User Tools

Site Tools


Past president of the Society for General Systems Research (now the International Society for the Systems Sciences), 1974.

Gordon Pask

by Ranulph Glanville


Gordon Pask's life and work may be interpreted through the filter of two themes: interaction and drama.

Those who knew him, or saw him make a presentation, will be acutely aware of the drama of the persona. And all those who know his work will be familiar with the use of the word interaction as well as words such as conversation, implying/embodying interaction.


Andrew Gordon Speedie Pask was born June 28, 1928 and died March 29, 1996, leaving a widow and two daughters. He was a sickly child not expected to attain adulthood. Spending much time alone and in bed, he developed a magical, mysterious world he populated with various animae. These animae remained with him throughout his life, and their traces may be found in much of his work.

He was also precocious: as he liked to say, he was not born but popped out of a champagne bottle. His early life, and his involvement with the educational establishment, is full of extraordinary events that help give a flavour of the man: chemical experiments with explosives blowing up his father's greenhouse, and his savaging of an anatomy table when, impatient at the lack of progress of his dissection, he used a fire axe to amputate a cadaver's leg, with shattering results. He later claimed also to have blown up a theatre and some inconveniently placed traffic bollards.

Early Work

From the earliest days Pask, with his lifelong collaborator Robin McKinnon-Wood, built astonishing machines. In the 1950s they built two revolutionary computing devices through their company, System Research (SR).

The Self Adaptive Keyboard Instructor (SAKI), was a machine that modified and generated exercises set trainee keyboard operators, compensating for their particular errors. Current keyboard trainers derive (probably unwittingly) from SAKI.

MusiColour (a light show) held “conversations” with musical performers. More than any equivalent (even in 1997), it built understandings of what the musicians were doing, and, when unchanging, modified its response through boredom (a wonderful quality by which to indicate intelligence). It interacted with musicians, encouraging change and development. Musicians described it as an extra band member. But it was not only the programming that was revolutionary. Pask developed computers in various different media, of which the chemical/dendritic computer is the most significant. This work on the mechanical embodiment and organisation of computing has a growing following in the Artificial Life movement.

Conversation Theory (CT)

These early works demonstrate an interest in interaction and the drama-resolved in learning-which was to bear greatest fruit in the development of Pask's masterpiece, Conversation Theory-a reflexive theory which is, in effect, a theory of theory building.

Pask recognised the way to confirm that a student had learnt, in a world of individual mentational processes, was to ask them to “teach it back” to the teacher. Therefore, he promulgated conversation between teacher and learner, in which both could negotiate agreement about what had been (and could be) learnt.

This teaching back was the basis of CT, itself developed through debate both at SR and in seminars with postgraduate students and colleagues at Brunel University, during his most creative period. CT may be briefly characterised as follows:

Characterisation of CT

There are, in a Subject Matter a learner might be interested in, Topics-which may be learnt. Topics relate to each other so that: each Topic is derived from at least two other Topics, or any one Topic entails at least two others (otherwise there would be no way of making difference between Topics) all Topics, arranged together to form a Subject Matter, are logically required to form a multiply linked mesh-ie, there is (global) cyclicity

If Topic A can be derived from Topics B and C, then Topic C can be derived from Topics A and B, and Topic B can be derived from Topics A and C (cyclicity at a local level)

The Topics to be studied support the making of Concepts (of Topics) by the conversational participants such that these Concepts become the participants' Understandings (they are known by the participants). The participants are named psychological (p-) Individuals (later alternatively psycho-social Individuals), embodied in mechanical (m-) Individuals. (P-individuals and m-individuals do not need to be in a one-to-one correspondence.)

Concepts and Topics, and those p-Individuals that understand them, share the formal property of cyclicity (shared, at a macro level, with the cyclic form of the Subject Matter): they are self-reproducing automata, or circular processes. In a sense, everything in a Paskian universe of knowables is a variety of p-Individual.

The process of learning (developing a coherent set of understandings of Topics previously not-understood by a student) is a process of conversation about and with Topics. Understandings may be expressed to other learners/teachers, or within the Subject Matter (in the form of the “Mesh” of its “Entailments”). Learning can begin anywhere, because of cyclicity. For instance, the student chooses starting points, pruning the Entailment Mesh (which is a circular hypertext) to form a heterarchical Entailment Structure that expresses the individuality of the learner and has starting points at the bottom, together with one or more end Topics.

In this context, a conversation is the process of the presentation and re-presentation of developing Understandings between p-Individuals forming an exchange, leading to a shared agreement (possibly to disagree): mutual concurrence. The process of conversation is interactive and, achieving resolution (for the student) where separate parts finally cohere to provide the outcome Understanding (or “knowing”), dramatic. Its formalisation into CT makes it available for computer implementation as a type of learning environment. CT also proposes a way of acting, which Pask tried to follow: rather than give information, he preferred to tease out an understanding, conversationally, with those he was interacting with.


Using the Conversation Theoretic arrangements of Subject Matters allowed Pask and his team to experiment in areas not previously open to precise investigation. Their early work on learning-to-learn and on learning styles and individual differences made a major contribution, impossible without the formalisms of CT. Pask was aware that his work contributed to the creation of Second-Order Cybernetics, along side work on Autopoiesis (Varela, Maturana, Uribe and, in a different manner, Beer); or von Foerster's Eigen-objects and Glanville's Theory of Objects; and Spencer Brown's Laws of Form and Bateson's Difference that makes a Difference. It is worth noting that all these people were friends of Pask. In many cases Pask's insights were the inspiration for their work, and CT is, without doubt, a founding Theory of Second-Order Cybernetics.

Later Work

Pask became restless with the limitations he saw in CT. He became determined to find the protologic or protolanguage (Lp) that defined entailments between Topics within Conversations, hoping, initially, to automate the development of Entailment Meshes through the addition of new Topics, but eventually conceiving of Lp as something even more universal: as the protolanguage that underlies all conceptualising and languaging. His work in this area is extremely difficult. Its first expression was as the computer program “Extend”, which moderated the addition of new Topics into an existing Entailment Mesh. He also became concerned with what he saw as the determination and finality of the Conversation: that it begins and ends, whereas he saw the universe as continuous and life as a dance. His animae were everywhere.

He spent his final years attempting to generalise from the particular event of the particular conversation to a more general, endless form of interaction he called the “Interaction of Actors Theory” (IA). IA is concerned with multiple, concurrent, sequential and con-sequential conversations, and with the continuity that exists across each particular conversation: at a universal level, it is concerned with a universe constituted of conversation. IA is more general than CT, closer to the immortal: a conversation is an embodiment of interaction. Time will tell how well he succeeded. But, regardless of this success, his contribution in the form of CT is indeed major.

Main Colleagues

In spite (or, perhaps, because) of the distinctiveness of his persona, Pask worked well with others. As well as his wife Elizabeth and Robin McKinnon-Wood, other long time co-workers at SR included Brian Lewis and Bernard Scott, George Mallen, Mike Elstob, Tony Watts, Dionysius Kallikourdis, Robin Bailey, Nick Green, Ytzhak Hayutman and Paul Pangaro. Pask, ever generous in inclusion, would also want his students included in this number. His own list would be very long indeed!


Pask's contributions in educational technology are enormous (see, for instance, the introduction to Diana Laurillard's book “Rethinking University Teaching”, London, Routledge, 1993), although sadly unsung. He was a pioneer before his time. He attempted to create, in a world of underdeveloped computation, applications of enormous significance, originality and humanity, through immense inventiveness and imagination. For these reasons, his applications must sometimes be seen as skeletal or prototypical embodiments of his theories. Nevertheless, even today, they often lead the field and are remarkably effective. But to grasp their potential, it may be helpful to think of them as poetic indicators rather than as literal embodiments. His writing presents other difficulties: it can be very technical, with complex diagrams, and suffers from Pask's wish to adumbrate (his term) everything within his position. Nevertheless, the quality of his thought and insight is profound and revolutionary. It is all too easy to miss the power, the vision and the magic.

Pask was a wonderful (and conversational) teacher who left his personal mark on many doctoral and other students around the world: he was generous with his time and concern for students, although he could snap ferociously at what he considered lazy or unworthy thinking. He had a particular affinity for architects and the Architectural Association School in London, which provided many of his doctoral students and became an academic refuge. The celebration of his life arranged by his widow Elizabeth was held there in June 1996. He was also closely associated with various universities, especially Brunel, the Open, Concordia and Old Dominion Universities, and latterly, the University of Amsterdam.


Pask's work is groundbreaking and still ahead of the field. It has laid foundations for the development of a more humane understanding of human action, and an intelligent interaction between humans and computers. There is no question that CT, and the later Lp and IA, are thoroughly constructivist theories, concerned at heart with the (necessarily) personal generation of Understandings of the world which are uniquely of each participant in any conversation.

Nevertheless, there may be some distance still to be travelled. For all that he would have wished otherwise, Pask's concept of his Topics seems to belong in a world of knowables that remains constant and unchanged in the face of the Understanding of each knower: ie, a world that retains quasi-realist elements. There are also problems associated with decidability in algorithmic processes, and the choice of terminology that, reflecting many sources and influences, sometimes lacks coherence to those well versed in only some of these sources. Perhaps pioneers can only take us so far: they define the need for a change in point of view, they point us in the right direction, but they are too caught in the ground from which they developed this point of view to be able to travel far down the line (Pask levelled this same criticism at Wiener).

Some may object to the amount of detail, in this account, of Pask as a person. Pask is a scientist whose work connects with his being far more, and far more obviously, than most. Indeed, it is, in certain respects, that very connection which is the subject of his scrutiny: Pask once defined the human as a “machine for learning”. For Pask, the study of human learning was also the study of himself.


My understanding of Pask's work comes, in the first place, from the generosity and love of Gordon himself. It is only proper to acknowledge that, and the patient coaching of Bobby McKinnon-Wood, at the outset. I would like to acknowledge help from the following, who read and contributed to the development of manuscript (in some cases with strong disagreement): Dr Bob Barbour, Prof Graham Barnes, Prof Gary Boyd, Prof Richard Jung, Mr Peter Paine, Dr Paul Pangaro, Dr Bernard Scott, Prof Gerard de Zeeuw. Nevertheless, what I wrote here is what I wrote, and what I did with their advice is what I chose to do. I accept all blame for inaccuracies as well as for my interpretation and telling of the story.

Ranulph Glanville, October 25, 1997

Finding out more:

There is a web site devoted to Pask, from which several links may be made:

Contributions to that site may be sent to:

Paul Pangaro generously maintains both an archive and a web site:

A festschrift for Gordon Pask, edited by Ranulph Glanville, was published in the Journal “Systems Research” (vol 10 no 3). A commemorative issue of the “Journal of Human-Computer Studies” is in the process of assembly, edited by Bernard Scott and Ranulph Glanville.

An archive of Paskiana is being considered by De Montfort University in Leicester, UK, and the Pask family currently has a private archive.

Some key works:

MusiColour and SAKI are not very well covered in publications. But see:

“MusiColour” in Good IJ (ed) (1962) “The Scientist Speculates”, London, Heineman

Appendix to “A comment, a case history and a plan” in Reichardt J (ed) (1971) “Cybernetics, Art and Ideas” Greenwich, New York Graphic Society Ltd and London, Studio Vista (includes an Appendix added in October 1968).

“SAKI-Twenty Five Years of Adaptive Training into the Micro Processor Era” (1982) International Journal of Man Machine Studies, vol 17

Conversation Theory and some of its extensions are covered in, for instance “Artificial Intelligence- a Preface and a Theory” in Negroponte N (1973)

“Machine Intelligence in Design” Cambridge, MIT Press, later (1979)

“Soft Architecture Machines” “CASTE: a System for Exhibiting Learning Strategies and Regulating Uncertainty) (with Scott BCE) (1973) International Journal of Man Machine Studies, vol 5

“A Theory of Conversations and Individuals (Exemplified by the Learning Process on CASTE” (with Scott BCE and Kallikourdis D) (1973) International Journal of Man Machine Studies, vol 5

“The Representation of Knowables” (with Scott BCE and Kallikourdis D) (1975) International Journal of Man Machine Studies, vol 7

“Conversational Techniques in the Study and Practice of Education” (1976) British Journal of Educational Psychology, vol 46 no 1

“Styles and Strategies of Learning” (1976) British Journal of Educational Psychology, vol 46 no 2

“Organisational Closure of Potentially Conscious Systems” (1981) in Zeleny, M (ed) “Autopoiesis” Amsterdam, Elsevier

“Review of Conversation Theory and a Protologic (or Protolanguage), Lp” (1984) ECTJ, vol 32 no 1

Pask's interest in and impatience with much Computer Aided Learning (especially Instruction) is expressed in “Machines that Teach” (1961) New Scientist, June 10

“Man as a System that Needs to Learn” (1968) in Stewart D (ed)

“Automata Theory and Learning Systems” London, Academic Press

“Adaptive Machines” (1968) reprinted in Davies IK and Hartley J (eds) “Contributions to an Educational Technology” London, Butterworth (1972)

“Anti-Hodmanship: a Report on the State and Prospect of CAI” (1972) Programmed Learning and Educational Technology, vol 9 no 5

His delight in Art, Architecture and the creation of intelligently interacting actors may be found in

“The Architectural Relevance of Cybernetics” (1969) Architectural Design, Sept “A Comment, a Case History and a Plan” in Reichardt J (ed) (1971) “Cybernetics, Art and Ideas” Greenwich, New York Graphic and London, Studio Vista

The Interaction of Actors Theory awaits proper publication. There is a paper awaiting processing that was left on Pask's desk. Otherwise, consult the following

“Interaction Between Individuals, its Stability and Style” (1971) Mathematical BioSciences, June

“Interaction of Actors Theory” (with de Zeeuw G, compiled and edited by Glanville R) (1991) OOC Program, University of Amsterdam

Pask's last paper, which attempts to place Conversation Theory and the Interaction of Actors Theory together in a shared context is

“Heinz von Foerster's Self-Organisation, the Progenitor of Conversation and Interaction Theories”, Systems Research, vol 13 no 3

Other preoccupations are developed in

“A Proposed Evolutionary Model” (1961) in von Foerster H and Zopf G (eds)

“Principles of Self-Organisation” London, Pergamon

“A Proposed Experimental Method for the Behavioural Sciences” (1964) in Wiener N and Schade J (eds)

“Progress in BioCybernetics” Amsterdam, Elsevier

“A Cybernetic Experimental Method and its Underlying Philosophy” (1972) International Journal of Man Machine Studies, Jan

“A Fresh Look at Cognition and the Individual” (1972) International Journal of Man Machine Studies, vol 4

“A Cybernetic Theory of Cognition and Learning” (1975) Journal of Cybernetics vol 5 no 1

“The Limits of Togetherness” (1980) in Lavington S (ed) “Progress in Information Processing 1980”, Amsterdam, North Holland

“One Kind of Immortality” (1991) in Glanville R and de Zeeuw G (eds) “Mutual Uses of Cybernetics and Science” Amsterdam, Thesis Publishers (2 Vols)

His own approach to Cybernetics may be found in the early, yet still apposite

“An Approach to Cybernetics” (1962) London, Hutchinson

His other books are

“The Cybernetics of Human Learning and Performance” (1975) London, Hutchinson

“Conversation, Cognition and Learning” (1975) Amsterdam, Elsevier

“Conversation Theory: Applications in Education and Epistemology” (1976) Amsterdam, Elsevier

“Calculator Saturnalia” (with Ranulph Glanville and Mike Robinson) (1980) London, Wildwood House

“Microman” (with Susan Curran) (1982) London, Century Publishing

Ranulph Glanville, independent academic
CybernEthics Research
52 Lawrence Road
Southsea Hants PO5 1NY UK

From the earliest days, Pask saw cybernetics as the approach that most captured the concepts of interaction and drama. Developed cybernetic machines in abundance from very early on. His work in learning (and learning machines) lead him to develop the most interactive of cybernetic theories, “conversation theory”, which he later tried to generalise to the “interaction of actors theory”, both of which firmly placed the onus and responsibility for (inter-)action on each participant. Pask was the first to welcome error as not only inevitable, but also constructive.

Add your contributions here!

gordon_pask.txt · Last modified: 2015/01/31 23:55 (external edit)