Fostering superficial learning

Different views of learning and what it means to follow rules are discussed. The activities, the language and context of expertise and knowledge have traditionally often been neglected and underestimated as something ancillary to the real knowledge which lies "beyond" such surface manifestations and viewed as at most means to learn the knowledge beyond these aspects. The primacy of these aspects are argued for. It is further argued that these should be more emphasized and fully appreciated. The goal of fostering "superficial" learning is suggested, i.e., an emphasis on practical activities, the use of the language of the domain to be learned and getting familiar with the new culture. A case study is presented and discussed in terms of the theoretical viewpoint. 1. Criticisms of Traditional Approaches to Cognition According to traditional cognitive science based on the representation hypothesis human cognition is described (a) in terms of representations describing a portion of the world, e.g., in terms of propositions, ideas, or mental images or (b) in terms of symbol manipulation (also non-descriptive ones). These representations have been ascribed the role of explaining cognitive human action. When observing a subject perform a cognitive task, a researcher has typically "explained" the subject's behavior by ascribing a mental representation of some sort to the subject. Since the end of the eighties, the traditional cognitivistic view on cognition and learning has been questioned by several theorists. A "situated" view, sometimes even called a situated "theory" of cognition, is a term that covers several diverse arguments against the more traditional view of cognition. These arguments often originate from one or more of the following themes (Holm & Karlgren, 1996): Firstly, there is no justification for assumptions that there are general, contextindependent cognitive mechanisms and procedures as opposed to viewing action as inherently situation specific. Secondly, that it overemphasizes reflective, intellectual, rational behavior and neglects instinctive action as well as bodily and emotional aspects. Thirdly, it is too individualistic, because representations are socially constructed and maintained and human thinking is often supported and influenced by socially constructed artefacts. Another issue is the set of various problems with mental representations, e.g., how to view the relation between mental representations and action. The traditional view has been criticized for being too mechanistic. Rather than viewing representations as something that determine actions as a controlling system, representations may be viewed as resources for actions (Suchman, 1993). Furthermore, the idea of having a science of the mental can be questioned in itself (Holm & Karlgren, 1996, Wittgenstein, 1958). 2. Learning and Rule-following According to traditional theories in cognitive science, learning something, e.g., a rule in arithmetic, Prolog programming or French grammar, consists in acquiring certain rules and representing these mentally in one way or another. The mental representation of what is learned has been viewed upon as a condition for being able to use the knowledge, to follow the rule, e.g., do some calculation or utter French sentences based on the rule. Once the learner has understood the rule and stored the representation, the representation can be applied more or less contextindependently, at least within a domain or area, or to problems that are considered to be ”the same”. A consequence of this view of knowledge and learning is that the goal of instruction is to support the learner in building an adequate representation of what is to be learned, the algebraic rule. The aim of instruction is to convey the rule, or its meaning, to the learner and support him/her in constructing a representation of it. The rule, or the representation of it, will then guide the person in future action. However, rules and representations may be ascribed a very different role and the relation between representations and actions may be viewed upon in different ways. When somebody comes across a certain rule, the rule may be interpreted in innumerable ways, as exemplified by Malcolm (Malcolm, 1986): "Someone might interpret a NO ENTRY road sign as applying only to motorists, not to cyclists; and someone else might interpret it as applying even to pedestrians. A child might understand the instruction 'Always close the door after you enter the house' as not meaning to keep it closed. A novice in cooking might take the words of a recipe for a cake, 'Add two eggs', as meaning that the whole eggs, including their shells, should be added." Malcolm, p. 158. However, there is an inclination to say that people learn to follow rules because they are guided by the meaning of the rules. Of course, there are times when rules and recipes do guide behavior. Sometimes these are even learned by heart. A rule can be expressed in various ways, in sentences, formulas or sign-posts, and these things can be interpreted in various ways. Somebody might retort that the meaning of a rule "guides", instructs, in how to act when confronted with new situations. Consider the example of learning the rule of addition discussed by Malcolm. How is it possible for a person to perform an addition of two numbers that has never been done by him/her before, or he/she has seen anybody else perform? Is it because the meaning of the rule "guides" the person, or instructs how to do the new addition? This would have to mean, that the rule in some sense contained all the future steps, that they were already there in the meaning of the rule. How else is it supposed to be of any help? And even if this were true, how would it help the person: where is it said in which way one is supposed to follow the rule and how to copy the steps taken by it? There are different ways of copying. Maintaining that learning to follow a rule means storing a mental representation of the rule, or the meaning of the rule, leads to serious philosophical problems. There is philosophical puzzlement over how one is able to follow a rule. There is a feeling that the understanding of a student must go beyond the given examples and exercises so that he/she can go on. Extensive discussions on what following a rule consists in, are found in Wittgenstein's writings, although he does not provide any other explanation to how a rule can be followed, other than that a certain training has been received (Malcolm, 1986). Instructions and examples are given, and then, at some point, the learner simply goes on, continues, in the same manner. The rule cannot instruct the learner in a definite way since a rule could be interpreted in various ways, and the learner could continue in various ways, but usually does not. Malcolm, following Wittgenstein, has attempted to show that appealing to concepts such as 'the meaning' of a rule, 'intuition', 'similarity' in situations, 'inner voices' etc. do not explain the capability of following rules. The meaning of rules, intuitions, similarities, inner voices can all be misinterpreted or misunderstood and do not explain the ability to follow rules. As Malcolm concludes, "[W]e go on, all agreeing, following rules ... in new cases – without guidance. Other than the past training, there is no explanation." (Malcolm, 1986). According to Wittgenstein, the meaning of a rule is fixed by "our customary way of applying the rule in particular cases ". There is a way of acting that is interpreted and called ’following the rule’. Indefinitely many other ways of acting are possible: but these are not interpreted as and called 'following the rule'" (Malcolm, 1986). An explanation for this agreement in action between people is often asked for (e.g., "they must share the same mental representation..."). But it is rather the other way – there is not an agreement because people have absorbed the meaning of the rule. Unless there is agreement in action, a rule would be meaningless. If half of the people act in one way when confronted with a new situation, and the other half in another way, the rule would no longer be a rule. In what way might these philosophical ideas relate to theories of learning and instruction? Many theorists on learning have emphasized the importance of concrete examples, opportunities to practice one's knowledge in real life, as opposed to teaching abstract knowledge. However, the role of practice and examples is often viewed to be a support to learn the true or deeper meaning, to learn some kind of "underlying" rule. Taking this view on rule-following and learning, the examples and praxis are not only a support for learning something else. The "meaning" of a rule cannot fully be abstracted from the praxis, or the customary way of applying a rule in particular cases, without which the rule would become meaningless. The knowledge to be learned is thus not something beyond the concrete examples and the exercises. Crucial for the learner is to get a hang of the praxis. The philosophical problems with explaining learning and knowledge in terms of rules or other representations can of course be generalized to other areas than those mentioned, e.g., rules about a scientific domain such as protein purification. This view on rules and learning is thus one form of the more general criticism directed towards traditional cognitive science for presupposing more or less context independent mental mechanisms, processes, representations and viewing these as explanations of behavior that is considered to be ”similar”. 2.1 Rational Reflection vs. Intuition Learning and expertise do not have to be viewed as primarily based on certain mental representations. Expertise does not only consist in knowing rules of the domain. Getting to know the rules is the first thing a novice does, and much more than learning rules is demanded of an expert. Although knowledge about facts and rules do play a role in expertise, the less explicit aspects of expertise are easily overlooked (Brown, 1990). Dreyfus and Dreyfus's model of learning and the development of expertise shows how learning rules may be important in the initial phases of learning, but what really characterizes expertise involves "intuition"; knowledge that is not so easily made explicit (Dreyfus & Dreyfus, 1986) Getting used to, and adopting the specific language usage of a new domain is one important part of learning and going through a "representation shift". This discussion is related to the discussion of the relation between representations and action: representations are traditionally regarded as a determining or controlling mechanism and such a view tends to neglect the importance of action and instinct for a learner to be able to construct representations (Holm & Karlgren, 1996, Malcolm, 1986). 2.2 Objectivistic and Constructivistic Views "Objectivism" is sometimes presented as the view that equates information and knowledge and views knowledge as something that each individual can come to know in the same way. Constructivists often assume that knowledge cannot be objectively defined and that it instead is individually constructed from what learners do in their experiential worlds (Akhras & Self, 1996). As a consequence, there is no "target knowledge" that could be pre-specified and transferred to the learner (Akhras & Self, 1996). Consider an example taken from Akhras and Self, although one can objectively specify the syntax of a programming language, knowing a programming language is not possessing this objective knowledge, the knowledge "comes out of actively transforming and interpreting an experiential world involving the use of that programming language. The meaning of the programming language elements can only be built by the learners from the cognitive structures that they manage to assimilate while interpreting their own experiences". The key words here should not be "cognitive structures" or even "actively interpreting own experiences". The key word is rather "use". Learning the meaning of the programming language elements can only be done by practicing to use these elements and taking part in activities in which these elements are actually employed. Constructivists emphasize individual reflection on knowledge which is a sound reaction against objectivistic and behaviorist approaches, but the meaning of programming language elements or any other concept is so much more than individual conceptions of these! What gives the elements of a programming language meaning is how these elements are used in programming activities rather than the individual conceptions programmers have of the elements. 2.3 Learning as Gaining Membership in a Social Community That language is essentially a social activity and that the meaning of words cannot be based on something private has been convincingly put forth by Wittgenstein (ref). Many of these, or similar, insights have lately been echoed by several different theorists and often in polemics towards "traditional cognitive science" (Holm & Karlgren, 1996). The importance of talking science and participating in scientific activities, rather than just hearing science has been emphasized by Pea who also draws upon arguments put forth by Wittgenstein (Pea, 1992). In the research area of learning theory, the traditional theories have during recent years been rivaled by theories that emphasize that knowledge is socially constructed and that central to learning is gaining membership in social communities. Rather than viewing expertise as independent facts learning and expertise is seen as a process of becoming a member in social communities and discussed in terms of apprenticeship (Brown et al., 1989, Lave & Wenger, 1991).