The ‘child’s best interests’ as an argumentative resource in family mediation sessions

We used Discursive Psychology to study the claims and arguments which occur when ‘the child’s best interests’ is produced as a resource in family mediation settings. Analysis draws on data from three pairs of separated or separating parents attempting to resolve child contact or residency disputes through mediation. Our analysis focuses on the tendency of claims to the abstract notion of the child’s best interests to exacerbate conflict, especially as parents drew on conflicting research in this area. Changing expectations of fathering could be observed in the men’s argumentative positioning, and this was taken up in different ways by ex-partners and by mediators. Participants aligned themselves with mediators’ statements by picking up details of mediators’ language, hampering mediators’ attempted neutrality. The problematic nature of acknowledging the intensity of emotions in this process was also highlighted.


Introduction
We take a Discursive Psychology approach to consider how children's well-being and interests concerning post-separation parenting are presented and responded to in family mediation sessions. In particular, we investigate how the concept of 'the child's best interests' is produced as a resource in mediation settings (by parents and mediators). At which points in the discussion is this invoked, and with what effect? We analyse extracts from family mediation sessions, selected from three pairs of separated or separating parents attempting to resolve child contact or residency disputes through mediation. How are children's needs and interests occasioned in mediation talk? When and how do parents refer to their child's interests, and/or to the more abstract concept of 'a child's best interests', and how is this responded to? We are interested in how the concept of the child's best interests becomes a resource for parental positions and arguments about contact, residency and parenting agreements after separation. Does the appeal to considering a child's best interests progress the mediation talk in a way which enables conflict resolution? Is the concept of a child's best interests a useful one for mediators to focus on?
One of the key stated aims of mediation, and of the law concerning post-separation arrangements for children (Children Act 1989 s 1) is a focus on a child's welfare or best interests. Studies of therapeutic processes have found that children's interests are routinely sidelined or represented (Hutchby and O'Reilly, 2010), and this has been found in mediation too. Dingwall and Greatbatch (1991), reporting on a study of UK mediation in the 1980s, argued that dispute resolution was a parent-centred process with parents and conciliators making few references to child welfare. However, a generation on, the findings from the 1980s study may not reflect current mediation practices, views of child well-being or parenting norms. Trinder et al. (2010) looked at in-court conciliation, using conversation analysis (CA) to study how the child's needs and wishes were discussed by parents and conciliators (sessions recorded in 2004-2005), concluding that 'the needs and wishes of these children are regularly invoked by adults, but as a strategic and contingent argumentative resource, rather than as the starting point for the decision-making process ' (p. 234). A consistent feature in their data was that parents focused in talk on their needs and wishes as parents -in contrast to the 'official' idea of the child's needs being focused on. Conciliators also attended to this and focused on the parents' feelings about not seeing their child. Trinder et al. conclude 'The challenge for the next decade is to find child-centred methods of routinely including children in dispute resolution processes'. A decade on from this research, despite consistent attention on the importance of prioritising children's needs in dispute resolution processes, children are rarely directly included Walker and Lake-Carroll, 2014), which leads to the child's voice or interest usually being represented by its parents, or brought up by the practitioner, in dispute resolution processes. It is therefore relevant to consider how children's views, needs, or interests are discussed in their absence in the mediation setting.
The data come from a three-year, Economic and Social Research Council-(ESRC) funded study 'Mapping Paths to Family Justice' undertaken by the Universities of Exeter and Kent. The project's central aim was to provide evidence about the awareness, usage, experience and outcomes of out-of-court Family Dispute Resolution (FDR) processes. We used several research methods to address these aims. First, we conducted a national survey of out-of-court FDR awareness and experiences, and second, an individual interview study with 95 people who had experienced an out-of-court FDR process between 1996 and 2013 as well as 40 individual interviews with FDR practitioners. For the final phase of the study, we recorded examples of each FDR process. We recorded five complete mediation processes (four concerning children's matters and one financial; nine sessions in total), as well as three collaborative law processes (11 sessions in total) and five first interviews within solicitor negotiation processes. The overall project findings have been reported elsewhere (Barlow et al., 2013Hunter et al., 2014). Here, we focus on the data from some of the recorded mediation sessions. Notably, despite our efforts to include experiences of child-inclusive mediation (both in interviews with practitioners and parties, and our recorded sessions), there were few actual instances of this -only two interviewed parties had experienced this. Although two-thirds of the mediators interviewed had trained in child-inclusive mediation, they either rarely practised it or had qualms about doing so (Ewing et al., 2015). For example, one mediator (and solicitor) said that 'I have never found the need to bring children into mediation because my practice seems to work quite well on the basis that the children's voice is heard but it's the parents that will bring that voice'. Another mediator and solicitor said, I guess I would be hoping that the parents could be reliable consultants with their own children and what their wishes are, because if they aren't reliable then the thing is fairly doomed from the start and there is a lot of work to be done.
So, the child's voice-or wishes-is assumed or hoped to be represented by the parents in the mediation process. In this article, we draw on data from the mediation processes involving children's matters -and in particular on data from three sets of parents -to investigate when and how talk about children's best interests is occasioned during mediation sessions, and what effect this has on the process.

Analytic approach
We used a Discursive Psychology approach, drawing on elements of critical discourse analysis and CA to understand how participants display understandings in everyday, or 'naturally-occurring', talk (Edwards and Potter, 2001;McHoul and Rapley, 2005). Discursive approaches have no recognisable 'method' in terms of formalised procedures, but share a set of assumptions about language practices -an 'analytic mentality' (Schenkein, 1978). A Discursive Psychology approach typically 'starts with the categories, constructions and orientations through which a sense of agency, say, or severe distress, or a moment of understanding are displayed in a piece of interaction in a particular setting' (Wiggins and Potter, 2007: 73). Language is both constructed and constructive; it is action oriented and situated within a particular sequential environment -words are understood according to what precedes and follows them. The analytic focus is on examining discursive practices in naturalistic materials, including institutional talk. We view mediation as an institutional practice in that talk in mediation settings is normally goal oriented with particular forms of talk permitted; moreover, participants attend to their institution-relevant identities -for example, mediator, parent, ex-partner (Drew and Heritage, 1992). In discursive approaches, a few cases, analysed in detail, are assumed to shed light on widespread practices and assumptions, and also possibly highlight future areas to be investigated on a larger scale (McHoul and Rapley, 2005). Discursive Psychology's emphasis on the construction of specific versions encourages the researcher to consider the practices that those versions are part of, and the particular work that they perform (Edwards and Potter, 2001). For this analysis, we consider how parents talk in mediation sessions about parenting, child residency, gender equality, without taking these claims as facts about the family or individuals' experience. How do people draw on resources, such as claims about parenting behaviour, or a child's best interests. How are claims built up? (Heritage, 2012). At what point in a discussion do participants draw on a resource, and what effect does this have on the progression of the talk? How does the mediation process -or the institution -produce certain forms of arguments and make other types of talk less likely?
Mediation processes have been the subject of a number of discursively oriented studies, and in particular from CA, which have focused on how mediators develop 'de-escalatory practices' (Greatbatch and Dingwall, 1997), how mediators attempt to maintain a stance of neutrality (Jacobs, 2002) and how neighbourhood mediators work to obtain a 'solution' . These studies noted gaps between normative ideals and actual practice, as mediators face multiple competing demands which lead to 'interactional difficulties'. So, while neutrality may be aimed for, CA analyses have demonstrated how the mediator manages this ideal in interaction, while simultaneously progressing the task of achieving outcomes and deflecting tension. Garcia (2010) demonstrated how participants actively align themselves with their mediator in neighbourhood disputes, using 'interactional competence', so while officially it is the mediator's job to balance power between disputants, in practice participants shift the power. A discursive analysis avoids making inferences about participants' inner beliefs, or moral judgements on participants' truthfulness, but views talk as a participatory activity. Parents' assertions that they want to do their best by their children are treated as statements produced to demonstrate competence (Dingwall and Greatbatch, 1993).

Analysis
Extracts were selected from the nine recorded mediation sessions on two related criteria. We looked for instances of parents talking about the 'best interests' of the child, and also for talk that included related concepts, in particular abstract ideas about child well-being, what is good for children in general. We use extracts from mediation sessions with three heterosexual couples. In this data, talk about a child's best interests cropped up when parents were in dispute about where children should live after parental separation, and how much contact each parent would have:

Extract 1. Separated couple, Jack and Kerry, 1 with a woman mediator
An hour into their first mediation session, the father (Jack) has been describing his idea of how their child will be cared for during time spent at his house. Here, the mother, Kerry, produces a strong statement, 'I have got to … ' repeated four times. 'but what I get nothing out of it'. The mediator responds with a direct rebuttal of Kerry's claim. 'You are not rolling over' and follows this up with a rebuke 'it's not about what you get out of it it's about Josh'. In the context of mediation, this exchange has two noteworthy aspects. The direct rebuttal of a participant's claim and moral rebuke are in contravention of the assumption of neutrality -the mediator backs up her strong rebuke with a claim about what mediation is and is not about: 'it's about Josh … '. The institutional business of mediation is shown by the mediator to be about the child, and about the child having 'a healthy relationship' with both parents. In this way, the centrality of the child's interests is attended to by the mediator, but in doing so she had to override core mediation values of neutrality and objectivity Dingwall, 1997, 1999). Kerry suggests that this means that 'everything I want for my son I can't have', and quickly reframes this as 'he can't have'. At this point Jack, the father, picks up on this and mirrors the mediator in rebuking Kerry for her focus on herself. Kerry's response to this explicitly links what she wants for her son with having 'my son's interests at heart' (capitals denote shouting). In fact, she claims, her wishes for her son are based on his interests. She offers evidence -her participation in the mediation process.
In this extract, the mediator introduces the focus on the child at a point of tension in the discussion, and the parents take this up as a resource, with Jack suggesting that Kerry is prioritising her own wishes, which rouses Kerry to a defence for which she uses the term 'my son's interest'. We see how here the mediator's attempt to focus the discussion on the child's interest is not a successful tool for calming the discussion, but rather the opposite-it becomes a challenge to the mother, and the father takes up the term enthusiastically as a strategic way of demonstrating his ex-partner's (inappropriate) focus on herself: In this extract, the mediator asks Jordan, what he thinks would 'work better for' the child, Harry. Jordan's response (Lines 3-4) is quite specific, a direct drawing on a presumption of shared residency (Fehlberg et al., 2011;Trinder, 2010). To this, Nicola asks whether this will be in 'Harry's best interest'. The presumption of shared residency is countered with the concept of the child's best interest. This is not a 'rights-based' discourse from Jordan, it is an 'evidence-based' discourse referring to child well-being research. The mediator questions this reading of shared residency (Line 9). Nicola counters with reference to 'the logistics' (Line 11) -not with reference to 50:50 parenting. In response, Jordan brings in an example of 'parental one-upmanship' -unlike Nicola, he implies, he has not moved around, he is therefore demonstrably a 'constant' in their child's life. We see in this extract several categories which participants link to the child's interest -doing school pickups (Line 14), being a 'constant' (Line 16).
The mediator intervenes with a focus on the practicalities (Line 19). While this might seem to be a neutral intervention, it links back to Nicola's challenge to think about the 'logistics'. In this way, an ostensibly neutral intervention often patterns or references one party's claim or terms. Jordan brings in a complaint about Nicola (Lines 22-27), finishing with another appeal to the child's best interest: Jordan thus positions himself as reasonable, in contrast to Nicola's unreliability in moving around, but also demonstrates his attention to the child's best interest. On being invited by the mediator to respond, Nicola 'categorically' disagrees with the shared residency. This extreme case formulation (Pomerantz, 1986) needs no elaboration or justification (Sidnell, 2004). The form of the statement closes down discussion. Antaki and Wetherell (1999) demonstrated how extreme case formulation can be used to subsequently 'show' concession, and Nicola's categorical refusal here may be a strategic move.
During this extract, first Nicola and then Jordan challenges the other, citing the child's interest (rather than their rights or preferences as parents) as a reason for their position. Jordan draws explicitly on the 'evidence base' for shared parenting; Nicola talks about 'logistics', the example of which is 'school pick-ups'. The concept of the child's best interest in this talk is invoked by Nicola as a response to a plea for shared residency, and by Jordan as a summary of his argument for the child spending equal (or as much as possible) time with both families.
In the third extract, a separating couple discuss a similar issue. Peter has proposed that both of their children have two separate but equal homes, spending one week in each. Roberta is unenthusiastic: with with the children (.) not having, not being (.) with the children not 3.
being with you as much as you both want? So so for example, for you, Roberta, um how would it feel for the (.) not to see them for (.) uh, a week, 4. and (.) how much is that dri:iving your decision making? 5. Roberta Um that um that's not driving it at a:all. Umm uumm (0.2) 6. Med Okay. And the same question, by the way, Peter. 7. Roberta It is purely about what I believe to be the best for them. They (.) 8.
um and (.) you know there may be some element of (.) umm, are they at 9.
are they with Peter, crying, and they want me. Um (.) and I will find that 10.
very difficult to deal with. Not I mean, not (.) obviously, I'll miss them if I 11.
don't have them for a week but i:it's (.) it's more to do with how I (0.1) 12.
would worry how they're feeling about things (.) and not being there for them would just feel (.) cut off and distanced a:and (.) um (.) you know, 21.
there's an element of (.) relegation uuh (.) to:o (.) not being able to 22. fulfil (.) the role that I have been fulfilling since they were born (.) Um 23.
(.) um (.) and I don't see why (.) that should happen (.) to the 24. detriment of the children. Because I (.) you know I um.. my belief um I 25.
believe that Roberta and I offer(.) different valuable things (.) to to each of 26.
our children and I don't see why (.) um they shouldn't benefit from that.
The mediator indirectly suggests here (Lines 1-4) that the parents in dispute about child contact and residency might not be focusing on the children's best interests, but perhaps (and reasonably, he implies) 'how would it feel . . . not to see them for . . . a week?'. Roberta denies this bluntly 'it's not that at all', and elaborates bringing in directly the concept of the children's best interests -'It is purely about what I believe to be the best for them'. This is another extreme case formulation -both hard to counteract, and also setting up possibilities for 'show' concessions later. Roberta insists that her opposition to the 50:50 week-with-each-parent proposal is based on the children's emotional needs. However, the justification for this claim is interesting as it hinges on her feelings and involvement: 'I would worry how they're feeling' and 'not being there for them'. Roberta's argument thus links the children's interests and needs with her presence. She is the parent most attentive to the children's emotional needs, and therefore, her feelings and emotions about the children are not separable from their best interests. This is a strong assertion about who has the right to make epistemic claims about the children's interests.
Here (Lines 13−18) the mediator, performing a symmetrical neutrality, repeats his earlier question to Peter, of his feelings about Roberta's earlier proposal that the children live with their father for alternate weekends, plus a midweek night (a common contact arrangement in the United Kingdom). Peter responds (Lines 19-20) 'I would just feel cut off and distanced' and refers to 'the role that I have been fulfilling since they were born' (Line 22). In this term, he manages to demonstrate his emotional involvement, but also his active longstanding involvement, concluding with a reference to the children's best interests: 'I don't see why that should happen to the detriment of the children.' He adds to this (Lines 25-26) that both parents offer 'different valuable things' which the children should benefit from.
Both parents resist the suggestion that their feelings are driving their preferences, and talk instead about the impact on the children. Peter additionally emphasises his long-term childcaring involvement, and draws on abstract notions of good parenting. The mediator attempts to talk about the parents' feelings, perhaps to challenge the either/or dichotomy of feelings versus evidence in parenting. The dispute between these parents over how much time the children will spend in each parent's home continues into the second mediation session (a few weeks later):

Extract 4: Peter and Roberta (Session 2)
1. Roberta I mean, really the children need consistency (.) and they need a base. They 2.
need to know where their house is. They need to know that this is when 3.
they are going to be at home. How would you like to li:ive in a hotel for six 4.
nights and then in another hotel for another six nights and just keep (.) 5.
shifting? It makes no sense. That's not normal. This is what I am ta:alking 6.
about, being rational to the point of lu lunacy. That is not (.) normal (.) for 7.
children. They need to have a home and then they will go and see you (.) 8.
for five nights out of fourteen.
(a few lines omitted) 9. Peter Okay. So (0.2) what you are saying about they need a home, (.) a base and 10.
then some other uh prop:erty too, this is wrong. The way you are speaking 11.
about it is wrong. We are two parents. We conceived them. We've brought 12.
them up for six years. Roberta talks (Lines [1][2][3][4][5][6] about what the children need: consistency, a base, to know where their house is. She is drawing on abstract notions, but then moves rhetorically into more personal territory 'How would you like to live in a hotel …?'. She contrasts the 'normal' experience she is proposing with the 'lunacy' of Peter's 'rational' 50:50 suggestion. She finishes her turn with an absolute 'They need to have a home and then they will go and see you (.) for five nights out of fourteen'. Here Roberta is in line with the current default practice in family law, whereby children usually live primarily with the mother. Roberta does not justify why she (rather than Peter) will be the parent with 9/14 nights; her argument rests on the children's need for 'a home' (Line 7) rather than two 'hotels', 'and then they will go and see you (.) for five nights out of fourteen'. This is similar to Nicola's 'categorical refusal' to contemplate 50:50 residency in Extract 2. Neither statement is taken by the speaker to require justification, which closes down discussion. Peter takes issue with the idea of 'a home, (.) a base, and then some other uh prop:erty'. He refers to equality of parenting (Lines 9−12). He then maintains that 'They will have two houses of equal standing (.) with two parents of equal STANDING' (Lines 12-13). The mediator intervenes and-while his turn has aspects of an impartial formulation: 'Is the reality however …'-this is a mild rebuttal of Peter's statement by suggesting that the exact language (home, house, property) is unimportant. However, he continues that 'they will have (.) a home with you both' -which is in line with Peter's terminology, and not Roberta's. Peter picks up on this and immediately agrees 'Yes. A home with us both' (Line 19). Here, the mediator uses standard current practice in how to talk about post-separation parenting (two homes, even if children spend more time in one). Despite the mediator's attempts to remain neutral, his lexical choices privilege one party's words or claims over the other.
For the next extract, we return to Nicola and Jordan, in the same mediation session as earlier, still discussing shared residency:

Med
But um can we just while we are on the issue I wasn't going to spend 2.
as much time on Christmas (.) but uh it seems to me that there is a 3.
major misunderstanding between you uh to put it mildly uh at the 4.
moment as to what's actually going to happen this Christmas [so let's 5. Nicola [Yeah 6.
well (.) that that's the reason that I wanted to cover it today because no 7.
conversation has been entered into regarding Christmas this year (0. The mediator (Lines 1-4) attempts to progress the talk on from the cyclical disagreement, downgrading the couple's argument to a 'major misunderstanding', though the following 'to put it mildly' suggests the mediator is aware of this being a reformulation. The reformulation of disagreement to 'misunderstanding' is routine in counselling talk. Edwards and Potter (2001) argue that such reformulations perform two functions. They downplay tension, suggesting no fault, while setting up relationship problems as a kind of puzzle that can be unravelled via counselling. Moreover, they might demonstrate to fighting ex-couples how to constructively talk about their differences in future.
The mediator stresses that the topic of Christmas is a participant-led concernwhich Nicola immediately ratifies 'because no conversation has been entered into regarding Christmas this year'. The use of formal, third-person language is a way of demonstrating reasonableness and capacity to step back from emotions (Edwards and Potter, 2001), particularly in interactions involving both professional and lay members (Atkinson, 1982). Nicola suggests that her working situation is relevant, but Jordan's response draws again on the concept of the child's interests: 'But this isn't about you, it is about Harry and what's best for him'. Nicola subsequently moves to a personal claim 'well I care for Harry'. This reference to maternal care is drawn on as relevant to considering the child's interests -as Roberta did in Extract 3. Nicola suggests a solution (Lines 14-15), but this is presented not as Jordan's right (or as Harry's best interest), but rather as a gracious relinquishment: 'therefore I am prepared'. The earlier extreme case formulation is here modified with a 'show' concession (Antaki and Wetherell, 1999). Jordan returns to the notion of shared residency, and the final extract is five minutes later in this session: Jordan has just produced an elaborate description of what he and his partner and their child can offer Harry, culminating with a claim based on an abstract norm: to increase the contact [with father, stepmother and half sibling] is 'only a good thing and not a bad thing' -another extreme case formulation. Nicola's response is similar to Roberta's in Extract 3, with a three-part list pattern: the child needs a base, stability, routine. Nicola repeats her earlier 'categorical disagreement' with shared residency, claiming that a child needs one stable base. Previously, Nicola drew on her knowledge of Harry's need for care (Extract 2). Here, she draws on general ideas of what is good for a child (a base, stability, a routine), though she talks specifically about Harry (not an abstract child) needing this.

Discussion
We have highlighted how parents drew on various argumentative resources to demonstrate their attention to good parenting, while negotiating shared care of their children. Investigating the details of the talk helps clarify how the mediation process, both in the goals and assumptions inherent in the institution, and in mediators' practices, produces certain forms of claims and arguments.
Our analysis adds to previous research with close attention to the resources that parents currently bring into the mediation sessions. In these extracts, we observe the gendered cultural norms of mothering and fathering being both displayed and contested. The role of fathers post-separation is more contested than in previous decades. While this fine-grained analysis focuses on just three couples over nine mediation sessions, the findings are consistent with a much larger set of observations compiled by a range of investigators (summarised by Dingwall, 2010). Moreover, the concerns and arguments demonstrated in these sessions mirror recent family law debates and shifts in expectations of post-separation parenting (Collier and Sheldon, 2008).
Parents come to mediation services with strategies and knowledge -but sometimes from contradictory sources, such as competing 'expert' positions on a child's wellbeing. They engage performatively in the mediation process, demonstrating parental competence, reasonableness and attention to the child's best interest. They also align themselves with mediators whose interventions may be intended neutrally but can be taken up by parents as argumentative resources (previously noted in Garcia's (2010) neighbourhood mediation participants). Both fathers and mothers refer to the 'child's best interests', directly and implicitly, as justification for their proposals about shared parenting arrangements. Two mothers in these extracts make links between their emotional state and the child's best interests, which is taken by the fathers (and sometimes the mediators) as the mothers attending to their own needs in contrast to focusing on the child's best interests. However, this contrast depends on an assumption of autonomy, in which the interests of all -in this case, parents and child -are independent. An alternative reading is possible and has been taken up by feminist scholars (e.g. Diduck, 2013) who argue that an inability to disentangle each individual's interests may only be a 'problem' because the current autonomy discourse has made it so.
In these extracts, the fathers do not invoke a parental rights discourse (though they do elsewhere), but refer to research which supports shared parenting, to the concept of the child's best interests and to their personal history of involved parenting. These are systematic (and predictable) differences in the resources from parenting literature which fathers and mothers use as justification, but here, we note how fathers and mothers in these studies both orient to the 'child's interests', rather than to their rights as parents, or their feelings. It seems that an acknowledgement of one's parental feelings is risky as it may lead to accusations of being selfish or irrational. This might be alleviated by more explicit focus during the process on the emotional impact of these topics, and development by mediation services of therapeutic skills for dealing with emotions (Lewis, 2008).
We can see a gender difference in the activities which fathers and mothers are performing. With all three couples, the mother is either currently, or assumes she is likely to become, the primary carer, with the child/children spending the majority of their time living in her house. The fathers are all arguing for a greater share of parenting -Jack, for two consecutive days, the other two for equally shared parenting. Nicola feels it is possible to 'categorically refuse' to contemplate 50:50 parenting arrangements (Extracts 2 and 5), and Roberta makes a similar categorical statement (Extract 4). We do not know the details of previous parenting in these couples, but in terms of the interactional positions in mediation, men are regularly in a position of challenging existing or assumed childcare arrangements while women are more often defending the status quo (also noted by Dingwall et al., 1998). Notably, fathers in these extracts feel justified in arguing for equal parenting time. These observable gender differences in mediation talk can be understood discursively in terms of the parenting position and legal options available, rather than necessarily an innate gender difference (Stokoe and Smithson, 2001).
Our interest in mediators' talk here is primarily on how their interventions attend to, privilege or avoid certain types of claim or evidence. We see how mediators' interventions might inadvertently privilege one party's argument, by lexical choice or mirroring talk. Greatbatch and Dingwall (1989) similarly highlighted how mediators may offer 'selective facilitation', encouraging parents towards a particular option rather than remaining resolutely neutral. Much time is taken up with arguing over lexical choices; choice of phrase can be crucial to the mediation process (Stokoe, 2013). Specifically, the fathers were insistent that their home should be known as one of the child's homes, and this might be as important as how many nights the child spends there. These arguments, both sides with evidence bases in parenting literature, reflect the related clash of paradigms in the legal system, and recent debates on the presumption of shared care versus the 'one home with primary carer' model. While parents in these mediation sessions orient in their talk to child's interests rather than parents' rights (more so than in the in-court conciliation talk studied by Trinder et al., 2010), this alone will not provide solutions to the details of shared care. An appeal to parents to consider the child's interests may refocus them on the child, but notions of the child's best interests can be drawn on in conflicting, but evidence-based ways. In this data, a reference to the child's best interests often exacerbated conflict.
We would make several recommendations as to how mediators might address the issues identified in this article. Briefly, these include the following (considered in more detail in Barlow et al., 2014;Ewing et al., 2015): • child-inclusive mediation (where children are old enough to be meaningfully included); • better initial screening to identify when parents hold strongly polarised norms; • joined-up support for mediation clients, particularly including counselling/therapeutic interventions; • steering parents away from oppositional use of abstract statements and research evidence and encouraging them to focus on the actual needs and interests of their particular children; • robust reality testing of parents' proposals, to determine how they will work in practice in the context of the particular family.