Do Organisations Learn? Some Implications for HRD of Bateson S Levels of Learning

Do Organisations Learn? Some Implications for HRD of Bateson’s Levels of Learning

Primary author:

Dr Paul Tosey

Senior Lecturer,

School of Management,

University of Surrey,

Guildford,

Surrey GU2 7XH,

UK

Tel. +44 1483 689763

e-mail:

Secondary author:

Dr Jane Mathison

e-mail:

Abstract

This article explores and appraises Gregory Bateson’s theory of `levels of learning’ (Bateson, 2000a) and its implications for Human Resource Development, especially with reference to issues of organisational learning.

In Part One, after briefly reviewing Bateson’s biography we describe the origins and contents of the theory. In Part Two, three particular features of the theory are explored, together with their practical and theoretical implications for HRD:

1.  The significance of the recursive relationship between the levels;

2.  Bateson’s theory is not a stage theory of learning; `higher’ levels of learning are neither superior to, nor necessarily more desirable than, lower levels;

3.  Bateson’s emphasis on the notion of context, which implies that the task of management involves sensitivity to such contexts.

In Part Three the discussion emphasises the holistic nature of Bateson’s theory, in that the levels of learning combine cognitive, embodied and aesthetic dimensions. We review some limitations of the theory, then conclude by considering its perspective on the question, `do organisations learn?’.

Keywords: epistemology; recursion; emergence; complexity.

Introduction

Gregory Bateson’s theory of `levels of learning’ (Bateson, 2000a)[i] appears potentially very relevant to Human Resource Development, on the basis that `understanding the dynamics of learning is what differentiates HRD from several closely related fields of practice’ (Yorks & Nicolaides, 2006:144). Bateson’s theory is discussed in diverse literatures, including brief psychotherapy and communication studies (Keeney, 1983; Watzlawick, Weakland, & Fisch, 1974); education (Bloom, 2004; Brockbank & McGill, 1998; Peterson, 1999; McWhinney & Markos, 2003); and organisational learning (Argyris & Schön, 1978; Bartunek & Moch, 1994; French & Bazalgette, 1996; Wijnhoven, 2001; Visser, 2003; Engeström, 2001; Engeström, 2001).

It is worthy of note that Argyris and Schön (1974:18-19) attribute their well-known conceptual distinction between single and double loop learning to Ashby, though in the same place cite Bateson as the textual source of the paradigmatic example of the thermostat as a cybernetic system. Their publications seldom mention Bateson or cite his work - an exception is Argyris and Schön (1978:8-29).

We have found Bateson’s theory both fruitful and enigmatic over many years. It appears to be misconstrued at times in the literature, a situation not helped by the fact that Bateson’s writing was primarily theoretical and lacks detailed empirical application. We have sought to understand Bateson’s theory better ourselves, and to explore its potential to conceptualise issues of learning. Here we focus on three issues:

1.  The `levels of learning’ may be treated as a simple hierarchy, which tends to ignore Bateson’s emphasis on the recursive relationship between the levels.

2.  It is not a stage theory of learning, by which one advances from lower to higher levels. `Higher’ orders of learning are neither necessarily superior to, nor more desirable than, lower levels.

3.  Bateson’s emphasis on the notion of context in learning is significant, and implies that the task of management involves sensitivity to such contexts.

Bateson’s theory implies that organisational learning is created by change in patterns of behaviour that emerge from changes in context (including changes in the interpretation or perceived significance of context). This follows from Bateson’s critique of the extent to which intentional action can produce the desired effects in human or natural systems. It contrasts with the view that actors learn new understandings and skills first, and then enact these in order to generate new organisational capabilities.

Part One: Bateson’s theory of levels of learning

Gregory Bateson: biography and influence

Gregory Bateson (1904 - 1980) was an English academic whose writing (Bateson, 1973; Bateson, 1979; Bateson, 1991; Bateson & Bateson, 1988; Bateson, 2000a), has influenced diverse fields including family therapy and communications studies (Hawkins, 2004; Harries-Jones, 1995; Lipset, 1980).

His cross-disciplinary perspective is an important feature of his work, which is concerned with natural and human systems in general. His biography illustrates this diversity. The son of the geneticist William Bateson, he attended Cambridge University where initially he read zoology before turning to anthropology in 1925. In 1936 he married Margaret Mead, with whom he conducted anthropological fieldwork in New Guinea and Bali, and took up residency in the USA in 1940 (Bateson and Mead divorced in 1950). Bateson was a prominent member of the Macy conferences, which began in 1946 and laid the foundations for the science of cybernetics[ii]. In the 1950’s, human communication became the main focus of his work, during which time he developed the `double bind’ theory of schizophrenia (Visser, 2003; Bateson, Jackson, Haley, & Weakland, 1956). From 1964-1972 he was Director of the Oceanic Institute, Hawaii (1964 – 1972). Bateson is also cited as a formative influence on Neuro-Linguistic Programming (Bostic St.Clair & Grinder, 2001) from his time at Santa Cruz in the 1970’s, where he was a member of the Board of Regents of the University of California.

Bateson’s best known work is his collection of essays, `Steps to an Ecology of Mind’ (Bateson, 2000a), originally published in 1972, which reflects this intellectual diversity and in which he addresses his central preoccupation with epistemology. Epistemology concerns the realm of knowing – how human beings perceive and construe their worlds - and is typically contrasted with ontology, which is to do with the nature of `reality’. Thus Bateson’s daughter, Mary Catherine Bateson, observes in her foreword to the 2000 edition of `Steps to an Ecology of Mind’ that; `The processes with which Gregory was concerned were essentially processes of knowing: perception, communication, coding and translation’. Crucially; `Bateson’s alternative epistemology emphasized that ecosystems were primarily informational, rather than material or energetic.’ (Harries-Jones, 1995:169).

Particularly because Bateson was attempting to address complex epistemological issues embedded in language and ways of thinking, his writing may present challenges to the reader. As noted by Brockman (1977:5-6); `Bateson is not clearly understood because his work is not an explanation, but a commission… what we must do is reprogram ourselves, train our intelligence and imagination to work according to radical configurations’.

Origins of the levels of learning

That issues of learning were important to Bateson is shown by the fact that no term has more entries in the index[iii] to `Steps’. Bateson’s writing emphasises the relational (Belenky, Clinchy, Goldberger, & Tarule, 1986) and contextual, rather than individual and intrapsychic, aspects of learning, yet from an epistemological rather than social perspective. Among other things he argues that `mind’ resides in connections and relations in systems, not in the brains of individual people.

In 1964 Bateson wrote `The Logical Categories of Learning and Communication’, an essay included in `Steps’, which merits a close reading. Bateson describes it as an attempt to illuminate `the barriers of misunderstanding which divide the various species of behavioural scientists… by an application of Russell’s Theory of Logical Types (Russell, 1921) to the concept of “learning”.’ (Bateson, 2000a:279). In summary, the theory of logical types distinguishes between a class and members of that class and, in order to avoid logical paradoxes, stipulates that a class cannot be a member of itself. For example, accountancy firms, publishers and car manufacturers are all members of the class, `business organisation’, but `business’ itself is not.

Bateson applied this principle to human interaction, arguing that it necessarily involves simultaneous, experientially fused but logically distinct, communication about both (a) content, and (b) the nature of the relationship between those communicating. In other words the latter type, which is often non-verbal, classifies the interaction. He called this `metacommunication’ (Bateson, 2000a:137; Watzlawick, Beavin, & Jackson, 1968).

As an example, if I wrestle my young son to the ground while at home, I can predict that he will immediately join in `playfighting’. There is no need for me to say `let’s have a playfight’, or even, `I’m about to wrestle you to the ground, when I do please be aware that this is playfight not a real fight’. If I did he might well tell me not to be so boring, and lose interest. This illustrates that making metacommunication explicit is not just superfluous, it can be counter-productive, puncturing the very context it is intended to create. However if I were to wrestle a business client to the ground in my workplace I would most likely be disciplined. The context and relationship are quite different, hence the content has a very different meaning. No amount of protesting, `it was just a playfight’, is likely to get me off the hook.

Generally, people understand metacommunication without needing to make it explicit because, through upbringing and experience, they have learned to read and respond to wide variety of social contexts. In workplaces, therefore, people have learnt how to deal with contradictions between what is espoused and what they understand to be `really’ going on. These subtleties are usually worked out tacitly, as making an overt challenge to apparent mismatches between the communication and metacommunication of someone in authority can be risky. The request from a boss to employees (perhaps on an awayday) to receive `completely honest’ feedback about their management style may represent a dilemma. In choosing to respond, an employee will probably review rapidly and tacitly their past experience with this person, the possible consequences of replying, and what would be appropriate in the given social setting. We suggest their response is likely to represent an intuitive but informed judgement about how honest the boss really wants the employee to be, and is unlikely to take the request for `complete honesty’ entirely at face value.

In formal logic, confusion between class and membership creates paradox, a problem that Russell’s theory of logical types was designed to address. Yet Bateson felt that; `Some of the most interesting aspects of communication may depend upon the use of contradictory messages at different logical levels…’ (Bredo, 1989:30). The way people experience and manage mismatches between communication and metacommunication forms the basis of Bateson’s `double bind’ theory. While he applied this principally to the aetiology of schizophrenia (Bateson, 2000a:206-12), Mary Catherine Bateson points out that the double bind is `part of the fabric of ordinary life’ (Bateson, 2005b:15). While we do not develop discussion of double bind theory here, it also has significance for organisational learning (Visser, 2003).

The story of the dolphin

While working at the Oceanic Institute, Hawaii in the 1960’s Bateson observed and contributed to the training of dolphins. In `Steps’, he describes (2000a:276-8)[iv] a particular event[v] that was very significant for the formulation of his theory. The influence of that event is also acknowledged as important by Argyris and Schön (1978) and Visser (2007).

In the standard training method, whenever the trainer noticed behaviour that she deemed desirable for the dolphin to repeat in front of an audience, it was marked by the reward of a fish. This type of learning could be thought of classical conditioning, in that the dolphin was learning to associate specific behaviours with the reward of food. However, once a particular behaviour was established in the dolphin’s repertoire, the reward would stop. Bateson observed that initially this appeared to cause the dolphin some discomfort and frustration. Then, perhaps to some extent as a result of its agitation, the dolphin would eventually produce a new behaviour, which the trainer would then reward before.

This training took place over a series of sessions. What struck Bateson so forcibly was the following incident:

`…each of the first fourteen sessions was characterized by many futile repetitions of whatever behaviour had been reinforced in the immediately preceding session. Seemingly only by accident did the animal provide a piece of different behaviour. In the time out between the fourteenth and fifteenth sessions, the dolphin appeared to be much excited; and when she came onstage for the fifteenth session, she put on an elaborate performance that included eight conspicuous pieces of behaviour of which four were new and never before observed in this species of animal. From the animal's point of view, there is a jump, a discontinuity, between the logical types.

In all such cases, the step from one logical type to the next higher is a step from information about an event to information about a class of events. Notably, in the case of the dolphin, it was impossible for her to learn from a single experience, whether of success or failure, that the context was one for exhibiting a new behaviour. The lesson about context could only have been learned from comparative information about a sample of contexts differing among themselves, in which her behaviour and the outcome differed from instance to instance. Within such a varied class, a regularity became perceptible, and the apparent contradiction could be transcended.’ (Bateson, 1979:137)

Bateson inferred from this incident that the dolphin’s frenzy, and the subsequent novel behaviour, indicated that the dolphin’s learning was of a different logical type. Specifically, it was learning about context (Bateson’s term) through repeated examples, such that the dolphin was acting as if able to distinguish between the class of behaviours that would be rewarded, and the class of behaviours that would no longer be rewarded.

As commented by Harries-Jones (1995:111), Bateson broke with the behaviourist tradition in that he: `considered the circumstances under which behaviour became `reinforced’, and argued that an important level of constraint, which behaviourism seemed to ignore, entered into the whole situation of learning. He would later label this constraint a repeatable “context”.’

It is important to note that this notion of context refers to an embodied, experiential reality: `for Bateson a context is the particular whole which a given part helps compose, not something separate from or abstracted from that part’ (Bredo, 1989:28-9). Later we consider the significance of Bateson’s notion of context for understanding issues of organisational learning. First, though, we describe Bateson’s levels of learning in more detail.

The levels of learning

Bateson posits five levels altogether, designated Learning 0, Learning I, Learning II, Learning III and Learning IV, as shown in table 1. Bateson scarcely discussed Learning IV, commenting that it `probably does not occur in any adult living organism on this earth’ (2000a:293), where he also indicates that it is likely to involve evolutionary change in a species. Since this has no direct relevance for organisational learning, we confine our discussion to Learning 0 through to Learning III.