Start with what the student knows or with what the student can imagine?

Kieran Egan




It is commonly argued that one of the securest findings of educational research is that new information, to be best understood, must be attached to knowledge the student already has. Formulations of this finding have been various, but it has been a staple of educational thinking from the time of John Dewey, and before him of Herbert Spencer, to the recently published National Research Council monograph How People Learn: Bridging research and practice (Donovan et al. 1999). In this paper I want to suggest that the common principle of "starting where the student is" may be both inadequate and restrictive in ways little observed.

In its place I will suggest we might sensibly adopt the principle of asking at any point "what the student can imagine" as a starting point for further inquiry.

I will present my argument in two parts. First, I will give reasons why we should be more skeptical than has been common about the first principle, and then show in what way the second has distinct advantages. The first principle has been used both as an element of teaching procedures, and also as a principle for structuring the curriculum. In both cases, I will suggest, the second principle offers us a direction to richer teaching practice and a richer curriculum. For the purposes of this paper, I will focus on the elementary years, though much the same arguments could be made about any stage of schooling.

To show the long influence of the first principle, I will consider its early articulation in the work of Herbert Spencer, and its use by John Dewey and others in shaping the curriculum, and in particular in giving a foundation for the Social Studies curriculum.

Spencer argued that children's early and simple experience had to form the basis for all future learning, and that there must be a regular and orderly progression from what is already familiar to what is slightly less familiar&emdash;an expansion "by slow degrees to impressions most nearly allied" (Spencer, 1861, p. 82). In my experience this principle is believed by nearly every teacher and professor of education I have encountered. Most people assume it is so obviously true that even to question it suggests a degree of nuttiness.

The educational reformers of the late nineteenth and early twentieth-centuries, encouraged teachers to rethink their ways of presenting new material to students in their particular curriculum areas in accordance with the principle&emdash;so elementary mathematics would begin from the experience children will have had with fruit or games such as marbles, and language studies would begin with the forms of expression with which children would already be familiar, rather than with abstract grammar. But those subjects were still not fully articulated with the meaningful daily interactions of children with their local environments--the paradigm of "where the student is". What was further proposed was a new, central curriculum area&emdash;the Social Studies&emdash;which would begin with the material of children's everyday experience, with themselves and their families, their neighborhoods and communities, and gradually expand learning from this meaningful core of students' experience to less familiar knowledge, until, in the end, the whole universe of knowledge could be understood as an expansion from what was most vivid and meaningful to the child. The Social Studies was designed to tie all the knowledge being learned in other curriculum areas to the child's experience.

So, we must start with what is most profoundly known by the student, and build new knowledge on that basis. "If I had to reduce all of educational psychology to just one principle, I would say this: The most important single factor influencing learning is what the learner already knows. Ascertain this and teach him accordingly" (Ausubel, 1968, p. 18). Apart from this slightly odd way of putting Spencer's principle, I think there are four reasons why we might be wary of accepting it. Here are four things that might cause us to worry about it:

First, if this is a fundamental principle of human learning, there is no way the process can begin.

Second, if novelty is the problem for human learners&emdash;i.e., things unconnected with what is already known&emdash;reducing the amount of the novelty doesn't solve the problem. And if we can manage some novelty, why can't we manage more? Perhaps Spencer's "most nearly allied" is far too conservative.

The third objection is less directed at the principle than at the way it has been, as far as I can see, invariably interpreted in education, and in the construction of the elementary Social Studies curriculum. It is assumed that what children know first and best is the details of their everyday social lives. That is, it is assumed that children's thinking is simple, concrete, and engaged with their local experience. But, as I will elaborate in the second half of the paper, children also have imaginations and emotions and these, too, connect with the world. If children's minds are supposed to be restricted to the everyday details of their social lives why are they full of monsters, talking middle-class rabbits, and titanic emotions? We cannot sensibly explain Peter Rabbit's appeal in terms of its "familiar family setting" (Applebee, 1978, p. 75), when it involves a safe forest and a dangerous cultivated garden, and death so close, and so on.

Fourth, and this appeals to old-fashioned intuition, a few moments' reflection should make clear that no-one's understanding of the world expanded and expands according to this principle of gradual content association.

Now, given the almost universal acceptance of this principle, and the fact that these four reasons will have little impact on most who believe it, I should add what seems to me the reason such ideas survive so tenaciously. The principle survives because, like many such knowledge claims in education, it is a mixture of analytic truth and empirical generalization. That is, at some level, the principle is true simply because people define its terms to mean something that can't be other than true. So it is understood to mean something like, you don't know whatever you don't know and if you learn something new it has to fit in with what you can find comprehensible. If you don't speak Chinese, and are told the solution to the three pagodas puzzle in Chinese, you will be unable to understand it. If you do speak Chinese, and know what pagodas are and understand the puzzle and other prerequisites, you will be able to understand the solution. At this level, the principle isn't very helpful. What would make it interesting are reliable empirical generalizations; that is, research showing conditions that constrain learning that are other than logical truths, and these are not thick on the ground.

Well, that's a bit clotted, trying to say too much too briefly. But it suggests why one might come to question the common interpretation of starting where the student is. If we mean it as simply an analytic truth, it is of no particular interest to educators. If we mean it as an empirical claim, as it has been taken at least since Spencer's time, then it is doubtful for the reasons given above.

The second part of the paper suggests that we consider what might be implied by taking the principle of what the student can imagine seriously. For obvious reasons there has not been much research on students' imaginations, and yet they are clearly central to students' learning. By largely ignoring the imagination because our research methods have difficulty coming to grips with it is somewhat self-defeating.

First, I will give reasons for considering the imagination as the ability to think of things as possibly being so (cf. White, 1990). This odd ability, which seems to be most energetically active in our early years, is clearly not some casual frothy part of our mind's functioning, to be blown away with the growth of rationality. It is a hard-working core of children's thinking (cf. Warnock, 1976).

Second, I want to look in some detail at just three features of children's imaginative engagement with the world, and show how they challenge the first principle that has had so significant an impact on teaching and curriculum thinking. The first feature concerns the story form, which children engage so enthusiastically very early in life, and in all cultures and in all times that we know about. The second concerns the kinds of abstract and affective concepts that give structure to the fantasy stories common across the world--such concepts as good/bad, brave/cowardly, rich/poor, anxiety/security, and so on, with all the freight they carry for human cultures. The third follows from the content of children's fantasy stories, and why they are full of creatures like Peter Rabbit, a talking, middle-class animal such as cannot exist.

Clearly storytellers don't pay much attention to the first principle, in the way that the designers and defenders of the elementary Social Studies curriculum do. They boldly begin with galaxies far far away and long long ago, if it suits them, confident that their audience will make sense of the content. They can introduce strange characters, and weird situations, like Harry Potter's or Gandalf's and Frodo's, as long as they build their narratives on the kind of abstract concepts students are familiar with, like good/bad, brave/cowardly, anxiety/fear, etc.

If we consider how children learn about the temperature continuum conceptually, they begin with the concepts 'hot' and' cold', as these are both logically and empirically the first concepts that can be meaningful--'hot' being hotter than the child's body and 'cold' being colder. They can elaborate or mediate these concepts by learning the concept 'warm', for example, and then learn an array of other temperature terms in the context created by the first terms that form the ends of the continuum. Later they will learn to use theoretic terms, like thermometer numbers, to refer to temperature. This procedure, which is very effective in dealing with the material world, creates problems when applied to concepts that have no mediating categories, like life/death, or nature/culture, or human/animal. The world of children's imaginative lives is, however, filled with generated mediations between these logically discrete concepts. If there is nothing between life and death, we invent ghosts, that are to 'life' and 'death' as 'warm' is to 'hot' and 'cold'. Between nature and culture we invent that menagerie of talking middle-class animals that fill the fantasy stories of children and the myth stories of the world.

Something serious is going on here that has little to do with starting where the student is and gradually building on the knowledge they have--except if we define these as analytic truths. The connections are made by metaphoric leaps, not by logical connections. My paper is a meditation on what we might need to give up of our assumptions about learning and curriculum if we acknowledge the importance of student' imaginations. For teaching practice, we need not become constrained by trying to make content associations with knowledge students already have--there are other ways of expanding knowledge. For the curriculum, we need no longer be constrained to tie knowledge to the everyday experience of students, which can be very dreary for them, but can recognize that their imaginations allow much freedom in how they can go about grasping the universe of knowledge.

It will be necessary to conclude by pointing out the obvious--that I am not arguing for ignoring the first principle and ignoring the student's prior knowledge and everyday experience. Rather I am arguing that these have been taken as implying greater restrictions on children's learning and curriculum possibilities that is warranted when we consider their imaginative lives. We can start with what they can imagine.




Applebee, Arthur N. (1978). The child's concept of story. Chicago: University of Chicago Press.


Ausubel, David P. (1968). Educational Psychology: A cognitive view. London: Holt, Reinhart, & Winston.


Donovan, Suzanne M., John D. Bransford, and James W. Pellegrino (eds.) (1999). How People Learn: Bridging Research and Practice. Washington DC: National Academy Press.


Spencer, Herbert. (1861). Education: Intellectual, moral and physical. London: G. Manwaring


Warnock, M. (1976). Imagination. London: Faber.


White, Alan R. (1990). The language of imagination. Oxford: Blackwell.


Return to Home Page