I have been in private practice since 1986 and The Institute was founded by me. The Hedenbro Institute offer couple therapy, family therapy, supervision, teaching and research. I have published several articles in both English and Swedish, written books in Swedish and produced eight films about interaction. As a researcher I have been associated to the Karolinska Institute in Stockholm Sweden since 1996. My thesis ” The family triad – the interaction between the child, its mother, and father from birth to the age of 4 years old” was presented 2007 but the study is still under work now when the children are 20 years old.
I am trained as a psychotherapist and started this training in Tavistock, London. My roots as a familytherapist started in London but continued at The Philadelphia Child Guidance Clinic where Salvador Minuchin was an important therapist.
Over the years I have been teaching and working in several other countries such as England, USA, the Scandinavian countries, Russia and Erbil in Irak.
Working with families with a systemic perspective is most important but we also use psychodynamic and cognitive/affect based methods.
Children´s communication and interaction is a specialty where Micro Analysis plays an important role to understand and help. Intervention is a method in which video filming and reviewing with clips from that film is used to give the parents or professionals guidance in how to support the child in development. It also helps the parents or caregiver to find ways in how to help the child in his or her difficulties and problematic situations. Attachment is important through our whole life and therefore it is most important to establish a secure attachment during infancy and early childhood. Also later in life for example with adopted children attachment based work is really important.
The realization that from the very beginning of his life a child is a social being has greatly contributed to an interactional view being taken of the relationship between biological and psycho sociological factors in the development of the child. The interplay between child and caretaker is thus central and the aim of early psycho sociological intervention is to strengthen, enrich and broaden the positive experiences in this interplay between the child and those who care for him.
Often the first step in turning a negative contact to a positive one, is to give the parent or caretaker a chance to feel competent and important to the child and its development here and now.
Those experiences of interplay which are important for the child to be able to create its ”representation of belonging” are often small everyday events, concrete and often repeated. These small events make up the core of the interaction. It can be difficult to evaluate the quality and intensity of these important little episodes and this can sometimes be a problem for us ”helpers”. We can be tempted to think of them symbolically and to verbalize about them even though these small episodes, as Daniel Stern puts it, symbolize nothing but simply are what they are.
From the very beginning the infant is capable of social interaction and the intimate interplay which normally occurs spontaneously between the caretaker and the child has shown that this early communication, composed of sounds, facial expressions and movements in a subtle interaction is a real dialogue. In this dialogue the signals between the adult and the child flow back and fore. When the child is reflected in this way it experiences that it is seen, heard and understood, the child in other words is confirmed -I exist, I am loved.
The child is constantly exposed to new experiences and phenomena which require an explanation and meaning. Parents need to help the child put a name to everything new and therefore the child directs its anxious questions again and again to the parents to dare to take everything in.
”We see you and we love you” is the refrain which gives the child the courage to receive and incorporate new things in its inner world and in its own language ¬new words, new understandings.
Already from the beginning the child is prepared to experience pleasure or pain, satisfaction, security and kinship or the opposite. When the child is given the opportunity to interact with caregivers, sensitive to its needs, it appears itself to be able to influence or guide its environment so that it triggers satisfying care and contact, for example through smiles, eye contact and movements which can be interpreted by the caregiver. The child’s own initiative and reactions can therefore be said to initiate and strengthen the sensitive and flexible adjustments shown by the parent, but on the other hand the absence of reactions from the child can also contribute to the development of unsatisfactory care.
The apparent absence of such reactions may have several different causes. The child may give weak signals and its reactions are therefore difficult for the parent to understand and respond to. The child may have special needs and require extra or a great deal of time and/or support, or may perhaps have a different tempo from its parents. For some reason or another a parent may need help to create positive feelings for the child. They are perhaps preoccupied with their own difficulties and troubles. Right from the start this may lead to a negative spiral with insecure relationships and mutually negative expectations between child and parents. This negative spiral can be difficult for the parents to break unaided.
A child´s development should be seen as a dynamic process of interaction capable of modifying the development and adjustment patterns of the child. This perspective makes early interplay an important point of entry for influencing the child’s development.
Interaction-based intervention takes place in the here and now -that which the parent and child can understand at that moment in time. It is in this relationship focused on the intersubjective meeting of parents and child, that the parents can discover and ”see” their child, its behaviors and needs, the signals it sends out, the initiative it takes and also its reactions. Intervention can also provide parents with concrete knowledge and information on what he or she can do to change and develop their parenting behavior so as to favor the development of their child.
The knowledge gained from interaction-based intervention and the possibilities thus given for changes in behavior which in turn alter the responses from the child can stimulate parents to rethink and change the inner representation they have of the child and of them-selves as parents. Those representations which have grown out of earlier interaction experiences, myths and legends in the parent’s family of origin act as a generalizing force/have significance in this respect.
Marte Meo is one of several methods for interaction-based intervention. It was developed by Maria Aarts in Holland and can be translated to ”on your own strong points”.
Marte Meo is used today around the world to help and support parents, foster parents, school and preschool teacher, staff of mother and childcare clinics, people caring for the mentally handicapped, in child and youth psychiatry etc. to develop and encourage communication and interaction with children and young people.
Its aim is to help and support parents and others in caring professions to see the needs and resources of the child and to understand that it is possible to react to the needs of the child in away designed to encourage the child’s development and improve the interplay between themselves and the child. The work is based on the problem definitions of the parent or caretaker.
Marte Meo therapists, irrespective of whether they work in the home environment, nursery or school, videofilm a short, everyday interaction sequence representative of the daily experiences between the child and caregiver.
The film material is then analyzed according to the Marte Meo principles for development enhancing communication. The main principles of supportive parental behaviors are
- Follow and affirm the initiative of the child
- Establish eye contact
- Share emotions
- Name with approval
- Take turns
- Lead Positively
A reviewing between parent and therapist is done in the absent of children and is focused on the pictures of the child.
The therapist highlights sequences of the film which can be further explored and shows how it is possible to improve and deepen interaction between child and adult so that the most development enhancing communication can be achieved. By stopping at particular points in the film, for example when the child is ex-pressing something verbally or non-verbally, taking the initiative or indicating a need, the parent may be able to recognize something which he or she was unable to do at the actual time. Perhaps the therapist can even show a golden moment when the parent succeeded in reacting correctly even if only for a second. For those with children who have difficulties in relating, or for parents who don’t see themselves as important in the child/parent relationship in can be very encouraging to see that the child is actively looking for the parent’s reaction or affirmation. Actually seeing the child’s face, its expression, its eyes, is what motivates the parent to be there for the child and resume supportive interaction. A close up of an infant can help the therapist to define the baby’s feelings and share this with the parents. It requires only a little positive feeling, just enough to make the parent feel good and supportive. In this way the parent can be helped to feel emotionally attached to the child and start responding to it.
The second reason for using a still picture of the child is to encourage the parent to start looking at the child in real life. This is the most natural thing that parents normally do with infants. They look at the child with both curiosity and love and in this way get to know the child and understand its needs. When a mother/parent concentrates too much on their own role in the relationship they tend to get to preoccupied with what they do as a parent. By showing revealing pictures of the infant the parent begins to understand that he/she needs to follow the child’s initiatives/need.
It is the responsibility of the therapist that the film are welcomed, shared and discussed. The principles for a development enhancing communication are taken from nature and are therefore experienced as so dear and simple that most of us take for granted that they are principles which we always use in communication with others.
It is only when these qualities are lacking or too weak that we realize how difficult it is to convey and give information about them. We tend to lose the most natural things when we are confronted with situations we are unable to control or recognize, when the difficult messages we receive are too different from those we expect.
Coparenting is something that has been focused on much to little. Already in the 90s John Gottman in Seattle had in his research about couples focused on spillover effects on the child. Some variables seem to be more important for the Child. As I had the oppurtunity to be a visiting researcher there he and his team has been very important to me in the way their work with couples. James McHale has focused on coparenting which is how you work together as parents. This is important part for the child not only that there is a cooperation for example in separations but also how you support each other as parents in daily life. This is focus is also taken in the systemic based work where allianes, different roles, cordination, joint focus and shared emotional engagement are worked on.
The research is linked to the Karolinska Institute The Department of Child and Youth psychiatric health. In the research Infant and child development has been combined with a systemic family system perspective. It is a longitudinal study that started 1996 and we are now collecting and analysed data from the children that have been in the study since they were newborn up to the the age of 15 years. Up to the age of 4 years it was a multicenterstudy where four differents sites took part.
In Sweden we now have a follow up when the ”children” are 20 years old. The families has taken part in a Lausanne Trilogue Play situation, a task in the family with siblings and questionnaires are filled out from parents , the youth her/himself and the teacher. The parents were interviewed during pregnancy and several times during the our research. In this text I will focus on instruments that I have developed or used. And the results presented is from the early years in the child´s development.
The study of child development has entered a fascinating phase where new knowledge on how actions, perception, cognition, and emotions develop is coming to light. Daniel Stern, in his book The Present Moment (2004), discusses the importance of the “micro moments of now” – how these moments might live for only a few seconds but still contain the individual’s experience of the past, the present moment in relation to others, and inner representations of the future, how the child thus develops, being influenced and formed in its interactions with others.
A systems approach in dyadic interaction and rhythm are two important aspects that are vital for infant-adult interaction/communication and behaviour and for adult conversation/communication and behaviour (Fogel 1992b): the infant influences the mother and the mother influences the infant (Beebe et al 2000). The systems approach enables the researcher to go beyond observations of the mother-infant dyad to understand child development and includes early communication/interaction in the mother-father-infant triad.
Daniel Stern ventures even farther in the preface of the second edition of the book The Interpersonal World of the Infant concerning child development. He asks, “Is the triad for an infant, a set of three interrelated dyads, or is it an entity in itself that can be represented”? (page 34). Parke (1988) postulates that studying the triad cannot be reduced to exploring various dyadic components, because the triad has a developmental trajectory of its own. In the dyad, coordination of communication is vital to finding a rhythm, the interaction flow, the “musicality of the dyad”, which also seems to exist in the triad. For an interaction to be considered well synchronised, it must have a certain flow, which allows space for all three participants to take part in a balanced way.
Elisabeth Fivaz-Depeursinge and her team in Lausanne (Fivaz-Depeursinge & Corboz-Warnery 1996) invited me to collaborate with them in a study on the mother-father-child triad using a systemic and developmental approach that matched the perspective I had from my clinical and professional work. They had developed Lausanne Triadic (later called Trilogue) Play (LTP), which captures not only various dyads but the triad as a whole with all its subsystems. It is a structured task in which the family plays and interacts in four phases: one parent and child, the other parent and child, all three, and finally the couple interact with each other. The non-active partner is simply present. The LTP session permits an assessment of alliance – defined as cooperation and coordination – and the method allowed me to make microanalytic observations of triadic interaction.
In a triadic interaction of this kind, sensitivity toward the child can be affected by the capacities of the individual partners and by marital relations between the parents as well. The combined interactions influence the family process in a sophisticated way. The need to further explore the effect such circumstances of co-parenting and family functioning have on triadic interaction and the child’s later socio-emotional development inspired me to learn from John Gottman and his group in Seattle and their focus on the marital relationship on a microanalytic level. In retrospect, this seems to be reasonable in light of current observations by Carneiro of a link between prenatal co-parenting and good postnatal family functioning at 3 months (Carneiro et al 2006).
Family policy and cultural issues
Fathers’ roles in Swedish families have changed dramatically over the past 40 years. National laws and the strong Swedish women’s liberation movement (which targets gender equality in higher education and the labour market) have helped to bring about these changes.
Today, the average Swedish mother is actively employed outside the home, and the average Swedish father takes a more active part in early childrearing than did fathers in previous generations. When the baby is born, the father is by law allowed 10 days leave from work and is paid 80% of his salary up to a certain income; highly paid parents are not fully reimbursed. In addition, the law grants each family 390 days at 80% reimbursement of the salary of the parent who chooses to stay home. Since 2002, each parent must take at least 60 of these days. This means that the father must take 60 days’ parental leave. If he does not, the family will forfeit them as the mother is not allowed to use “his” days.
Even though we refer to mothers and fathers in the reserach familysystems is not only consisting of mother and father but can be two mothers, two fathers, single mothers or larger systems.
Cross-cultural research can have important clinical implications, particularly since some theorists have argued that culture organises behaviour in fundamentally unique ways (Rogoff 1990). Lamb and his colleagues (1982a, 1982b) found that, regardless of whether the family was traditional (mother as primary care-taker) or non-traditional (father as primary care-taker for 1 month or more), Swedish fathers differed from traditional US fathers in that US fathers played more with their infants. Swedish fathers were not more likely to engage in physically stimulating play with their babies than mothers. This disagrees with the bulk of the literature on the unique role of fathers in play with infants (Parke & Brott 1999). A comparison of this research with Lamb’s research revealed that Swedish parents held and played with their babies less than US parents (Lamb et al 1983). Hwang (1986) found that when both parents were present in Swedish families, mothers were more likely to show affection for their infants and make bids for their attention than fathers.
Based on these findings, that US parents overall and US fathers in particular tend to play with their infants more than Swedish parents, it seems plausible that cultural differences in triadic interaction may exist.
Development – systemic circular thinking
The research is based on system theory where the following components constitute the theoretical framework. Based on these concepts, variables were defined and used to assess various aspects of interaction in the triads in qualitative and quantitative ways.
In the 1970s a tool new to the investigation of mother-infant interaction – microanalysis – became the method of choice (Brazelton, Kozlowski & Main 1974; Stern 1971; Trevarthen 1974). The theory of child development at that time effected a shift from the “one-way influence model” where the parent affects the child to a bi-directional model of interactive exchange in which the child and the parent influence each other (Bell 1968, Thomas & Martin 1976). This model led to the hypothesis of a balance between simultaneous interactive regulation and self-regulation (Beebe & McCrorie 1999, Tronick 1989). Incapacity or loss of flexibility on either side could result in pathology as described by Tronick (1989) in a study of infants to mothers with maternal depression. Tronick found that infants withdrew and became preoccupied with self-regulation when mutual regulation exchange failed to develop.
From birth, babies have the capacity to respond to stimuli from the outside world, such as voices, different sounds, facial expressions, hugs and touching, and emotional reactions from the parents. Early imitation has been proposed to be an early form of intersubjectivity (Meltzoff & Moore 1977, 1999), and many researchers agree that infants are born with minds that are especially attuned to other minds as manifested through their behaviour (e.g. Stern 1971, Tronick 1989). The proto conversations between young infants and their parents and the tight mutual coordination of the timing of movements and facial expressions in infant-mother interaction enable the parents to mirror the infants’ affect, and by this, the “sense of shared experience”, which is part of “primary intersubjectivity” (Rochat & Striano 1999, Trevarthen & Hubley 1978, Trevarthen 2000).
When the child is around 9 months old, it begins to use the child-parent dyad – or a dyad with another adult – to focus on something outside and share this focus. This age has been called the 9-month revolutionary age (Tomasello 1999), when the capacity of mentalisation, “Theory of Mind”, also starts. When the child begins to develop its understanding, others act as “intentional agents” – which means they have goals and intentions –the child can acknowledge and understand this in others. The child use gestures to indicate to the other member of the dyad that he or she should “tune in”, with a focus on the outside event. The term “joint attention” is usually used to describe this capacity. But this ability to use the dyad as the base for exploring “the third” also happens in the triad with another person as the third.
Intersubjectivity continues to develop into other forms as the child grows older, and Daniel Stern has termed it the intersubjective matrix; a crucible in which interacting minds take on their current forms. Two minds create intersubjectivity as, equally, intersubjectivity shapes the two minds. Intrapsychic has developed into intrasubjective (Stern 2004).
The discovery of mirror capacity, probably due to “mirror neurons” in the brain, has prepared the ground for new hypotheses. Imitation is part of this function, and seeing somebody smiling makes us smile more. It seems as if this happens only when we observe someone else doing something, and it has been proposed that our brain experiences the action as if we were doing it ourselves (Damasio 1999, Gallese 2001).
If we live in such a complicated intersubjective world, Daniel Stern asks, how do we keep from being totally and continuously influenced by other’s experiences and feelings? He outlines three possible answers: by being selective, by making sure that the activation of mirror neurons does not constantly spill over to the motor neurons so that reflexive imitation is an on-going behaviour, and by dosing resonance with others (Stern 2004).
Rhythm in interaction
According to Trevarthen (2000), by the age of 2 months, infants should be able to adjust their interactions with their parents and other people and have true mutual communication with others. Studies of interaction between infants and their mothers indicate that the dyad can be regarded as a unit with rhythmic temporal synchrony. Early communication between parents and their babies seems to involve pausing for a response on the part of the parents and the baby. These pauses before responding and the rhythm of interaction when talking and answering may be important in early reciprocal interactions.
The coordination of timing between parent and infant starts to develop long before there are words. There is a “melody” to build on when the verbal develops. Stern (1985) speaks of this as learning and forming non-verbal ways of “being with”. It is a process by which we anticipate the partner’s pattern in relation to our own.
This early interactional rhythm obviously defined various ways of “being with”, and the patterns of vocal rhythm organise the infant’s experience of relatedness and development. Patterns of timing coordination could be looked upon as a “life-span” measure. The patterns form one base for the infant’s representations of self-in-relation-to-other (Beebe, Jaffe & Lachmann 1992; Stern 1985).
Regulatory disorders usually have underlying deficits in self-regulation, and sensory processing is coupled with inattention, poor emotional and behavioural control, and feeding problems. Infants with regulatory problems can develop normally in infancy, but symptoms like emotional and behavioural control evolve over time (DeGangi et al 2000).
The infant has inborn brain functions for processing information and finding patterns and order (Schore 1994). This capacity helps the infant not only to engage in activity that stimulates itself but also to build expectancies and act on those expectancies (Haith, Hazan & Goodman 1988).
In summary. The person’s behaviour – memorised behaviour – and the partner’s behaviour are supposed to form an interaction. In the study of communication, communication cannot be viewed as a simple process but as “moment-to-moment units” that are nested within larger units (Beebe 2000). Self- and interactive regulation are concurrent and reciprocal processes, each one affecting the other (Gianino & Tronick 1988). When infants loose eye contact with the parents, the parents use tactics like calling the baby, gestures, or facial expressions to bring the baby back into the parents-infant interaction. It may be that some level of probing for the infant’s attention by the parent may be important for repair. Positive emotional communication involves sensitivity to the infant’s cues and to the need for breaks and for gentle repair to bring them back into interaction (Tronick & Cohn 1989). The infant supposedly desires coordinated interaction and also seems to have strategies for repairing an interaction after an episode of mis-coordination.
Minuchin (1974, 1985) described the importance of the family system and the relevance of co-parenting for child development and socialisation. The systemic perspective basically says that the whole is more than the sum of its parts (Öquist 1992). A newborn family is an entire system and these parts: child-mother, child-father, child-mother-father, and the couple dyad. These systems regulate and mutually influence each other. Each system is sensitive to changes and developments in the other systems.
Minuchin describes the family’s combined interaction in all systems as forming a process on a higher level, because what is between the partners will be an important part of how the entire system forms, develops, and functions. We could perhaps add a fifth part to the newborn family system: the parent “mother and father”. It could be hypothesised that the mother-father dyad is different than the man-wife dyad; the parent dyad begins to form with the pregnancy and birth of the first child. Mother- and fatherhood constellations begin to develop, and parenthood identities form.
Each individual and each subsystem plays a significant role in how triadic interaction develops. It is not enough to observe and describe these various processes to be able to understand the process of development that the child undergoes. Fivaz-Depeursinge (1996) suggests that the relations that are established in a triad have their own features and that development of the different relations must be studied in the context of the entire child-mother-father system. Triadic play interactions evoke interactive skills that differ from dyadic parent-child interactions (McHale & Fivaz-Depeursinge 1999). Until the interactions of the entire family unit are assessed, we cannot know how the various subsystems – co-parental, marital, parent-infant, and individual – function.
Transition to parenthood
The transition from marriage to parenthood is considered to be one of the most difficult transitions in life and is often referred to as a “crisis”. In many couples, the wife experiences the first decline in marital satisfaction after the arrival of the first baby (Cowan & Cowan 1992). But the lowest point in marital satisfaction was found to be a year after the birth of the first child (Shapiro, Gottman & Carrere 2000). It is important to mention that remaining childless does not ensure marital stability. The divorce rate for childless couples was reported to be 50% and for couples who became parents only 25% (Cowan & Cowan 1992).
When a couple is expecting their first-born, the primary task of forming a dyad as a couple is now challenged and the relationship must make room for two subsystems: the marital and the parental. The time of pregnancy becomes not only the period in which the baby develops and prepares for birth but also a period where the parents prepare to be mothers, fathers, and parents. Studies reveal how parents – usually the mother – build an inner picture of the baby that is most clear around the 6th month of pregnancy and then becomes more vague with time, perhaps as a protective function to be able to welcome the real baby.
Maternal psychopathology has been associated with difficulties in parenting, but the impact of paternal psychopathology has received less attention. Studies of the father found that paternal psychopathology also affected child and family development during pregnancy (Perren et al 2003, 2005). During preparation for parenting, the parents were interviewed about the imaginary baby and family; parent psychopathology was found to influence both factors. Even though mothers had higher stress and depressive symptoms during pregnancy and early parenthood than fathers, fathers also reported more symptoms than before pregnancy. This emphasises that we must acknowledge not only that the newborn child is influenced by the mother-child or father-child dyad but that the child is a member of a triad consisting of the mother, the father, and the child.
The inner world of the expectations of the mother and father of how life as a family will be – the partner as a parent – are important questions that research has begun to focus on. Studies of marital satisfaction have mainly focused on women’s unhappiness with men’s lack of involvement in the household (McHale 1997). Men’s and women’s worries about difficulties in their co-parental relationship were associated with current marital quality. Marital dissatisfaction could arise in the early postpartum months and years when perceptions of an unequal division of labour were expressed prenatally (McHale et al 2004). In marriages with higher marital quality it was possible to voice concerns.
A few studies have focused on fathers’ representations and formed the hypothesis that the fathers’ inner representation of being three is one of the most important variables for whether the newborn family will be a harmonious threesome (von Klitzing et al 1999). McHale et al (2003) demonstrated significant associations between parents’ prenatal expectations about the future family process and observed co-parental functioning in trilogue interactions after birth. In interviews where mothers and fathers were more pessimistic about their future family process, there was less family warmth, less cooperation, and more disagreement than shown by the more optimistic parents in LTP several months after birth.
Another study of marital interaction prenatally found that the couple’s interaction in the first few months of their marriage predicted how much their first-born infant cried during parent-infant play, and the couple’s relationship in the newlywed and pregnancy period predicted later quality of father involvement (Shapiro 2005). Early infancy is especially interesting to study, because co-parenting and family patterns seem to be maintained by both parents and are stable over time (Fivaz-Depeursinge & Corboz-Warnery 1999).
Studies that focus on couples and begin before pregnancy are rare but important for our understanding of what in the couple relation buffers them during the transition to parenthood. What predicted stable or increasing marital satisfaction were whether spouses expressed fondness and admiration for their partners, how aware the spouses were of their partner’s world, and what degree of unity each spouse expressed (Shapiro et al 2000).
Marital relationship and co-parenting
As mentioned above, the transition to parenthood is a challenge for the couple relationship. Studies such as Cowan & Cowan’s (1992) found, in general, a marked decrease in positive exchanges between spouses and an increase in conflict after the child’s birth. Those couples with high marital satisfaction before birth still had high satisfaction after birth. Especially important protective factors were the expression of affection and complicity between the partners (Shapiro, Gottman & Carrere 2000).
Co-parenting processes include how parents interact with each other when the child is present – how they work together and coordinate their roles as parents. But strong co-parental alliances are those in which the co-parents provide support and solidarity for one another’s parenting efforts with the children (McHale 2004). Other studies propose that the mother’s sanctioning of the father’s participation in family life and the father’s proactive family engagement are particularly important for co-parenting (Allen & Hawkins 1999).
Since most cultures place responsibility for the child on the mother, she can also decide what access the man will have to the children (Lamb & Oppenheim 1989). The mother’s role of gatekeeper to their partners’ involvement with the children could be related to the withdrawal of the father. Talbot and McHale (2004) discussed this role of the mother in their study of the individual features of mother and father. Several studies have found links between marital quality and co-parenting behaviour as well as the particular child, depending on child age, gender, and birth order (Lewis 1988, McHale 1995, McConell & Kerig 2003).
But even if the marital relationship affects the quality of parenting, not all couples who have a stressed relation act out their conflict in their co-parenting. And some couples with co-parenting problems are still happy with their marriage. In microanalysis using the LTP situation it has been observed how parents can be good at co-parenting, even when marital communication is poor (Fivaz-Depeursinge & Corboz-Warnery, 1999).
The couple relation and the parenting relation are really two different systems that, although there is partial overlap, are independent of each other. Observations of family interactions with both parents allow an evaluation of whether parents support or compete with each other’s initiatives with the child. The Child-Parents-Interaction Coding System (CPICS) is a tool with which such initiatives and turn-taking sequences can be analysed (Hedenbro & Lidén 2002).
Research in co-parenting began in divorced families (Hetherington, Cox & Cox 1979). Later, studies were made in intact families, and although patterns in the families vary, there are some basic elements, of which one is the involvement of interaction. One general supposition is that happily married couples are capable of greater warmth and sensitivity when interacting as a family with their children. McHale and Cowan (1996) have studied how marital conflict is transferred to co-parenting, rendering the system more hostile and competitive. They emphasise that the context in which family patterns are assessed is important to take into consideration. One parent’s adjustment to the other parent in triadic interaction plays an important role in how each parent acts. It is not at all certain that interaction in the mother-child dyad will be the same as when the father is also present.
McHale & Rasmussen (1998) have suggested that asking parents about their co-parenting would be a valuable tool that would complement investigations of the marital relationship. For example, women’s reports of interparental conflict were found to be associated with high levels of observed antagonism and low warmth and cooperation during family group interaction (McHale, et al 2000). These significant associations remained even after controlling for marital quality.
The couple relationship is especially strongly related to involvement, satisfaction, and quality of fathering (Parke 1999). Men in unhappy marriages who withdraw from the marriage also withdraw from the children (Gottman 1994). More recent studies have shown that husbands’ communication with their wives during pregnancy in the conflict situation at 6 months of pregnancy not only predicted the father’s involvement with their 3-month-old babies but was also positively correlated with the infants’ smiling and crying (Shapiro 2005). There is also some evidence that infants to fathers in unhappy marital relationships did not refer socially to their fathers in novel situations (Dickstein & Parke 1998).
The association between marital difficulties and problematic parent-child relationships is much weaker for mothers than for fathers (Margolin, Gordis & John 2001). It could be hypothesised that the differentiation between marital and parental subsystems is stronger for mothers. Mothers seem to engage in parenting regardless of their marital satisfaction. Mothers tend to get closer to their children when the marriage is unsatisfactory while fathers withdraw.
In summary. In the threesome, the infant is “embedded” in the parents’ relationship. The “space” given to the infant and the atmosphere that the infant is becoming part of and affecting are phenomena that require investigation. Marriage affects children, and parenting seems to be the key factor in this (Gottman & Katz 1989, Katz & Gottman 1997). A harmonious, supportive marriage makes parents more sensitive, responsive, and involved. The quality of interaction in the parents’ relationship has a “spill over effect” on the child (Katz, Gottman 1996). Spill over effects concern the transfer of moods, emotions, or behaviour in good or bad ways from one subsystem, for example, the husband-wife, to another subsystem, for example, the mother-child. In this way, spill over effects affect child development.
Family triangular interactions and child functioning
Triadic interactions have mainly been described in situations when the infant focuses on an object; signals interest, pleasure, or frustration; turns to the parent who is following the action; and shares a joint focus with the parent. Toward the end of the first year, these interactions have become part of “secondary intersubjectivity” because the child shows the capacity to share attention and intention with others (Tomasello 1995).
But from the moment of birth, the infant is exposed and embedded in larger contexts and interactions than strictly dyadic interaction, so triangular interactions become important for development (Dunn 1991, Schaffer 1984, Fivaz-Depeursinge et al 2004). Longitudinal observations have found differences between infants growing up with parents who support each other in their parental function versus parents in conflict. The infants from good co-parenting relationships engaged more easily in triangular interaction and received more sensitive and adjusted responses than infants who grew up with conflictual parenting. (Fivaz-Depuersinge & Corboz-Warnery 1999, McHale & Cowan 1996).
In the development of family therapy, triangulation was considered to be a situation in which one person was positioned to ease tension between two other people (Bowen 1972). This has been considered a risk situation for the child’s emotional development because this could lead to a situation where the child acts as a go-between or becomes the centre of its parents’ focus (Minuchin 1974). In a competitive relationship, parents could compete for the infant’s attention and not support each other’ contributions, thereby being intrusive and overstimulating (McHale & Rasmussen 1998).
Studies indicate how the influence of early family factors in the first year of life, such as warmth – defined as affection and tenderness between the parents and toward the child (McHale & Cowan 1996, McHale & Fivaz-Depeursinge 1999) – correlate with children’s socio-emotional development and later with peer interactions (Behar 1977; McHale & Cowan 1996; McHale, Krasnow & Slavick, 1997, McHale & Fivaz-Depeursinge 1999).
Methods for studying triadic interactions between children and their parents
When I was planning the study, I visited the Lausanne group in Switzerland and the Seattle group in the US and I joined the “Trilogue group” that they had started for research on triadic interaction. Both the Seattle and the Lausanne groups had developed methods for describing triadic interactions; a third method was used in the Swiss Basel group. Despite the merits of these methods, they lacked variables for making microanalytic analyses of various dyadic subsystems, that is, events such as initiatives and turn-taking in triadic interactions. In my clinical experience it was important to be able to measure events that were immeasurable with these (or any other) methods; a system that could describe the affective level and different partners’ participation in dialogues and trilogues in a microanalytical way – “moment by moment” – was needed.
Study design and data collection
The design of the study was longitudinal and prospective. Data is collected over 20 years to cover various aspects of the parents’ situation and their children’s/young adults development. In this study, the focus is on triadic interaction and the development of communication. Data describing interactions until the children reached 48 months of age were used for the thesis but data and analyses up to 16 years of age has been published .
Pregnancy week 20–25
The couples were interviewed in the studio in a situation structured just enough to make comparisons possible; otherwise, as little structure as possible was used so that it would be easier to access the parental inner world. The interviews were conducted by well-trained clinicians and videotaped. Interview topics were their own childhood experiences; their history as a couple; pregnancy and emotions related to this; and inner representations of themselves as a mother or father, of the child, of the future family, of their partner as a parent, and of relationships with the future grandparents. As a complement to the interview, each partner drew a symbol on a paper representing the family when the child is born. The parents were given the information about the task and, within a few seconds, they drew the symbol. This procedure was designed to try to catch non-verbal inner representations. At this meeting each partner independently filled out the ENRICH (Wadsby 1998), which covers ten areas: Personality, Communication, Conflict Solving, Economy, Free Time, Closeness and Sexuality, Children and Parenting, Family and Friends, Role of Man and Woman in the Household, Beliefs. The couple also filled out the Symptom Checklist-90 (SCL-90) (Derogatis et al 1973).
Infant aged 4 weeks
The family visited the studio for the second interview. This interview focused on the pregnancy period since the last meeting, the delivery, and the first few weeks as a family. It was a semi-structured interview. The families were happy to come and eager to talk of their experiences. The interview last 1.5–2 hours.
Child aged 2, 3, 8, 9, 18, and 48 months
The family visited the studio and the LTP Situation was used. This is a videotaped, semi-structured situation that facilitates the examination of the triad as a whole and an organisation of its parts (Fivaz-Depeursinge, Frascarolo & Corboz-Warnery 1996). Parents were instructed on how to use the baby seat and encouraged to adjust it for their baby’s comfort. Parents were asked to schedule their studio visit for a time that was good for their baby, when they were likely to be alert and recently fed. Toys were not present until 18 and 48 months. At 18 and 48 months a toddler chair replaced the baby seat.
Child aged 3 and 9 months
One week after the studio visit the families were videofilmed in their homes. The three members were instructed to play together for 10 minutes; no further instructions were given. The purpose of this was to observe whether the patterns in the studio were similar to those at home. It became obvious that the patterns were similar but that the family shifted faster and more often between the four LTP phases during play at home. During videofilming in the families’ homes at 9 months, each parent was also filmed in two dyadic situations with the infant and without the other parent present: one play situation and one where the parent changed the child’s diaper. At 9 months, the parents were also asked to fill out the ENRICH questionnaire again.
Child aged 18 months
The parents took part in a “meta-emotion” interview (Gottman 1996). This is an interview that is conducted separately with the man and woman, and its purpose is to understand how they think and feel about emotions. It is an extended interview and covers questions about childhood, the child’s current emotional life, as this parent understands it. The interview focused on past and present emotions of love, pride, anger, sadness, and joy and what the parents think and feel about their child’s emotions. The parents also fill out the Toddler Behaviour Questionnaire (TBQ) that measures temperament factors (Hagekull 1985).
In addition, instead of filling out the ENRICH questionnaire again – since this instrument was too extensive for the families to fill out with the child present – the family filled out a questionnaire concerning “family climate”. This new instrument captures the most important issues in family life (Hansson 1989).
Child aged 20 months
Three children and their parents came at the same time to the studio. The parents were invited to sit in the corner of the room and have a cup of coffee; in another corner, a set of toys were placed so that the children could play. The session took 1 hour. The purpose was to observe and understand how the children created triadic play; how they solved conflicts; and how they used their parents in social referencing, going to them or exploring and playing themselves. The families were grouped in the order they had been admitted to the study, so the groups could contain both boys and girls. Due to sickness, three groups had only two children.
Child aged 48 months
The children were again invited to participate in an LTP situation similar to the one described earlier. I chose to see the children again at age 48 months, which was the age that the Lausanne team used. Much time had passed: four families had divorced and one family had moved from the area. The other 15 families came to the studio. The mother, father, and preschool teacher filled out the Preschool Behaviour Questionnaire (PBQ) (Hagekull & Bohlin 1996). The PBQ covers eight areas: Ego Strength; Internalizing or Externalizing Problems; Concentration Problems; Aggressiveness; and Social, Peer and Adult Competence.
Child aged 5 years
The children were filmed in their preschool together with at least two friends. The purpose of this was to see whether any patterns of interaction with peers could be linked back to early interactions in the family.
Child aged 7 years
A home visit was made to videofilm interaction in the entire family at a dinner table. One family had two more children, while one family still had only the child that had taken part in the study. All other families had one more child. Two of the families who were divorced also participated, but only one of the parents was with the child. The parents were also asked to fill out the Child Behaviour Checklist (CBCL) questionnaire (Achenbach & Edelbrock 1983).
Selection of variables for the longitudinal and prospective analysis
When the children were 3 months and the first LTP sessions were analysed, the data were published in a research report (Hedenbro 2006) and used to evaluate which data were important and suitable for statistical analysis of the longitudinal data set. Four quantitative and two qualitative variables (defined in the CPICS manual) were extracted:
Turns and turn-takings
These are the more important results from the thesis.
Number of contributions in triadic interactions
In most families (17/20), the children made more contributions than the parents. Children often made several different contributions at the same time, such as a neutral facial expression combined with vocalisation.
The rate of contribution was higher for the fathers than the mothers in 11/20 families. In 2 of these 11 families, the mothers took no initiatives in triadic interaction. In 7 families the mothers took the initiative more frequently than the fathers.
Frequencies and kinds of contributions made by children
The children made 11 kinds of initiatives/contributions: positive facial expression; negative facial expression; neutral facial expression; seeking eye contact; physical movement; directing attention toward an object; directing attention toward one of the parents; emitting positive vocalisation; emitting negative vocalisation; neutral vocalisation; and coughing, sighing, and hiccupping. All initiatives/contributions were categorised as positive, negative, or physical.
More than one kind of contribution was usually used at the same time. This was especially true for neutral facial expression, which was the most frequent contribution. This contribution was followed most frequently by physical movement, whereas seeking eye contact and positive vocalisation were rarely used. Number 11 was added to the list when it was obvious that coughing, sighing, and hiccupping were very common at this early age and that the parents responded to these actions so that it became a starting point for communication.
Turns and turn-takings (quantitative)
Percentage of contributions in each triad that leads to turn-taking sequences
Following the child’s contribution, the mother had a turn-taking sequence in 16/20 families and the father in 14/20 families. When the mother initiated communication, it was followed by turn-taking sequences in 10 families as opposed to in 6 families when the fathers took the initiative. In 1 family, no turn-taking sequence at all were observed. In another 2 families, only in one of the four possible subsystems was the contribution followed by turn-taking, and these turns were in response to the child’s contribution. In 1 of these families, the mother was involved in the turn-taking and in the other, the father was involved.
In 2 families, contributions led to turn-taking sequences in all four possible subsystems. When the child made the initial contribution, sometimes the mother responded and sometimes the father responded, and when either the father or the mother made the initial contribution, the child responded.
Mean per cent of initial contributions followed by turn-taking sequences in the triad
Turn-taking sequences took place following 31% of the child’s contributions, 28% of the mother’s initiatives, and 21% of the father’s initiatives.
Mean per cent of turns within turn-taking sequences in the triad
If the child started the turn-taking sequence with a contribution, the mean number of turns that followed in the same sequence was 3.55. If the mother started the sequence, an average of 3.51 turns followed, and if the father started the sequence, an average of 3.97 turns followed.
Mean values of verbal and non-verbal affirmation
The parents used verbal and non-verbal methods to affirm and support contributions from the children. The means of the parents’ use of verbal or non-verbal affirmations were assessed. Mothers and fathers used non-verbal affirmation more often than verbal affirmation. Verbal affirmations seemed to be used to reinforce non-verbal affirmations. Parents often affirmed children’s contributions more than once. Sometimes both parents affirmed the same contribution.
In nine families the mothers tended to use non-verbal affirmations more often than the fathers; in seven families the opposite trend was found. A similar situation was found in relation to the use of verbal affirmations.
Use of clarifications related to number of parental initiatives
Parents who started a turn-taking sequence seemed to have various methods of encouraging children to join in and take a turn. The parents used clarification as one of the tools to achieve this. Clarification is the repetition of a verbal or non-verbal initiative or contribution. The clarifier may repeat the original initiative or contribution exactly or in a slightly varied form; for example, words may be stronger or clearer in tone or gestures larger or more emphatic. Clarification often involves more than one repetition. In most families, the mother and the father used clarification. But in three families the fathers did not use clarification, and in two families the mothers did not use clarification.
Fathers’ mean number of initiatives was 6.10; their mean number of clarifications was 6.45. For the mothers the corresponding means were 5.63 and 6.16, respectively.
Synchronisation was used as per Kantor & Lehr (1975) who defined synchronisation as the process of regulating and directing the way in which time is being used. It involves five sub mechanisms: control, prioritisation, a structure of guidelines, coordination, and reminding. Does a dance of interaction arise when the parties are allowed space to express themselves? Do the turns of the parties interlace or do they appear to be one-way dialogues? Both child and parents contribute to the synchronization of the dyad and the triad. Are all dyads in the triad activated? Synchronisation is qualitatively assessed and coded on a scale with four categories (1–4): Five families were classified as “1”, nine as “2”, four as “3”, and two as “4” (not being at all synchronised).
Co-parenting demands a high degree of sensitivity. The partners must have a sense of respect for each other’s contributions and a sensitivity that supports each other’s interaction or contribution. Inclusion occurs when each parent gives space to the other parent’s interaction/activity with the child. Exclusion occurs when one parent physically excludes the other parent from interaction/activity with the child. Inclusion: Mothers in 17 families included the fathers, while fathers in 18 families included the mothers.
We observed that the tempo of interaction can vary for each partner; however, when sensitivity to each person’s contribution exists, it helps the family make the early adjustments necessary to find a common tempo. A common way participants use to help each other in triadic interaction is to show an interest in the contributions of the others so that this will result in a turn-taking sequence.
Two important steps in turn-taking sequences were observed with great regularity in most triads. Step one typically occurs when the baby either makes a contribution or reacts to a parental contribution. Step two is often verbal or non-verbal affirmation by one or both parents.
Although non-verbal affirmation is much more common, verbal affirmation is also often used. When a parent initiates the sequence, clarification often follows the contribution. This helps get the infant’s attention and encourages the infant to join in the interaction. It is also similar to the normal behaviour of a child who repeats its contribution to get a parent’s attention.
It seems that as early as age 3 months, babies have developed an internal sense of the threesome or triad besides a sense of the twosome or dyad. The baby knows that the three members of the triad are interacting together and is already able to focus attention on both parents in an LTP setting. For example, when one parent makes an initiative or contribution, the baby may look at that parent and then almost immediately glance at the other parent as well.
It is essential to triadic interaction that parents include each other via non-verbal and verbal signs. Out of 20 families we found only 3 mothers that excluded the fathers and 2 fathers that excluded the mothers. To further investigate how inclusion or exclusion of a partner can affect triadic interaction, we looked at turns that followed contributions made by the child, mother, or father. It is interesting to note that in two of the families in which exclusion was practiced, both parents excluded each other; in one of these families, very few turn-taking sequences occurred while in the other family no turns were taken. In the third family, it was the mother who excluded the father and the outcome was that the father’s contributions were not followed by turn-taking sequences, whereas the mother’s contributions were followed by turn-taking sequences. Following the child’s contribution, there was no turn-taking at all.
Sensitivity to the partner’s contribution and on-going interaction may be essential to the flow of complex triadic interaction; triadic interaction tends to occur less often when one partner excludes the other or when both partners exclude each other. More turns followed the child’s contributions in families with well-synchronised interaction than in families without well-synchronised interaction. It can also be postulated that the parents’ sensitivity to each other’s contributions, their tempo, and their cooperation may influence their interaction with their newborn children. Good cooperation between parents who have a positive relationship might be more likely to support a baby’s participation and training in turn-taking sequences.
The observations indicated that 3-month-old infants, supported by their parents, are capable of taking an active part in triadic interaction. Parents use non-verbal, vocal, and verbal forms of communication to set the structure for early communication and turn-taking sequences that will later develop into dialogues and trilogues. The observed trilogue patterns seem to be similar to those seen in dyadic communication.
In supportive triadic communication it seems to be important for the parents to balance their initiatives to make space for the child’s participation. Babies are individuals who have differing levels of ability to adapt to the flow of interaction in triads and to literally and figuratively make their voices heard. These differing levels of ability seem to influence the interaction of parents with the child and with each other. This is in line with the suggestions from Parke (1988) and Fivaz-Depeursinge et al (1996, 1999).
Because of the complexity of these data, the analyses were run in cooperation with Statisticon AB, who are skilled in statistics and mathematics.
Qualitative and quantitative data were analysed in cooperation with Statisticon AB. The analyses were performed in the statistical programs SAS (SAS Institute, Inc.) and R (http://www.r-project.org/).
The statistical analyses were mainly descriptive. Means and standard deviations were calculated. Inter-rater agreement was measured with Pearson’s correlation coefficient; some tests were also checked with the Intra Class correlation (ICC) coefficient. Friedman’s test was used to evaluate changes over time within a group (children, mothers, or fathers). The Mann-Whitney two-sample test was used to evaluate differences between groups. Spearman’s rank correlation was used to analyse associations between continuous variables. A p-value < 0.05 was considered significant.
Developing a method and a manual for analysing triadic interaction
Paper I in my thesis describes the CPICS, the method by which LTP sessions can be evaluated in a quantitative and qualitative way where each partner’s contribution is in focus on a micro level.
The CPICS was developed to assess the results of videotaped LTP sessions in a systematic way. As already described, the instrument and its categories are based upon existing theories and hypotheses for describing communication between the newborn child and its parents. The CPICS can be used with quantitative or qualitative parameters, for clinical purposes, and for research. It was developed for cross-sectional comparisons and for monitoring development of communication in the triad (by repeated assessments of the same triad) in projects using prospective longitudinal designs.
This procedure is used: Each LTP part is videotaped. The video-counter is used to register the start and stop of events and the length of time for single events and sequences. Assessments are made from the tapes, and 10–20 minutes of interaction are analysed. The observers should be trained using the manual and an instruction videotape.
The reliability of the assessments of various observers was tested in 80 families: 20 Swedish families, 20 US families, and 20 German-speaking and 20 French-speaking Swiss families. The children participating in the assessments were on average 3 months of age.
Two well-trained Swedish observers (none the author of this thesis) independently evaluated the videotapes. Observer agreement was measured using Pearson correlation coefficients. Significant observer agreements were found in several comparisons, while low correlation coefficients were found in others.
When the CPICS was used to assess 3-month-old children, the hypothesis was that communication and assessment categories would become more easy to describe as the child and its abilities developed. In the longitudinal prospective follow-up at age 9 months, 18 months, and 48 months, new calculations of inter-rater agreement (Pearson correlations between two coders) were made. At each age, the CPICS protocols were independently coded by two persons, none the author of this thesis. With two exceptions, reliability was satisfactory (Table 2).
The CPICS system was found to be an instrument with good observer agreement and easier to use when the children became older. Until more studies with the system have been published, variables with observer agreement > 0.69 are suggested as acceptable for scientific studies of groups. In clinical use, all categories should be used with caution until reassessments have verified their reliability and validity.
Explaining how dyadic and triadic interaction – communication and conversation – develop in a non-clinical sample over time between parents and children aged 3–48 months (paper III)
The development of dyadic and triadic interaction was analysed over time when the children were 3, 9, 18 and 48 months old using the variables contributions, turn-taking sequences, turns, and affirmations. One of the children, a girl who was later diagnosed with an autism spectrum disorder, was found to be an “outlier” in the statistical analyses and was excluded from analyses. The group of families in the study was reduced to 19.
While differences in frequency of contributions between the four assessed ages of the children were non-significant for the parents, differences for the children were significant (p = 0.000). The frequency of contributions was significantly higher at 9 than at 3, 18, and 48 months (p = 0.003, 0.000, and 0.000, respectively). The frequency of contributions was also significantly higher (p = 0.001) at 3 months than at 48 months. Children’s contributions were significantly more frequent than parents’ at each age assessed (p for the differences at all ages = 0.000), while differences between the parents were non-significant (Figure 1).
Types of contributions made by children
Parents reacted to 11 types of child contributions with affirmation.
More than one kind of contribution was commonly used at the same time. This was especially true of neutral facial expressions, the most frequent contribution, which was often combined with another, more active contribution, such as physical movement or vocalisation.
At 3 months, a neutral facial expression was followed in frequency by physical movement, whereas seeking eye contact and positive vocalisation were rarely used. When the infant was 9 months, the most common contribution was attention directed toward an object, followed by neutral facial expression. Physical movements were also frequent, as was attention directed toward a parent. Positive and negative vocalisations were now more frequent than neutral vocalisations.
At 18 months, toys were part of play and interaction, and it was not surprising that contributions that include attention directed toward an object were most common. Seeking eye contact decreased to 0, which might be because of the toys, and the mean incidents of coughing, sighing, and hiccupping decreased to 0.1. Positive vocalisation and neutral vocalisation increased while negative vocalisation decreased in frequency.
At 48 months of age, positive vocalisation continued to increase as did neutral vocalisation. Coughing, sighing, and hiccupping – definitely non-intentional behaviours – were almost non-existent at 18 months. Clear communicative contributions – verbal and non-verbal – were now more common (Table 3).
Turn-taking sequences per contribution
The overall difference in frequency of turn-taking sequences over time for the child (p = 0.006) was significant and for parents non-significant. Turn-taking sequences were more frequent at 18 and 48 months.
At 9 months, the number of turn-taking sequences for the child was significantly lower than for the parents (p = 0.017). No other significant differences were found. To determine whether variation in length of recordings of the LTP situation (LTP permits parents to decide the length of time that they will interact) had biased the results, the means for contributions per minute and turn-taking sequences per affirmation were calculated for each age after cases under and over the median length of time at the respective ages were selected. For children, there were small, inconsistent differences between shorter and longer recordings for both variables. For parents’ contributions, there were slightly lower frequencies during longer recordings, and for turn-taking sequences, there were slightly higher frequencies for longer recordings.
For results in the later studies when the children has been growing I refer to my later published articles.