Ference Marton: In The International Encyclopedia of Education. Secon edition , Volume 8. Eds. Torsten Husén & T. Neville Postlethwaite. Pergamon 1994, pp. 4424 - 4429.

Phenomenography

Page I (II)

Origin
Phenomenography is a research specialisation with its roots in a set of studies of learning among university students carried out at the University of Göteborg, Sweden, in the early 1970s. The point of departure for these studies was one of the simplest observations that can be made about learning, namely that some people are better at learning than others. This straightforward observation led to the first question which was to be investigated empirically:

1. What does it mean, that some people are better at learning than others?

which in its turn led to the second question:

2. Why are some people better at learning than others?

There was an ambition from the very start to take as little for granted as possible. Learning was studied under comparatively natural conditions and the aim was to describe it through the eyes of the learner. The studies were conducted by holding individual sessions in which a student was asked to read a text which was either taken from a textbook or had that character. The students were informed that after reading the text they were going to discuss their understanding of it with the experimenter. Thus, after completing their reading, the students were duly interviewed about what they understood the text to have been about. Sometimes more specific details were also taken up. In addition, they were asked to give as full an account of the text as possible. After that, the interview continued with questions about their experience of the situation and they were specifically asked how they had gone about learning the text.

All the interviews were tape-recorded and subsequently transcribed verbatim. On scrutinising the transcripts of the students' accounts of how they had understood and remembered the text as a whole, a limited number of distinctively different ways of understanding what the text was about could be identified. Furthermore, these different ways of understanding the text were seen to be in logical relationship to one another - of inclusion, or exclusion, for instance. Each of the different understandings was described very carefully, to bring out its special characteristics in relation to the others, thus forming a set of what came to be called categories of description. By drawing on the logical relationships found between the different ways of understanding the text, a hierarchy was established between categories of description. Such an hierarchically ordered set of categories is called the outcome space. The outcome space thus depicted the different ways in which the text had been understood; by referring to this outcome space the categories of description could be compared with one another to judge how appropriate, in relation to specified criteria, was the understanding they represented. This line of reasoning applies, of course, not only to the understanding of the text as a whole but also to the various topics dealt with in the text.

The outcome space provided us with an instrument for characterising - in qualitative terms - how well learners succeed with their learning task. We had thus arrived at a way of answering the question, "What does it mean that some people are better at learning than others?" (Marton, Hounsell, Entwistle, 1984).

The characterisation of the qualitative differences in the outcome of learning was based on the students' accounts of their understanding and remembering of the text as a whole or of certain parts of it. When the transcripts of the students' accounts of how they had experienced the situation and of the way in which they had gone about the learning task were analysed, again, some striking differences were found. For some of the students the text they were reading was transparent, in a manner of speaking, in that they were focusing on what the text referred to; they were trying to understand what it was about. Other students - who recounted experiencing the situation such that they were expected to recall the text after reading it - focused on the text as such, trying to move it into their heads, as it were. The former way of relating to the learning situation was called the deep approach and the latter the surface approach. It was found that the deep approach was closely associated with "higher" categories of outcome (i.e. better understanding of the text) while the surface approach was associated with "lower" categories of outcome (i.e. more shallow understanding of the text). There was thus a strong relationship between the way in which the students understood the content of learning (the text) on the one hand and the way in which they experienced the learning situation (and their own act of learning), on the other. The two aspects of learning, the content aspect and the act aspect, are, of course, two aspects of the same whole.

Thus the second question posed at the outset of this research enterprise could be answered, at least in part. "Why are some people better at learning than others?" Because people differ in their approach to learning tasks (Marton et al., 1984).

Further research demonstrated that the relationship between approaches to learning on the one hand and the qualities of the outcomes of learning on the other is invariant across forms of learning other than learning by reading, even if the specific natures of both the approaches and the outcomes vary both with the type of learning activity - for example, essay writing (Hounsell, 1984), listening to lectures (Hodgson, 1984), problem solving (Laurillard, 1984) - and with the specific content.

At the focus of this first set of studies was the set of different understandings of some specific content which learners developed in a certain situation; sense was made of these in terms of differences in the approaches the learners adopted to the specific learning task - i.e. in terms of differences in their way of experiencing the specific situation. The second step in developing the phenomenographic research orientation was to shift the focus of interest away from that which emerges in a specific situation and towards the learners' preconceived ideas about the phenomena dealt with in the specific situations. The way in which children understand numbers, for instance, is of vital importance for the way in which they deal with arithmetic problems (Neuman, 1987; Marton & Neuman, 1990); the way in which students understand matter is of vital importance as far as their understanding of chemical reaction is concerned (Renström et al 1990) and so on. Detailed knowledge of the ways in which learners understand the central phenomena, concepts and principles within a domain prior to study is believed to be critical for developing their understanding of the central phenomena, concepts and principles, and hence for their mastery of the domain (Bowden et al, 1992).

That branch of development of phenomenography dealt with the content aspect of learning. The act aspect of learning was also considered in that, similarly, learners' conceptions of what learning actually is are crucial for the way in which they experience the act of learning, and thus for what approach they adopt in relation to specific learning tasks (Säljö, 1982; Marton, Dall'Alba, Beaty, 1992).

The recurring principle in all the investigations quoted here is: whatever phenomenon or situation people encounter, we can identify a limited number of qualitatively different and logically interrelated ways in which the phenomenon or the situation is experienced or understood. Naturally enough, in subsequent studies this principle was found to be applicable to phenomena and situations well outside the educational context in which the above initial studies, described above, had been carried out. Theman, (1983) explored conceptions of political power, Wenestam (1984) investigated ideas of death and Marton, Fensham and Chaiklin, (1992) studied Nobel laureates' views of scientific intuition, for instance.

From some empirical studies of learning in higher education phenomenography thus evolved as a research specialisation aimed at "describing conceptions of the world around us" (Marton, 1981). This research specialisation will now be characterised in terms of its object of research on the one hand and in terms of the methods used when studying this research object, on the other.
Object of research

Phenomenography is the empirical study of the differing ways in which people experience, perceive, apprehend, understand, conceptualise various phenomena in and aspects of the world around us. The words experience, perceive ... etc., are used interchangeably. The point is not to deny that there are differences in what these terms refer to, but to suggest that the limited number of ways in which a certain phenomenon appears to us can be found, for instance, regardless whether they are embedded in immediate experience of the phenomenon or in reflected thought about the same phenomenon. The different ways in which a phenomenon can be experienced, perceived, apprehended, understood, conceptualised etc., according to our way of describing them, are thus independent of the differences between experience, perception, apprehension, understanding, conceptualisation etc.

This point can be illustrated by an example taken from a piece of phenomenographic research. One of the ways in which young children experience numbers is as what Neuman (1987) calls "finger-numbers". According to her, children frequently "lay" the numbers 1 to 10 on their fingers, calling one of the little fingers (usually on the left hand) "1", the ring-finger "2" and so on. Numbers larger than 5 are then understood as 5+ some fingers. In carrying out simple arithmetic tasks children try to keep "the undivided 5" together. Hence when solving problems like 2+7=? they reverse the addends and transform the problem to 7+2=?, where 7 is "undivided 5"+2, and the problem as a whole becomes (5+2)+2=?

What, then, is a conception of something - or a way of experiencing something? (Here, it is noted, the two expressions are being used interchangeably.) It is not a mental representation or a cognitive structure. It is a way of being aware of something. One might be aware of 7 when one perceives it as 5+2 when one looks at one's hands (or as 6+1 or 4+3), it might be an immediate experience of the number 7 or it might be the result of reflection, or there are still other possibilities. In all cases, however, 7 is seen as a sum of two parts, 5 and 2 (or 6 and 1, or 4 and 3). Awareness is a relation between subject and object. Furthermore, when something is the object of attention it is always seen, or thought about or whatever, in some way, by somebody. We simply cannot deal with an object without experiencing or conceptualising it in some way. In this sense subject and object are not independent, but they form a unity; there is a relation between them which can be called their internal relation. Subject and object are what they are in relation to each other. Following from this, a way of experiencing or understanding a phenomenon says as much about the experienced, understood phenomenon as it says about the experiencing, understanding subject.

Asplund, Marton & Halász (1992), studying the qualitatively different ways in which secondary school students understood one of Franz Kafka´s short stories, argued that their work not only illuminated how young people make sense of literary texts, but was in fact a contribution to research on the interpretation of Kafka's work. In a similar way Lybeck, Marton, Strömdahl & Tullberg (1988) argued that they have made a contribution to the characterisation of "the mole-concept" in Chemistry through their study of secondary school students' differing understanding of that concept.

The nature of experience
An experience or a conception of a phenomenon - the internal relation between subject and object - is a way of delimiting an object from its context and relating it to the same or other contexts and it is a way of delimiting component parts of the phenomenon and relating them to each other and to the whole (Svensson, 1984). The delimitation from and relating to a context is the external horizon of the phenomenon. The delimitation and relating of parts is the internal horizon of the phenomenon. The external and the internal horizons together make up the structural aspect of the experience. There is a corresponding referential aspect in the meaning inherent in the experience. Let us consider an example taken from Neuman's (1987) work, where the relation between the structural and the referential aspects of two different ways of understanding numbers is illustrated through the description of a change from one to the other. A seven years old, just started at school, solved the problem "If you have two kronor (crowns) and you get seven more - how much money have you then altogether?" in the following way.

He starts off with a so called counting-on procedure saying "2 ... 3, 4 ...".The idea is of course to add 7 units of which "3" is the first, "4" is the second and so on. As he does not use any keeping track procedure for he could not possibly know when he has uttered exactly seven number words. He simply can not hear "the seven-ness" of seven. Now, what in actual fact happens is that he pauses upon saying "7" and then says "8,9" and declares that the result is 9 (the full sequence reads as follows: "2 ... 3, 4, 5, 6, 7 ... 8, 9). An interpretation - in our view highly reasonable - is that although this little boy is trying to add seven units to two to begin with, when he says "7" he realises all of a sudden that this "7" can be seen as the last unit in the addend "7" if only "7" is placed as the first addend, instead of seeing the same unit as being some way through the second addend.

We can see how the structure of the sum as experienced changes from "1,2... 3,4,5,6,7,8,9" to "1,2,3,4,5,6,7, ... 8,9". Here it is internal structure of the sum that changes. (No obvious change in the external horizon can be noticed. The little boy is probably delimiting this problem from the situation at large and relates it to other number problems). Corresponding to the change in the structural aspect of the experience there is a corresponding change in its referential aspect . The meaning of each number changes, the meaning of "1" and "2" changes from being the first and second unit in 2 to being the first and second unit in 7. The meaning of "3", "4" and so on change from being the first, second and later corresponding units in 7 to being the third, fourth and so on units in 7. The meaning of "8" and "9" changes from being the last two units in 7 to being the two units in 2.

The structural changes cannot come about without the changes in meaning. Nor can changes in meaning come about without changes in structure. The structural and the referential aspects thus dialectically constitute each other. Neither is prior to the other.

Page II (II)