Care and the shadow of the fourth age: how does home care get caught up in it and how does it stay away from it?

Abstract This article examines how care encounters at the elders’ homes are forged, and how the way these encounters are forged avoids or evokes the social imaginary of the fourth age. Data were gathered in Portugal from elders receiving home care (16 cases), their care workers (eight cases) and family carers (six cases), through participant observation and informal conversations (conducted at the elders’ homes), as well as focus groups. The collected data were analysed according to the procedures of Framework Analysis. This study found five forms of care encounters – marked by conflict, infantilisation, burden, harmony and indifference – the harmony form being the only one found to maintain the fourth age at a distance. It concludes that home care has a Janus-like nature in relation to the fourth age, and that the way home care encounters are forged depends on the conditions of the care settings and the actions of all participants in care encounters. It also concludes that it is difficult to maintain the social imaginary of the fourth age at a distance when the elders exhibit high levels of infirmity. Finally, it concludes that family carers play a crucial role in the way care encounters unfold. Implications for practice and policy include vocational training regarding the relational component of care, and information and educational programmes for family carers.


Introduction
The provision of long-term care at elders' own homes is expanding in most countries around the world (Fujisawa and Colombo, 2009), and the care arrangements tend to be mixed, including formal care workers and informal carers, mainly family carers (Saraceno and Keck, 2010).
Care has a relational dimension, as it implies a relationship between the carereceiver and at least one care provider. The care relationship itself has been one line of research on the topic of home care for older people. In parallel, during the last decades theoretical work on care has proliferated, and recently we have witnessed a theorisation of later life that differentiates the third age (later life without care, associated with consumption and agency) from the fourth age (later life with care, associated with loss of agency and decay). However, to the best of my knowledge we do not have empirical evidence about how care relates to the fourth age. More precisely, we do not know to what extent care maintains the social imaginary of the fourth age at a distance or, in contrast, makes it a conspicuous presence.
This article reports a study conducted in Portugal that purported to understand how care is deployed at the elders' home, as well as to examine to what extent the way care is deployed protects the participants in the care encounters from the ignominy of the fourth age or, on the contrary, exposes them to it.

Theorising the fourth age and care
The fourth age paradigm Higgs and Gilleard (2015a) argue that in late modernity, particularly in affluent societies, we have witnessed a fragmentation of later life. This fragmentation is translated into the emergence of the third age as a 'cultural field', characterised by autonomy, choice and leisure, and the emergence of the fourth age as a 'social imaginary', associated with 'real' old age, i.e. decline and decay. The third age and the fourth age are inter-connected, as the more the third age with its 'anti-ageing' narratives and 'age-denying' practices expands in a given society, the more the fourth age is darkened and pushed to the ending point of later life Gilleard, 2015a, 2016).
The fourth age does not correspond to a stage of life nor to a chronological category, but rather to a 'collective representation of all that is feared and found most distasteful about old age' (Higgs and Gilleard, 2016: 121). The fourth age is constituted, according to Gilleard (2015a, 2016), by four elements: frailty, abjection, care and loss of agency. These authors conceive frailty as both mental and physical infirmity but also as a status associated with a sense of 'otherness' (being different from us) and 'lessness' (being less than ourselves). In this vein, frailty entails both material and social vulnerability. In turn, abjection refers to those aspects of frailty that are socially constructed as the most distasteful and repugnant. It is associated with the inability to look after oneself and it foregrounds a social category, a class of abject people which may also include care providers. Frailty and abjection are accompanied by the need for care. It is in and through care that the fourth age is 'socially realised as both narrative and practice' (Higgs and Gilleard, 2016: 74). Finally, loss of agency means, in general, losing the power to think and act autonomously and independently.
It is important to add that Higgs and Gilleard (2016) conceive care as having a 'Janus-like' nature, i.e. as being like a 'double-edged sword' in relation to the social imaginary of the fourth age. They claim that care has, paradoxically, the potential to protect older people from the ignominy of the fourth age, promoting, amongst other things, what Lloyd et al. (2014) designate 'relational autonomy', and, on the contrary, to expose them to it, producing an erosive effect on agency and identity. In this vein, Higgs and Gilleard (2016: 138) advocate for an understanding of 'how those deemed frail are frequently further failed and how … we can learn to "fail better"'.
The same authors suggest that when a relationship has little or no reciprocity from the part of the care-receiver it becomes a 'relationship of one', and consequently some narratives and practices of care, such as, for example, the narrative of compassionate care and its associated practices, become difficult to sustain (Higgs and Gilleard, 2016). Higgs and Gilleard (2016) call attention to the fact that the concept of personhood, in which the person-centred care approach is based, is connected to the concept of the fourth age, given that it purports to combat frailty and abjection. Kitwood (1997: 8), a leading theorist on personhood in the context of dementia, defines personhood as the 'standing or status that is bestowed upon one human being by others, in the context of relationships and social beings'. According to Higgs and Gilleard (2016), the Kitwood perspective on personhood confounds its metaphysical aspects (personhood as agency, consciousness, rationality and second-order reflexivity) with its moral aspects (personhood as a moral status, demanding certain rights), and advocates that the personhood of adults with dementia is dependent on the relationships established with them rather than on their own capabilities. Following the same authors, Kitwood's assertion places the responsibility of preserving the personhood of those with dementiaboth in the sense of demonstrating moral concern for them and sustaining their capabilitiesonly on the shoulders of the care providers. However, Higgs and Gilleard (2016) call our attention to the fact that many people with dementia and others in need of care do not have the necessary capabilities to constitute metaphysical personhood, and that these deficits increase with time. Furthermore, they emphasise that most care providers demonstrate moral concern for people with dementia but 'their care practices may either deepen or lighten the darkness of the fourth age' (Higgs and Gilleard 2016: 133). In this vein, Higgs and Gilleard (2016) advocate basing a 'philosophy of care' on the fourth age perspective rather than on the concept of personhood, because keeping the social imaginary of the fourth age at a distance 'requires no assertions about the "personhood" of people with dementia, but simply the recognition of a common humanity and the taking of due care by carers as moral agents' (Higgs and Gilleard, 2016: 26).
The ethics of care Tronto (1993Tronto ( , 2013 clarifies the nature of care, arguing that it is both a disposition (e.g. being concerned) and an activity (e.g. feeding). She also emphasises that care is inherently relational, involving interactions between two or more persons, and that it is put into practice through a complex process. Tronto (1993) initially identified four phases of the process of care: caring about (recognising unmet needs), taking care of (taking responsibility for meeting the needs), care-giving (undertaking concrete actions in order to meet the needs) and care-receiving (responding to the care received and making judgements about it). In addition, Tronto (1993) identified the ethical/moral qualities necessary to carry out each phase of the process of care. They are, respectively: attentiveness (being attentive to unmet needs), responsibility (taking responsibility for meeting the needs), competence (providing skilful and appropriate care) and responsiveness (evaluating the effectiveness of the care provided). Tronto (1993) advocates that 'good care' is achieved if each phase of the process of care is carried out according to the respective ethical quality. Later, Tronto (2013), when developing a Theory of Caring Democracy, added a fifth phase to the process of care: caring with. In this phase, care needs should be ideally met in accordance with democratic commitments to justice (assignment of care responsibilities and other responsibilities in a non-dominated way), equality (having equal voice) and freedom (absence of domination). Accordingly, solidarity and trust are the ethical qualities necessary to accomplish the 'caring with' phase.
Care as 'body work' and 'dirty work' Twigg and colleagues (Twigg, 2000;Twigg et al., 2011) also contribute to the clarification of the nature of care, arguing that providing health and social care is an activity with a special nature: it is a form of 'body work', which is commonly 'dirty work'. Body work consists of working on the bodies of patients/users and it incorporates two dimensions, an instrumental dimension (a practice) and an expressive dimension (involving emotional work). In turn, body work is dirty work in two senses. First, it deals with the 'negativities of the body' (Twigg, 2000: 7), such as faeces, urine and the like. The presence of odours, sights and textures produce a feeling of disgust in care providers and a disruptive effect on close relationships (Isaksen, 2002). It is suggested that in the most extreme cases the care-receiver's home may also be dirty, and a 'dirty workplace' may have negative impacts upon care workers 'through unofficially increasing their workload, further devaluing their work and risking their wellbeing' (Wibberley, 2013: 156). Second, body work is dirty work because it is pushed to private realms by a society that puts a high public value on youth and success and has an aversion to decay, failure and death.
The empirical study of the care relationship in the context of home care Research on home care for older people has addressed several themes. Higgs and Gilleard (2016: 106) identified four themes: (a) 'the social policy issues of funding and organising these sectors of domiciliary care…'; (b) 'the motivations and moral identities of care workers'; (c) 'the tasks, challenges and stresses home care workers face'; and (d) 'the care relationship itself, whether contrasted explicitly with informal, family care, or explored directly'. Considering the aim of the present article, only the literature that focuses on the theme of the care relationship will be reviewed, in particular that which includes both home care workers and family carers.
Some of studies do not focus directly on the care relationship, but on how the materiality and meaning of the 'home' contributes to shape the care relationships, specifically the power dynamics between care workers and care-receivers (e.g. Trojan and Yonge, 1993;Twigg, 2000;England and Dyck, 2011). Nonetheless, some of these studies emphasise that home care is also shaped by factors beyond the home, namely by the culture and rationality of organisations and services (e.g. England and Dyck, 2011).
Studies which have focused more directly on the care relationship have privileged three issues: negotiations between elders and care providers (Vivian and Wilcox, 2000;Spiers, 2002;Valokivi, 2005;Breitholtz et al., 2012;Sundler et al., 2016); quality of, or satisfaction with, the care relationship (Eustis and Fischer, 1991;Karner, 1998;Fujiwara et al., 2003;Graham and Bassett, 2006;Chon, 2015;Ayalon and Roziner, 2016); and factors associated with (in)dignity in care (Trojan and Yonge, 1993;Tadd and Calnan, 2009;Stewart and McVittie, 2011;Moe et al., 2013;Lloyd et al., 2014). Studies which have focused on the first issue converge on the idea that care relationships involve complex interpersonal negotiations and exchanges imbued with important issues, such as dependency and power. This has been underlined by other authors (e.g. Rummery and Fine, 2012;Walsh and Shutes, 2013). In turn, studies which have explored the second issue show that there are positive and negative perceptions about the quality of, or satisfaction with, the care relationship, and that in some cases these evaluations diverge between those involved in the care relationships. Finally, studies which have addressed the third issue identify several factors associated with (un)dignified care. Respect, trust and elders' autonomy are the most mentioned factors associated with dignified care; other studies show that feeling like a burden also compromises dignity.
In sum, this literature has called our attention to the fragile, dynamic and fluctuating nature of the care relationship in the context of home care, which derives from several underlying negotiations, ambiguities and ambivalences. This literature has also explored the issue of quality and satisfaction in care relationships and the factors which promote, or undermine, dignity.
Considering that the theoretical work on the fourth age is recent, it is not surprising that the issue of the fourth age is explored, albeit partially, in only one of the studies reviewed above (Lloyd et al., 2014). This study emphasises that social relationships play a key role in maintaining the elders' identity and, consequently, the fourth age at a distance.
Home care for older people in Portugal Formal home care in Portugal is provided mainly by the voluntary sector, although the care market (private for-profit sector) has been growing during recent decades. The state has a minor role in home care provision, being mainly a regulator of the care sector as a whole and a funder of the voluntary sector.
In most countries of the world, home care workers have low qualifications and earn low wages (Bettio and Verashchagina, 2010), and this is particularly true in Portugal. With respect to qualifications, the ministry responsible for home care in Portugal establishes that vocational training is a prerequisite for the practice of home care, but local services may exempt the worker from this prerequisite. Moreover, training programmes neglect the relational component of home care, that is, the relationship between care workers and service users. In addition, there are no educational programmes for family carers in Portugal.
In Portugal all people who need care and support with activities of daily living are eligible to receive home care provided by the voluntary and public sectors; in the voluntary sector the services are paid according to the income of users and their families. Home care includes services such as personal care, meals, laundry and house cleaning. Personal care includes incontinence care and assistance with feeding, dressing, bathing and toileting. In the voluntary and public sectors, each care visit takes about 20 minutes.

Research questions and methodological approach
The present study addresses the following research questions: • How is care forged by older persons and care providers at the elders' homes?
• How is the social imaginary of the fourth age resisted or, on the contrary, reproduced by the way home care is forged?
The first question seeks to gather a deep understanding regarding how the giving and receiving of care is deployed in the various contexts of home care encounters. The second question seeks to examine how the way home care is forged protects both the older persons and their care providers from the social imaginary of the fourth age (by not exacerbating frailty, and not producing abjection and loss of agency), or exposes them to it by contrary actions. Considering the research aims and questions, a qualitative approachwith its emphasis on studying 'things in their natural settings, attempting to make sense of, or interpret, phenomena in terms of the meanings people bring to them' (Denzin and Lincoln, 1994: 2)was used to collect and analyse data.

Research participants and sampling
The participants in the present study were elders who were receiving home care, their home care workers and, in some cases, also their family carers. The elders were selected according to the following criteria: aged 65 or older; receiving formal home care; able to maintain a conversation in Portuguese; and able to provide verbal informed consent. No criteria were used to select the care workers and the family carers, as they were automatically selected through the selection of the elders. However, informed consent was obtained from all research participants, including care workers and family carers.
A not-for-profit institution that provides social care services for older adults in the region of Algarve participated in the process of selection and recruitment of the elders, by identifying the elders who satisfied the study criteria and granting formal access to elders and respective care workers. There was concern about ensuring, as much as possible, a diversity of cases in terms of the socio-demographic characteristics of the elders. Convenience sampling (Ritchie et al., 2014) and theoretical sampling (Glaser, 1998) were used to select the cases.
A total of 16 cases were included in the final sample. In ten of these cases the care encounters had the participation of the elders and care workers (always two care workers), while in the remaining cases they had the participation of elders, the two care workers and family carers. In the former cases there is a clear predominance of women (eight women, all over 80 years old, and two men, both under 80 years of age), while in the latter cases the opposite is true (five men, all over 80 years old, and one woman under 80 years of age). Some social characteristics of the elders are presented in Table 1. The group of care workers is composed of eight women, all over the age of 40, and with nine years of schooling. In turn, the group of family care-givers who participated in the observed care encounters is composed of one daughter, one son, one daughter-in-law and three wives. The first three are over the age of 40 and have ten years of schooling, while the last three are over the age of 70 and have no more than four years of schooling.

Data collection and analysis
Oral informed consent was obtained from all research participants before the start of data collection. 1 The data were gathered between 2014 and 2016 through participant observation, informal conversations and focus groups (of care workers). Only the author of this article undertook the whole process of data collection and analysis. The observations and informal conversations were conducted in the elders' homes. The researcher accompanied the home care workers on their visits. The observations aimed to capture 'what is going on' in home care encounters, that is, what the elders and care providers say and do during the care workers visits (including non-verbal communication). In turn, the informal conversations (with the elders, care workers and family carers) aimed to clarify some aspects of the observed actions and interactions, as well as some analytical puzzles. Together, participant observation and informal conversations allowed the properties and features that shape care encounters to be captured. About every two months the cases were distributed to other pairs of care workers, in a rotation mechanism between the teams of care workers. This means that the same pair of care workers conducted more than one visit to all the elders who participated in the study. Finally, focus groups were selected as a method to collect data, as during the observations and informal conversations at the elders' homes the care workers expressed sometimes their conceptions about what is a 'good home', a 'good user' and the like. It was considered that these conceptions deserved to be better explored in a systematic way through the focus group method, which is particularly suitable for capturing the culture or shared understandings of a group (Bloor et al., 2000). The focus groups took place at the premises of the institution that collaborated with the study. Two focus groups were conducted with the participation of the researcher and eight care workers (the same care workers who had participated in the observed care encounters), both lasting about one hour.
The data collected through the observations and informal conversations were recorded through jotted notes and later these notes were converted with more detail via a word processor. A total of 248 observations of home visits were carried out (an average of 16 observations in each home), each lasting, on average, 20 minutes, complemented by a large number of informal conversations, altogether producing 245 pages of typed notes. Data collection ended when theoretical saturation (Glaser, 1998) was reached. The focus groups were recorded and transcribed verbatim.
Data analysis was performed according to the procedures of Framework Analysis, a technique of thematic analysis, in which constant comparison between cases and categories (Ritchie et al., 2014) was conducted to enhance the quality of the analysis. A typology of forms of care encounters was constructed on the basis of the emerging categories. Data analysis was conducted with the help of NVivo 9 software. The credibility of the findings was explored through the technique of member checking (participant or respondent validation), but only with the care workers. Considering the content of the findings, it was considered that the inclusion of the elders and their family carers in member checking could be disturbing for them.

Findings
The findings of the focus groups with care workers, in particular those related to the social constructions/representations shared by the participants, are presented throughout the description of the forms of the care encounters. Nonetheless, it is necessary to note another finding of these focus groups. The care workers revealed that they only had initial vocational training, but this training did not include the relational component of care, nor any guidelines from their co-ordinators as to how they should relate to the elders.

Forms of care encounters
In order to account for the different ways through which care is forged, the concept of social form, as defined by Allan (2009), was used: 'A form is a patterned mode of interaction through which people meet personal and group goals' (Allan, 2009: 240). Five forms of care encounters were identified, which are marked by conflict (two cases), infantilisation (three cases), burden (three cases), harmony (three cases) and indifference (five cases). There is no clear association between a particular form of care encounter and a particular team of care workers; each form of care encounter maintains the same basic characteristics regardless of the teams of care workers.

Conflict
The care encounters included in this form are represented by two cases, with each involving the elders, two care workers and one family carer (in one case, a son and, in the other case, a daughter). The elders, both male, are bedridden, and exhibit some physical strength and memory. Although they are able to talk, they do it only when they are questioned about something. They receive personal care, including incontinence care and bathing.
In both cases the care encounters take place in the elders' bedrooms, which are very dirty (walls with faeces, garbage on the floor, intense odour), under-furnished and depersonalised (they resemble a prison cell, completely empty of personal possessions). The care workers express aversion in relation to the conditions of these care settings: It is really difficult to work here (accompanied by non-verbal communication showing difficulties in withstanding the strong odour). (Care worker 3, during a home visit) The objects and utensils that the care workers need to undertake their work are also scarce in these care settings. On the whole, these care settings fit the social construction of a 'bad home' which is shared by the care workers who participated in the focus groups. A 'bad home' is a home that is dirty and that does not have the things that the care workers need to undertake their work properly, such as towels, soaps and diapers. Apart from this, the elders are also often 'full of urine and faeces' when the care workers arrive at their homes (urine and/or faeces in several parts of the body, as well as in pyjamas and bed sheets, due to the overloaded and unchanged diapers). This reinforces in the care workers the feeling of disgust and the idea that the family carers are negligent. In this respect, the care workers expect that family carers are attentive to and responsible for the incontinence care outside the periods covered by the home visits.
When we look at the way the care encounters unfold in these settings, we find that the actions taken by the care workers and the family members are task-oriented (although the family members neglect most of the elders' needs, as we will see later), that is, they are focused on the practical care tasks rather than on the elders themselves.
It is also observable that the care workers, while performing the care tasks, tend to ignore the elders, given that they talk with each other as if the elders were not there. Furthermore, the care workers, and also the family carers, do not protect the elders' privacy. For example, the care workers do not knock on the door before entering the bedroom and do not ask for permission to take off the diaper. The care providers (care workers and family carers) also do not promote the elders' autonomy, as all the decisions about care are taken without the elders' participation.
However, the most important feature of these care encounters is that they are substantially marked by conflict. There is a harsh relationship between the care workers and the elders and also, although with lower intensity, between the family carers and the elders. On one hand, the harshness between the care workers and the elders is revealed by interactions with the following characteristics: the elders do not co-operate with the care workers (e.g. not complying with the care workers' requests, making the body rigid) and they often show aggressiveness. In turn, the care workers ask them to do (or not to do) certain things with an authoritarian tone of voice, touch the elders' bodies in a brisk manner and sometimes even make threats, such as that they will put them in a nursing home. These interactions seem like a 'fight' between two opponents. This was explicitly recognised by one of the care workers before entering the elder's home when she said that she was going to another 'battle'. On the other hand, the harshness between the elders and their family carers is found mainly in the sour tone of voice they use to talk with each other and their non-verbal communication (indicating mutual impatience and annoyance). It should be added that the elders' behaviour is in line with the social construction of a 'bad user' that is shared by the care workers. Not co-operating, being aggressive and not manifesting trust in the care workers are the characteristics of a 'bad user'.
Conflictual relationships are not limited to the interactions between the elders and their care providers. The interactions between the care workers and the family members are even more problematic, as on several occasions they were fraught with bitterness. During one home visit there was a disagreement between the care workers and the elder's daughter about the best way to change the elder's diaper, and this disagreement ended up in an overt and strong dispute. In this respect, it should be noticed that the care workers, in both cases, have a negative image of the family carers. They think that they neglect the elders' needs and that they do not help them in the care provision.
Not providing the necessary conditions so that the care workers can do their work properly (supplying the objects and utensils needed for care provision), as well as not having the house and the elder clean, are the characteristics of a 'bad family', a social construction shared by all the care workers who participated in the focus groups.
Participant observation revealed many indicators of family negligence. Apart from the dirtiness of the bedrooms and many occasions on which the elder had not yet had breakfast at the time of the home visit (around 11 am), the researcher witnessed an episode in which the elder was lying on the floor, completely naked and covered with faeces and urine, while his daughter was dealing with other household chores.
Most likely as a consequence of these patterns of interaction and the conditions of the care settings in which they take place, the elders express sadness through verbal (e.g. saying that they feel sad) and non-verbal communication (e.g. having a sad look in their eyes). They even say that they want to die: I'm very sad with all of this, I want to die, please give me death … Life does mean anything to me anymore! (84-year-old man)

Infantilisation
In this form of care encounterrepresented by three casesthere is no participation of family carers, although they exist. The care workers have the idea that the family carers neglect the elders in two cases. The elders, two women and one man, are bedridden and their physical strength and memory is very low (they would not be able to make their bodies rigid in order to make the care workers' job more difficult and they are not able to remember the last meal they had). In addition, one elder is not able to talk and the other two have great difficulty in talking. It is important to emphasise that higher levels of physical and mental infirmity are found in this form of care encounter. These elders receive personal care, including bathing and incontinence care. The care settings are the elders' bedrooms, all of them pleasant (clean, with natural light, not under-or over-furnished), although one has an infantile decoration.
As in the first form of care encounters, here the actions of the care workers are also task-oriented and they sometimes ignore the elders, do not protect their privacy and do not promote their autonomy. However, contrary to the conflict form, in this form, the care workers display tenderness while providing care, by caressing and smiling. Yet, while providing care they also use 'baby talk': using diminutives, shortening sentences, simplifying vocabulary and using a childish tone of voice. In general, they interact with the elders as if they were babies, neglecting what they are trying to say or devaluing their wishes, as in this field note: The care worker said: My little girl, how are you today? Next the care worker asked the elder if she would like to have baby formula using baby talk. While this care worker was feeding the elder, the other went to get a teddy bear and used it to play with the elder. She said: people of this age like to have a teddy bear. After this, the care workers laid the elder in bed and began to undertake her personal hygiene. When I asked the elder if she would prefer that I leave the room, one of the care workers said: she does not care about that anymore. (Visit to a 91-year-old woman) In these care encounters the elders adopt a passive posture in relation to the care workers, as they fully co-operate with them by accepting the requests and orientations of the care workers. Due to their difficulties in talking, they are not demanding nor making requests to care workers.

Burden
This form of care encounterrepresented by three caseshas the participation of family carers (the elders' wives in two cases and the elder's daughter-in-law in the other). Two elders, both men, are bedridden, and the third one, a woman, is not bedridden but is confined to home. All of them still have physical strength, good memory and they communicate without difficulties. All receive incontinence care and assistance with bathing. The care encounters of the men take place in their bedrooms and in the case of the woman it takes place in the bedroom and bathroom. One of these care settings is pleasant, while the others less so, given that one is overfurnished (about which the care workers complain) and the other has no natural light.
The interactions between the elders and the care workers are marked by tenderness, although the care workers' actions are task-oriented. In contrast, the interactions between the elders and their family carers exhibit some coldness and insensitivity, particularly visible in the family carers' actions and attitudes, and also some impatience or intolerance mainly on the part of the elders. The following interaction between the elder (75-year-old man) and his wife is illustrative of coldness and insensitivity on the part of the latter: Elder: I have here my saint [making reference to his wife]. Elder's wife: There are no saints here, saints are in heaven. Be quiet, don't talk nonsense.
In addition, there is a distinctive pattern of interaction between the family carers and the elders in this form of care encounter: on one hand, the family carers emphasise the elders' physical and mental infirmity and, on the other hand, they make them feel like a burden. They systematically emphasise, in the presence of the elders, that they are a 'dead body' or that they do not have 'mental sanity anymore', and they complain that providing care implies a lot of work which they do unaided most of the time. The following testimonies are illustrative: Certainly as a consequence of these actions, the elders feel like a burden, and this also makes them feel bad and sad: Elder: You see the work that is needed! I need two persons to get me out of my bed. It's a lot of work! Researcher: How do you feel about that? Elder: I feel bad and sad as you can imagine, I feel that I demand a lot of work. (75-year-old man) It should also be stressed that the elders abdicate from exercising autonomy, letting their family carers decide for them.

Harmony
These care encounters include three cases where participants are the elders and the care workers, although these elders also receive care from family members who were not present in the observations. Nevertheless, it should be noted that the care workers have a very good image of the family carers. As one of the care workers said: 'Here there is a lot of family support, they buy everything the elders need' (Care worker 7, during a home visit). All the elders receive personal care, a bedridden woman receives incontinence care and bathing, and the remaining ones, a woman and a man, both not bedridden but confined to their homes, only receive help with bathing. All of them still have some physical strength and good memory, and they communicate without difficulties. The care settingsthe elder's bedroom in the case of the bedridden woman and the bedroom and bathroom in the remaining casesare pleasant. The elders' possessions (e.g. perfumes, hats, combing brushes) are visible in these settings. In addition, these settings have all the things the care workers need to undertake their work (towels, soap and diapers). These care settings fit the social construction of a 'good home', that is, clean and equipped with all the things the care workers need to provide care properly.
The care workers' actions are not fully task-oriented, as they talk with the elders about their past lives while they are providing care and they offer additional time after they have finished their work to talk about their biographies, or provide company, although they have strict guidelines about the time spent in each visit. The interactions between the care workers and the elders are dominated by tenderness, which is reciprocated. In addition to tenderness, the care workers show concern about the elders' needs and the elders show gratitude towards the care workers. They also protect the elders' privacy, acknowledge their presence (rather than ignoring them) and they respect their wishes. Furthermore, the care workers promote the elders' autonomy, and encourage them to take decisions about their care, as this field note shows: The care workers talk with the elder with a smile on their faces and with a pleasant tone of voice (in a sweet but not infantilising tone). They touch her body gently and ask her if she would like clean bed sheets. While they are performing her personal hygiene (washing and cleaning her), they talk with her about her last profession, something much appreciated by the elder. The elder also asked one of the care workers how her daughter was doing and they talked a bit about her. When the care workers were combing the elder's hair, with the elder's combing brush, they praised her hair and she liked to hear that. At the end, the care workers put some perfume on the elder, something that the elder was used to in the past, and they caressed her hand. The elder responded warmly to this. (Visit to an 82-year-old woman) Finally, it is important to mention that these elders meet the criteria of a 'good user', as conceived by the care workers during focus groups, which includes co-operating, not expressing aggressiveness and trusting the care workers.

Indifference
These five cases only have the participation of the elders and the care workers, with the exception of one case in which a family carer is also present (the elder's wife). However, all elders included in this form have family carers also. The care workers have the idea that the family carers do the minimum in relation to the elders; while they think that they do not neglect them, they do not think they provide strong support. Two elders, both women, are bedridden and the remaining three, two women and one man, are not bedridden but confined to home. They have some physical strength and good memory, and they communicate without difficulties. All of them receive personal care, including incontinence care and bathing.
The care settings are the elders' bedrooms, although in two cases they also include the living rooms. All the care settings are pleasant, with the exception of one in which there are cockroaches in the bedroom. This situation generates revulsion in one of the care workers.
The interactions between the care providers (care workers and the family carer) and the elders are task-oriented, and marked by indifference, that is, there is no tenderness but there is also no harshness. In one care encounter, the conversation between the care worker and the elder was very short and circumscribed to the practicalities of care: Care worker: Good morning! Elder: Good morning! Care worker: Lift your leg. Okay. Today it's done. Take care. Elder: Bye! (Visit to a 91-year-old woman) The care providers, in particular the care workers, provide care in a very 'mechanical' and cool mode (neither warm or cold): they undertake the care tasks in a quick way, talking very little with the elders, and when they talk the subject is related to the care tasks. The care relationship is formal, with mutual respect, including respect for the elders' autonomy but, at the same time, with emotional distance or detachment, as in the following field note: The care workers started doing their work very quickly (incontinence care) and they talked with the elder very rarely. This was the only conversation between the care workers and the elder: Care worker -Today this is bad (referring to the fact of the amount of faeces in the diaper). Elder -Heh, today this is bad! Care worker -Let's turn aside, please. Yes, that's it. Ok. It's done. Elder -Don't over tighten the diaper. Care worker -Okay, I'm not going to tighten it too much. There were no smiles and no caresses between the care workers and the elder. The care workers took 12 minutes with this visit, which was marked by emotional distance and coolness. (Visit to a 72-year-old man) Overall, the care settings, the elders and the family carers do not fit the social constructions of 'bad home', 'bad user' and 'bad family', but they also do not fully fit the opposite social constructions.
Forms of care encounters and the avoidance-evocation of the fourth age As described earlier, the social imaginary of the fourth age is constituted by frailty, abjection, care and loss of agency. Frailty is present in all forms of care encounters, although the components of 'otherness' and 'lessness' are not present in all forms, nor with the same intensity. Care is also present in all forms of care encounters but the way it is deployed varies according to the forms of care encounter. For example, if we consider the ethical/moral qualities that Tronto (1993Tronto ( , 2013 associated with each phase of the process of careattentiveness, responsibility, competence, responsiveness, solidarity and trustwe verify that these qualities were not present, or fully present, in all forms of care encounter. With regard to abjection and loss of agency, these elements are not clearly found in all forms of care encounter, nor are they necessarily linked to the same patterns of interactions and the same characteristics of the care settings. At a glance, we can see that the fourth age is resisted or maintained at a distance in the form of care encounters marked by harmony, and that it is brought to centrestage in the remaining forms of care encounter, although perhaps in a less pronounced manner in the form marked by indifference. In the harmony form of care encounter, the fourth age is kept at a distance, given that the care settings are pleasant and personalised, they are cosy and have the possessions that the elders need to sustain a respectable self-image (e.g. cosmetics and clothes). Goffman (1961: 20) designated these personal possessions as 'identity kit'. A pleasant and personalised care setting contributes to hold back the emergence of abjection and loss of identity. Furthermore, the interactions between the elders and the care workers are also pleasant, as they are characterised by mutual tenderness and respect. In addition, the care relationship is balanced in terms of power, as the elders and the care workers manifest interest in the lives (or past lives) of each other, the care workers do not make the elders feel as different (otherness) or with less value (lessness), because they do not ignore them, they protect their privacy and promote their autonomy. We can say that care is deployed with respect for the ethical/moral qualities described by Tronto (1993Tronto ( , 2013. All in all, the care workers focus their attention not only on the care tasks but also on the elders as persons with biographies, aspirations and preferences. This does not exacerbate frailty and does not produce abjection and loss of agency. Using Goffman's (1971) terminology, this form of care encounter preserves some 'territories of the self' which are essential to sustain a dignified self-image.
In the remaining forms of care encounter, the fourth age is brought to the centre-stage and this acquires its maximum expression in the form marked by conflict. In this form, the extreme dirtiness of the care settings and elders' bodies contributes inevitably to exacerbate frailty and to produce dehumanisation and abjection, which affects not only the elders but also the care providers. Frailty and abjection are accentuated by care providers' actions such as negligence (on the part of family carers), ignoring, not protecting privacy and not promoting autonomy. Care is deployed without the necessary qualities to ensure 'good care', as defined by Tronto (1993Tronto ( , 2013. Ultimately, this whole picture ends up curtailing the elders' agency and dignity, although they try to have some control over the care relationship when they do not co-operate with the care workers. Nevertheless, power imbalance is evident in these care encounters to the disfavour of the elders. Hence, the fourth age is omnipresent in this form of care encounter, due to the fact that conflict is not compatible with 'good care' (Tronto, 1993(Tronto, , 2013. Care implies compassion and 'tuning in' with the persons' needs (Heijst, 2011), something that is absent in these care encounters, which are dominated by separateness and animosity. The 'territories of the self' are seriously violated, leading to what Goffman (1961) designated 'mortification of the self'.
The fourth age is also evident in the infantilisation form of care encounter. This form has many similarities with the conflict form, but instead is dominated by actions taken by the care workers which infantilise the elders and by concomitant actions taken by the latter, manifesting subordination to the former. This kind of interaction devalues the social status of the elders, equating them with the status of babies who are characterised by total dependence on their care providers. This inevitably activates the components of 'otherness' and 'lessness', accentuating frailty and producing abjection. Another consequence of treating the elders as if they were babies has another negative consequence: the erosion of the elders' agency. In these care encounters 'good care', as defined by Tronto (1993Tronto ( , 2013, is also absent, and the power imbalances which disfavour the elders are also clear.
The exacerbation of frailty and the concomitant production of abjection are also found in the burdensome form of care encounter, although through different dynamics and processes. In these care encounters, actions performed by the family carers, such as emphasising the elders' physical and mental infirmity and the burdens of the care situation, exacerbate frailty and produce abjection. The elders explicitly state that they feel that they have little value and that they demand a lot of work. Therefore, family carers are far away from 'good care' (Tronto, 1993(Tronto, , 2013, although the care workers maintain a tender relationship with the elders. Another element of the fourth ageloss of agencyis also found in these care encounters, as the elders transfer the decisions about care to their family carers. Therefore, power imbalances are also evident in these care encounters which disadvantage the elders. Finally, in the form of care encounter marked by indifference, the interactions between the care providers and the elders are task-oriented, mechanical and cool. The care providers only take the time that is absolutely necessary to carry out the care tasks. Although there is mutual respect in these care encounters, the care workers are not fully attentive to the elders' needs, nor fully concerned about the effectiveness of the care they provide. In this vein, they do not meet all the qualities of 'good care' (Tronto, 1993(Tronto, , 2013. It is suggested by Tronto (2013) that 'good care' requires sufficient time and proximity between the carers and the elders, something that is absent in these care encounters. But perhaps what is more important regarding this form of care encounter is that indifference ends up producing some level of objectification, which affects not only the elders but also the care providers. This objectification ends up producing a sense of 'otherness' and 'lessness' and, consequently, abjection.

Discussion
Starting with the big picture, the findings reveal several forms of care encounters, being the result of different combinations of aspects related to the care settings and with the actions and interactions of the persons involved in the encounters. Most of these aspects were already identified by previous studies on the care relationship in the context of home care, such as dirtiness/cleanliness of the care setting and the care-receiver's body (e.g. Twigg, 2000;Isaksen, 2002;Wibberley, 2013), respect and trust (Trojan and Yonge, 1993;Tadd and Calnan, 2009), ignoring/not ignoring (e.g. Tadd and Calnan, 2009), privacy (e.g. Tadd andCalnan, 2009), autonomy (e.g. Valokivi, 2005;Tadd and Calnan, 2009;Lloyd et al., 2014) and feeling a burden (e.g. Tadd and Calnan, 2009). Other aspects less explored (or not explored at all) by the aforementioned studies were also identified in the present study, such as family negligence, task-oriented versus person-oriented actions, and the manner through which the interactions are established (showing harshness, coolness, tenderness).
Still looking at the big picture, we verify that the social imaginary of the fourth age is brought to centre-stage by all forms of care encounters, with the exception of the harmony form which keeps this imaginary at a distance. The fourth age pervades almost all forms of care encounter because conflict, infantilisation, burdensomeness and indifference compromise 'good care' (Tronto, 1993(Tronto, , 2013 and produce a sense of 'otherness' and 'lessness', abjection and loss of agency. The fourth age is kept at a distance in the harmony form, because the attention is not focused on the elders' infirmity, but rather on the elders, as persons with biographies and current aspirations and preferences. There is warmth, respect and trust, as well as reciprocity and a balanced distribution of power in these care relationships. Looking now at the more specific components of the results, we can verify that the way through which care encounters are forged results from micro-dynamics and structures. It was demonstrated that the conditions of the care setting (cleanliness/dirtiness, having the things that the care workers need) play an important role. The homes involved in the conflict form of care encounter fit the social construction of a 'bad home' and this negatively interferes with the care workers' actions. As suggested by other studies, dirtiness produces a disruptive effect on care relationships (Isaksen, 2002) and puts at risk the care workers' wellbeing (Wibberley, 2013). Regarding this, it is important to underline that in one of the two cases included in the conflict form, the cleanliness of the care setting improved during the last months of field work, and at the same time the care workers' disposition and mood also improved with positive repercussions on the manner of care provision. This confirms the importance of the care setting and conditions to the manner in which care workers perform their work. Still at the micro-level, the results also show that all participants in the care encounters have an important role in the way care is done. With respect to the elders, despite the power imbalance in their disfavour that exists in any care relationship (Twigg, 2000), this study demonstrates that their actions (or lack of action) contribute to the way the care encounters unfold. The relevance of the elders' actions is particularly evident, for example, in the conflict form of care encounter. Interestingly, the findings also reveal that the elders who have high levels of physical and mental infirmity, particularly those who have great communication difficulties, are involved, exclusively, in the infantilisation form. This also confirms that it is difficult to keep the social imaginary of the fourth age at a distance when the care relationship becomes what Higgs and Gilleard (2016) call a 'relationship of one'. In this respect, Higgs and Gilleard (2015b: 1) argue that care 'should concentrate less on ambiguous and somewhat abstract terms such as personhood and focus instead on supporting people's existing capabilities, while minimising the harmful consequences of their incapacities'.
In what concerns care providers, the findings show that the care workers play an important role in the way the care encounters unfold, and ultimately in the possibility of maintaining, or not, the social imaginary of the fourth age at a distance. This is evident in all forms of care encounter, albeit less decisively in the burden form. However, the findings indicate that family carers play an even more decisive role in this respect, whether or not they are present when the care workers make their visits. This becomes evident when we compare the conflict form and the harmony form. In the conflict form, the family negligence towards the care settings and the elders clearly interferes negatively with the care workers' disposition and mood, and most likely also on the elders' disposition and mood. This, in turn, interferes negatively with the relationship between the elders and the care workers. The harmony form of care encounter shows us the opposite situation. The crucial role of the family carers is also clearly evident in the burden form. Therefore, in the cases here studied the family carers' role is more decisive than the care workers' role, because they can interfere not only with the conditions of the care settings and the elders that care workers encounter when they arrive at each house, but also through the interactions that take place during home visits. In fact, the influence of family carers in care encounters does not even require their presence in these encounters.
Nonetheless, these micro-dynamics and structures are embedded in meso-and macro-dynamics and structures, more specifically in the organisational context of home care and the public policy context. For example, if the care workers had had vocational training regarding the relational component of care, they may not have reacted negatively to the aggressiveness of the elders, nor have infantilised them or interacted with them with indifference. Previous studies suggested that the culture and rationality of home care services are inevitably taken into the home visits (England and Dyck, 2011). In turn, if the family carers had had information and education about dignity in care, family negligence (of the care settings and the elders), family insensitivity and family indifference may not exist.

Strengths and limitations
The prolonged engagement in the field and sustained observation are two strengths of the present study, as a significant number of observations were conducted over two years in each home. This enhanced the richness of data and the robustness of analysis. Nevertheless, the present study also exhibits some limitations. First, the findings refer to social contexts characterised by low incomes and low levels of education. Second, focus groups were conducted only with the care workers, as it was not possible to schedule focus groups with family carers. Third, member checking was conducted only with the care workers, which may have some implications for the trustworthiness of the findings.

Conclusions
This study contributes to developing empirical applications of the theoretical perspective of the fourth age. Care encounters at the elders' homes can take different forms and these contribute differently to the avoidance or evocation of the social imaginary of the fourth age. This confirms the 'Janus-like' nature of care.
The way home care encounters are forged, and the associated presence, or not, of the social imaginary of the fourth age, depend on the conditions of the care settings and the actions of all participants. The present study supports the claim made by Higgs and Gilleard (2016) that it is difficult to maintain the social imaginary of the fourth age at a distance when the elders exhibit high levels of infirmity. It also reveals that in the context of home care, family carers play a crucial role in the way care encounters unfold. However, these micro-dynamics and structures are embedded in meso-and macro-dynamics and structures, among which the lack of vocational training regarding the relational component of care and educational programmes for family carers about dignity in care will play a significant role. These have implications for practice and policy.
Note 1 In social sciences in Portugal, there are no research ethics committees at universities, research organisations and social care services, although this situation is changing. Hence, the study was conducted in accordance with the code of ethics of the Portuguese Sociological Association and the ethical sensitivity of the researcher. The initial informed consent from the elders was updated during the fieldwork, especially at times when the investigator was present when intimate care was provided. Nevertheless, the observations at the elders' homes were conducted with discretion to protect the elders' privacy. During the fieldwork, the researcher witnessed some episodes of elders being neglected by their family carers. Faced with this, the researcher suggested to the care workers that they should report the situation to the co-ordinators of the home care services, who made the necessary interventions to prevent negligence in the future.
Financial support. This work was supported by National Funds provided by the Foundation for Science and Technology (FCT) (project UID/SOC/04020/2013).