Psychotherapy / Meditation

A Buddhist Middle Way Approach in Therapy Mark O’Donoghue

Mark O’Donoghue. Originally published in Australian and New Zealand Journal of Family Therapy, Volume 23, Number 4, 2002, pp. 196-201. Mark O’Donoghue is a counsellor at Adelaide University and is a Buddhist practitioner and meditator (since 1987). Address for correspondence: Adelaide University, SA 5005, Australia or email:

For many years I have been practising Buddhist meditation in my personal life. In this article, I present one way in which Buddhism has influenced my work. People who seek counselling often swing between two unsustainable or unsatisfactory alternatives, for example being ‘full on’ or doing nothing. Polarisations that trap people can be lived, remembered or imagined.

Buddhism proposes the middle way as a path out of these extremes. I present ways to deconstruct dichotomies, allowing people to find more effective alternatives. I show how to actualise a ‘middle way’ approach that is appropriate to Western therapy. I will also examine times when a middle way is not an ethically acceptable response.


This article is my attempt to adapt Buddhist ideas and practices, in a careful and critical way, to therapy. Buddhism and therapy have two very different aims. Therapy is to help people live their day-to-day lives more effectively and perhaps with less pain. The central problem of Buddhism is how to deal with the inevitability of sickness, old age, death and anguish in general. Someone may have a life that does not require or desire therapy, and yet may not have confronted the central concerns of Buddhism. Equally a person may have resolved the central concerns of Buddhism, but on a more ordinary level, may not be living a good life (Kornfield, 1998). Kornfield shows that many meditation practitioners and teachers have been left with serious problems such as alcohol abuse, sexual abuse, or misuse of power and money. Over the last ten years, many stories of this nature have been coming out and I have observed some myself. Kornfield also notes that even when these larger problems do not exist, meditators may still need to sort ordinary parts of their lives, e.g. how to relate to people, hold a job or have intimate relationships. So in bringing Buddhist ideas and practices into the therapy domain, I do not wish to deny the aims of therapy. If therapy tries to persuade people that they should become Buddhists, that is not useful. However, I believe Buddhism can contribute to therapy.

There are two ways that Buddhism can be related to therapy. Firstly, as transpersonal therapists do, by developing a therapy that fully embraces the ‘spiritual dimension’, such as Assagioli’s Psychosynthesis (Assagioli, 1971), which incorporates the Hindu Higher Self, and Fenner’s approach (1995) which draws on Buddhist notions of emptiness. The limitation of these approaches is that if one’s spiritual values are not consistent with those being incorporated, then there will not be an appropriate fit. It makes sense to seek out a therapy that aligns with one’s spiritual commitments. The second way is to maintain a non-spiritual therapeutic position, but to include the useful insights from spiritual traditions, as I will do here with the Buddhist middle way.

The Buddhist Middle Way 

The Buddha was an historical person whose life and teachings gave rise to the religion called Buddhism. The story as we know it describes a person brought up as a prince, living an indulgent life that was protected from anguish. He then for the first time encountered a sick person, an old person and a dead person, conditions which had been unknown to him. He realised that his princely life had not prepared him to deal with sickness, old age and death. He rejected all worldly pleasures, adopting an ascetic’s life of mortification and meditation. This was the second extreme, the first being his former life of pleasure. This second life still did not prepare him to deal with sickness, old age and death. He continued to meditate, but modified his ascetic practices and this provided a suitable foundation for his enlightenment. He concluded that neither the extreme of sensual pleasure (indulgence) nor physical deprivation (mortification) (Bachelor, 1997) was the way to go, and thus proposed the ‘middle way’ between these two. The middle way was subsequently prescribed as ‘the way’. Advice from a middle way perspective on how to write this paper would go something like this: do not make the article too long or too short; do not go into too much detail nor too little detail; do not write too forcefully nor too softly. Thus a space of non-extremes would be created. Many Buddhist writers since have defined ‘right behaviour’ in this way. A concern with a middle way, or with ‘being balanced’ is not unique to Buddhism; it is also a part of the ‘common sense’ of our own culture. A journalist captures this: ‘The balance is difficult to maintain — too much influence is seen as prescriptive and patronising, and too little as indifferent’ (McPhedran, 10.6.2000: 69). Although lay wisdom includes a middle way approach, it also includes other approaches, exemplified by phrases such as ‘Go for the max’, ‘Pull out all the stops’ and ‘Crash through or crash’, that are not consistent with a middle way approach. Within Buddhism, the middle way is given a more central place, and is drawn from more consistently.

My Use of the Buddhist Middle Way

If therapy becomes simply a prescription for a space of non-extremes, it becomes yet another standard against which people’s lives can be measured and to which they must conform, rather than assisting people to find out the way that will best suit, and serve, their lives. In the Buddhist and lay wisdom, the middle way often becomes an ethical prescription. People can be accused of ‘going too far’ or ‘not far enough’. Recently I saw someone on television describe their position as the ‘middle of the road’ in contrast to the ‘mad greenies’. These are all ways of saying the speaker is right because they have the ‘middle ground’. They are claims to moral superiority, often deviously so, because they avoid debate about the ethics of different positions. This is very similar to the tactic of declaring one’s position to be ‘neutral’ and thus morally superior and beyond question. There are many extremes I find it important to support. For example, to talk of a middle way between ‘too much’ and ‘too little’ abuse, rape or assault is totally against my values. The middle way I have found useful for therapy is a way out of polarities rather than the prescription for an ethical position. I have found a middle way approach most useful when people swing between two unsatisfactory or unsustainable extremes. This is a very common reason for people coming to therapy. Some dichotomies that have trapped people who have come to counselling with me are: trust vs distrust; optimism vs pessimism; positivity vs negativity; idealised happy self vs depression; over-indulgence vs self hatred; perfection vs imperfection; total control vs no control; and the possession of very good friends vs friendlessness. Sometimes people have found a way to do things that works, but as time goes on, this position ‘drifts’ towards an extreme, so that it becomes unsustainable. In these cases, a brief reference to the extreme usually suffices to situate change within the middle way.

A Simple Case Example 

When I see people who are very stressed, I will often ask them to imagine a stress scale from zero to ten. When they are at the ten end of that scale, they are so stressed that they cannot work effectively. When I suggest that they need to reduce their stress a bit so that they can begin to work effectively again, they agree with me. One day I discovered that Mario and I had not been talking a common language: on closer examination I found that even though Mario (not his real name) said ‘Yes’, he was thinking ‘No’ because he thought I was saying his stress should be at zero, i.e. he should give up work entirely. I was talking a language that has a series of graded steps between zero and ten, whereas Mario was hearing only zero and ten.

Kelly’s ‘Personal Constructs’ 

The Buddha advised us to take the middle way rather than either extreme. However, he did not elaborate a theory to account for the polarised nature of our thinking. For such a theory I have turned to George Kelly and to Buddhist Madhyamika philosophy (Fenner, 1990, 1995). George Kelly founded his Psychology of Personal Constructs on one postulate: ‘A person’s processes are psychologically channelized by the ways in which he [sic] anticipates events’ (1955: 46); and eleven corollaries. For Middle Way therapy, the most relevant corollary is the Dichotomy Corollary: ‘A person’s construction system is composed of a finite number of dichotomous constructs’ (56). Kelly ascribes ‘… a dichotomous quality to all human thinking’ (1955: 109). The conceptualisation of constructs as polarised is contrasted with ‘concepts’ and ‘categories’, both of which are unipolar ways of describing thought (Bannister & Fransella, 1980: 21).

Expanding on this, Kelly wrote:

Each construct involves two poles, one at each end of its dichotomy. The elements associated with each pole are like each other with respect to the construct and are unlike the element at the other pole (Kelly, 1955: 137). Polarities can be either explicit or implicit (Bannister & Fransella, 1980: 22). Explicit polarities are obvious to most people. For example, ‘small’, ‘slow’, ‘short’ imply the corresponding polarities ‘big’, ‘fast’ and ‘tall’. Implicit polarities are not so obvious. Kelly argues that all statements are at least implicitly linked to their opposite, for example, ‘red’ and ‘not red’, ‘talk’ and ‘not talk’ and ‘sitting’ and ‘standing’. The ‘common abstraction’ that defines a construct is itself defined in contrast to other constructs, and so on. For example, when we look at the construct ‘chairs’, we might have at one pole lounge chairs, and at another, kitchen chairs. ‘Chairness’ is the aspect they have in common, that makes possible the comparison of two types of chairs. Again, chairs might be one pole of the construct ‘furniture’.

The Madhyamika (Middle Path) Theory of Dichotomies

In the Middle Path tradition of the Buddhist writers Nagarjuna (2nd Century) and Candrakirti (7th Century), ‘analytic’ meditation is used to achieve a state of liberation or unconditional freedom (Fenner, 1995: 29). Analytic meditations take many forms, one of which is to collapse two polarities into each other, thus dissolving the polarity. An example of this in ordinary life is to be neither aggressive nor passive. The result is dissolution of the conceptual category that is made up by the dichotomies. Taking this route results in a cessation of conceptuality, resulting in a direct perception of reality unmediated by conceptuality, which is the Buddhist enlightenment. This is not the aim of therapy, at least as most people (including myself) practise it, although the various Transpersonal Therapies may constitute an exception. However, the Middle Path meditators’ understanding of how conceptuality works provides another useful answer to the question: ‘Why dichotomies?’ Within a Middle Path account of dichotomies, there are five developmental steps: (1) undifferentiated objects; (2) characterised objects; (3) concepts; (4) thoughts and (5) language. According to Middle Path theory, the world is made up of undifferentiated ‘objects’. The undifferentiated objects of the world are divided up into separate objects by a conceptual process of the ‘characterisation of objects (via dichotomies)’. Based on a characterisation of objects, concepts are formed and these are in turn the basis for thoughts and language. Concepts are formed by a process of defining what the concept is and is not, for example, ‘good’ and ‘not good’ define the concept ‘good’. According to this theory, ‘good’ is not an intrinsic quality but is a relative one, because ‘good’ only exists in relation to ‘not good’. In thought and language we refer to one end of the dichotomyon on its own. This gives the illusion that the concept (e.g. ‘good’) exists in and of itself, isolated from its opposite. However, the two ends remain connected in some way, which continues to influence their use. It is this connection that results in all the forms of polarisation and their consequences. Thus if we identify ‘good’, then implicitly there must be ‘bad’, and vice versa. Because the two poles are connected in this way, people’s judgements may shift from one pole to another if one pole is not sustainable. To demonstrate how this theory works, I will use the example of a ‘table’ and show the evolution from an undifferentiated ‘object’ to defining an object as existing separately from all other objects (Fenner, 1995: 37–38). The world is made up of undifferentiated objects. Through a conceptual process, the undifferentiated ‘objects’ are divided up into objects that have been ascribed the relevant characteristics of tables and those that do not. Tables and not tables are explicitly linked because they are defined in opposition to each other. We then go on to use the word ‘table’ as if it is not defined in opposition to ‘not tables’. We cease to be aware of the other pole, that is, ‘not tables’, but of course, the polarity between tables and not tables still structures our thinking.

Therapy and Middle Path Buddhists

In the Middle Path analysis, dichotomies are brought together to generate a paradox and thus produce liberation by ‘destructur(ing) conceptuality” (Fenner, 1995: 44). Middle Path meditators aim for a state where all conceptuality, which is necessarily bipolar, is destructured, leaving a non-conceptual state that ends anguish, since anguish is itself a concept. In contrast, the Middle Way approach to therapy that I am proposing can begin with a deconstruction of particular concepts which will be replaced by other concepts — ones that are less painful or more in alignment with what the people who see me want, but of course still bipolar. When people are caught in a swing between two poles, the dichotomous structure of the concept shifts them between the two poles as if they were in a maze in which they are trapped. When one pole is unsustainable or unsatisfactory people tend to shift to the other pole, which leaves them no room to display any other behaviour, feeling or thought. The role of the counsellor is to locate this trap and help people out of it. When one pole is sustainable, workable or satisfactory people will not swing to the pole to which it is linked. Bipolarities can be divided into two types, one that is necessarily always present and another that exists within the former but has varying degrees of bipolarity or different bipolarities. These are sometimes called ‘bipolar’ and ‘scalar’ (Bannister & Fransella, 1980: 22). Scalar constructs exist as a product of bipolar constructs. For example, black and white can be two poles of a bipolar construct and shades of grey is a scalar construct that can exist between these two poles. Another example may help. There are two types of choices that we can be given. One involves a bipolarity that is either 0 or 10 and we must choose which number we want, giving us only two choices. The second set of bipolarities is 0 &1, 1 & 2, 2 & 3 … 9 & 10. In the second set, we can choose between ten separate dichotomies. If people believe they have only the first set of choices (as Mario did in my earlier case example) and this is unsatisfactory for them in some way, we can assist them to have access to the second set, thus giving them more and different choices.


When I describe the way thinking is dichotomised, and how the middle way leads out of that situation, the solution looks obvious. However the way out does not look obvious to a person caught in such a polarity. The middle way is obvious from a middle way paradigm but not from a bipolar paradigm, in the same way that visual gestalts (like the picture which is ‘both’ an old woman ‘and’ a young girl simultaneously) are obvious once you see them, but not if you cannot see them. We therapists get trapped in the obviousness, which can block us from the creative ways to assist people to step out of such dichotomies. For example, just saying to someone that they need to be balanced has little benefit, since the obviousness of this statement renders it ineffective. What we need is a systematic way to bring about this change.

A Middle Way Therapy 

I have distinguished four types of middle way, which I will demonstrate by using the example of how a person might deal with someone with whom they are in conflict. Firstly, for a scalar construct, a middle way is half way between the two defining poles. Passive or aggressive communications are two poles that people often get caught between. A common middle way out of these dichotomies is assertive communication. Halfway along a scale is probably ‘the’ middle way as defined within the original Buddhist middle way presentation. However, subsequent developments in Buddhist philosophy and practice defined the middle way as neither a nor b, thus including the scalar middle way but also other possibilities. This second type of middle way includes all points along the scale between the two poles, not just the half way point. Then, the option ‘neither passive nor aggressive’ could be replaced by a large range of possibilities including: less aggressive, slightly aggressive, assertive, less passive or slightly passive — anywhere along the scale other than the two extremes. A third type of middle way can be obtained by stepping out of a specific construct altogether. This is appropriate when both extremes are located within a non-helpful or non-preferred construct, for example if the construct is aggression and the extremes are verbal aggression and physical aggression. Such a construct might make a person swing between two unsatisfactory types of aggression that are dichotomously linked. We might question the construct’s usefulness or preferability, and perhaps locate another more useful construct, for example, ‘communication’. Thus a person could be neither verbally nor physically aggressive, but respond in a way that is not within the construct of aggression, but perhaps can be located within the construct of ‘communication’. Fourthly, following Madhyamika philosophy, we can devise another category that steps out of both poles but does not step into another construct, thus moving to a non-conceptual realm outside of all constructs. Such a state cannot really be described by language, because by definition language is founded on dichotomies; it involves a direct non conceptual experience of reality as it is. Such a person would be in (Fenner, 1995: 29) ‘a state of liberation’ and ‘unconditional freedom’. Such a person might have a more fluid sense of reality, such that things and states form and dissolve continuously, removing the likelihood of being caught in any one state.

Lived, Remembered and Imagined Polarities

So far I have written about polarities in peoples’ lives, and will call them ‘lived polarities’. Polarities can also be remembered as lived in the past or imagined as possibilities for the future; both of these can also trap us. The memory of a past extreme can result in people living at the alternative extreme. The memories of abused trust can lead people to be overly distrustful; recalling that they failed because they did not work hard enough can induce people to try so hard that they get exhausted and fail again, and memories of times when they have been naively idealistic can lead people to be overly cynical. A polarity can also be imagined. Sometimes people can imagine only an extreme and unsatisfactory alternative and this keeps them in their current and unsustainable extreme. An example is students who work too hard and come to me in a stressed condition. The only alternative they can imagine is not to work at all. Because they have worked so hard for so long, this becomes a very appealing alternative for them. Their thinking becomes trapped in this polarity. People can also worry about being too kind or too harsh. For some this has never happened, but they imagine it and know that the imagined extreme is not workable. The imagined extreme becomes the reference point that keeps them ‘stuck’ at the other polarity. For example, when counselling someone who is very harsh with her or him self, we might look to kindness as a way out. Such people often comment that you can be too kind, and I agree. When kindness is located between harshness and being too kind, we have a more precise definition of appropriate kindness that is not trapped in a dichotomy.

How I Present this Information 

Earlier in my career, I used to present this concept very dramatically. On two sides of a white board, I would write a list of the two extremes and all the things that would go with them. I also might ask about the voice tone, colour, feeling and phrases associated with these ways of thinking or looking or feeling. I would explore consequences of that belief and why they were not acceptable or working for the person. This is very useful in locating statements, beliefs, actions and feelings within these extremes. Once I had done this and the person was clear why each did not work, the swings stopped or were deconstructed. I would then dramatically draw a circle in the middle and ask if this was the way out. People would usually agree that it was but would be at a loss to know what this middle way would look like. This was not surprising, because their thinking and responses had been structured by a dichotomy, and some creative work was needed to create a middle position. At that time I was more influenced by the strategic approach to family therapy as exemplified by Haley (1973), Erickson (Rosen, 1982) and the earlier White (1979, 1986). I now mostly prefer a more ordinary and less dramatic presentation.

Samantha: An Illustration from My Work 

Samantha came to counselling to deal with a range of problems. For the purposes of this article I will focus only on the problem we called ‘stewing’. Stewing would involve Samantha going over and over events in her life, which she was not happy about. As she kept going over these events, she would feel worse and worse until she was like a rabbit caught in a spotlight unable to escape. For her the solution was the non-existence of stewing.

In the first week of meeting we discovered a contrast between ‘thinking’ and ‘stewing’ as responses to a supervisor with whom Samantha had difficulties. When she stewed on the interactions with the supervisor she became angrier and could not study. Thinking put these difficulties in context and again gave her a perspective that allowed her to continue studying and working with her supervisor. During the next week, we discovered a number of ways Samantha had used to get out of stewing. These included recognising that stewing was happening, using thinking to reduce the intensity of stewing, laughing at herself for stewing and thinking about roller blading.

Samantha objected that ‘I cannot always laugh’ as a way to stop stewing. The benefit of thinking about roller blading was that ‘Stewing was sitting on the side’. When measured against the extreme of the non-existence of stewing, however, these did not count as ways out of stewing. These two responses to two of the potential ways out of stewing both indicate that Samantha has in mind some state where there is no stewing, rather than steps to reduce stewing. We have a polarisation between ‘stewing’ and ‘non existence of stewing’. The non-existence of stewing is an idealised, non-sustainable and imagined extreme and was destined to fail as a goal.

I have (at least) two ways of responding to such a polarity. Firstly, I use a process of questioning that deconstructs the extreme preference. This then allows for the discovery of other ways out of stewing that are not extreme. Secondly, I present some counter ideas and examples for consideration.

I ask questions such as ‘Is it possible for there ever to be totally no stewing?’ Samantha replied: ‘NO’. (If the answer had been ‘Yes’, I could have asked, ‘If someone is to get to a point of no stewing do you think there will be steps along the way, which will include some stewing? Is half stewing better than all stewing? Is three quarters stewing better than all stewing?’)

With Samantha, I talked about change being a process of small steps and contrasted this with trying to take one big step. I drew a diagram on my white board to represent this (see Diag. 1) in which a person swings between ‘all at once and going nowhere’ and ‘giving up and going nowhere’. I then contrasted these two extremes with ‘one step at a time and getting there’ (Diagram 2) which is a middle way for the polarities of ‘all at once and going nowhere’. If one step is too large, we can break it up into smaller steps and if those are too large, then we break them down into smaller steps and so on, until a step is found that can be taken. I also like to suggest that we zoom in as if we have a telephoto lens and keep doing this until we find steps that can be taken.

Figure 1
‘All at once’ or ‘Giving up’

Figure 2
‘One Step at a Time and getting There’

During week three, Samantha spent one weekend off with her family, did not stew and during the following week studied for a few days. A new criterion began to emerge against which these developments could be measured and could contribute towards reducing stewing. We then had a concept of degrees of stewing. Samantha could see changes and answer ‘Yes’ when asked if these new ways out of stewing were desirable. The first polarisation was between stewing and non-stewing and a second polarisation was uncovered between there being no solutions and only one solution. We discussed the possibility of laughter as a way to overcome stewing and Samantha said, ‘I cannot always laugh every time when I am stewing’. Her statement was based on the presupposition that laugher could be the only response to overcoming stewing. I asked if it was possible that only one thing would solve the problem, or was it more likely that a diversity of approaches would help, so that she could move from one to the other if one approach did not work. I drew on the analogy of a tool bag filled with many different tools for dealing with stewing. I also drew a diagram with many lines converging on one point to suggest a multitude of approaches to solving one problem. Before using a Middle Way approach, I would have focused much of our effort on thinking as a way out of stewing. Within the domain of thinking, we would have had both thinking that was a way out of stewing and thinking that was a polarisation of stewing, which would have inevitably flipped back to the other pole, stewing. Now I can better help her to focus the domain of ‘thinking’ so that we separate thinking that is useful and thinking that is an extreme and unsustainable. I still ask people to assess whether ‘thinking’ is useful, but they have different criteria against which to measure it.


A Middle Way approach to therapy can deal with the pervasive tendency to polarise and swing between two unsustainable extremes. The Buddhist middle way made this clear to me, and the work of Fenner and Kelly explained why it is so. Over the years I have considered grounding this work in a number of alternative theories including: Hegel’s dialectics; Russell’s work on logical types and Bateson’s subsequent use of that work; Derrida may also have relevant things to say from a post modern perspective and Trialectics has also been suggested as a compatible approach. Rather than focus on these authors I decided to focus on Buddhism’s contribution and Kelly’s theory. I chose Kelly because his work provided a clear and close fit to my practice. However, I wrote this article from a Buddhist perspective for two reasons. Firstly, Buddhism was the source of inspiration for my middle way approach. Secondly, it provided a new, inspiring and, I believe, interesting source of ideas from which to draw. In my work as a therapist, I am always attending to the influence of dichotomies either in the background or explicitly. It has helped the people who consult me to step out of the polarisations that trap and limit their lives. I have tried to avoid naively or inappropriately pushing Buddhism into the western therapeutic tradition, and I have endeavoured to engage critically with the Buddhist approach.


I would like to thank people who have made comments on earlier drafts of the article: Jane Copeland, Natalie Fuller, Sally Hebenstreit, Greg Smith and Denis Heath. I have presented versions of this paper at The South Australian ANZSSZ Conference, 2000, and The Pan Pacific Family Therapy Conference, Melbourne, 2001.


Assagioli, R., 1971. Psychosynthesis: A Manual of Principles and Techniques, NY, Viking. Bachelor, S., 1997. Buddhism without Belief: A Contemporary Guide to Awakening, London, Bloomsbury Publishing. Bannister, D. & Fransella, F., 1980. Inquiring Man: The Psychology of Personal Constructs, 2nd edn, Harmondsworth, Penguin. Fenner, P., 1990. The Ontology of the Middle Way, Dordrecht/Boston/London, Kluwer Academic. Fenner, P., 1995. Reasoning into Reality: A System-Cybernetics Model and Therapeutic Interpretation of Buddhist Middle Path Analysis, Boston, Wisdom. Haley, J., 1973. Uncommon Therapy: The Psychiatric Techniques of Milton H. Erickson, M.D., NY, London, Norton. Kelly, G., 1955. The Psychology of Personal Constructs, Vols 1 and 2. 2nd edn. London, Routledge & Kegan Paul. Kornfield, J., 1998. Even the Best Meditators Have Old Wounds to Heal, Psychotherapy in Australia, 4, 3: 37–40. McPhedran, I., Trouble in Paradise, The Advertiser (Adelaide), 10.6.2000: 69. Rosen, S., 1982. My Voice will Go With You: The Teaching Tales of Milton H. Erickson, M.D., NY, London, Norton. White, M., 1979. Structural and Strategic Approaches to Psychosomatic Families, Family Process, 18, 3: 303–314. White, M., 1986. Negative Explanation, Restraint and Double Description: A Template for Family Therapy, Family Process, 25, 2: 169–186. Mark is a counsellor at The University of Adelaide and OCAR Services. Address for correspondence: The University of Adelaide, South Australia 5005. Fenner writes about ‘destructuring’ which is different from deconstructing. To destructure is to undo and not replace with another concept. To deconstruct is to undo but to replace that concept with another one. This distinction hinges on whether you believe unmediated experience is possible, as Madhyamika Buddhists do. I often draw simple diagrams in counselling sessions because people find them useful. I suspect other counsellors do the same but we do not write or tak about this so I have decided to include some in this paper. It might be a useful thing for others to do.