Towards a Philosophy of Instruction
J. Michael Spector
I had the pleasure to participate in two meetings in 1999 and a third one in 2000 in which the topic of the future of instructional design was the focus. The first occurred in March at Twente University in the Netherlands, organized by Sanne Dijkstra and partly sponsored by the Instructional Design Special Interest Group of the European Association for Research on Learning and Instruction (EARLI) (Spector, 1999). The second was a follow-on to that meeting, held at the University of Bergen in Norway (Spector & Anderson, in press). The third was held in New Orleans and sponsored by the Structural Learning, Instructional Systems and Intelligent Tutors Special Interest Group of the American Education Research Association (AERA). All three meetings involved prominent instructional design researchers from Europe and North America. All were concerned with the impact of technology on our discipline. All included specific discussions concerned with epistemology and ontology. Several themes emerged from these meetings.
There appears to be renewed interest in foundation issues, especially with regard to philosophical foundations. There is clear interest in how to support learning in and about complex domains. There is renewed interest in integrated and holistic perspectives of instructional design and as well as of learning (the emphasis on the latter in the constructivist literature is perhaps more well known). Finally, there is a concern that in order to make progress it is vital to conduct experiments and establish viable links between - and not merely encouraging words about - theory and practice.
The purpose of this short essay is to explore the landscape of such concerns in the form of a set of tentative First Principles for Instructional Design. These might become the basis for further debate and discussion that could eventually lead to a well elaborated philosophy of instruction. That is my hope. If I manage to continue the discussions already underway and provoke them a bit more, then I shall count this effort a success. I realize that much work along these lines has already been done, most obvious in two volumes edited by Reigeluth (1983, 1999). There is much yet to be done.
Definitions and Progress in Instructional Design
In this essay, the key terms that I adopt are ‘instruction’, ‘learning’, ‘knowledge’ and ‘philosophy’ and I use them in the following way. Instruction is that which supports or facilitates learning. This implies that the kind of learning that is of interest is planned or intentional learning, which involves a goal of some kind. Learning essentially involves a change in abilities, attitudes, beliefs, capabilities, knowledge, mental models, patterns of interaction or skills. These changes may be localized within an individual, a group of individuals, an organization, or perhaps even a society. To be properly called learning, such changes should tend to persist over time and across a variety of distractions. The fact that changes have occurred should be either directly or indirectly observable. I shall call this definition the Learning Principle (L), since it might be challenged by some, and it is possible to investigate its implications for the design of instruction.
Much of the instructional design literature is devoted to distinguishing among the kinds of things that can be learned and the kinds of changes that can occur. The list of things which might change (abilities, attitudes, and so on) is intentionally long and intended to be broad. The items in this list are drawn from both learning metaphors presented by Sfard (1998): learning as the acquisition of knowledge and expertise, and learning as participating in a community of practice. She argues that there is a place for both perspectives. The view that these two broad perspectives are intertwined and necessary constitutes what I call an integrated perspective (this also forms the basis for the Integration Principle discussed below; see Spector, 1994, 1995).
Philosophers have historically differentiated knowledge from belief. Knowledge was historically reserved for those things about which we could be certain (such as the conclusions of valid mathematical proofs). Beliefs, then, were only more or less probable and likely to be subject to much change and instability. This use does not fit common usage very well and is no longer the only acceptable philosophical view. Indeed, the more modern notion is that we build up our knowledge in stages, in bits and pieces. This perspective is generally regarded as a kind of naturalistic epistemology in contrast to the rational epistemology that has dominated the history of philosophy. Naturalistic epistemology found its way into educational research through the works of Dewey (1916), Piaget (1937, 1970) and Vygotsky (1978). These philosophers of education examined the development of human thinking in its many and varied natural settings. In short, it is modern educational researchers who have taken the empiricist imperative - human experience must be the starting point for understanding - to heart. This natural and close connection to experience, both from a design perspective and from a learning perspective, provides the starting point for a philosophy of instruction. I call this starting point the Experience Principle (E): understanding begins in and is based upon human experience. This principle has clear and specific implications for design as well as for learning and assessment of outcomes. The arbiter of what works in instruction is not good intentions or high ideals but, rather, a careful and continuous assessment of outcomes and attempts to link observable effects to underlying causes.
Gagné and Merrill (1990) published a landmark article that recognized an important theme and trend that is more typically associated with constructivists (e.g., Duffy & Jonassen, 1992) and advocates of situated learning (e.g., Lave, 1988). The theme in that paper is that the proper object of instruction is most typically and most often an integrated and purposeful human activity - an enterprise. Of course this theme is not new. It can be traced back to earlier research, such as Piaget's (1937, 1970) work concerning the experiential setting in which children developed understanding and Vygotsky's (1962, 1978) examination of the social context in which conceptual understanding developed. A broad expression of this notion is the Context Principle (C): Context is relevant for learning and the construction of meaning. Therefore, context must be taken into explicit consideration when planning instruction.
Most instructional designers accept some version of the Context Principle either explictly or tacitly. Instructional designers, in other words, do have at least an implicit philosophy of instructional design and that philosophy is related to identifiable philosophical principles (such as might be found in the literature on naturalistic epistemology, pragmatism, and so on). Differences involve what might be considered a relevant context. For example, relevant to mastery of a step in a procedure might be an understanding of the general purpose for the whole procedure and a general view of how it is performed. This is why scenario-based and situated approaches have long been used in training complex technical skills (Dijsktra, 1997; van Merriënboer & Dijkstra, 1997).
Closely associated with the Context Principle and one that has been explicitly addressed by many instructional design researchers (Duffy & Jonassen, 1992; Gagné & Merrill, 1990; Merrill, 1993; Spector, 1994, 1995) is what I shall call the Integration Principle (I): human experience does not come neatly compartmentalized. It must be considered and understood in the context of a setting that typically includes the following: other individuals, a variety of goals (some of which might not be explicit and might conflict with others), various artifacts and technologies, activities, and cultural, organizational and societal influences. One implication of the Integration Principle is that a broad and possibly multiple approaches to learning and instruction is necessary (see Sfard, 1998, for example). A second implication is that many of the distinctions that instructional designers have been inclined to make over the years are not necessarily helpful in designing, facilitating or evaluating learning. Goodyear (2000), for example, argues that the following recurring distinctions can lead us to adopt exaggerated positions and forget that the starting and ending point should be human experience:
Figure 1. Recurring distinctions questioned by Goodyear.
The Integration Principle has implications for design as well. As van Merriënboer and Dijsktra (1997) argue, whole task activities are more likely to be supportive of learning than part task activities (they identify exceptions to this, of course, such as when particular sub-skills must be practiced in order to acquire automatic performance). The most common distinction between declarative and procedural knowledge is called into question by this principle. Most typically, human activities involve both kinds of knowledge and completely separate treatment in an instructional sequence is therefore likely to be sub-optimal. A background principle often cited by designers is that people will learn what they do. What people most typically do is apply both declarative and procedural knowledge to accomplish a particular task. Therefore, it is worth designing instruction to include both types of knowledge brought to bear in an explicit fashion in a learning activity.
The final principle that I would like to propose for this initial set of principles is the Uncertainty Principle (U). Basically, this principle is a reminder that the instructional enterprise of instructional is complex and our knowledge of relevant aspects is incomplete. We do not know in a complete or comprehensive all of the factors and mechanisms involved in learning. Educational research in general produces provisional and tentative conclusions. If one adopts Reigeluth’s (1983) view that instructional design is primarily a prescriptive enterprise aimed at a set of principles to be used to guide the development of optimal learning solutions, then instructional research is fundamentally about what works best in various situations and circumstances. While we are inclined to generalize findings, we should realize that there are inherent limits to the generality and certainty of our findings. In short, we may know less about learning and instruction than we are inclined to believe.
Concluding Remarks: A First Set of Principles
I have presented a rough sketch to support a set of five basic principles to consider as a foundation for a philosophy of instruction. There are as follows:
Whether or not such principles are useful for the design of instruction or have any observable impact on learning outcomes is perhaps worth some discussion. I do not know if such principles are useful but I do believe that some first principles are worth discussing. The instructional design research community has been through a period of unnecessarily divisive debate. We have seen new technologies emerge and a great deal of enthusiasm for the potential of learning technology. Many exemplary projects exist to illustrate what is possible. However, there is too little research being done to establish links between instructional design and learning outcomes. Studies to replicate previous research are quite rare. In my opinion, the instructional technology research community is in jeopardy of losing its scientific underpinnings. Some may celebrate this as a long overdue outcome; I would regard it as a serious loss.