A different kind of practice? Meanings attached by practitioners to the idea of ‘adult protection’

Summary This article began with an observation made during re-analysis of the dataset from a study of Scottish adult support and protection practice. Namely, different practitioners could mean different things when they said they had been doing ‘adult protection’ work. For this article, therefore, practitioners’ conceptualisations of adult protection work were hypothesised inductively from the dataset, and refined into categories through a process of constant comparison, open and axial coding. The dataset comprised material from case files and practitioner interviews relating to 23 ‘adults at risk’. Findings When practitioners said they had been doing ‘adult protection’ work, they could mean that: (a) they perceived themselves to have been responding to abuse, harm or high risk; and/or (b) they were adopting a certain formal, even coercive tone of work; and/or (c) they had formally flagged the work as adult protection, including through the use of procedures. Practitioners drew on and combined these meanings in different ways. They also varied in the extent to which they saw adult protection as a distinctive type of practice and/or as different from what had gone before. Applications Support and protection work involves nuanced judgements about the nature and intentions of policy, about how these relate to particular situations and about how best to translate them into practice. Time and support is required for this complex work. Further research is needed to deepen understandings of the practice context, particularly with respect to risk. Policy-makers should engage with such research as they continue to develop adult protection/safeguarding policies.


Introduction
Since the 1990s, a distinctive category of social work and inter-agency activity has been emerging in UK social policies, which was originally referred to as 'adult protection' or the 'protection of vulnerable adults' (Department of Health, 2000;National Assembly for Wales, 2000), but is now increasingly termed 'adult support and protection' in Scotland (Scottish Government, 2008a) and 'adult safeguarding' in England (ADSS, 2005). Over the same period, several research studies have investigated processes and barriers in the implementation of these policies and related procedures and guidelines (Brown & Keating, 1998;Hogg, Johnson, Daniel, & Ferguson, 2009;McCreadie, Mathew, Filinson, & Askham, 2008;Manthorpe et al., 2010;Mathew, Brown, Kingston, McCreadie, & Askham, 2002;Northway, Davies, Mansell, & Jenkins, 2007). The focus of such studies has tended to lie elsewhere than the specific nature of the project of 'adult protection/safeguarding' itself. That is, the remits of the policies under study, and hence of the practitioners charged with implementing them, have been delineated in terms of the need to respond to 'abuse' and/or 'neglect' of adults rendered 'vulnerable' by a certain range of factors, and have not been explored beyond this to any significant degree.
This article reports on a further analysis of the dataset from one of these studies, the Scottish Adult Support and Protection (ASP) study (Hogg et al., 2009)'', on which I worked between 2006 and 2008. Based on initial observations that the idea of 'adult protection' might not have one shared meaning, this further analysis explores what practitioners meant, when they considered that they were doing adult protection work. In outlining the practice meanings of 'adult protection' emerging from the dataset, particular attention is given to the reference points used by practitioners to construct these meanings, either through contrast or through continuity. That is, did practitioners consider adult protection to be continuous with or different from the other work that they do? Did they consider it to be continuous with or different from the work that they would have done, prior to the rise of adult protection policies and procedures? The answers to these questions have implications for the ways practitioners interpret their new and/or ongoing roles. They also have implications for the effectiveness of two-way communication between policymakers, practitioners and others, in contemporary Scotland and elsewhere.
The article sits within a theoretical tradition concerned with the interpretations formed by social actors. Our actions arise, from this perspective, not directly from the situations in which we act, but always also from the meanings we attach to those situations (Blumer, 1962). Shared meanings are common within a given society, underpinning successful interactions (Blumer, 1962;Garfat, 2004). However, personal interpretations and frameworks of interpretation can also be in tension, complicating social life (Garfat, 2004;Goffman, 1974). This certainly appears to be true of adult protection, from the indicative findings I set out below. Moreover, whilst Scottish policy has moved on since the ASP study, the complexity of the discourses of 'adult protection', 'safeguarding' and 'risk' remains, as does the complexity of the relationship between policy and practice. The concluding section of the article draws out some implications of this.

Background
Policies and procedures for adult protection, then the accepted terminology, were introduced in most Scottish local authorities from around 10 years ago. Whereas similar developments in England and Wales were underpinned by national guidance (Department of Health, 2000;National Assembly for Wales, 2000), no equivalent centralised guidance for adult protection existed in Scotland at that time. Nevertheless, Scotland was equally as affected by the growing climate of concern about 'abuse of vulnerable adults', and its early protection policies often drew on guidance from elsewhere in the UK. Arguably, therefore, the ASP study was conducted in a Scottish policy context more directly comparable with other UK countries than is the case today.
The existence of adult protection policies in Scotland, and the attention that they gave to identifying circumstances of 'abuse' or 'neglect', implied that adult protection work could be distinguished from other work. That is, some activities of social work and interagency teams fell under the policies whilst others did not (Hogg et al., 2009). This is not to say that policies necessarily promoted a different kind of practice in adult protection cases, either from other types of cases or from ways of practising that had gone before. Nevertheless, evidence from England and Wales, where most UK research has been conducted, does suggest some practice differences attendant on the introduction of comparable policies and procedures. These have usually been framed as positive advances: for instance a greater capacity to root out abuse (Rees & Manthorpe, 2010); or a fuller focus on investigation combined with a new duty and power to protect potential future victims of an identified abuser or abusive service (Brown & Keating, 1998). However, other implied changes are more negatively framed. For instance, the increased use of risk assessment for the protection of workers as opposed to service users (McCreadie et al., 2008); increased risk aversion amongst support staff, which has 'taken the humanity out of what we do' (Rees & Manthorpe, 2010, p. 520); or the fear of being overwhelmed by bureaucratic requirements and/or the need to prioritise the investigation of allegations, in the context of multiple demands and finite resources (Brown & Keating, 1998;Rees & Manthorpe, 2010).
There is also contrasting evidence, both from Scotland and elsewhere, which questions the association of emerging policies with new and/or distinctive types of practice on the ground. For instance, some practitioners with recently formalised adult protection responsibilities have argued that they are 'doing it already' (Brown & Keating, 1998;McCreadie et al., 2008). Meanwhile, from a different angle, some studies have reported practitioner compliance with adult protection requirements only where these conform to previously-held norms: in the formalised responses to predatory abusers rather than where fragile relationships with family carers may be at stake, for example (Hogg et al., 2009); or by social workers more often than health workers, the latter operating within agency cultures that are experienced as less compatible with the official requirements (McCreadie et al., 2008).
These issues of difference, continuity and change are all richly present in the ASP study dataset. I introduce this study next. The further analysis reported here is then described, followed by the presentation and discussion of its findings.

The ASP study
The Scottish Adult Support and Protection (ASP) study was commissioned by the (then) Scottish Executive to identify strengths and weaknesses in Scottish interagency adult protection practice prior to the implementation of the Adult Support and Protection (Scotland) Act 2007. Supplementary aims of this study were to provide baseline data and to pilot methodology for post-Act comparisons at some point in the future. The fieldwork was conducted in four local authorities selected to represent some demographic contrasts, though not necessarily representative of all 32 local authorities in Scotland.
The project used case study methodology to examine multi-professional adult protection interventions with 23 individual adults. Each adult represented a single case, together with the network of supports surrounding her or him. Potential cases were first identified by senior social work managers or their delegated representatives as having involved adult protection issues. If they thought it was appropriate, for instance if they were confident it would not cause distress, a member of local authority staff known to each individual then approached them, their legal guardian and/or other representative to request consent for the case to be studied. Details of cases were passed to the research team when and if an approach had been made and consent had been granted. Given our reliance on this series of gatekeepers, the final sample was more akin to a convenience than a representative sample. However, older adults, learning and/or physically disabled adults and adults with mental health problems living alone, with family or in residential settings were all represented at least to some degree.
Each case study began with the analysis of documentation, primarily social work files. Where individuals had been known to social work already, as was frequently the case, this analysis aimed to focus on the period of adult protection concerns and their immediate context. However, the first point of relevance to these concerns was often difficult to pinpoint, and several months' or years' worth of files were sometimes read, because they seemed continuously or periodically relevant. Following this, interviews were conducted with all professionals involved in each case, as far as this was possible. This included social work, social care, health, police, housing, advocacy and others. Interviews aimed to clarify information about the nature and chronology of events and interventions, and to explore each participant's reasons for their actions and decisions, as well as their reflections on the case as a whole (Hogg et al., 2009). This fieldwork took place between mid-2006 and mid-2008. Therefore, at the time of their research interview, all practitioners were working to local authoritylevel adult protection policies and procedures, but many could remember a period of their working lives when these had not been in place. Some cases were ongoing when the policies and procedures had been introduced. As a consequence, the dataset is extraordinarily rich from the perspective of examining continuities and changes in practice and representations of practice, in the context of policy change. Conversely, all the cases pre-dated implementation of the Adult Support and Protection (Scotland) Act 2007, as noted, and as discussed further below.

The analysis reported here
For the purpose of the present analysis, every set of circumstances and/or piece of work was identified across the dataset, whose status as 'adult protection' had been assumed, decided or at least discussed by practitioners, as opposed to by the managers involved in identifying cases for the research project. Forty-one such circumstances or pieces of work were identified overall. There were more of these than there were cases, because practitioners often situated elements of the same case differently with respect to their understandings of adult protection activity. For instance, a learning disabled woman might have been under-stimulated at her day centre, verbally taunted by her peers and physically assaulted by her sister. One practitioner might only have characterised responses to the physical assault as adult protection when reflecting on them in her research interview, whilst another practitioner might have considered responses to all three circumstances to be adult protection. Also, a particular piece of work could be characterised as adult protection in one sense, but not in another. For instance, a third practitioner might have reflected on the ways in which she considered professional responses to the verbal taunts to constitute adult protection, whilst maintaining that they did not fit this category in the sense of requiring enactment of procedures.
The interpretation(s) of adult protection which underpinned each communal or minority practitioner perspective on each circumstance or piece of work was hypothesised inductively and refined into a set of categories through a process of constant comparison, open and axial coding (Gibbs, 2007). Evidence both from case file recording and practitioner interviews was used to inform this process. Despite the multi-agency focus of the ASP study as a whole, it should be noted that, where issues of definition were explicitly reflected on in the research interviews, this was almost exclusively by social work or social care professionals. This was possibly because these professionals were more frequently in a position to give an overview of a particular case, and over time had both adult protection and nonadult protection functions in relation to it. Even within this relatively homogeneous group the concept of 'adult protection' had multiple meanings and functions, however, as set out below.

Findings
This section explores practitioners' representations of adult protection in three broad subsections. The first subsection explores representations based on the types of situation to which adult protection was understood to respond. The second explores representations based on the type of practice which adult protection was understood to be. It then shows how this way of thinking interrelated with and differed from the first way. The third subsection explores representations of adult protection as a type of formal system, again showing how this way of thinking interrelated with and could differ from the other ways presented.
The findings should be read with the following in mind. First, the three themes were not mutually exclusive. On the contrary, some could overlap whilst others were in tension, as explained below. Second, all the themes recurred frequently across the dataset. On the basis of these data, however, it is not meaningful to quantify their relative frequency, because of the nature of the sample and some cautions attendant on secondary analysis (see for e.g. Thorne, 1994). Third, whilst examples are drawn upon to illustrate the findings, identifying details have been changed or removed from these.
1. Adult protection as a response to specific types of circumstance A common characterisation of adult protection amongst professionals was as a response to actual or alleged abuse. The equivalence between situations of abuse and situations requiring an adult protection response went unquestioned in a number of conventionally abusive situations: for instance the serial sexual exploitation of learning disabled adults by a member of support staff, or the physical assault of a man with dementia by a younger relative. However, this equivalence could be assumed in other types of circumstance as well. For instance, professionals in one case felt a mother was extremely, detrimentally over-protective of her disabled daughter. When asked by researchers why this was an adult protection case, the advocacy worker for the daughter stated that denial of opportunities, particularly through withdrawal from services as had happened here, was a form of emotional abuse. Clearly, this interpretation of the situation only functions adequately as an answer to this question within a discourse which understands adult protection to follow logically from abuse (Willig, 2001).
Adult protection could also be conceptualised as the management of risk of abuse. The role of the practitioner in this discourse was conceptualised as preemptive because the feared outcome had not yet necessarily occurred. For instance, one learning disabled woman who had been sexually abused by an uncle before she was known to services, and before services had had adult protection policies and procedures in place, was identified to require adult protection interventions and monitoring on an ongoing basis with the advent of these policies, and long after her severing of contact with this uncle. This was because her own use of alcohol and her sexual behaviour was deemed to place her at risk of further exploitation and assault. In another case, a young woman with mental health difficulties and fluctuating capacity began to co-habit with a man with a record of sexual offences. This prompted grave professional concerns that she would be harmed by her partner, on the one hand, and on the other hand that her freedom to make choices and build relationships would be curtailed disproportionately by her parents, who wished to remove her to a residential setting. Intervention was through care management, increased supports and monitoring, with workers never finding persuasive evidence that the partner had acted in ways to justify their further intervention. Formal procedures were never invoked; however front-line workers were clear that this still constituted adult protection work for them.
'Adult protection' could also refer to responses to harms and risks of harms quite far removed from the relevant practitioners' definitions of abuse. This is partially exemplified in the case described above, to the extent that it centred on risks to the woman's independence posed by her parents' intentions, which no practitioner described as potential 'abuse' in this case, but which some did conceptualise as the primary reason for adult protection involvement. In a further example, suitable accommodation and services had been difficult to secure over an extended period for another woman with suspected mental health problems whose behaviour was experienced by others as disturbing and bizarre. The woman's lifestyle was unhealthy to the point of being dangerous given her many physical health conditions and her social worker, supported by her seniors within the council, called an adult protection case conference on these grounds. The social worker described risk of harm as the most important factor in prompting this development, because the woman could well die without support.

Adult protection as a way of working
A second way that adult protection could be characterised was as a type of work with a specific orientation or tone. This type of characterisation was often clearest where practitioners were explaining why a piece of work did not constitute adult protection for them. For instance, one man with dementia lived with and was cared for by his daughter. He alleged that his daughter shouted at him and on occasion pushed or hit him. His daughter reported exhaustion with her caring role. A social worker offered increased home support, and cited their acceptance of this support as the reason adult protection action would not be required. Another man with dementia was neglected by his co-resident nephew, who was increasingly struggling with addiction problems. A respite carer remarked that, 'It's difficult to define as adult protection because his [the nephew's] intentions are good'. Implicit in these types of reasoning, especially common but not exclusive to family situations, was an understanding of adult protection as starkly differentiated from more typical ways of working in social work and care. Adult protection was conceptualised here as more formalised, heavy-handed or even punitive than these other ways of working, and certainly as possessing greater potential for curtailing the perpetrator's autonomy, if not the victim's too. Importantly, however, practitioners characterised the presenting concerns as 'abuse' and/or 'neglect' in both the above examples. Where adult protection was thought of as a way of working, then, it was possible for practitioners to conceive of themselves as dealing with abuse without their doing adult protection work. Indeed, it was possible for them to conceive of adult protection as a proposed new way of working, in situations actually best addressed in another way.
The idea of adult protection as a new way of working does not necessarily correlate with ideas of adult protection as a response to abuse or risk of harm, then, because responses to abuse or risk of harm may have been entirely based on long-established ways of working. As another example, some older or longerstanding cases had previously drawn on other formalised processes prior to the advent of adult protection policies and procedures. Key examples include circumstances involving criminal allegations, in which the police became involved, and circumstances involving recourse to the Adults with Incapacity (Scotland) Act 2000 (the AWIA) and/or to similar powers in older legislation, because an adult was deemed incapable of making particular decisions. Often, practitioners spoke about these activities as adult protection activities in their research interviews, though this terminology may not have been current at the time this work began and had not been used in case files. These practitioners may possibly have thought of adult protection as a way of working which was distinctive from their other practice, then, but they did not present it as a way of working which was new.
One such case in particular highlights the potential for different views on the distinctiveness, if not the newness, of adult protection as a way of working. This case involved a young, profoundly disabled man who was suspected of being neglected and financially exploited by his family. The local Community Learning Disability Team, headed by a senior social worker and psychiatrist, was monitoring the man's needs and co-ordinating a response through an integrated model within the mental health system known as the Care Programme Approach (CPA). Additionally, an application was being prepared for guardianship under the AWIA to confer legal decision-making powers about the man's care onto the local authority. At the time of their research interviews, both the psychiatrist and the social worker considered that these measures were adult protection measures as well as CPA and AWIA measures. Indeed, when asked by researchers which meetings had been adult protection case conferences, because none had been flagged as such in case files, the psychiatrist stated that he 'would consider the Care Programme Approach and adult protection to be merged'. However, a number of involved and extremely concerned workers based elsewhere repeatedly pressed for specific adult protection actions to be taken in response to the alleged abuse. These workers clearly did not consider other ways of working to amount to adult protection, but conceived of adult protection as a different thing entirely.

Adult protection as a formal categorisation and procedure
The different representations discussed so far could overlap in different ways. For instance, a practitioner might represent adult protection as a more or less distinctive type of intervention, more or less heavy-handed and formalised in comparison with other types of intervention, and which might or might not extend its remit further than conventional definitions of abuse. A third type of representation, that of adult protection as a formal categorisation and procedure, could also overlap with these others in a range of ways. Where it was understood as a procedure, three broad types of perceived relationship between adult protection and practitioners' established ways of working were discerned across the dataset. Presented first below are representations of adult protection as a procedure which maps onto existing best practice. Presented second are representations of adult protection as a procedure which does not always map fully or well onto existing or best practice. Presented third are representations of adult protection as a procedure available for use by practice, potentially changing the nature of that practice in the process.
Some representations of adult protection as a formal procedure mapping onto existing best practice coincided closely with the above examples of adult protection as a way of working, consistent with existing ways of working. For instance, some practitioners who were dealing formally with concerns through criminal, CPA and/ or AWIA processes prior to the rise of adult protection policies and procedures, but who identified these in retrospect as adult protection processes, presumably situated adult protection developments at the level of terminology. That is, they would continue to do what they had done before, but would distinguish some of their procedures as 'adult protection' now. However, it should be noted that the range of pre-existing practices which practitioners now considered to be encompassed by adult protection procedures varied. For instance, some practitioners considered themselves to be working under procedures only whilst investigating suspicious injuries or allegations, and to have set them aside if intent to harm could not be established and/or when supportive rather than more heavy-handed measures were deemed to be the best response. This is in line with the narrower understandings of adult protection as a distinctive way of working and/or as a response to abuse as relatively narrowly defined. Other practitioners considered themselves to be working under procedures when providing broader types of support to counter broader types of harms or risks as well, and would not necessarily consider such procedures to carry overtones of coerciveness or blame.
As an alternative to the merging of the two, there were sometimes distinctions drawn between activities considered conceptually to qualify as adult protection activities, for instance because they were working to alleviate abuse, and adult protection procedures, with the latter perceived as narrower and sometimes inappropriate. For example, the man with dementia described above, who had previously been cared for by his nephew with addiction problems, was found a place in a care home by a social worker. The nephew argued vociferously that this should be a temporary arrangement. He talked about taking his uncle home, though he had made no attempt to do so yet. The social worker explained that, though the case was in essence one of protecting a vulnerable person, it had not been made a 'formal adult protection case'. This would have seemed too 'dramatic', she explained, given the nephew was currently complying with professionals. However, adult protection procedures were held in reserve in case the nephew should make an attempt at removal in future.
Alternatively, professionals might have felt themselves to be working with adult protection issues conceptually, and might have wanted to flag them as adult protection formally, but might have felt unable to do so without an appropriate 'trigger' for procedures. For instance, they might have felt their own suspicions of financial exploitation or neglect were insufficient grounds on which to instigate procedures without an allegation from the service user or an observable injury. Sometimes, a single such allegation or injury appeared to function eventually as a pretext, allowing months' or years' worth of previous concerns to be revisited and more robustly addressed once procedures had been triggered. Prior to this point, however, there was a sense that procedures did not fit well with what practitioners felt would have been most helpful practice in these cases.
Standing in contrast to examples like these was one case of a man with dementia who had a close but extremely volatile relationship with his co-resident daughter. The man frequently made allegations of arguments and physical 'scuffles' with this daughter, for instance over household decisions or the safety of his going out alone. The daughter tended to report that the incidents had involved necessary restraint following safety concerns, though she resisted suggestions of other ways to handle these. On the occasion of each allegation a social worker interviewed both parties, establishing that the man wanted no action taken, but not always precisely what had happened. This social worker recommended 'no further adult protection action' but increased support following each such interview, particularly relationship-building work with the daughter to increase her receptiveness to help. The social worker described the case in her research interview as 'adult protection apparently, but in reality more to do with family relationships'. This might be interpreted as a further example of adult protection understood as a type of work: namely, investigative only. Alternatively, however, there was perhaps a sense here of adult protection as a procedure which is too broad, encompassing things which weren't thought, conceptually, to fit.
Adult protection as a formal system could also be conceptualised as a resource to be used by practice. One way some practitioners put this resource to use was to emphasise the severity of their concerns, attempting to shift other professionals' perspectives in order to effect a more robust response. The argument by some practitioners mentioned above that a mother's over-protection of her daughter is abuse and requires an adult protection response, made in the face of other professionals' more cautious and conciliatory approaches, had something of this tenor. In another case a care home was found to have dangerously failing standards, and the local authority organised a series of adult protection multi-agency meetings and direct interventions to manage the situation in the short term. The specifics of the actions taken took limited direction from adult protection procedures in place at that time, which were more focused on interventions with specific individuals. However, the label of 'adult protection' certainly mobilised a range of agencies to act together urgently. Related to this use of adult protection was its use in an attempt to unlock resources. The perspective of the social worker with the woman mentioned above, who had suspected mental health problems, challenging behaviour and a dangerously unhealthy lifestyle, is a good example here. This social worker convened an adult protection meeting to address the 'unavailability of [a] care package either in-house or external': a care package was made available as an outcome of this meeting.
Formal categorisation as adult protection was not only perceived to have the potential to influence other professionals, but also by some participants to have potential influence with perpetrators. For instance, whereas the social worker with the man with dementia, who alleged frequent 'scuffles' with his daughter, wanted to deal with the issues entirely outside of formal systems, the day centre manager thought an adult protection case conference should be called. She felt this would bring home to the daughter the level of professionals' concerns about her care practices and her non-cooperation with the support offered to improve them, hopefully effecting change. In examples like these, as in the professional examples above, there is the suggestion of adult protection procedures used as a resource to change practice, making it more robust and influential in relation to concerns that professionals would have liked to influence previously, but felt disempowered to do so. In this sense adult protection was characterised as new.
A final example of adult protection represented as a resource for practitioners is different, in that the ends to be achieved are not solely the protection and support of service users. Namely, some practitioners spoke about formal categorisation of adult protection cases as a way to establish their own accountability and/or to shield themselves from blame. For instance, in one case procedures were activated after concerns about neglect had accumulated over several months in the context of a carer who was losing capacity. When asked why procedures were activated at this point, one practitioner explained that he 'had to be seen to respond appropriately' given the level of risk, both to the service user and to the carer, which had now been reached. Certainly, the function of procedures here was in large part to alleviate these risks, but the practitioner's turn of phrase suggests that this was not their only function. Another practitioner went further, suggesting that more and more cases were now being categorised as adult protection cases because practitioners were 'covering their backs'. The sense that practice is changing with the advent of adult protection is particularly prominent in statements such as these.

Discussion
These findings have shown that different practitioners could mean different things, when each of them talked about having been doing adult protection work. Three groups of meanings have been identified, though these are not necessarily exhaustive, nor are they mutually exclusive. Namely, practitioners could mean that: a) they perceived themselves to have been responding to 'abuse' or 'risk of harm', however they defined this; and/or b) they were adopting a certain formal tone of work; and/or c) they were following adult protection procedures, which were defined more or less broadly, thought more or less appropriate to these particular circumstances, and were applied for more or less strategic ends.
Through these meanings, different understandings emerged about the extent to which adult protection represented a distinctive way of working, and indeed a new departure for social work with adults.
These findings are significant for their demonstration of the distance that can develop between a policy that frames itself as a clear response to an identified problem, and its practice manifestations, where there is inevitable uncertainty about the nature of particular presenting situations and the consequent bearing on these of policy and ethical imperatives (Taylor & White, 2006). The findings might be viewed as a partial reflection of these practitioners' situated judgements in the face of such uncertainties. Inconsistent use of the term 'adult protection' in the course of describing these judgements is not the focus of interest here in and of itself. However, variations in understandings of the nature and intent of adult protection policies are of considerable interest, because these raise questions about the effectiveness of communication between policy and practice, and about the manner and consistency of policy implementation (McCreadie et al., 2008;Northway et al., 2007).
The study of the relationship between policy and practice has often been divided into two broad camps. On one hand is a model of policy as top-down prescription. This way of thinking about policy would tend to link ambiguities in its interpretation or substantive variations in its implementation either with errors of design and/or with poor practice of its implementers. On the other hand is a model of policy defined on the ground by the contingent judgements and interactions of reflexive agents (Becker & Bryman, 2004;Prior & Barnes, 2011). The usefulness of the latter model for understanding social policy has long been acknowledged and elaborated (Ellis, 2011;Lipsky, 1980), and its inevitability in practice has been emphasised, particularly within a field such as social work. This is because delivery of policies via the medium of human interaction and relationships is argued to involve reflexive, contextualised judgements, necessarily more nuanced than those for which policies might offer prescriptions (Broadhurst, Hall, Wastell, White, & Pithouse, 2010;Prior & Barnes, 2011). Grounded as it is in meaning-making as a necessary mediator between situations and professionals' responses to them (Blumer, 1962;Garfat, 2004), this article also tends towards the latter model. Accordingly, it offers an account of the findings below that focuses on the complexities of meaning-making, first in relation to abuse and second in relation to risk.
The first way that some practitioners defined adult protection, to reiterate, was as a response to a phenomenon which they termed 'abuse'. However, this definition was not as unanimous nor as internally consistent as policies might perhaps have predicted. The notorious difficulty of arriving at a shared definition of abuse (Brammer & Biggs, 1998;Dixon et al., 2010) offers part of an explanation for this. Indeed, it has long been argued that the concept of 'abuse' is a social construction, perennially susceptible to contestation and change (Hacking, 1991;Harbison & Morrow, 1998). More telling still, however, is the lack of consensus, in these findings and others (Johnson, 2012a), that this concept is or can be central to adult protection. With reference to the same issue in child protection work, Buckley (2003, p. 15) describes a 'divergence between the way in which the work [is] actually experienced by practitioners and service users and the perceptions which [are] held by other stakeholders'. Specifically, the closer a study gets to what child protection workers actually do, the less fitting seem official representations of the task as, first, identifying an objectively distinct phenomenon and/or threshold which has been crossed and, second, acting accordingly. Child protection work is argued to be more continuously fluid, intuitive and uncertain than this; it's actually central task of 'making sure by the means available . . . that for practical purposes a child is 'all right'' (Parton, Thorpe, & Wattam, 1997, p. 83) not to be helpfully guided by so abstract a concept as 'abuse' (Buckley, 2003;Parton et al., 1997). The same argument could be made for adult protection, with the implication that continued policy adherence to this or a comparable model will contribute to a similar sense of 'divergence' between policy and the experience of practice.
Second, practitioners were making sense of policy developments in an immensely complex landscape with respect with risk. A considerable amount has been written about the effects on social work of ever-proliferating risks and/or perceptions of risks, coupled with an assumption that these can and ought to be averted (Kemshall, Parton, Walsh, & Waterson, 1997;Webb, 2006). Some key weaknesses of child protection systems, in particular, have been linked with the massive investment of resources into attempts to identify high risk, and the parallel focus on documenting accountability in a prevailing 'blame-culture', to the detriment of support for children and families (Munro, 2010;Parton, 1998). Traces of similar approaches to risk have been evidenced here, either in descriptions participants gave of their own or others' practice, or in their characterisations of adult protection policies: for instance, descriptions of invoking policy as a way to 'cover your back', or of procedures as a poor fit with the 'real', relationship-work that some situations were thought to require (c.f. Broadhurst et al., 2010).
Notwithstanding this, previous empirical work has also highlighted great scope for variety in 'risk rationalities' (Kemshall, 2010), along a number of dimensions and at a number of levels. For instance, local authorities can vary in their ways of conceptualising and managing risk (Brown & Stein, 1998), as can welfare agencies, as can individuals (Stalker, 2003). Again within the more developed field of child protection work, Broadhurst et al. (2010) highlight a particularly broad set of contrasts between the 'formal logic' whereby policies seek to manage risk and the 'informal logics' of practitioners on the ground. The present findings include a number of comparable examples. For instance, judgements that the label of 'adult protection' will unlock resources within a particular local authority context, about the scope for relationship-building to alleviate specific, highly concerning situations, and about the usefulness or otherwise of formalised procedures based on the ways particular families might be expected to interpret these, are all informal strategies for managing contextualised risks. Following Broadhurst et al. (2010), it is worth noting that some such strategies may turn out in retrospect to have been misjudgements. Nonetheless, a full appreciation by policy-makers, senior managers and others that this type of work necessarily deals in situated judgements is vital, if they are to nurture rather than erode this core of practice (Broadhurst et al., 2010;Taylor & White, 2006). Furthermore, a critical appreciation amongst policymakers of the types of factors on which practice judgements are contingent will help them to avert unwanted policy outcomes, for instance increasing risk-aversion as an outcome of practice cultures perceived as unsupportive when things have 'gone wrong'.
In summary, whilst the findings highlight considerable variation in practitioners' characterisations of the intended outcomes of adult protection policies, and whether and how these might be achieved, they also support other research which argues that there are limits to the certainty to be imposed on practice. Indeed, helpful policy responses to this complex terrain include protected time and support for practitioners to work sensitively and confidently with uncertainty (Parton, 1998;Taylor & White, 2006). However, this is not to argue that the development and clarification of policy intentions would not also be helpful. These particular findings suggest that such clarifications should engage specifically with the multi-faceted landscape of risk and the evidently mixed expectations that adult protection should constitute a different kind of practice.
One additional implication of these findings is the need for further, close ethnographic study of adult protection practice. This would assist policy-makers to engage in as nuanced a way as possible with the practice context, and would address the fact that these findings are limited in a number of respects, being based on retrospective accounts collected for another purpose, rather than realtime accounts and observations of meaning-making on the ground. Furthermore, in Scotland at least, the policy and legislative context has now been altered by the Adult Support and Protection (Scotland) Act 2007 (the ASPSA). Changes include the re-conceptualisation of 'adult protection' as 'adult support and protection', to emphasise that it encompasses holistic support and preventative measures alongside crisis interventions, and the replacement of the concept of 'abuse' with the concept of 'risk of harm', to more easily encompass unintentional actions and omissions, self-harm, self-neglect (Scottish Parliament Information Centre, 2007), and the risk of each of these.
The breadth of the concept of 'risk of harm' appears to herald a significant broadening of the field of Scottish adult (support and) protection, at least as defined in the ASPSA and related official materials. Some activities previously Notwithstanding this, and as this article has shown, the contents and coherence, or otherwise, of policy are only one part of what constitutes practice on the ground. More research is required to explore continuities and developments in practice, in order to evaluate policy and the lived effects of policy, in Scotland and indeed elsewhere.

Ethics
The ASP study was referred by the Dundee University Ethics Committee to the NHS Tayside Ethics Committee, which in turn delegated ethical permissions to individual Caldicott guardians because it classified the work as audit. Permissions were therefore obtained from the three Caldicott guardians, four local authorities and two police forces in the areas in which the work took place.

Funding
The Scottish Adult Support and Protection study was funded by the Scottish Government and Capability Scotland, and was undertaken by the White Top Research Unit, University of Dundee in collaboration with Age Concern Scotland. The PhD of which this publication forms a part was supported by the Economic and Social Research Council (grant number ES/G013853/1). However, the views expressed in this article are the author's own.