ABSTRACT: The current collection of theories of instructional design assembled and discussed by Reigeluth (1998a) serve as a backdrop to this attempt to lay out what we might consider as the requirements for a full theory of instructional design. Such a theory would show characteristics of fullness: comprehensiveness (its coverage of all domains), abstractness (encompassing all processes), utility (wide applicability), validity (grounded in psychology). Theory building involves optimal selections among these characteristics, with trade-offs being made. The ideal of a full theory is suggested and steps to take to build such a theory are laid out.
A theory of instructional design is an organized set of prescriptions that assists in the preparation of instruction. On the face of it, a theory is but a descriptive tool that underlies a procedural model of design for the preparation of instruction. It is such a model or methodology that gets used by the instructional designer or the instructor for readying the actual execution of instruction.
But a theory is also much more than that as well. As Reigeluth brings out excellently in his book on theories of instructional design (Reigeluth, 1998a), a theory involves choices that are rooted in deep beliefs, beliefs about values to underlie education for instance. A theory, then, is a synthesis of one's deep-seated beliefs about the educational process, particularly about what is important to learn and about how learning occurs. A theory is in essence a set of theoretical choices, a stance taken and proposed.
A theory is a collection of choices made about what is important to consider.
Two core issues arise out of this fact. First, can the choices be overcome--or expressed another way, why must choices be made in the first place? Second, what is the best way to propose a theory--or again, what is an appropriate form for a theory? It is these two issues that are explored here. I want to first question our current theory building practice, thereby identifying serious limitations that hamper us in instructional design theory, and then ponder whether it is possible to create the framework for a Physics-like grand theory of instructional design that would unify the current elements of disparity that we witness.
I shall not be proposing such a theory, only setting the scene for it to be proposed by the more creative theorists, young and older, that may be attracted to this challenge. Prolegomena are the underlying elements that can help in the crafting of theory. They constitute a prior discussion of the framework in which theory is crafted .
The initial impression one gets from the new batch of theories anthologized by Reigeluth (1998a) is one of diversity: so many theories and so at odds with one another that one is left perplexed. My unease arises from the question, Is this a healthy situation, one where creativity is blossoming, or one where Babel reigns instead?--an admittedly value-laden question. I believe that the latter is more representative of the situation than the former.
The current set of theorists set forth frameworks for designing instruction that they believe to be optimal. Emphasis is placed here by one, there by another, and so on--that is after all the choice-making that theorizing involves. Each theorist says in effect "This is what is important to pay attention to, this is what needs to be dealt with for successful instruction." For instance, elaboration theory (Reigeluth, 1998b) emphasizes contextualizing knowledge elements within their more holistic topical structures, whereas constructivism (as represented by Jonassen, 1998) emphasizes authentic problem-solving. Along the same lines, but from a different angle, some of the theories focus on conceptual outcomes of learning, while others deal with affective outcomes, and so on.
Emphasis in theory-building is fine as long as it does not inadvertently reduce the scope of the theory. This is a crucial matter: how a theory is expressed in terms of scope is of the utmost importance. Consider the following cases. The attitudinal theory crafted by Kamradt & Kamradt (1998) is, for all to see, restricted to the attitudinal domain of learning outcomes. It is highly explicit about its scope. That is not nearly the case, however, with elaboration theory, which does not express its scope in a highly explicit manner (it is applicable to complex content). Nor with most of the theories we are dealing with here, this despite Reigeluth's (1998a) grouping of the theories in broad categories of applicability and his editorial remarks.
Theory building is perhaps naturally open to applicability fuzziness.
In effect, it is generally up to the user to figure out the scope of the theory. This is a major theoretical failure, one which undermines progress in this field. As Tennyson & Schott (1997) put it in discussing the current diversification of instructional design theory and practice, "it can lead to fruitless separations in our discipline so that specialists tend to forget the forest when looking to their trees of interest" (p. 13).
Despite this, they predict that the pursuit of a single, all-encompassing instructional design theory will be abandoned in favor of an interactive network of theories, such as those represented in their book (Tennyson, Schott, Seel, & Dijkstra, 1997). The same applies to the set of theories brought together by Reigeluth (1998a), who explains the diversity of theories in terms of their situational specificity and the role of deep-seated values in orienting choices.
I want to argue here that the very direction of instructional design theorizing as foreseen in these important contemporary collections is mis-oriented. We need to seek the forest (out from behind the trees)--that is our mission. It is not a holy grail, but something that should be eminently do-able, even if not easy. Taxonomies can help of course, as suggested by Snelbecker (1998) and discussed by Reigeluth (1998c), but they merely indicate the way to greater integration--they are a first step in laying out the landscape.
Theoretical integration should currently be our prime challenge in instructional design.
Instructional design is an artificial science (in contrast to natural sciences) in that it is oriented to human goals (see Reigeluth, 1998c). It is a technology in this respect, a way of accomplishing something. As a science, it has only one aim: specifying the means towards goal satisfaction. The underlying scientific assumption is that given a well-specified goal, there is arguably one best way to achieve it. The proposition that means and ends interrelate is not at issue here: different ends will require different means. At issue, rather, is the pluralism factor: is there one best way?--or is it that "for any given goal, there is almost always more than one method that can be used to attain it" (Reigeluth, 1998c)?
The answer to this issue most likely depends on the granularity of the analysis. At one level, two methods may well be considered to be roughly equivalent, but if we were to analyze in great detail all the factors involved in a given situation, one of the two would come to the fore. The question is a practical one: is it worthwhile (or even possible) to conduct that detailed analysis for the anticipated gain in effectiveness or efficiency? In practice, perhaps not (and instructional design is a practical science); but in theory (the goal being theoretical), it remains true that there is but one best way.
Pluralism may well be a mistake of grand proportion.
I explore below three issues that can perhaps help us see the problem in a larger context and thus lead us to overcome it. I take a strongly critical stance in these meta-theoretical matters. Others will likely have other views.
The crux of the problem I see can be expressed thus: instructional design theorists typically (and quite naturally, one should add) develop their theory of instructional design in accord with their selected educational philosophy, even though this philosophy is often not explicit. The force of Reigeluth's argument that instructional design theories are grounded in values--both regarding what should be taught and how it should be taught--lies in this recognition of the diversity of theories. This leads to their islandship (their relative incommunicandor, whereby they largely remain isolated from one another, with little communication in terms of criticism and cross-pollination).
This problem has many facets: a political one, related to curriculum development, a psychological one, related to the issue of learning, and a sociological one, concerned with scientific community. Let us see what each involves.
What I call the political facet concerns who decides what is to be taught. This is the realm of curriculum theory. It appears that theory in this arena is also rather unsteady at the moment. Egan (1997) suggests very convincingly that much of the ills of education today find their source in the schizophrenic mission (my term) that education gives itself. He proposes essentially that the three standard aims of education are at root incompatible, to the point of working against each other, thus leading to confusion and inherent educational incoherence. The three standard aims are socializing the young, teaching them forms of rational knowledge, and helping them realize their unique potential.
The socialization aim involves a certain homogenization of children (the social shaping that collective life demands), the rationalization aim enters children into "the great cultural conversation [via] the hard academic disciplines that will make them knowledgeable, discriminating, and skeptical" (Egan, 1997), and the Rousseauean aim of fulfilling the individual potential focuses on the individuality of experience. It is easy to see how each of these aims can undermine the others and how they can, individually, lead to divergent theories of instruction, and hence divergent theories of instructional design.
Incoherence of goals in education starts everything off on the wrong foot.
Egan's theoretical analysis illustrates well the type of decision-making that occurs to create the values underlying what we teach. As in other realms of analysis, its impact lies in the explication of the underlying assumptions of the different views. All this is prolegomena to curriculum theory.
But what I really want to get at is this: instructional design theorists need not only make very explicit their underlying educational values, they need also refrain from curriculum theorizing. Mixing the two realms, particularly when it is covert, deeply muddles the result. This is best illustrated by constructivist thought, although it appears pervasively in instructional design theorizing.
Refraining from dealing with the aims of instruction is very demanding, I know. But it is essential if we are to develop some science of ID. Here is the point: any IDT should be able to take whatever the client comes up with as the goal of instruction and address that goal in some proper fashion. Whether the goal is to enhance one's creativity in some general way, or whether it is to learn the multiplication table, a well-formed instructional design theory should handle it. Is this asking too much? Yes, if we look at instructional design theory today. No, if we are serious about doing instructional design theory.
It is imperative to refrain from goal-setting, which is a main muddling factor.
The reason to refrain from goal-setting is simple: that activity is the province and responsibility of other parties, namely the particular stakeholders in a given educational context, or those acting on their behalf. Instructional designers, and instructional design theory, should have no say in what is taught. Their say, an immensely valuable and respected one, is in how to best prepare to teach whatever is determined should be taught. The politics of that determination is not in the realm of ID, and instructional design theory usurps its rights and privileges in engaging in these politics.
Instructional design is not unlike architecture in that respect: it builds to the client's specifications, but does not determine those specifications itself. Architects (and instructional designers) might assist a client (or a client group) in suggesting models to examine or processes to engage in to determine the goals, but the choices are not fundamentally theirs. This is not to say that goals are unimportant to education, or that anything goes! We all agree on the great importance of curriculum dialogue and decision-making in education. We simply need to accept the legitimacy of the political process and not attempt to circumvent it through our own proselytizing.
What this stance amounts to is a repudiation of the place of values in instructional design theory. It should strive to establish itself as beyond values. Remaining stuck in the value-laden politics of education will prevent it from attaining its full potential as an artificial science.
Moving now from the political (the what should be) to the psychological, we enter yet another domain in need of revolution (is it a "fin de siecle" phenomenon we are witnessing?). I speak of learning theory and refer you to the summative account of Mayer (1996) among others (e.g., Sfard, 1998). Mayer describes well the three conceptions of learning that currently occupy the principal theoretical scene in educational psychology: associationist, cognitive, and constructivist learning.
The muddled scene comes from the fact that these three "facets" of learning are not integrated into a comprehensive theory that could deal with their relationships. Instead, in a mirror-like way to instructional design theories, they are generally treated as alternative theories of learning, theories that might have some restricted domain of applicability of typically unspecified scope. Now, the fact that the learning theorists have not performed properly is truly vexing, given that any instructional design theory builds on that educational psychology foundation, and it may be cause for excuse for instructional design theorists as a consequence. It certainly adds to the burden of instructional design theory since it must now add specificity of its own regarding learning processes. Skiing and the multiplication table may have a lot in common (both are skills) and not much in common with genetic models or statistical analyses (rather more complex), but all have to be dealt with by instructional design theory, building on how we assume each is learned. We simply cannot wait for psychology to consolidate its own learning theories--we need to charge ahead.
Learning theory is itself in need of strong consolidation.
The issue of concern is that an instructional design theorist's own inclinations with respect to learning theory will likely play a major role in theory formulation and must therefore be made explicit. The fuzziness of learning theory makes that difficult of course, as does also the link between content and learning processes. Mathematical procedures for instance (as indeed many things) can be approached in two distinct ways: by rote or through modeling. The learning processes are distinct. But while the instructional design theorist must remain aloof of content decisions, she or he cannot remain uncommitted in terms of learning.
An additional difficulty, strongly related to learning theory, is the view one takes of knowledge. I refer here not to epistemological issues of the sort raised by constructivism (I consider this epistemology to be a red herring for instructional design theory--the true issue being task authenticity rather than objectivity), but rather to the ways in which we categorize knowledge. Bloom's taxonomy of learning (1956) and Gagné's categories of learning (1985) have been influential but have forgone much of the role they once played in knowledge analysis within ID, particularly since the emergence of constructivism and collaborative learning within contemporary educational thinking.
While categorizing knowledge is highly philosophical, learning is not so, even though the former informs the latter. Learning theory is not a prescriptive science, but a natural science aimed at describing the natural processes involved in learning. In principle, there is therefore scope for potential unification within the study of that phenomenon, more strongly so, again in principle, than within instructional design theory. Should such a consolidation develop, it could be a tremendous boost for a parallel consolidation within the fractionalized realm of instructional design theory.
And now to the sociological facet, the most controversial perhaps. I believe a major problem in instructional design theory is an evident lack of critical comparisons between theories. It is as if we have so many theories that cross each other in the dark and are afraid to ask each other the time of day. More interchange--critical at that--is sorely needed (Snelbecker, 1998, makes a similar plea). Bring back the days of Skinner and Chomsky! The field of instructional design theory cannot afford to have the panoply of theories it does go unchallenged, and it is mainly the role of theorists themselves to define their theories in contrast to those of others.
Some would say that strong theories themselves are dangerous, as Thomas (1997) has done in arguing that such theories inadvertently put us in blinders and restrain creative adaptations. There is surely a danger there, but it would seem to be highly avoidable by the development of a critical tradition within the field. I believe we currently have too many, rather than overwhelming theories. Consolidation through critical analysis is highly needed.
Theorizing in our field is not so much a process of bringing the existing facts together (as the classical view of theory would portray it), as an analytic process of developing interplaying models to describe and promote a view of design. It is an area of work that is eminently open to critical analysis and intellectual confrontation.
The strong confrontational exchanges over the value of constructivism that took place in the first half of this decade (in conference panels, journals, and magazines--see Molenda, 1997), had they remained gentlemanly and more focused, could have served as a process model in this respect. The philosophical underpinnings of the debate was to undermine that process, however, with more passages in the dark than true theoretical evolution. It just may be that the sociological set-up of scientific endeavor in the social sciences (and perhaps more broadly as well) favors such an outcome. Anderson (1996, chapter 6), in examining communication theory, discusses how scholarly arguments must go about these processes of claim making in order to impact the field. The socio-psychological aspects of scientific theorizing heavily constrain the needed dialogue.
Theoretical shyness at critical interchange must be overcome.
And yet, if we idealistically hope to advance instructional design theory, we must at some point start engaging in more cross-fertilization and dialogue. The new anthology put together by Reigeluth, as well as the book by Tennyson, Schott, Seel, & Dijkstra (1997), merely start this process. The theorists themselves must confront each other more--and we must find ways to help them along in this respect.
Now that we have considered these three facets of instructional design theory (viz. the political, psychological and sociological), we need to turn directly to instruction in order to consider its essential functions and then finally consider what is involved in building a full theory of ID.
It was argued earlier that a good part of the diversity we witness in instructional design theory arises out of the issue of what to teach. In parallel to this, Reigeluth (1998c) ascribes the main differences in perspectives to the underlying values theorists bring to their theorizing. Apart from content, values relate to assumptions about learning (as discussed above), and to views about teaching itself of course (instruction).
Given that instructional design theory has the goal of specifying the design of instructional products and processes, it is essential to ask ourselves what the nature of instruction is. Not in its details, but fundamentally, in its functional core meaning. This is an essential part of a prolegomenon to instructional design theory.
At its most basic, instruction is an intentional effort to influence the thinking of another person. This is a very broad scope for a definition, but it is needed in order to focus on the core elements of instruction. In more elaborate terms, this effort to influence is typically undertaken by an instructor (teacher, professor...) and directed to a specific audience (learners, students...) with the consequent influenced thinking resulting in learning.
Instruction, at its core, is an intentional effort to influence.
Let us first consider the intentionality of instruction. People learn all the time, on their own, without instruction, for instance by reading books, exploring the web, talking to colleagues, etc. Instruction goes further: it is a deliberate attempt to organize learning in a focused way. Instruction is inherently goal-oriented and that is why much is made in instructional design about the need for strong goal determination and goal specification.
The proposition that instruction is an attempt to influence another's thinking follows from that goal orientation. We all agree that the act of influence is benevolent, i.e., in the best interest of the learner. The instructor is always well-intentioned in relation to the students, even when forcibly acting to bring about learning. The intentional influence of instruction is particularly evident in the shaping of attitudes, but it can also be seen to operate in all instruction, even the most seemingly neutral in values (arithmetic for instance) or the most professedly liberal or open-minded. Once again, learning occurs all the time on its own, but instruction is a principal way to influence that learning.
Two main processes are fashioning motivation and structuring the learning situation.
A recognized aspect of the element of influence is the motivational one--the importance of focusing the learner's attention and keeping engagement high (task focus) throughout the instructional event. One of the most influential theories of instruction, that of Robert Gagné (1985), included motivation as the first of its events of instruction. And Keller's (1983) theory focuses principally on motivation.
Yet the motivational aspect of influence should be controversial. Indeed, it is heavily intertwined with content selection (remember those boring classes each of us has experienced?) and with the situatedness (or lack thereof) of the instruction itself. Just-in-time instruction (increasingly found in corporate training), highly contextualized instruction (as found in problem-based and case-based learning), naturalistic instruction (using realistic settings and problems, simulations and other high impact instructional media) are modern developments in instructional approaches that emphasize authenticity (Jonassen, 1998). They seek to capitalize on the learner's natural curiosity and current intellectual needs, thus doing away with artificial means of motivating students. Ultimately, we might even think that any need for artificial motivating elements within instruction might be well an indication of failure--that instruction has not been properly (or at least optimally) prepared.
Motivation must be an essential consideration in instructional design theory (as it is very explicitly so in Keller's theory (1983) and should be more so in others as well) and to the extent that it is, it should not enter instruction per se. This may appear hopelessly unrealistic, but should remain a driving force behind instructional design theory. Indeed, it is perhaps the greatest challenge we face in instructional design theory and certainly an area in need of much more attention.
Motivation is essential to consider in instructional design in order to eliminate it from instruction.
Another aspect of influence in instruction is the age-old issue of how tightly controlled or open-ended one wishes instruction to be. This structural aspect of instruction also interacts strongly with content selection, but goes beyond it as well. Detailed goals (operational procedures to accomplish something, for instance) might profit best from strong sequencing and high control in instruction; broad goals, on the other hand (for example anything dealing with appreciation) are often optimized by more fluid, less rigid instruction. Notice my hedging in these statements, however. The complexity of instructional situations precludes the formulation of strong rules in this regard. Nevertheless, we would generally all agree that we would not give a lecture in order to teach the multiplication table, or employ programmed instruction for an understanding of Kant and Hegel. The key instructional decision that needs to be formulated might well be the one regarding when to switch from a directive style of instruction to a more open, constructivist one (Hannafin, 1997)--in other words, what is the rule that applies here?
While precise rules of instruction would be nice to have (and those outside of education, and not aware of its nuances, are often surprised by this lack), it is just too much to expect to have such a set of rules. The reason, of course, is the complexity of the task in relation to the variability within the environment. Just consider the difficulties encountered in the specification of pedagogical rules in the field of intelligent tutoring systems (Farquhar & Orey, 1998), despite the strong task-focus of most of these systems. Or consider the potential rigidity that can result from a strong rule-oriented design approach such as Merrill's (1998) if the approach is applied without finesse.
That core difficulty in specifying with any precision the rules of instruction, even at the level of how tight or open instruction should be, should however not be ascribed to choices in educational philosophy or to values. These choices just cover up our inability to determine the rules or to come to grips with the complexity, and thus give us a way to by-pass the issue. But that remains the principal challenge for a theory of instruction, and thus too for instructional design theory.
Our values perhaps cover up our lack of a desirable set of rules of instruction.
Reigeluth (1998c) makes the case in this regard that the shift we are witnessing in education from a teacher- to a learner-centered paradigm (from strongly directed to more open learning) flows naturally from changes in society--moving from values appropriate to the dying industrial age to values more aligned with our current information age. How quickly this shift is occurring may be an issue in itself, but it definitely is taking place. Once again, however, it is not that different educators have different values in this regard, but rather that what we value as a society is changing--from more mundane elements of knowledge and compliant attitudes to higher forms of problem solving and initiative. Open learning and constructivist environments like those favored by Hannafin (1998) and Jonassen (1998) are just more appealing for this type of content. That is the crux of it: it is really the content of education--what we teach--that has changed. But remember that content decisions do not belong to the instructional designer. The meta-rules of instructional design theory must deal with this content approach to instruction interaction, a tough challenge given the intertwining of knowledge.
A final aspect of the nature of instruction concerns the stated goal of the influence that constitutes instruction, namely the thinking of another person (the learner). The decision to remain so broad in this aspect of the definition rests on a conception of learning as fundamentally involving an interaction with information in one's environment, along the lines of Piagét's assimilation/accommodation processes for instance (Piagét, 1970). Instruction is nothing other than the crafting of that interaction in a purposeful way.
Instruction is the crafting of information interactions for the benefit of the learner.
It is one's thinking, via the interaction with information in the instructional environment (including "live" information from a teacher), that leads to learning. Whether we are dealing with either of the two forms of learning--associative or structural--does not change the fundamental nature of the interaction with information in the environment. The high level of abstraction of this definition may appear impractical, but it is appropriate in our search for the fundamental nature of instruction.
A final matter in this regard, essential for instructional design theory, is whether to best consider instruction as product or process. Teaching in a classroom is a process (an event unfolding in time), whereas instructional software on a computer is a product (generally packaged and distributed as such). And yet, the software product unfolds and manipulates an instructional event in the same sense that the teacher does in the classroom. Likewise, the teacher's instructional process is not merely spontaneous (even if at times highly reactive to the situation at hand), for it follows a plan and involves a set of materials--a product view of teaching.
Instruction can be viewed both ways and instructional design theory needs to consider both as well. In effect, learning is an interaction with information products and occurs over time in an organized sequence fashioned as an event. All products unfold events, just as all events involve products, no matter how spontaneous or ephemeral these sometimes are. Both views are required for instructional design theory to reach towards generality.
The buttons that appear below will be found at the top and bottom of each page of the discussion. The first button will take you to the previous page (in this case, to the beginning of paper #27). The middle button will take you to the ITForum home page. The last button takes you foward in the discussion (in this case, to the remainder of Duchastel's paper).