The Therapeutic, Musical Relationship: a Two-Sided Affair?

A Consideration of the Therapist's Significance of the Musical Input in Co-Improvisation

By Simon Procter

Editors note: The article presented here is republished from British Journal of Music Therapy Vol. 13, no. 1, 1999 with the kind permission from the publisher and the author.

|Abstract| |Introduction| |Sources of Evidence| |An Examination of Clinical Practice| ||Discussion - the Significance of the Significance| |Acknowledgement| |References| |Figures|

Abstract

Many improvisational models of music therapy involve therapist and client improvising spontaneously together. This is widely described as a form of musical relationship. However whilst there is much discussion of the client's musical input in the literature, the therapist's music attracts less attention. The author considers reasons why this may be the case and seeks evidence, from the music therapy literature and beyond, as to whether the therapist's musical input is of significance for the therapeutic process. An example of detailed analysis of the author's own clinical work is presented, in order to establish whether the therapist's musical input has an observable impact on the shared musical experience and might thus be judged to be significant for the therapeutic musical relationship. Concluding that it is indeed significant, the author goes on to consider the implications for the way music therapists consider their work.

Introduction

Whilst surveying the British music therapy literature for a recent research project (outlined later in this article), I was struck by the mismatch between the way music therapists typically refer to their clients' input and the way they refer to their own.

Where writers focus on musical events within therapy sessions, they typically describe the client's musical input ("J sang a nursery rhyme", "X played the metallophone", "P's playing suggested great anger" etc.) rather than the therapist's. Indeed, in many cases, there is no corresponding description of the therapist's music, even though the client is playing as part of a co-improvisation with the therapist. Thus the client's music seems to be considered out of its context of musical relationship rooted in the experience of co-improvisation.

Before going on to consider possible reasons for this perceived discrepancy, it is worth briefly considering the twin concepts of co-improvisation and musical relationship, since these are central to the tenor of this article.

Co-Improvisation

Music therapy in the UK has a strong tradition of active co-improvisation involving both therapist and client(s) making music together. The development of skills in this area is an important part of all the recognised music therapy trainings in the UK. This contrasts with the situation in many trainings elsewhere, within which, for example, students may be instructed in the use of receptive music therapy or where music is used as a form of behavioural stimulus and/or reward. This suggests that music therapists in the UK attach particular importance to active co-improvisation as at least part of the therapeutic process.

Musical Relationship

Just as interpersonal therapy in general values relationship between therapist and client, so in British music therapy there is an acceptance of the fundamental importance of musical relationship arising in co-improvisation to the therapeutic process. Influential pioneers as diverse as Nordoff and Robbins and Priestley emphasise its centrality to their work as well as the importance of its musical nature:

"The child-therapists relationship, as it develops in therapy is, in fact, an interrelationship. It results from and takes expression in the musical and personal interresponsiveness of child and therapists..." (Nordoff and Robbins, 1977, p. 180).

"An essential aspect of this method of using improvised music communicatively is that the therapists make no effort to establish a relationship with the child other than on the basis of musical expression and musical activity" (Nordoff and Robbins, 1977, p.189).

"The musical relationship between therapist and client, and the verbal relationship, are two quite different things, having totally different characters. They resemble the difference between two judokas bowing to each other before fighting in the dojo, and actually fighting, or a husband and wife discussing the garden and the same pair making love. Verbal communication can be a cold, lonely business. When one person speaks the other must be silent or his speech will annihilate that of the first. It is not a united expression like music and fighting and lovemaking. But if one person improvises and a second enfolds her sound expression in his own, she is instantly contained in a greater whole. She expresses herself completely and yet feels that she is a part of something greater. Aware and whole and yet a part, it is a good feeling" (Priestley, 1975, 223).

Presumably this concept of musical relationship or interrelationship indicates a mutual relationship - one which is moulded and re-moulded by the musical contributions of both therapist and client and within which the musical inputs of both therapist and client are therefore of significance. Yet where writers do consider the therapist's input, it is most commonly their verbal input which is described ("I wondered if this might correspond to other areas of insecurity in her life" , "I acknowledged her pain", "I chose to say nothing"). If distinguished practitioners neglect to describe their own part in the shared music-making, focusing instead on their verbal input, does this throw into question the presumed significance of the therapist's musical input for the musical relationship? And if so, where does that leave the concept of musical relationship - can it really be a genuinely two-sided relationship between therapist and client?

Before assuming such doubt, however, it is perhaps worth considering alternative explanations for this apparent mismatch:

• It might be assumed that the therapist is a neutral factor, always providing an optimum experience, supporting and challenging as appropriate.
• The therapist's musical role might be seen as too indefinite or elusive to be usefully described.
• Such a focus on the therapist's input might be regarded as potentially dangerous - revealing unchallenged assumptions and musical shortcomings in the clinician and author.

Of these possible alternative explanations, none seems particularly satisfactory - the first suggests a naivety and lack of self-awareness incompatible with the practice of music therapy, the second a jack of analytical rigour and the third an unwillingness to engage with reality. So we are left with the same question: is the therapist's musical input of significance within the musical relationship? Or, to put it another way, can the musical relationship as experienced in co-improvisation be regarded as genuinely two-sided?

Sources of Evidence

In an attempt to answer these questions arising from the music therapy literature, I have turned to a broader literature of music and in particular of co-improvisation.

As music therapists we are musicians who seek to engage clients in musical interaction for clinically definable therapeutic purposes, our musical actions within this interaction being informed by those purposes. Such interaction can thus be viewed in both musical and therapeutic terms - the two views not being at all exclusive, rather, each being capable of informing the other.

Thus it seems appropriate to look first to the musical literature relating to co-improvisation and then to return more specifically to those parts of the music therapy literature which do explicitly consider the therapist's musical input. For clarity, each of these is divided into two sections - theory and practice, even though such a boundary is necessarily artificial. Finally, there is a brief excursion into the possible parallel insights on offer from the field of mother-infant interaction research.

Music Theory and Psychology of Co-Improvisation

The influence of the therapist's music within therapeutic interaction has been specifically identified as an area of music therapy research to which music psychologists could very usefully contribute by focusing on real clinical work in conjunction with music therapists (Bunt et al., 1988, p. 69). At present, however, evidence has to be transplanted from theoretical writings on co-improvisation in general.

Pressing (1988, p. 134, p.135) emphasises the central role of feedback in ensemble improvisation, which he therefore describes as a "more open" skill than solo improvisation, because each performer is required to modify his output in the light of the output of his coperformers. In other words, co-improvisers do not simply play simultaneously: they also attend to and react to each other's playing and to the reactions of others to their own playing. This kind of attentive coimprovisation between two people must therefore be seen as a two-way process in which the inputs of both players are of significance. The question remains, however, is the kind of co-improvisation found in music therapy this kind of attentive co-improvisation, or at least can it be?

Music Practice

Musicians as diverse as Lionel Salter and Paco Peña are clear about the fundamental two-way nature of their very different fields of co-improvisation. Lionel Salter, as a baroque harpsichordist, has this to say about improvising the continuo:

"It is a two-way thing. The violinists, and the other string players in the group, spurred the harpsichordist on to invent something and vice versa... the harpsichordist might then think of something first and they would follow him." (Bailey 1992, p. 21)

Paco peña, the flamenco guitarist, quoted in Bailey (1992, p. 63), describes moments in co-improvisation when singer, dancer and guitarist somehow 'combine' to reach a new "level of performance" and calls this duende - a concept which is familiar to all improvisers, and to many of us in our work as music therapists. Similarly, Kjell Oversand describes moments when "improvisation is like the language which spontaneously originates between lovers, and what is usually called erotics" (quoted in Ruud 1995, p. 106). Much more mutual it is hard to imagine!

Music Therapy Theory and Research

Pavlicevic argues explicitly for a view of clinical coimprovisation as reciprocal: "It provides the space for a highly dynamic and reciprocal interaction between the therapist and the person... Clinical improvisation offers the musical partners the opportunity to apprehend one another directly through the music: they share a sense of themselves in relation to one another" (Pavlicevic 1995, p. 168).

A recent study (Neugebauer & Aldridge, 1998) shows a clear parallel between the way therapist and client relate musically moment by moment in co-improvisation and the changing relation between their heart rates. The results produced in this study demonstrate with remarkable clarity not only that there is a physiological basis to musical relatedness, but also that nonverbal communication in co-improvisation is mutual (i.e. that each person's playing or singing has an effect on what the other does) - what Neugebauer and Aldridge refer to as "mutualities of influence" (p. 51).

Music Therapy Practice

The writings and case studies of Nordoff and Robbins (1971, 1977) emphasise the therapist's awareness of, and musical response to, the client's musical presentation. The therapist should be able to meet the client, to offer an alternative way of being and, where appropriate, to lead the client into more organised or flexible playing.

Creative Music Therapy as outlined by Ansdell (1995) - itself an extrapolation of the approach of Nordoff and Robbins into new clinical areas - is underpinned by the understanding that client and therapist share in a creative musical relationship. Music therapists working within the Creative Music Therapy approach talk of key moments in their work when they are particularly aware of the two-way nature of the co-improvisation, comparable with Peña's notion of duende. Descriptions of this include "connecting with", "quickening", "catching", "meeting" and "moving" the client (all taken from Ansdell 1995). Sometimes this is even visible: Ansdell describes how a client's facial expressions changes as she becomes "connected" with the therapist in the musical experience (p. 44), and how another is physically liberated from pain (pp. 76-80). In such a setting, founded on listening and response, the therapist's own input to the musical relationship must be significant. Thus it seems fair to describe the relationship as two-way, featuring discernible moments of "connectedness".

In the most fully described case study yet to have come out of Creative Music Therapy, Colin Lee's client describes his feelings about improvising together as opposed to solo improvisation: "one picks up the other person's perceptions, one latches on and is fed by the dialogue and enriched. The interplay of the two enhances... It's a dual process... of linking together the listening musical mind, and learning and being moved by another person's musicianship" (Lee 1996, p. 86). This does seem to describe the kind of attentive co-improvisation outlined by Pressing above.

Parallels in Mother-Infant Interaction

In recent years, many music therapists (perhaps most notably Pavlicevic,1990) have pointed to parallels between mother-infant interaction and the sort of musical interaction that can be observed between music therapist and client. Whilst it is important to remember that the focus here is on observed parallels between two different forms of interaction, rather than necessarily on two manifestations of the same form of interaction, it is nevertheless also worth reflecting on the significance of the mother's contribution for the mother-infant dialogue, in order to consider any possible parallel for us as music therapists in the musical therapeutic relationship.

The research literature makes clear that mother-infant interaction is two-way -- that the cross-modal inputs of both mother and infant are significant for the interaction and hence for their shared experience of the interaction. Where mothers are not responsive or respond in an unexpected way, this has an observable impact on the infant's behaviour. Where the child cannot respond fully and adequately, this has an observable impact on the mother's behaviour (e.g. see Stern, 1977).

Thus, as Pavlicevic points out, mother-infant interaction is characterised by "dynamic interplay" which takes the form of reciprocal, mutual musical interaction between mother and infant. Pavlicevic uses this as a descriptive allegory for the moment-by-moment process of co-improvisation in music therapy. If such parallels with the practice of music therapy are as convincing as many music therapists believe, then this must at least impel us to consider whether such mutuality and interdependence is observable in our own clinical practice, within whichever clinical area.

An Examination of Clinical Practice

Background

In the earliest stages of my work with adults with chronic mental health problems, I found that my own musical input was different in this setting than elsewhere. In the joint, spontaneous musical improvisations, I seemed to be using more repetition both of small-scale material and of whole themes and melodies. I was also using many straightforward and explicit harmonic progressions. Cadence points were built up to over a lang period and reiterated many times and it frequently felt very obvious what was going to happen next. I wondered whether this predictability was a result of musical limitations on my part, or whether it might be offering clinically relevant musical opportunities to this particular client group. In order for this to be the case, my musical input would have to be regarded as significant for the ongoing co-improvisation.

On initial indexing of sessions, it seemed that times of perceived maximal connectedness in the musical relationship between the client and myself coincided with a high lever of what I regarded as predictability in my input. Similarly, when my playing lacked this quality of predictability, or when I took a wrong turning and suddenly became unpredictable, there often seemed to be a corresponding drop in connectedness between the client and me.

I thus undertook a qualitative research project (reported fully in Procter, 1997) to explore whether there was same observable basis to these impressions and whether changes in my musical input might really be therapeutically relevant to this particular client group.

Methodology

As part of the study, I developed a methodology for analysing data from clinical work in order to trace the relationship between perceived predictability in the therapist's musical input and perceived connectedness in the co-improvisation between the clients and the therapist. A brief summary of this methodology is presented in Figure 1, along with examples of the application of this methodology to a small part of one of the extracts analysed in the project (Figures 2-5). (The extract is taken from my third session with an adult diagnosed with chronic schizophrenia.)

Conclusions Drawn

In the graph display in Figure 5, predictability and connectedness seem to shadow each other closely. Where there is a significant movement in one, there is a corresponding movement in the other, although it is not possible to deduce a general causal relationship between the two.

Such a pattern, replicated in my analyses of other contrasting extracts, suggest that in the clinical work I was examining, there was a link between perceived predictability in the therapist's musical input and the musical connectedness between therapist and client. A fall in one is associated with a fall in the other, and a rise in one is associated with a rise in the other. Although predictability often seems to lead the way, this is not always the case, and it seems that in some cases a fall in connectedness may trigger a decline in predictability in the therapist's input. Thus the relationship may be described as being characterised by interdependency. Further and much more detailed research would be required to establish the exact nature of this relationship.

The significance of the therapist's musical input

The association which I noted in my own work between predictability in the therapist's musical input and the musical connectedness between therapist and client suggests that both parties influence the course of the coimprovisation and hence the musical-therapeutic relationship. This ties in with the earlier suggestion that co-improvisation in music therapy is a two-way process in which the therapist's musical input has a significant effect on the ensuing co-improvisation - the shared musical experience - and thus is of significance for the musical therapeutic relationship.

Discussion - the Significance of the Significance

If it is accepted that the therapist's musical input to the therapeutic co-improvisation is of significance for the ongoing improvisation, it follows that therapeutic coimprovisation is indeed characterised by dynamic interplay. It is thus a place of two-sided interaction where meeting, reflecting, challenging etc. can all take place within the process of co-improvisation itself. This is the kind of improvisation that Ruud describes when he writes: 'Improvisation means to change in relation to other human beings, phenomena, situations' (Ruud 1995, p. 93). Here is the very stuff of therapy! The musical relationship is to be found in the act of co-improvisation - of making music together. Client and therapist each experience the musical presence of the other and also experience themselves in active, changing relationship with the other. Thus it is possible for the dynamics of therapy to happen and to be observed within coimprovisation.

Of course, co-improvisation in music therapy is not always truly mutual. Indeed, the pathology of the sustain musical mutuality. In such cases, the therapist may work within the extant, non-mutual, musical relationship, in order to enable the client to experience such mutuality. This is a common thread in the case studies of Nordoff and Robbins (1971, 1977) as well as in the writings of many music therapists since, particularly those who identify themselves with this approach. For example Pavlicevic (1990) describes how the therapist "tests the interactive potential" of the improvisation. The fact that a client may not initially be able consciously or unconsciously to respond to the therapist's musical input does not invalidate the significance of that input for the musical relationship any more than the difficulty a client in psychotherapy might have engaging with his psychotherapist invalidates the significance of the psychotherapist's verbal input for the verbal relationship.

As music therapists we have something quite extraordinary to offer - a therapeutic relationship in music which, unbounded by words, is consequently quite other in possibilities than a verbal relationship could ever be. But that relationship, in order to realise its therapeutic potential, must be two-way. It demands that the therapist's input - i.e. his or her musical actions within co-improvisation - be regarded as significant for the relationship. Consequently, we must hold ourselves clinically accountable for the what, the how and the why of everything we do in music - each note, each breath, each harmonic implication, each nuance. When we neglect to do this, we abandon aur claim to therapeutic awareness, thus failing aur clients, ourselves and music therapy.

Acknowledgement

I would like to thank Mercédès Pavlicevic for her encouragement and her comments on an earlier draft of this article.

References

Ansdell, Gary (1995). Music for Life. London: Jessica Kingsley Publishers.

Bailey, D. (1992). Improvisation - its Nature and Practice in Music. London: The British Library National Sound Archive.

Bunt, L., Cross, I., Clarke, E. and Hoskyns, S. (1988). A discussion on the relationships between music therapy and the psychology of music. Psychology of Music 16, 62-70.

Lee, Colin (1996). Music at the Edge. London: Routledge.

Neugebauer, Lutz and Aldridge, David (1998). Communication, heart rate and the musical dialogue. British Journal of Music Therapy 12 (2), 46-52.

Nordoff, Paul & Robbins, Clive (1971). Therapy in Music for Handicapped Children. London: Gollancz.

Nordoff, Paul & Robbins, Clive (1977). Creative Music Therapy. New York: John Day.

Pavlicevic, Mercédès (1990). Dynamic interplay in clinical improvisation. Journal of British Music Therapy 4 (2): 5-9.

Pavlicevic, Mercédès (1995). Interpersonal processes in clinical improvisation: towards a subjectively objective systematic definition. In Wigram, Tony, Saperston, Bruce & West, Robert (Eds.) The Art and Science of Music Therapy. Chur, Switserland: Harwood Academic Publishers.

Pressing, J. (1988). Improvisation: methods and models. In Sloboda, John A (Ed.), Generative Processes in Music. Oxford University Press.

Priestley, Mary (1975). Music Therapy in Action. London: Constable.

Procter, Simon (1997). Predictability within the therapist's musical input: its significance for the clinical process in music therapy with adults with mental health problems. Master's Thesis: The Nordoff-Robbins Music Therapy Centre, London, UK.

Ruud, Even (1995) Improvisation as a liminal experience: jazz and music therapy as modern "rites de passage". In C.B. Kenny (Ed.) Listening, Playing, Creating - Essays on the Power of Sound. Albany: State University of New York.

Sloboda, John A. (1985). The Musical Mind. Oxford University Press.

Stern, Daniel (1977). The First Relationship. Harvard University Press.

Figures

Fig. 1. - A Brief Summary of the Methodology

Fig. 2. Musical Transcription and Segmentation

 View full size image 8.5666kB

Fig. 3. Verbal Description

Fig. 4. Categorisation

"++/--"

Fig. 5 Predictability/Connectedness Correlation

Since the two lines on each graph display (representing predictability and connectedness) are describing different commodities, and since each line is drawn on the basis of an informed impression rather than any arithmetical or objectively quantitative system, the points of crossover are of no significance. Nevertheless, the relative contours and particularly the relative changes of contour can be said to convey information as to the strength of the perceived predictability and connectedness with relation to musical time.