Candy (1991) defines self-direction in these terms: Being able to pursue a learning goal with equal vigour and determination without being adversely affected by external factors including the increase or decrease of rewards for pursuing or attaining the goal (…) p. 1

And further: Being aware of alternative choices, both as to learning strategies and to interpretations or value positions being expressed, and making reasoned choices about the route to follow in accordance with personally significant ideas and purposes. p. 1

Interestingly, these quotes point to very different aspects of learner autonomy, the central one being the control that the learner exercises over the various aspects of learning, beginning with the decision to learn or not to learn. But as Candy points out, there are other specific areas where learner-control can be exercised. p. 1

The first are the motivational-intentional forces that drive the learner to apply some determination (or “vigour”) to the act of learning. This area embodies the conative functions of learning which are at the foundation of learner initiative, motivation and p. 1

personal involvement. Most often, adult learners harbor life-goals that are related, but distinct from the actual learning goals (e.g. career advancement, good parenting or better health), as part of the conative baggage they carry. Other possible drives include the pleasure one derives from the act of learning in itself, and the satisfaction obtained from becoming part of a particular culture of knowledge (Houle, 1961). p. 2

The second set of elements that define learner autonomy involve control over the “nuts-and-bolts” of the act of learning, such as defining learning goals, deciding on a learning sequence, choosing a workable pacing of learning activities, and selecting learning resources (Hrimech & Bouchard, 1998). These elements can be grouped under the heading of the algorithmic aspects of learning. In traditional learning environments, most of the algorithms are the responsibility of a teacher or a teaching institution. Learning goals, student workload and methods of evaluation are usually stipulated at the outset and little participation in their formulation is expected from the learner. Derogation from this approach entails devolving to the learner, on top of the expected “learning tasks”, at least some of the “teaching tasks” normally reserved for the instructor. In this sense, we can say that autonomy is directly related to the number and magnitude of the “teaching tasks” that are appropriated by the learner (Tough, 1965). Most mediated learning environments require such participation from learners, albeit to different degrees and with varying results. p. 2

Until recently, learner control was necessarily limited to these two sets of features, conative and algorithmic. After deciding whether, what, and how to learn, one had covered all areas where it was conceivably possible to exercise some degree of learner autonomy. Now with the proliferation of learning environments that include mediated instruction materials, exponentially available learning resources, new means of communication, and a marketplace literally exploding with learning opportunities, two other components of learning emerge as possible areas where learner control may exercised, namely the semantics employed by the new media, and the emerging economics of the knowledge marketplace. p. 2

Because each environment offers its own set of communication pragmatics and its own approach to using them, we can say that the semantic components are an integral part of the learner’s experience, and as such offer opportunities to enhance or deter learner autonomy. p. 2

Based on these considerations, learners must not only decide why and what to learn, but also where to learn it. environment, but most importantly the choice rests on the perceived cost-benefit ratio and opportunity costs which are generated by each alternative. In this context, we can observe that the economics of learning are emerging as an important component of any learning environment. p. 3

METHODOLOGY Open-ended interviews were conducted with 13 adult students registered in online courses in Psychology, Finance, Education, and Political Science. Questions were formulated to explore the four areas of learner autonomy, using everyday terminology familiar to the students p. 3

FINDINGS p. 3

Predictably, about half of the students said it took some effort to prevent inertia when facing the prospect of doing tasks online, in the absence of an imposed schedule. p. 3

In the groups where there was less pressure to ‘participate’, students felt that the interaction was more meaningful and that they were more in ‘control’ of the environment. p. 3

Reasons for choosing the on-line version of the course were mostly linked to personal, family and work situations. Somewhat ironically, the same factors were identified as barriers to achievement in the on-line course. p. 5