Educational technology is seen as key for lifelong learning, but it has yet to live up to expectation. We argue that current learning environments are typically oriented too much towards structured learning to meet the needs of the lifelong learner. Environments for lifelong learning demand a higher degree of autonomy for the learner, must be open to eclectic sources, support soft informal learning activity, and accommodate evolution both in the experience of the learner and in the context in which this occurs. We propose sense-making through the construction of suitable interactive artefacts as a core activity for lifelong learning, and discuss and illustrate how this can be supported using Empirical Modelling. The merits of Empirical Modelling as a constructivist approach are assessed with reference to a criterion recently proposed by Bruno Latour, namely, the extent to which it strengthens five guarantees, taken together.

Context: Both Luhmann and Pask have developed detailed theories of social systems that include accounts of the role of learning. Problem: Rather than see the theories as competing, we believe it is worthwhile to seek ways in which a useful synthesis of the two approaches may be developed. Method: We compare the two approaches by identifying key similarities and differences. Results: We show it is possible to make useful mappings between key concepts in the two theories. Implications: We believe it is worthwhile for social scientists to be familiar with the two theories and that it is not a case of “either/or,” rather, it is a case of “both/and.”

Key words: learning, conversation, interaction, expectation, Niklas Luhmann, Gordon Pask

This article deals with the problem of how operationally closed systems can construct a reality and therefore get their bearings in the world. But rather than looking for new theoretical solutions, it suggests going back to the empirical philosophical tradition of early modernity, in order to find a solution. Following a suggestion by the leaders of both first- and second-order cybernetics, Wiener and Foerster, this article reframes Hume’s theory of causal inference in order to make the case not only that Hume anticipated second-order cybernetics in interesting ways, but also that modern cognitive sciences can use Hume and second-order cybernetics to inform each other leading to a better understanding of both. Starting from the statement according to which the problem of causality represents ‘one of the most sublime questions in philosophy, ’ the article goes deeply inside the problem of causality in order to argue that the modern approach to epistemology has to be conceived of as a process of internalization of cognitive facts. This search path leads to casting a new light on the paramount concept of sign, conceived of as the possibility that certain environmental events or data again set off the self-reference of a cognitive system, which thus switches from memory to expectation. The aim of this article is finally to show that the main results of an interdisciplinary theory of cognition such as second-order cybernetics are particularly congruent with the speculations of the Scottish philosopher, and that Hume’s reflections maintain an extraordinary relevance regarding the most advanced elaboration of the main epistemological problems

This article deals with the problem of how operationally closed systems can construct a reality and therefore get their bearings in the world. But rather than looking for new theoretical solutions, it suggests going back to the empirical philosophical tradition of early modernity, in order to find a solution. Following a suggestion by the leaders of both firstand second-order cybernetics, Wiener and Foerster, this article reframes Hume’s theory of causal inference in order to make the case not only that Hume anticipated second-order cybernetics in interesting ways, but also that modern cognitive sciences can use Hume and second-order cybernetics to inform each other leading to a better understanding of both. Starting from the statement according to which the problem of causality represents ‘one of the most sublime questions in philosophy, ’ the article goes deeply inside the problem of causality in order to argue that the modern approach to epistemology has to be conceived of as a process of internalization of cognitive facts. This search path leads to casting a new light on the paramount concept of sign, conceived of as the possibility that certain environmental events or data again set off the self-reference of a cognitive system, which thus switches from memory to expectation. The aim of this article is finally to show that the main results of an interdisciplinary theory of cognition such as second-order cybernetics are particularly congruent with the speculations of the Scottish philosopher, and that Hume’s reflections maintain an extraordinary relevance regarding the most advanced elaboration of the main epistemological problems

Brains, it has recently been argued, are essentially prediction machines. They are bundles of cells that support perception and action by constantly attempting to match incoming sensory inputs with top-down expectations or predictions. This is achieved using a hierarchical generative model that aims to minimize prediction error within a bidirectional cascade of cortical processing. Such accounts offer a unifying model of perception and action, illuminate the functional role of attention, and may neatly capture the special contribution of cortical processing to adaptive success. This target article critically examines this “hierarchical prediction machine” approach, concluding that it offers the best clue yet to the shape of a unified science of mind and action. Sections 1 and 2 lay out the key elements and implications of the approach. Section 3 explores a variety of pitfalls and challenges, spanning the evidential, the methodological, and the more properly conceptual. The paper ends (sections 4 and 5) by asking how such approaches might impact our more general vision of mind, experience, and agency.

Key words: action, attention, bayesian brain, expectation, generative model, hierarchy, perception, precision, predictive coding, prediction, prediction error, top-down processing.

Excerpt: In this chapter, I have argued that radical constructivist interpretations of constructivism differ from other interpretations in that radical constructivism is an epistemological theory based in viability. It is suggested that viability commits one to the expectation of and support for diversity in the classroom. Moreover, it obliges the radical constructivist to also reinterpret the mathematical meaning of concepts in light of the students’ inventions. To do this effectively, the radical constructivist must learn techniques of close listening and follow these by the articulation of student voice and the examination of the changes in his or her own perspective. It is the “voice-perspective” relationship which makes radical constructivism capable of deep reform in mathematics instruction.

A case for a pluralistic approach to cognitive science is sketched. It is argued that cognitive scientists should take seriously the possibility that a single, unified framework for all of cognition is an unrealistic expectation for its diverse interdisciplinary goals and subject matter. A pluralistic approach instead seeks ways of integrating the multiple perspectives that have provided explanatory success in loosely interconnected sub-domains of cognitive phenomena. Research strategies recommended by this approach are discussed, with review of research currently carrying out such strategies and others that may hold promise for the future. The article ends with a discussion of seeking closer integration of the inquirer into consideration of which explanatory framework to choose. A systematic exploration of this transactional approach to cognitive science may grant coherence to pluralism even as it embraces diverse schemes of explanation.

Key words: pluralism, cognitive science, complexity, dynamics, computation, pragmatism

Perception and perceptual decision-making are strongly facilitated by prior knowledge about the probabilistic structure of the world. While the computational benefits of using prior expectation in perception are clear, there are myriad ways in which this computation can be realized. We review here recent advances in our understanding of the neural sources and targets of expectations in perception. Furthermore, we discuss Bayesian theories of perception that prescribe how an agent should integrate prior knowledge and sensory information, and investigate how current and future empirical data can inform and constrain computational frameworks that implement such probabilistic integration in perception.

Key words: prediction, perception, sensory processing, bayesian inference, predictive coding, perceptual inference.

In the course of a long-term research project aimed at enabling a computer to analyse and code (-understand”) the meaning of ordinary English sentences, it became clear that much of the information necessary to understand many, if not most, sentencesis not to be found la the sentences themselves but must he supplied from another source. For the human reader this source is the fund of “knowledge” he has previously accumulated, both through living experience and through linguistic experience. This fund or knowledge is here visualised as a conceptual network onto which the incomplete information supplied by a sentence can be mapped, thus making it possible for the reader to fill in the missing pieces of the conceptual situation designated by the sentence. Such a conceptual network would seem to he the source, also of the various kinds of expectation (concerning the contents of those parts of the sentence which the reader has not yet read) which help the reader to resolve lexical and relational ambiguities. – It is suggested that a greater awareness of this function of the reader’s conceptual universe might lead to an improvement of instructional and remedial strategies for the teaching of the interpretive language skills.

Key words: language, cognition

Open peer commentary on the article “Applying Radical Constructivism to Machine Learning: A Pilot Study in Assistive Robotics” by Markus Nowak, Claudio Castellini & Carlo Massironi. Upshot: I point out that from a sociocultural perspective, repeated experiential interaction loops are not enough for constructing new context-dependent knowledge: the loops must be grounded in specific social practices, which are either culturally or historically situated. Also, to tightly connect human user and interactive machine-learning system, triple-loop learning needs to be used as well as criteria for validating an expectation’s confirmation.