomenographic research is best understood as placed within the research methodology and research approach of contextual analysis (Svensson, 2021).