Post-Human Families? Dog-Human Relations in the Domestic Sphere

Introduction

1.1 I begin this article with 3 contrasting views on dogs and the practices whereby their lives are entangled with those of humans. In his memoir, Sleeping with Dogs, Brian Sewell writes of Penny, the dog with whom he had 'grown from boy to man':

She had never been my toy, nor had there been anything of the brotherliness of my relationship with Prince, nor did I ever think of her as an adopted child satisfying a frustrated parental urge; she was my dog and we were man and dog in a primeval bond, and that was that – though I believe that I was a kinder and more considerate boy for having her, and a more compassionate man. (Sewell 2013: 25)

1.2 In this passage Sewell asserts a connection which recognises the differences between dogs and humans and involves compassion, kindness and consideration – positive qualities that are a direct outcome of closeness between him and his dog. He also criticises the idea that dogs can be commodities (therefore disposable) or substitute children.

1.3 The other views appear in the Huffington Post and take opposing positions on the relationship between companion animals and their humans: one takes exception to the idea that humans are 'parents' to 'pets' while the other sees nothing wrong with it. Steven Kurlander asserts that, 'It's really stupid to equate parenting with pet ownership… Pets, and animals too, are not humans, and should not be considered or treated as such' (Kurlander 2015). Marie Carter, in contrast, argues that 'Our pets should be classified as members of our family and not bundled in with the rest of our property. We are pet parents and not pet owners, after all!' (Carter 2015). Kurlander objects to the humanising of animals and asserts that the pet-human relationship is one of ownership rather than parenting. Carter, in contrast, asserts that it is time this legal definition was overturned; she backs up her argument with evidence that people regard their pets as members of their families.

1.4 In this paper I address some of the issues raised in these extracts. I explore the ways in which dogs and other companion species^[1] become family members and engage with the argument that this indicates the emergence of post-human families. In order to do this I first review the literature on families and companion animals, situating it in the context of an increasing interest in the post-human. I then ask what a post-human family would look like and whether it makes sense to talk in such terms, before presenting my own empirical data on the ways in which humans and dogs live with each other and the 'daily practices of kinship' which constitute them as kin (Charles and Davies 2011: 89). Finally I reflect on the usefulness of the term 'post-human' to understand the practices of human-animal kinship.

Post-human families?

2.1 There is an increasing interest in post-humanism as a philosophical approach which challenges both Humanism and the meaning of the category 'human'. It is particularly influential amongst researchers exploring human-animal and human-technology relations (see for e.g. Haraway 2008) and can broadly be understood as a displacement of the centrality of the human and a recognition that the non-human is an essential part of (post)-human life. This approach challenges anthropocentrism and the idea that humans are superior to and different from all other living creatures. It also unsettles those disciplines which look at human activities as if they take place in a purely human sphere. Sociology is no exception: for most of the 20^th century it took society as its focus without recognising that animals are incorporated into social relations with humans, that 'human' societies would not have developed in the way they have if this had not been the case (Shipman 2011), and that the anthropocentric social relations in which animals are entangled systematically disadvantage them (Carter and Charles 2013). It also regarded families as purely human affairs.

2.2 Post-humanist approaches to families, in contrast, recognise that animals are integral to many families and ask whether the emergence of post-human kinship relations can be identified empirically. Thus scholars have suggested that the existence of multi-species households and the increasingly widespread practice of 'pet keeping' in post-industrial societies are indicative of a shift to post-human sensibilities and kinship practices (Franklin 1999;Cudworth 2011). There has undoubtedly been a rise in popularity of pet-keeping and, to take dogs as an example, the dog population in the UK almost doubled from 4.4 to 8.4 million between 1963 and 2016 (Franklin 1999:89;PFMA 2016). There has also been a move towards keeping animals inside rather than outside the house and it is now commonplace for dogs to share intimate spaces within the home including beds (Thomson et al 2014). And even though households with children at home are more likely than other types of household to include companion animals, anthropomorphism and levels of attachment to these animals are higher in single-person households and those where children have left home (Hart 1995).

2.3 There is also a growing body of research showing that companion animals are increasingly defined as family members (Charles 2014). An Australian study (Franklin 2006) found that 88% of respondents ascribed family status to their companion animals while in the US the proportion is 91% (Harris 2011). Qualitative studies have found that animals are considered family because they provide emotional support and because they need humans to care for them (Charles and Davies 2011; Power 2008). Children especially regard pets as kin (Tipper 2011) and recent research shows that children are 'more likely to confide in pets than siblings' (Coughlan 2015). Indeed people of all ages report feeling closer to their dogs than to other family members (Cohen 2002; Pew 2006) and dogs are particularly important in providing support at levels 'comparable with' that provided by 'human relationships' (Bonas et al 2000:232; Enders-Slegers 2000; Hart 1995).

2.4 It is not clear, however, that an increase in pet keeping and the emotionally close relationships that can develop between humans and companion animals indicate the emergence of post-human families. Moreover, there is little agreement about how kinship relations can be defined as post-human. Here I look at some examples of how this has been attempted to tease out the different approaches.

2.5 Part of the argument that post-human families can be identified empirically rests on the alleged newness of the human-animal relations characterising contemporary, multi-species households. Franklin, for instance, argues that, since the 1970s, there has been a move to new forms of intimacy between humans and companion animals which results in 'hybrid' households (Franklin 2006). It is, however, contentious to claim that pets being 'companionate family members' is a new phenomenon (Franklin 2006:138); there have been intimate connections between humans and other animals, especially dogs, for hundreds if not thousands of years. Indeed, '[i]t was not until the sixteenth and seventeenth centuries that the farmhouse became an exclusively human residence' and, prior to this 'there was … an intimacy and interconnection between humans and animals (wild and domesticated) which it is now hard to comprehend' (Benton 1993: 73). And during the 18^th century the practice of giving pets human names and keeping them indoors became widespread amongst the middle classes, especially in urban areas (Thomas 1984). Franklin argues, however, that since the 1970s the family 'has affected a significant shift from a humanist to a post-humanist form' (Franklin 2006:139). In similar vein, Smith writes about her mode of living with rabbits and her attempts to perform human-pet relations that are not based on dominance and subordination (Smith 2003).

2.6 In Franklin's and Smith's studies we find two different definitions of post-human families: one that is based on the existence of new, 'companionate' relations between humans and animals (Franklin 2006) and the other that explores the possibility of performing human-animal relations so that the needs of both species are met (Smith 2003). These ways of understanding human relations with companion animals are part of an attempt to grasp changes in human-animal relations which are said to be characteristic of post-modernity (Franklin 1999). It is argued that we are now in an era where human exceptionalism is being challenged, a more emotional attachment between humans and other animals is emerging and that this is reflected in family practices. For Franklin, the claim that companion animals are family indicates both a breakdown in the species barrier and the 'surrogacy of animals for significant human relationships' (Franklin 2006: 142).

2.7 Power builds on these understandings of post-human households, distinguishing between humanist and 'more-than-human' 'models of family' (Power 2008: 541). For her, humanist models shape and mould dogs to fit in; dogs are either incorporated into families within a caring, parental relationship or as pack members and, in either case, relations of dominance and subordination prevail. The 'more-than-human' model entails an accommodation of the 'dogginess of canine companions' (Power 2008: 541) with dogs actively participating in the 'everyday practices' of family and thereby contributing to the shaping of home (Power 2008: 536). It also involves a change in everyday practices of kinship such that 'both people and dogs [are] altered' and their needs are catered for (Power 2008: 552). For Power, dogs are incorporated into families as dogs rather than because they are seen as 'furry humans' but the ways in which this is accomplished both reinforces and challenges anthropocentrism; it therefore does not necessarily mean a 'post-humanist abandonment of binaries', although it may trouble them (Charles and Davies 2011: 73).

2.8 Such troubling is associated with companion animals occupying a liminal space, sometimes being regarded as human and sometimes animal (Fox 2008). When they are regarded as animals they become disposable and a dog may be ejected from the family if domestic circumstances change or they behave in ways that humans deem to be unacceptable (Power 2008). This not only sits awkwardly with their status as kin (Fudge 2008:109) but also renders that status provisional (Shir-Vertesh 2012) and reveals the power differentials characterising human-animal relationships (Power 2008). Similarly, it has been argued that while pets are grievable they are replaceable, thus confirming their ambiguous status as both human and non-human and how they 'can be regarded as persons included in a human moral community while human exceptionalism remains intact' (Redmalm 2015: 32).

2.9 Emerging from these discussions are 3 arguments that are important for the question of whether we are witnessing the emergence of post-human families: (1) that close connections with companion animals disrupt human exceptionalism because companion animals occupy a liminal space between human and animal; they are understood as possessing 'human' qualities and are also regarded as 'other' and therefore unknowable. Animals' possession of 'human' qualities questions the uniqueness of these qualities as human while their unknowability undermines the idea of an all-knowing human subject - both, in different ways, question human superiority (Fudge 2008); (2) that companion animals are precariously incorporated into families and this precarity or disposability reveals the power relations underpinning their incorporation and the persistence of human exceptionalism (Redmalm 2015); (3) that extending the category of the human to include animals is merely an extension of humanism and does not challenge human superiority because whatever is included within the category 'human' is superior to the excluded 'other' (Braidotti 2013). From this it would seem that post-humanism does not simply mean multi-species households, if it did there would be no question that increasing numbers of families in western societies are post-human. More importantly it means establishing relations with companion animals that are not based on a human-other distinction. In what follows I argue that practices of kinship blur the human-animal boundary, that this blurring takes place in the context of unequal power relations which are an inevitable consequence of dogs' incorporation into families as dependents (Carter and Charles 2013; Smith 2003; Power 2008), and that, although post-human kinship practices can be identified, it is not clear that they give rise to a post-human form of family.

The study

3.1 To develop this argument I draw on two sets of data: responses to a Mass Observation directive on Animals and Humans which was sent out in the summer of 2009, and 21 interviews carried out with people who shared their domestic space with companion animals. The Mass Observation Project is based at the University of Sussex, UK. It was established in 1937 and its present incarnation dates from 1981. A panel of around 500 correspondents^[2] respond to 3 directives a year which prompt them to write about different topics based on their own experiences and observation. It is 'part history project, part anthropology, part auto/biography and part social commentary' and is written by 'ordinary people' rather than experts (Sheridan et al 2000: 12). Responses are housed in the Mass Observation Archive in Sussex. In the summer of 2009 a directive called 'Animals and Humans' was sent out and in this article I draw on the 244 responses. Two thirds were from women and one third from men with the ages of correspondents ranging from 16 to 90; this reflects the composition of the MO panel at the time (MO 2009). I subsequently (over an 8-month period between 2011 and 2012) conducted 21 semi-structured interviews with 19 women and 12 men whose ages ranged from the early 20s to 80. Interviewees were contacted through a veterinary practice, the Dogs Trust (a national shelter for homeless dogs), Guide Dogs for the Blind and by means of a snowballing sample in a local city. In the interviews, which were semi-structured, I followed up issues that had been raised in the responses to the Animals and Humans directive, focussing particularly on what it is that makes animals family.

Kinship practices

4.1 One of the ways in which animals become family is through practices of care (Power 2008; Charles and Davies 2011). Indeed the wellbeing of companion animals 'is acknowledged to be an object of direct moral obligation on the part of human members of the household' (Benton 1993: 64). Caring involves responsibility for another living being and this sense of responsibility contributed to animals becoming family members. One of the correspondents to the MO directive wrote about the difference between a pet and other animals:

A pet is different from other animals because it usually shares your home and garden and is part of your everyday life. Also you are responsible for it – you have to care for it and feed it and take care of its health. (J1890, F)

4.2 The idea that having responsibility for another living being and having to care for them was what made them family was widely echoed in the interviews. This responsibility meant that the needs of an animal had to be taken into account. One woman, for instance, spoke about the dog they had when the children were small.

[He] was almost like a third child in a way I suppose because he grew up with them, the two children and him, and … when we went out we would still have to think about where he went and if he came with us or not. (K005)

4.3 He became 'like a third child' not in the sense that he was a child substitute but in the sense that his needs as well as those of the children had to be catered for in family arrangements. Others spoke about changing their holiday practices once a dog became part of their family.

And, you know, before with just the two of us and obviously no children at home you could just pretty much do what you wanted when you wanted, but obviously with a dog you can't. And actually we were saying, you know, it would be quite nice next time we go abroad, you know, maybe to take him with us and that would mean getting the ferry and driving which we haven't, wouldn't normally do because obviously that takes time and it's not quite so easy, but we think we'd do that. Because he is part of the family really so he should be with us as it were, not all the time but if possible. And we've certainly had a couple of short breaks away where we deliberately chose to go somewhere where dogs are accepted… . (K002M)

4.4 The responsibility of caring for a dog, the need to take another living being's needs into account, and the simple fact that many people prefer to go on holiday with rather than without their dogs, changes their family practices so that they become inclusive of their canine family member.

4.5 The dependence of dogs led some to draw analogies between them and children and suggests that there are elements of a parental relationship between humans and their dogs (Power 2008).

Dog parents?

5.1 That the relationship can be understood as parental^[3] is supported by evidence that dogs could fill the gap created when children left home. One woman wrote that with her children gone she felt 'very lonely' and 'decided it would be nice to have a dog for company'. She recounted the story of how two poodles came to live with her and then wrote:

The three of us were very happy for about 10 years. We all used to go to bed together. They slept together by the side of my bed. We were all the best of friends. …When they got very ill and their lives came to an end it was two of the saddest days of my life. I had as much love for those little dogs as I did for my children. (S496, F)

5.2 In this case the dogs may have been surrogate children but others made it clear that even though their children leaving home had resulted in a changed relationship with their dogs, the dogs had not taken the children's place. One woman told me how, when her children grew up and left home, there was a shift in the relationships in her family with the dogs moving up the hierarchy (cf Fox 2008).

I mean perhaps not they become more of the family but they almost become your, I think you have to be careful they don't become too much your children in a way, I think it shifts a little bit when you've only got the dogs because when the children were living at home all the time the dogs were beneath the children and I think once the children had gone they come up a level actually in your thinking. (K005)

5.3 Care is taken here not to position the dogs as substitute children. This is also apparent in the practices surrounding birthdays and Christmas.

[W]e don't mark their birthdays at all, Christmas they usually get one present, a chew or something wrapped up, I think we started that with Tom [dog] really and we've ended up, I think because of the children and it was always, but certainly no, we didn't mark birthdays, I think that's part of them not being, I don't know, I mean I'm sure people do but we've never done. …They're dogs. (K005)

5.4 Practices, however, varied considerably. One woman recounted the lengths she had gone to in order to estimate her rescue dog's birthday so that it could be celebrated, and another heterosexual couple told me that they gave all their animals birthday cards.

5.5 These practices incorporate dogs into families as dependents for whom adults have responsibility in a way analogous to children but many stressed that dogs are valued because they are dogs and several said that they preferred dogs to children.

I think it's part of the family group, but I think it's still important that they're dogs and they're not children, but they still have an important role but in a different way, I don't hold with all this, you know all this stuff of putting, I don't know, these clothes on, all - these funny things that make them not dogs. (K005M)

5.6 This relationship which is one of responsibility for another creature who is dependent on you implies that 'differentials of power and dependency are intrinsic to the relation between a pet-owner and a pet' (Benton 1993: 146). Indeed, as Tuan argues, pet-owner relationships share similarities with parent-child relationships in so far as both are marked by dominance and affection (Tuan 1984).

Blurring boundaries

6.1 It could be argued that in order for families to be considered post-human, these power differentials between companion animals and their humans would need to be challenged and the distinction between human and other animals, which reinforces human superiority, questioned. But to what extent is a lack of distinction reflected in the way humans relate to dogs who are family members? Are dogs regarded as similar to their human companions or is the relation one of difference? And if it is one of difference does this imply human domination? The dependency of dogs and the requirement, both practical and moral, that they are cared for appropriately, suggests that although they are family members they occupy a subordinate position. The situation is, however, subtle and nuanced. One woman wrote that her dog was not the equal of her husband and herself, and yet she also writes of the dog owning their shared home.

We just have the one dog who we share our home with, although actually he seems to be the owner. He has the run of the house with the exceptions of the bedrooms and bathrooms … The relationship is different than that of mine to my husband, we try to treat each other as equals and certainly the dog is not that. (P1796, F)

6.2 This implies an ambivalence about the relative status of humans and dogs; in some ways they are not equal while in other ways they are like the 'owner', a reference, perhaps, to how caring for the needs of another may also be akin to serving them. Indeed, she writes at the end of her response that her 'eldest son does get quite cross with us sometimes for the way we treat Dillon he says but Mum he's only a dog!' (P1796, F). Their treatment of the dog, in the eyes of their son, elevates his status, i.e. it raises him up the family hierarchy, which their son thinks is inappropriate. There is also an implication that he thinks his parents are humanising Dillon.

6.3 It has been argued that anthropomorphism is essential for meaningful relationships to have developed between humans and animals (Serpell 2005) and it is certainly the case that people attribute so-called human characteristics to their dogs and other companion animals (Sanders 1999). However there is great care taken by both the MO correspondents and the interviewees to distance themselves from anthropomorphism, part of which involves keeping a distance – physical and emotional - between humans and animals.

6.4 For one of the MO correspondents, it was important to maintain the human-animal distinction and she did this by having her dogs sleep outside.

They [the dogs] are allowed in the house (but not upstairs). However, they sleep in an outhouse/ stable block. They are happier not being in a stuffy house; their coats are thick and glossy – and the distinction is maintained between animals and humans. (M3412, F)

6.5 This supports the idea that animals sharing intimate spaces with humans challenges species boundaries. Interviewees reported their attempts not to let dogs get on chairs or beds in an effort to maintain the human-animal distinction. In one interview, the blurring of boundaries was explicitly mentioned.

To me it's fairer on you and on the dog if you try and maintain that distinction, it can get quite - of course - particularly if you're sort of, you know, sitting on the sofa and the dog's alongside of you and you're sort of stroking the dog and you find yourselves talking to the dog about what's on the TV, that's quite interesting sort of, a bit of blurring. (K002M)

6.6 Clearly these family practices – watching TV together and commenting on the programme – blurred the animal-human boundary which this interviewee was keen to maintain. This was also apparent when the same couple were asked whether they ever gave or received cards on behalf of the dog. He said that they did not but she reminded him – saying 'liar, liar, pants on fire!' – that he had in fact given her a card on her recent birthday signed from the dog. There is a clear feeling that such practices could be judged as being of dubious moral value (or insufficiently masculine) and that it may be better not to admit to them, precisely because they blur the boundary between dog and human.

6.7 By the same token, efforts to avoid what is seen as anthropomorphism were not always effective. The same couple reported that they had tried to avoid a human name for their dog because 'it's not a human … and yet we ended up choosing Dylan but it just seemed right' (K002M). The man went on to stress that although the dog was part of the family he was not human.

He's one of the family but he's not a human. …He's a dog, yeah, I mean I think that's quite important to keep that distinction because I think if you treat him as a dog he will be happier and he will fit into the family better. He's not a human and I think that's quite an important distinction. (K002M)

6.8 He wanted to 'respect him as a dog and not [try] to make him something he isn't' (K002M). This entails respect for his dog-ness and could be seen as displaying a post-human sensibility. On the other hand it could be seen as a rationalisation of their relationship with their dog in light of the perceived moral disapprobation attracted by the admission of too close a relationship with companion animals and the blurring of boundaries that he has already admitted to in their kinship practices. This sort of response was more common in interviews than in the written MO responses where there was more openness about the intimacy of relationships with companion animals (Charles 2016).

6.9 Many, however, maintained an uncompromising belief in human superiority. One MO correspondent, for instance, commenting on the idea that a dog is a person's best friend, wrote:

I regard humans as being superior to animals – so given the choice of a dog or a man as my best friend – it will be the man. (C3603, M)

6.10 There were differing views about the superiority of humans over dogs but many recognised that family practices could either reinforce or blur the dog-human boundary. Keeping a dog outside reinforces boundaries while letting them on the settee blurs them. And in the interviews especially people were at pains to distance themselves from anthropomorphism. There seems to be a difference between what is publicly acceptable and practices that go on in private, reminiscent of Cornwell's distinction between public and private accounts (Cornwell 1984), and while kinship practices often blurred the human-animal boundary, in their public accounts many interviewees claimed to be maintaining it.

Selfhood and agency

7.1 One of the things that disrupts the category human is evidence that animals share many qualities that have been defined as uniquely human. Thus it is argued that dogs (and other animals) 'have elements of a core self that becomes present to us through interaction with them' (Irvine 2004:3).^[4] They are seen as 'having a mind, beliefs, and desires, just as we do' (Irvine 2009: 332) and, because they 'have agency and the other dimensions of the core self, they can choose courses of action' (Irvine 2009: 337). Dogs are related to as 'quasi-persons' (Benton 1993); they are individuated, experienced as actors exercising agency and as having an understanding of other creatures' intentionality. This way of engaging with animals can be understood as post-human in so far as so-called human characteristics are not confined to biologically human beings and was a characteristic of the descriptions of dog-human interactions in my samples.

7.2 In these descriptions dogs appeared as individuals with distinct characters and personalities which were observed in detail. One interviewee talked about his dog's personality with the following words: laidback, calm, friendly, mischievous, inquisitive, playful, obedient, stubborn - terms which could just as easily be applied to a human person (K002M) and which are 'vitality affects', 'important vehicles of the core self' (Irvine 2009:335). Another interviewee described Mollie who,

was old, cantankerous … did what she liked, …she was not a dog, she thought she was human. (K001M)

7.3 The idea that Mollie 'thought she was human' implies that she had human characteristics and perhaps was entitled to be treated as a human. It also points to the liminal spaces that companion animals can occupy, being neither human nor animal or being human and animal at different times (cf Fox 2008). Referring to animals as 'almost human' operates in a similar way. One woman wrote about animals wanting to do everything that human family members do and, in the process, becoming 'almost human'.

Animals that you love are definitely part of the family. They just gradually become part of the family so that in the end they are almost human! They want to do everything that you do and even sometimes to share your bed! (J1890, F).

7.4 Some suggested that it was because dogs had characters and personalities, that is they were persons, that they were part of the family. One MO correspondent wrote about the pets she lived with as a child:

They might not have been equal members of the family but they were still part of it. …my family and I have often felt that our pets have been part of our family. However this has only really happened with our larger pets, i.e. our cats and dogs. I don't know why this is, probably because you get more of a reaction from them and they all seem to have their own characters. (G3988. F)

7.5 This comment alerts us to the importance of an animal's ability to respond. Their ability to communicate is an aspect of their personhood and an indication of the interior life that many who live with dogs and other animals recognise (Fudge 2008).

7.6 Many not only experienced dogs as persons in their own right but also as having a sense of self. Irvine argues that a sense of self requires agency which 'implies subjectivity, in that an agentic being… has desires, wishes, and intentions, along with a sense of having those things' (Irvine 2009:332). There were many reports of dogs shaping interactions with their humans. This was commented on by a woman who had a small, rescue dog, Danny.

He's got to be with you. I mean if you're sitting having your breakfast in the kitchen he's got to be on your lap which is a blooming nuisance when you're trying to eat breakfast, say 'no Danny wait' and then he bumps you with his nose if he doesn't get your attention. No, he's got to be on your lap which is very strange but, oh no, so he has to go out for his walk now, 'no, I'm having my breakfast, you just wait'. (K011)

7.7 This quote points to the active engagement of dogs with their human companions in order to get what they want and supports the claim that they have a sense of self which involves agency, intentionality and mind (Irvine 2004). This small dog 'bumps you with his nose if he doesn't get your attention' and insists on being on his human's lap when she is eating breakfast even though this is something that she finds a nuisance. There were many examples like this, in both the written responses and in the interviews. Several interviewees spoke about how their dogs did not like it when voices were raised; this usually resulted in them trying not to have arguments in front of their dog and, in some cases, dogs would intervene in order to stop the argument (see Carter and Charles 2013).

7.8 This suggests that dogs have desires and wishes that they make known. They are active agents in their engagements with their human companions though they exercise this agency in the context of social relations which place them as dependent on their more powerful human companions (Carter and Charles 2013).

Emotional connections

8.1 Fudge argues that relationships with pets are 'potentially compassionate' and that imagination is required to feel compassion for another living being (Fudge 2008:67). This is reminiscent of the ideology of kindness which emerged in 19^th century America as a means of civilising children; ensuring that they were kind to animals was one way of doing this (Grier 2006). It has also been argued that the connection with animals has contributed to the development of qualities such as 'compassion, empathy and communication skills' and that they are endangered when humans do not live with animals (Shipman 2011:275). The view that animals make 'us' better people was expressed by MO correspondents and interviewees as well as Sewell who was quoted at the beginning of this paper.

I love animals! I think they are vital as family pets and teach us a lot about love, loyalty and responsibility. (M3412, F, 50)

8.2 As well as making humans better people, dogs often provided emotional support which is one of the elements that constitutes kinship (Charles and Davies 2011). A young man wrote:

In my late twenties I went through some mental and physical trouble which I tried to hide from my family. But I used to talk to Sandy [his dog] about my issues and she saw the real me. I think she knew I wasn't very well at the time. I lost interest in life and my well being for a few years but my devotion to her never wavered. Her love back was unstinting as always – I keep a picture of her in my bedroom. (H3784, M)

8.3 Here we have the idea that Sandy knows him better than his human family does – 'she saw the real me' – a view which was echoed in other accounts. And a man who had separated from his wife wrote about the support he got from his dog.

The best relationship I ever had with an animal was my dog. …He helped me get over the break up of my marriage just by being around while it was all happening… Relationships with animals are different from those with people because you can rarely trust people absolutely. You always know what a dog is thinking. Dogs have no hidden agenda. (M4463, M)

8.4 There is also evidence in these accounts that animals feel an emotional connection to their humans. The young man above talked about Sandy's love for him and another man recounted a heartrending story of how a German Shepherd who had accompanied him and his wife on a move from Latin America to Britain subsequently pined to death. 'Sadly' the writer moved to a new posting in the Far East and had to leave the dog behind with his mother.

I was told that the moment my wife and son left the house he lay beside some luggage we had left behind and simply gave up on life, dead within the month. I have since grieved over the death of that same child, but scarcely more deeply than I did, and do, over Lucky. (R2143, M)

8.5 Such understandings of dogs loving their humans are supported by scientific research which makes it difficult to dismiss them as anthropomorphism. Ethologists have shown that dogs form affective bonds with particular humans which demonstrates the 'reciprocal nature of human-animal interactions' (Topal and Gacsi 2012: 182). Furthermore, they suggest that families are 'mixed-species groups [which] should be regarded as natural entities. Attachment between dogs and their caregivers is an indispensable characteristic of this social system that has a bi-directional nature' (Topal and Gacsi 2012: 181).

8.6 Emotional connection means not only that dogs pine and die because their people leave them but also that people grieve when dogs die. One of the MO correspondents wrote:

Dogs especially can easily become part of the family and it is for this reason that we can sometimes grieve so much when they go to the great lamppost in the sky. We talk to them, we understand their thoughts and intentions in the same way that they often understand ours, and some people (my late father included) allow them to share the bed (on top of the blankets, I hasten to add) and lick their faces. (D1602, M)

8.7 But as well as grieving a dog's death, people may also get another dog. This is understood by some as indicating that dogs are replaceable and that this contradicts their status as family members. For many, however, an individual dog is not replaceable; it may as well be argued that someone whose human partner dies and subsequently re-partners is replacing them. The fact that a human partner is 'replaced' does not make that partner any less human but, in the case of dogs, it is said to demonstrate their liminal status (Redmalm 2015). Moreover dogs are often not 'replaceable' as this man's response indicates.

Our family had a dog (a nondescript mongrel) when I was a lad. Officially it was my sister's dog – but I spent most time with it. I talk [sic] it walks on a daily basis – and fed it. Sadly, it started to have fits – and had to be put down. I took it on its last walk to the vets. Although that was over 50 years ago – I still remember the anguish I felt.

It was so horrible to be parted from 'Scamp' that I have never owned a dog since then. I love dogs – but cannot bear the thought that they don't live for ever. (C3603, M)

8.8 Furthermore, there were examples of dogs 'replacing' humans who have died which seems to have happened in the following case.

When I was growing up my mother had 2 pugs – Poppy and her less appealing son Percy. Poppy had been given to my mother to cheer her up after the death of my older sister and she, the pug, was an incredibly important member of the family… Poppy lived to 17 which is very old for a pug. My siblings and I were on holiday with my father when she died and we were all devastated – even my father. She is buried in the orchard at my childhood home – it was one of the things my mother most minded about when she sold the house. (A3434, F)

8.9 Here the emotional attachment to the dog can be seen not only as an attachment to her but as a connection with the dead daughter and the selling of the house is so upsetting because it symbolises the severing of that connection.

8.10 Sometimes dogs are expendable even though they are part of the family. This is explained in terms of the relationship with them being different from the relationship with a human family member.

In a sense our animals were an extension of the family, but always secondary to it and when the dachshund developed incontinent [sic] problems, there was no question that she had to go, since the safety of our two young children was paramount. Of course it is a different relationship. (W2322, M, 65)

8.11 And, as a cursory glance at a dog rescue site will show, changed family circumstances are often the reason that dogs are looking for new homes. But often they are not expendable, even when they are difficult. As one interviewer said:

[Y]ou can't choose your family, you can choose your friends and it's a bit like that with, well, certainly with Flo … she's a nightmare but that's our responsibility. (K001M)

8.12 Flo was family and was not expendable. But, despite the adage of not choosing your family, there is evidence that people do choose who counts as family and that 'the boundaries between relationships that are 'given', in terms of consanguineal and/or affinal links, and those that are 'chosen' are not necessarily salient in understanding how definitions of family and kin are constructed' (Charles and Davies 2011:88). So replaceability and expendability may not only be applicable to animals but when experienced by humans do not raise questions about their human status.

Discussion

9.1 The evidence that emerges from these written and spoken accounts shows that the practices through which dogs are incorporated into families are contradictory and variable. While caring practices incorporate dogs into families as dependants and analogies are drawn between dogs and children who never grow up, dogs are not necessarily regarded as surrogate children. However, the analogy alerts us to the context of unequal power relations which structure relations of care; both children and dogs are dependent on and subordinate to adult family members. Secondly, although kinship practices blur the boundaries between human and animal, people recognise that dogs are both like and not like humans and, in attempting to avoid anthropomorphism, often reassert the human-animal binary. Thirdly dogs respond to humans and engage actively in relationships with them. This disrupts the species barrier because animals are experienced as having capacities which have hitherto been defined as quintessentially human. And the establishing of relationships with other animals makes humans better people – kinder and more compassionate. So the issues raised at the beginning of this chapter are present in the way people talk and write about their relationships with dogs: a relationship of care and responsibility, a distinction between humans and animals that is blurred in the practices of sharing domestic spaces, efforts to maintain that distinction by resisting anthropomorphism, and both respecting the dog-ness of dogs and asserting the superiority of humans.

9.2 So where does this leave the idea of post-human families? Earlier I suggested that there were 3 arguments to consider when reflecting on whether families could be characterised as post-human: (1) that kinship practices involving companion animals disrupt the human-animal binary; (2) that dogs and other animals are precariously incorporated into families; and (3) that extending the category human to certain categories of animals does not challenge human exceptionalism and cannot be characterised as post-human. The first argument, that human-animal boundaries are blurred by many of the practices involved in sharing a home with dogs, is supported by my data. These practices can be seen as post-human because they undermine the human-animal binary and associated notions of human superiority. The second argument, that dogs' replaceability means that human exceptionalism remains intact, has been shown to be problematic because the notion of replaceability can be applied to humans as well as animals and if it is used only with reference to animals it reinforces the human-animal binary. The third argument appears not to be relevant for understanding cross-species kinship practices. This is because boundaries are blurred rather than shifted and, although dogs share many human characteristics and may be regarded as 'almost human', their difference is recognised and respected. They are not incorporated into families as 'furry humans' (cf. Power 2008).

9.3 This evidence shows that kinship practices construct a world where dogs and humans are part of the same social group. Some of these practices can be understood as post-human in the sense that they blur the human-animal boundary but they exist alongside others which reinforce it. It is therefore difficult to conclude that a post-human family form can be identified empirically. Perhaps it is more fruitful to recognise that taking a post-humanist approach enables us to identify those kinship practices that disrupt the idea that there is a clear boundary between humans and other animals. It shows that the reality of entangled lives is messy and categories of separation are difficult to uphold in the daily practices of domestic life even when attempts are made to do so. In this way it highlights the instability of the category human and alerts us both to the deep connections between humans and other animals and to the power differentials that underpin human-dog relations, even when they are loved as unique, agentic individuals and seem to rule the roost.

---

Notes

¹ In this paper I use the terms 'companion animal' and 'pet' interchangeably. I also use the terms 'animal', 'other animals' and 'non-human animals' interchangeably. A discussion of terminology and my reasons for this usage can be found in Charles and Davies 2008: footnote 2 and Charles 2016.

² This is the term used by the Mass Observation Project to refer to panel members and I follow this usage here.

³ It is important to note that idioms of kinship and friendship are often used to describe close relationships with animals without necessarily suggesting a straightforward equivalence (Charles 2014).

⁴ Irvine, in her ethnographic study of an animal shelter in the US, argues both that people gain a sense of self and identity through their interactions with companion animals and that the way dogs interact with their human companions suggests the existence of a core self (Irvine 2004). Her argument that a sense of self is constructed through interactions between humans and their companion animals is analogous to arguments made about how a sense of self is formed in interaction with human others although she does not claim that dogs' sense of self is developed to the same extent as it is in adult humans. The question of whether dogs and other animals have a sense of self is, however, contentious. Thus Jerolmack takes issue with Irvine's position, arguing that we do not need to claim inter-subjectivity to be able to explain the mutual interactions of humans and animals (Jerolmack 2009), while Ingold argues that it is better to talk about inter-agentivity when conceptualising the interactions between human and animal others because this encompasses the whole embodied being rather than prioritising the engagement of minds (Ingold 2012).

---

References