Introduction

1For those interested in gaining an insight into the complexity of contemporary family relationships, we recommend watching the film Un air de famille (Family Resemblances), by Cédric Klapish, Agnès Jaoui and Jean-Pierre Bacri. The characters (a mother and her three children, all adults) meet for a birthday celebration dinner, over the course of which the resentments that each has accumulated over the years resurface and transform what was intended to be a friendly gathering into a psychodrama. This comprehensive portrait of a contemporary family – as cruel as it is realistic – even includes the dog Caruso, an old Labrador paralyzed by osteoarthritis, whose presence gives rise to a number of wonderful exchanges that reveal much about what unites and divides the family. Posters of this cinematic adaptation of a play feature Caruso prominently as a significant part of the story, allowing spectators to identify with the situation and project themselves into it, or find a "family resemblance" of their own in Family Resemblances.

2What can animals tell us about family relationships ? In what ways are they constituent entities of what is today considered a "family" ? How are they very specifically integrated into homes ? Are they considered "family members" ? If so, which members in particular ? These questions are at the heart of this issue of Enfances Familles Générations. They will not all be answered, mainly because the subject is relatively new, and there is not yet sufficient research addressing it to classify the subject as a field in its own right (as we will discuss in our review of the literature in the first part of this introduction). However, the articles gathered here make it possible to highlight the scientific issues involved in analyzing the place of animals in the family, to identify the empirical realities to which this subject refers, and in short, to lay the foundations for a research program that would initiate and consolidate a dialogue between the sociology of human-animal and the sociology of the family (as we will examine in the second part of this text).

Points of contact

3This section presents an overview of the literature on the place of animals in families by showing what the research teaches us about 1) the effect of animals on families and (2) the effect of families on animals.

The effect of animals on families

4Among the countless potential modes of connection between human and animal families, a review of the literature leads us to distinguish three : integration, assimilation and substitution. These three modes form a gradient that brings into play different concepts of the status of animals, both among researchers analyzing the relationships between families and animals and among the actors they are observing.

Integration : animals in families (belonging)

5First, there are the studies that focus on how animals are integrated into family groups. This integration can take many different forms, not necessarily involving assimilation or substitution, which we will describe below. By "integration," we mean the fact of belonging to a group without necessarily involving equivalence issues. Two types of research explore this question.

6First, studies derived from the sociology of consumption tend to regard animals in the home as consumer goods whose particular characteristics need to be understood. "Belonging" should be understood here in the narrow sense of "economic possession." Nicolas Herpin and Daniel Verger (1992) identify a number of differences between capital goods and family pets. The animals are not expendable goods ; unlike electronic devices or household appliances, they are domesticated living beings that will be affected if they do not receive constant care. They can be purchased, and they are often given or passed on by others, with no commercial transaction involved. Despite these significant differences, the acquisition of an animal is "the result of a decision that is not fundamentally different from that made by consumers when deciding between durable goods from which they expect certain services" (Herpin and Verger, 1992 : 262). Other studies on the reasons that consumers acquire pets distinguish between animals perceived as toys (Hickrod and Schmitt, 1982), or as a mark of social status (Wood et al., 2017) or as "brands" (Beverland et al., 2008), and animals perceived as friends (Bardina, 2017 ; Hirschman, 1994), as loving companions or leisure companions (Herpin and Verger, 1992).

7Other studies seek to measure and interpret the statements of many pet owners who explicitly regard their animals as "family members." In most of this research, being a "family member" means that the animal is thought of as a "person" rather than a mere possession. This form of integration is observed in situations where, for example, dog owners “routinely used their day-to-day experience with their dogs to define their animals as minded social actors and as having, at least, a ‘person- like’ status. Caretakers typically saw their dog as reciprocating partners in an honest, non-demanding and rewarding social relationship” (Sanders, 1993 : 211). This integration is also manifested in the redistribution of domestic spaces allowing animals to access areas of the house that were once reserved for humans, such as the bathroom, the bedroom and even the bed (Franklin, 2006 ; Gabb, 2008 ; Irvine and Cilia, 2017 ; Serpell, 1996).

8Depending on the families, the animals or the particular sociological approaches being used, animals may be considered to be either "objects" or "subjects" (Blouin, 2012 ; Greenebaum, 2010 ; Irvine, 2009a ; Sanders, 1993, 1999, 2003 ; Tuan, 1984). Belonging to the family can therefore have two meanings for the animal : as an attribute or as a member. The first involves a simple insertion of the animal into the domestic sphere occupied by family groups (being part of the furniture). In the second case, it is possible to talk about the integration of animals into family groups (being a family member). An equivalence now begins to emerge between humans and animals, which leads us to see integration as the beginnings of assimilation.

Assimilation : animals as family (equivalence)

9Although some authors point out that certain owners are careful not to anthropomorphize their pets and to keep in mind that they are "only animals" (Blouin, 2013 ; Charles, 2016 ; Irvine and Cilia, 2017 ; Power, 2008), a significant proportion of publications focus on the operations by which individuals construct or perceive a similarity between animal and human entities in the family. The place and effect of animals in families are then seen to a certain extent as narrow issues of structural, symbolic, functional, statutory or practical equivalence. The question then arises : if the animal is considered a family member, which member is it ?

10Most research emphasizes the parallels with children (Blouin, 2012 ; Franklin, 1999). The attraction to pets would thus seem to lie in the child-like qualities that their owners attribute to them (Archer, 1997 ; Blouin, 2012 ; Charles, 2016 ; Greenebaum, 2004). This process of assimilation, as with a child would be manifested, among other things, by the amount of time devoted to the care of the pets (Power, 2008) and by the owners acknowledging a mutual affection similar to that which unites parents with their young children (Shir‐Vertesh, 2012). Recent publications have also proposed a cognitive and physiological interpretation of this thesis of assimilation by showing that when a mother sees images of her child and her dog, a common network of brain areas related to emotion, reward and affiliation is activated (Stoeckel et al., 2014). In addition, dogs and their owners both produce a hormone (oxytocin) known to be responsible for maternal care and for strengthening the bond between a mother and her child when they look each other in the eye (Nagasawa et al., 2015).

11As well as quantitative studies measuring people's propensity to assimilate their animals with a family member, many studies describe the concrete practices that reflect and contribute to these assimilation processes. As with children, people impose special restrictions and privileges on their pets (Hirschman, 1994). They stay home to take care of them when they are sick (Hickrod and Schmitt, 1982). They celebrate their birthdays (Archer, 1997 ; Sanders, 1993) and give them gifts (Schaeffer, 2009). Among these practices that contribute to the assimilation of animals into families, three are the subject of particular attention.

12First, some authors focus on the ways in which pets are named. Lucy Jen Huang Hickrod and Raymond L. Schmitt point out that the name establishes the pet as a family member “by transforming it into an interactional object” (Hickrod and Schmitt, 1982 : 61). It gives the pet a particular personality that encourages certain inter-individual interactions and allows the family to speak about the animal as a member of the family group (Beck and Katcher, 2003 ; Brandes, 2010 ; Phillips, 1994 ; Tannen, 2004) or to consider it a confidant (Cain, 1983). Starting with the observation that one-third of dogs in France have a human first name, Baptiste Coulmont (2016) reconsidered Claude Lévi-Strauss's comments about "dogs, to which no human first name is given without causing a feeling of unease, if not a slight scandal" (Lévi-Strauss, 1962 : 270). The author thus describes the dynamics of the transformation of social and symbolic boundaries that are at play in attributing dog and human first names. Such ground-breaking work is also observed in the concrete interactions between humans and animals. Using an interactional linguistic approach, Chloé Mondémé (2018) shows, for example, that the ways in which owners address their pets can be similar to those used with very young children.

13Second, a number of studies deal with the evolution of the ways in which the death of a pet is handled. According to Svetlana Bardina, "nowadays, pets are often buried, mourned, and remembered as family members” (Bardina, 2017 : 416). Cemeteries for pets are becoming increasingly common. In France, the oldest such cemetery prohibits any practice "that appears to mimic human burials" (Regulations of the Société anonyme française du cimetière pour chiens et autres animaux domestiques, 1899 ; cited by (Gaillemin, 2009 : 495). But since, "pet cemeteries are starting to resemble human burial grounds" (Bardina, 2017 : 415). The study of these burial sites shows the development of a religious and familial identity conferred on pets whose deaths may be commemorated in relatively similar ways to those of close relatives (Brandes, 2010). Some older studies compare the psychological and emotional consequences of the loss of a human or animal loved one, generating controversy about their possible equivalence (Archer, 1997 ; Gage and Holcomb, 1991 ; Rajaram et al., 1993).

14Lastly, it is also worth noting the studies, particularly those carried out in North America, that establish links between domestic violence and animal abuse. For some authors, writing mainly from the standpoint of psychology, animal abuse can be a precursor to domestic violence. Although controversial, such statements are relatively common in the literature because, as explained by Leslie Irvine and Laurent Cilia, "The notion that animal cruelty leads to other forms of violence serves an important ideological function in explaining violent crime" (Irvine and Cilia, 2017 : 6). Other researchers measure the equivalence of violence by comparing rates of animal, marital and child abuse in families. For example, between half and three-quarters of women who are beaten by their partners say that their abuser also threatened or abused their pets (Ascione et al., 2007 ; Faver et Strand, 2007 ; Flynn, 1999 ; Volant et al., 2008) and, according to Sarah DeGue and David Dilillo, "the identification of animal cruelty in a home (perpetrated by parents or children) may serve as a reliable red flag for the presence of child maltreatment or severe domestic violence" (DeGue and Dilillo, 2009 : 1053).

15In these various works, it is the equivalence between animals and family members that is explored, without this equivalence implying something lacking on the human side, or a gap or unfulfilled role that the animal would fill by default. Other studies, presented below, deal more frankly with this dimension. We now move from assimilation to substitution.

Substitution : animals in place of family members (replacement)

16A number of studies seek to understand to what extent the integration of animals into families is a process of replacing a human family member (father/mother, brother/sister, spouse) with a pet. This thesis of substitution springs from situations where the animal is constructed as an equivalent of the human (as in assimilation), and that are characterized by an absence or, more commonly, by a gap that the animal is called upon to fill completely, sometimes but not always successfully.

17Studies on substitution generally start with the observation that attachment to pets is less pronounced among parents with young children (Blouin, 2008) and greater among unattached, divorced, widowed and remarried individuals, couples without children, newlyweds or those who no longer have dependent children at home (Albert and Bulcroft, 1987 ; Taylor et al., 2006 ; Turner, 2005). Therefore, in certain cases of family breakdown (Blanchard, 2015 ; Cohen, 2002 ; Kurdek, 2008, 2009 ; Taylor et al., 2006), animals are defined as temporary or permanent substitutes for children (Laurent-Simpson, 2017), parents (Bodsworth and Coleman, 2001), siblings (Tipper, 2011), a spouse (Zasloff and Kidd, 1994), or even an entire family.

18This substitution thesis is discussed and criticized by some authors who do not look at it from the angle of attachments, but rather who compare pet ownership rates as a function of household composition (Guillo, 2009 ; Héran, 1988 ; Herpin and Verger, 1992 ; McNicholas and Collis, 2000). According to Herpin and Verger, who worked on the situation in France, "This thesis would be verified if couples without children, and a fortiori single people without children, owned family pets more frequently than families with children. However, the statistical observation is the opposite." (Herpin and Verger, 2016 : 433). Thus, they write, "the family pet does not appear as a substitute for the child, but rather as a complement" (Herpin and Verger, 1992 : 274). These criticisms are to be understood as responses to some authors who view the substitution of humans by pets as a particularly acute symptom of a type of "bonding crisis" affecting contemporary Western societies. According to Paul Yonnet, "The proliferation of dogs and cats [is] a metaphor for the crisis of contemporary education" (Yonnet, 1983 : 8). The development of love and compassion for animal companions, as evidenced by the evolution of their place in families, is thus seen as a driving force behind misanthropy and the erosion of humanism in Western societies (Digard, 2012). For other authors, particularly those of Anglo-Saxon origin, substitution is more often presented as a remedy for the fragmentation of social and especially familial ties. In a situation of increasing "ontological insecurity" in Western societies, where social ties are more ephemeral and uncertain than ever before, pets can be perceived as providing more stability and coherence than human family members (Franklin, 1999). They are thus considered a source of emotional support, comfort and security for their human owners.

The effect of families on animals

19A consensus is emerging in the literature around the idea that the integration of animals into families implies a change in animal status. Within the family sphere, animals can become objects or people or they may remain animals. However, two types of work can be distinguished in this consensus : "fixist" and "flexible." In the fixist approaches, while various statuses can be assigned to a pet, once assigned the status will not vary, and it will have to be aligned with certain external structural variables, which are also intangible. The family is here conceived in a static way, as a perimeter that one enters or leaves, and that is not fundamentally changed by the integration of the animal. "Flexible" approaches, on the other hand, focus on the constant change in the status of pets within the family sphere : even if the pet is treated (or perceived) as a family member, it can just as easily be considered an object or "just an animal." Taken together, these approaches provide a dynamic vision of the family – as an entity whose perimeter is not fixist in advance and is subject to negotiation.

Fixist approaches : family and animal properties as discrete variables

20Within the fixist approaches, there are two kinds of studies : demographic and functionalist. In demographic studies, the main goal is to count and locate animals in the household. The family is mainly understood as a statistical entity (household) whose characterization is essentially limited to the effects of structural and external variables (social class, housing, etc.). In an analysis of survey data conducted in France by INSEE, François Héran (1988) shows that there are considerable variations in the social distribution of dogs and cats. The author uses the composition of the household as a starting point to explain the ownership rate of these animals. Based on an analysis of structural variables such as social class, place of residence and housing morphology, Héran shows that cats and dogs form a structural opposition system, revealing "a homology between the space of social classes and the space of animal species" (Héran, 1988 : 421). Herpin and Verger (2016) also use INSEE data, but longitudinally, to analyze the evolution of the place of pets over more than twenty years. These studies confirm the importance of structural variables such as housing, social background and household composition in understanding pet ownership (and the "forms" it can take), and also in describing a change in owners' motivations. According to Herpin and Verger, the reasons for acquiring a pet are less utilitarian and less related to the services rendered by the animal than to what its companionship can bring.

21Among the studies that examine the function of the animal in families (Staats et al., 2008), we highlight those that examine the role of pets through each stage of the family's life cycle. As individuals progress through this cycle, the role of animals may change, depending on the needs and expectations of family members (Staats et al., 2008 ; Turner, 2005). The way pets are integrated into families is then modelled on the phases of the human family life cycle, whose content, form and phasing remain unchanged by the insertion of the animal. The family life cycle thus seems to be reified, and animals must adapt and take their place in it according to the functions that the family group assigns to them.

22Several functions are generally assigned to animals in families. The first focuses on their positive influences on family life. They may perform relatively technical functions such as watchdog, protecting the home or children when they are home alone (Herpin and Verger, 1992). They help in staying active (walking, games, etc.) and in preventing loneliness (company) (Staats et al., 2006 ; Staats et al., 2008). They are intermediaries between family members (Tannen, 2004) and thus contribute to the emotional balance of the family group (Leow, 2018 ; Walsh, 2009b), as well as to the physical and mental health of its members (Albert and Anderson, 1997 ; Eckstein, 2000 ; Herzog, 2010 ; Walsh, 2009a).

23As we can see, fixist approaches involve studies that take a structural and functionalist perspective of the family (Parsons and Bales, 1955) and that, as such, explore the integration of animals into families in relation to their functions, particularly based on the phases of the family life cycle.

Flexible approaches : cross-redefinition of families and animals

24An important part of our literature review (rather, part of the animal studies field) addresses how animals fit into everyday family practices and are given various statuses (Blouin, 2008 ; Greenebaum, 2010 ; Irvine, 2009b ; Sanders, 1993, 1999). Sometimes regarded as alter egos, sometimes as "radical others," pets thus come, as a result of their "liminal" status, to challenge the boundaries between humanity and animality (Charles, 2016). The place of animals in families is approached in a "flexible" way, emphasizing the variety of "ontological framing" to which they can be subjected. Above all, these flexible approaches make it possible to view the same animal in very different ways based on the situation : as subjects when they cuddle up on the sofa and as objects that can be abandoned if they become annoying (because they have become too restrictive, too expensive, potentially dangerous or are no longer considered useful). In this sense animals are perceived as "flexible persons" or "emotional commodities" (Shir‐Vertesh, 2012). They are loved and incorporated into human lives, but may, at any time, be demoted and moved out of the home and the family (Irvine, 2009a ; Shir‐Vertesh, 2012). The ontological flexibility of animals then appears to be intrinsically linked to their status as "subordinates" (Tuan, 1984).

25Family and family ties are approached as vectors for breaking the species barriers : in other words, it is not just that animals are "humanized" by being integrated into the family, but that the family itself is profoundly redefined when it includes animals. Thus, we speak of "post-human families" (Charles, 2016 ; Smith, 2003) or "more-than-human families" (Irvine and Cilia, 2017 ; Power, 2008) to designate collectives composed of beings belonging to different species, but within which the boundaries between species are less important than the relationships of attachment and closeness. As a result, descriptions of this type of collective no longer dwell exclusively on the dynamics of assimilation of animals with humans. The animalization of humans and the hybridization of categories are also discussed : dogs are referred to as "fur babies" (Power, 2008) while family groups are redefined as "packs" (Franklin, 2006).

26In these studies, animals raise important questions about the concept of family : what does it mean to be a member of a family (Irvine and Cilia, 2017) ? How are new family roles (Greenebaum, 2004) and new kinship relationships built (Charles, 2016 ; Haraway, 2016) ? Lastly, how do the perimeters of the contemporary family bond emerge ? Based on the ontological flexibility of animals, this research will describe the family as a fluid agencement, subject to constant negotiation and redefinition.

Representing or naturalizing the family resemblance ?

27A general observation emerges from the bibliographical analysis of social science research on the relationships between families and animals : The literature pays scant attention to the history and diversity of contemporary and socio-historical situations where family and animal transformations intersect. We will see in this section that the studies, taken as a whole, generate a naturalization effect on the very specific zoological, sociological and spatio-temporal realities studied : The place and effect of animals in families are often restricted to those of "companion" dogs in contemporary middle-class Western homes. We will then see that this naturalization effect is also linked to the way certain semantic categories are used : The concepts of "companion animal," "family pet" or simply "pet," which are implicit in animals integrated into families, are rarely questioned or explained. Finally, we will see that this naturalization effect is partly the consequence of particular regimes that produce sociological knowledge about the place of animals in families : Many of these studies are part of an ethical or commercial framework of the particular reality that they describe and help to establish.

A focus on particular realities

28First of all, the focus on some very specific animals is obvious. The dog is undoubtedly the most studied species in its relationship with human families. More precisely, it is the "companion" dog that occupies the central place. Cats can also appear in the research. More rarely mentioned are relationships with fish, birds, horses, rodents (Smith, 2003) and wild animals (Jaclin, 2013 ; Jaclin and Shine, 2014 ; Kirksey, 2015). Research into "new pets" (insects, reptiles, etc.) is strangely absent from the literature. While this special interest in "companion" dogs reflects the pervasive presence of this species in Western homes, it should be noted that dogs are not as ubiquitous in other geographical and cultural areas, and that not all dogs in homes, even in the West, are "companion" or "family" dogs (Blouin, 2012). Dogs owned for utilitarian reasons (guarding, protection) remain largely outside the scope of these studies.

29Most research on the place of dogs in families also obscures (or ignores) the fact that the association between "dogs" and "companion dogs" is the result of a particular socio-historical process that has been underway in Western countries since the 19th century (Baldin, 2014 ; Baratay, 2008) and that in many respects intersects with the history of socio-historical transformations in the family. Before the end of the 18th century, the owning of pets was an exception and was characteristic mainly of the nobility (Irvine, 2004 ; Ritvo, 1987). Members of the aristocratic elite of ancient Greece and Rome and of medieval Europe often owned pets (mostly dogs, but sometimes cats). The practice gradually spread throughout the European aristocracy of the 17th and 18th centuries. It then became strongly associated with women : Lap-dogs shared in the idle and domestic existence of the nobility as did "ladies in waiting." In the 19th century, the phenomenon spread to the bourgeoisie with the adoption of the "boudoir" (Kete, 1994) or "salon" pet. During the Victorian period in Britain, dog care was considered a healthy occupation for married women, who were consigned to the home, deprived of contact with their husbands and obliged to train to become good mothers (Flegel, 2015 ; Traïni, 2011). Their husbands might also have relationships with dogs and other animals, but in completely different spatio-temporal situations, such as when hunting or at dog or horse races. This history of the "companion" dog has helped to shape the image of a relationship associated with domestic space, intimacy, and lack of professional activity, as well as with a form of social distinction. The Victorian family model still permeates representations of the integration of dogs into families today in relation to the nuclear family model indexed to the household.

30The association between "animals" and "family" that has developed with the "companion" dog must therefore be seen as a social construct inherited from a definition of the family that connects the private space of the home with femininity and the public space with masculinity. The political and regulatory dynamics that gradually moved dogs from the public to the domestic sphere (the home) and to private ownership must also be taken into account : the removal of stray dogs, setting up of dog pounds, imposition of taxes on dogs (Baratay, 2011), the obligation to keep dogs on a leash, the development of animal protection, the conversion of stray dogs into companion dogs (Howell, 2015), etc. Lastly, this association must be understood in the light of practices that mould animals based on the expectations and domestic constraints of future owners or, conversely, that define families suited to acquiring a pet : selecting breeds with morphological and behavioural characteristics compatible with specific situations of human-animal cohabitation ; supporting the dog's integration into the family by using "dog trainers" ; the work of animal shelters (Balcom and Arluke, 2001 ; Michalon, 2013), which involves the selection of families and supporting them during the so-called "adoption, " etc. Yasmine Debarge's article on this issue is an important contribution to the analysis of animal "socialization" processes in families. By analyzing the specific case of guide dogs, the author shows how the integration of these animals into families implies and generates a profound reconfiguration of the relationships between family members of the visually impaired.

31Beyond further study of the Western socio-historical dynamics of association between family and animals that centre on the "companion dog," it seems important to extend the analysis of the place and effect of animals in families to other empirical realities. For example, studies might be carried out on other species that participate, for better or for worse, in family life : hens, new kind of pets, and livestock on family farms, but also lice, mosquitoes, etc. It is also surprising that the literature has not paid more attention to cats, which are increasingly present in households, and whose integration into the family can be assumed to be, in some respects, similar to that of dogs (obligations of ownership) but sufficiently different to be of interest (possibility of evolving more freely in the public space). It should also be noted that the integration of horses into families could be addressed from the perspective of the development of sports and leisure riding activities and their effects on family life. The horse's entry into the leisure sphere is often considered to be the first step in the horse's gradual shift from farm animal to companion animal status (Digard, 1999, 2004, 2009 ; Tourre-Malen, 2009). The study of such a species could help to deepen the understanding of the links between family transformations and changes in the relationships with animals. It would also provide a better understanding of how the emergence of the middle classes, consumer society, leisure time and leisure activities have contributed to the "democratization" of pet ownership in many Western countries. As well as the Western realities that largely dominate the literature on the relationship between families and animals, it also seems important to look at other cultures in order to denaturalize the specific zoological, sociological and historical realities that exist there. The history of relationships between dogs and families, marked in the West by the introduction of the "companion" dog, is very different in other cultures ; for example, where the categories of "companion animals" or "work animals" are not used and are of little relevance for researchers who wish to describe these societies. Francis Lévesque's article in this issue analyzing the varying roles of the qimmiq (dog) in Inuit families in the Canadian central Arctic demonstrates this point brilliantly.

Loaded and rarely questioned concepts

32Another way to mitigate the naturalization effects of the specific realities described in the literature would be to further explore and clarify the concepts of "companion animals," "family pets" and simply "pets." As these designations are the result of a long history in which representations of the family and animals intersect and intertwine, they are loaded with implications that partly condition the ways in which animals are integrated into the family.

33Let's start with "companion animals." As we have seen above, the term "companion" historically relates to the aristocracy, and to the "lady's companion" who often shared in the idle pursuits of noble women. "Keeping company" was a role that certain humans and certain animals (dogs) took on. From the outset, we noted the functionalist connotation of the term "companion." It is therefore surprising that Jean-Pierre Digard (1999) defined the "companion animal" as "useless". For the anthropologist, what characterizes a companion animal is precisely the fact that it is being used, if only for the pleasure of its owner. Such a (re)definition of the term "companion animal" resonates with contemporary transformations of the family into a space for freedom, for absence of labour, and for unconditional emotional ties. However, this (re)definition should be put into perspective. Empirically speaking, excluding companion animals from the realm of utility raises certain issues : for example, the social utility of assistance dogs is generally connected with the so-called companion relationship they have with their owners, but the owner's approval is itself a form of utility, whose scope, nature and benefits are for the owner to define. This implies less a complete exclusion from the realm of utility than a reconfiguration of the concept of the utility of companion animals. Such a reconfiguration is central to the studies on the integration of animals into families, which often seek to understand, even justify and valorize, the roles of "companion" animals in families.

34The notion of a "family pet" seems intuitively the most appropriate to describe the integration of animals into the family sphere. The adjective "family" evokes the word "familiar," which in this context refers to a form of relationship with things or beings that are well known (Thevenot, 1994) and with whom social demands are relaxed. "Familiarity" also resonates with "intimacy." The literature provides several examples of how humans evoke the intimate relationships that bind them to their pets (Charles and Wolkowitz, 2019) and the fact that their pets "love them as they are" (Charles, 2014) unconditionally, without judging them. This discourse on the unconditional love of animals, on their absence of judgment, brings us back to the idea of familiarity that refers to situations where the demands of sociability are less onerous. In short, you can really "be yourself" with your dog.

35The English language contains another meaning of the term "familiar animals. " It is used in particular to refer to witches' companion animals (cats, rats, or owls for example). Here, the ownership of "familiar" animals is synonymous with a chosen avoidance of all family life, particularly associated with celibacy. Feminist readings of witch-hunting periods emphasize that it is precisely their independence and autonomy from men that led some single women to be defined and persecuted as witches (Chollet, 2018). In this configuration, the company of certain animals has served as a "symptom," or even a symbol, of aspirations to independence. The feminist reinterpretation of the image of the witch therefore leads us to approach "familiar," "family" and "authenticity" in a different way : In the witch model, it is only outside the family that freedom from oppressive social norms can be found, and that the authenticity of the individual can be expressed.

36Lastly, let us take a look at the English term "pet." The word appeared in the 16th century and derived from the French word "petit" or small. It was first used in reference to children, then to certain animals, mainly the lap-dogs mentioned above. From the outset, the term was associated with the family sphere. It has an elective dimension since "pet" also means "favourite." Our "pet" is the child we prefer, the one we cherish, the one we spoil more than the others. Understood in this way, the integration of animals into families is conceived as a privilege reserved for certain animals, specifically chosen by a human being from a mass of anonymous animals. "Pets" implies the inclusion of particular animals into a space of proximity to humans who are expected to provide them with protection, care and consideration. The family is then defined as a privileged, positive and rewarding universe, sheltered from other social activities. However, the term "pets" can also be used as a criticism of a form of family upbringing. "Pets" initially meant children who were spoiled, poorly educated, pampered or treated with excessive care. Applied to animals, this idea is now widely disseminated by the media, which regularly point to the "excessive" and "irrational" behaviour of animal owners, their anthropomorphism, and the excessive emotional closeness they maintain with their animals.

37This brief overview of the history and etymology of the concepts of "companion animal," "family pet" and "pet" highlights the specific concepts of family and animals that underlie them. Without a basic analysis and explanation of the chosen concepts, authors risk producing a tautological analysis of the place and effect of animals in families. For example, the use of the term "pet," which implies affection and positivity ("favourite"), can lead in some investigative situations to researchers asking their subjects if they love the animals they love. Taken to an extreme, if all the terms analyzed above ultimately refer to a way of integrating an animal into the family, one can sometimes get the impression that studies exploring this issue without terminological rigour obviously arrive at the conclusion that family pets are considered family members. This lack of critical reflection on the terminology seems to be related to the regimes of knowledge production with which the issue of the integration of animals into families has been examined so far.

Two Regimes of Knowledge Production

38What are the epistemic and political conditions for the emergence and development of academic knowledge on the place and effect of animals in families ? A thorough and systematic analysis is required to answer this question. We will limit ourselves here to stating some observations from our literature review, and suggesting some explanatory avenues.

39A surprising initial observation is the virtual absence of the sociology of the family in our corpus. In the literature we consulted, we found few specialized journals, principal authors, or primary theoretical or methodological frameworks. Similarly, none of the authors interviewed claimed to be doing research in family sociology. Adrian Franklin made a similar observation in 2006, insisting that "the sociology of the family and the sociology of housing need a new post-humanist makeover, [...] because neither families, households or housing can be thought of any longer as humans among themselves" (Franklin, 2006 : 137). While there is no shortage of social science studies focusing on the place of animals in families, it must be noted that they are not written by family specialists. As we will see, they can be grouped into two large and relatively distinct regimes, although they are part of the same dynamics of the valorization of animals in families.

40The majority of the publications listed on the subject can be generally linked to consumer studies. Before the mid-2000s, studies on the preferences of consumers and the structure of the households of the owners of animal companions predominated. Like Elizabeth C. Hirschman's article on "the socialization of consumers' companion animal preference patterns" (Hirschman, 1994 : 619), many studies on the place of animals in families are published in marketing journals. The conditions for the production of sociological knowledge on the subject appear to be closely linked to the development of companion animals. The scholarly understanding of the "pet" phenomenon is thus literally based on the actors in these markets who provide researchers with most of their study tools. Some public institutions (as is the case in France with INSEE or INED) occasionally produce statistical data on "pets," rates of ownership, the socio-demographic description of owners, etc., which are of sociological value (see the work of Héran, as well as of Herpin and Verger). However it is mainly the professional organizations of the pet industry that produce this type of data. Thus, the most complete statistics are always to be found among industry companies and trade unions. Although these data are difficult to access and potentially very expensive (Haraway, 2007), they provide an empirical basis for a large number of academic publications on the place of animals in families. These publications complement the broad dissemination, in the form of press releases, of executive summaries of these figures, through which the pet industry tirelessly emphasizes the huge number of pets in households and the increased spending on them. It is as if a significant proportion of academic authors researching the place of animals in families worked alongside the pet industry to publicly represent its clientele, thus acting as spokespersons for a phenomenon that, without them, would simply not exist in the form in which we know it.

42It is therefore possible to identify two regimes of knowledge production on the place of animals in families, distinct in terms both of the scientific fields in which they are deployed (reviews, concepts, approaches) and of their political and moral significance. Consumer studies approach the "pet" phenomenon as a market phenomenon, while animal studies consider it more from the point of view of an ethics of caring or empathy. But beyond these differences, the two regimes each in their own way exist as part of a perspective on the moral, political and market valorization of pets, or rather of the company of animals. Therefore a real political economy of the affects underpins the literature on the place of animals in families. This leads us to make two remarks about the research mentioned here.

43The first concerns the almost exclusive research interest in the positive dimensions of the integration of animals into families. When a few difficulties of cohabitation are noted in some articles, it is generally only to underline how they hinder the integration of animals into families, which is seen as a necessarily desirable outcome. An entire field in animal studies remains unexplored, specifically an analysis of the place of animals in terms of describing the potential problems, negotiations and uncertainty of integrating them into families. Second, in a world where the pet industry fully exploits the affective relationships between humans and animals, and in return gives them greater social legitimacy (Haraway, 2007), we question how this research contributes to the framing of the integration of animals into families as an important, positive and legitimate social phenomenon centred on responsibility, care and affection. It seems important to us not only to be aware of the risks of the instrumentalization and "capture" of research by economic interests, but also to develop a true economic sociology of pet markets, which is currently lacking.

Sociologizing the study of family- animal relationships

44Moving beyond this critical interpretation, we will now sketch out a few lines of research to complete and rebalance the particular image provided by the current literature on the dynamics of integrating animals into families.

Line 1 - Symbolic construction of animals in families : status, functions and cohabitation

45First, we could explore the cultural construct of an imaginary world in which animals are part of the human family sphere. When one thinks of Walt Disney's cartoons, toys, comic books, illustrations, children's literature and video games (Goubault, 2018), it becomes clear to what extent Western youth culture is populated by animal representations. These representations have become important elements in family education. The article by Sophie Michon-Chassaing and Georges Gonzales in this thematic issue illustrates how animal presence permeates family socialization, and how animal representations can be useful a posteriori in the construct of symbolic kinship. However, it remains to be documented how they are used to help children understand their place in the family group. Represented animals are often anthropomorphized figures, with social (especially family) relationships modelled on human ones. The family dimension of these representations of "humanized" animals could be the subject of numerous analyses, usefully contributing to an understanding of the emergence of sociological definitions of the family. To continue in this vein, it would be important also to document the dissemination of this cultural model outside its emergence in the Western home. A stimulating research project might then consist in understanding how certain representations of the family are "imported" into countries such as China or Japan, through the animal representations embedded in these symbolic goods. Just as we ask ourselves how the Disney culture contributes to the worldwide dissemination of an anthropomorphic vision of animals, we might ask how it also contributes to a homogenization of the representations of the human family.

46Another way to approach the symbolic dimension of the integration of animals into families is to expand the socio-historical and semantic analysis (outlined above) of the categories "companion animals," "family pets" and "pets" in relation to the evolution of concepts of the family as a place of intimacy, private life, leisure activities and free time. A discussion could then be initiated with recent studies that highlight the need to think of relationships with animals first and foremost as working relationships (Coulter, 2016 ; Porcher, 2011 ; Porcher and Schmitt, 2010). How can we understand the place of animals in families in such a framework ? Are "companion" pets characterized by the fact that they are exempt from the obligation to work or, on the contrary, should they be seen as performing a particular type of work (emotional, educational) within the family itself ? In any case, it is precisely these new forms of animal utility (disability assistance, emotional support, educational support) that seem to be emerging in response to the integration of animals into families that research should explore.

Line 2 - Material conditions for the integration of animals into families : new anthropozootechnical agencements ?

47In light of the various historical elements mentioned above, it seems clear that the integration of animals into families is linked to "lifestyle" changes. Within this vague term concrete realities are discernible, such as types of housing, their location, their thermal and acoustic comforts, their appliances, their furniture, the availability of a garden, patio, etc. The literature approaches these elements in a fairly general way, as major variables that favour or restrict the integration of animals into families. More interactionist perspectives are lacking that would take into account the way in which the physical environment of a dwelling specifically frames cohabitation between humans and animals. This in turn would give rise to a list of specific agencements (leaving doors and windows open to allow the movement of cats, for example) or facilities (cat houses, baskets, niches, litterbox space, etc.). The focus would then be on the physical details that make the relationship possible (and not just the ones that constrain it). In this context it will also be necessary to consider dog restraint and guidance devices (leashes, collars, harnesses, muzzles) and animal identification devices (tattoos, electronic identification microchips and corresponding databases) that, in many countries, enable an animal to move about freely in public spaces and therefore make possible a family relationship with animals outside the home. Researchers could also explore the effects of the many technical innovations that have facilitated the keeping of animals in a household : kibble and processed foods of all kinds, electronic cat doors, ultra-absorbent litter, flea treatments, electronic identification microchips, scratching posts, pet toys, video surveillance cameras, etc. These various technical objects are aimed at a market whose sociological analysis seems inevitable.

Line 3 – How do animals participate in the internal construction and external socialization of family groups ?

48Acting as mediators (Tannen, 2004), animals can participate fully in the formation and internal cohesion of a family group's identity. The question of the socialization of animals within human families must therefore be explored in its diversity : Is the identity construction of a family the same whether an animal is adopted or purchased ? How does this play out when the animal – young or old – arrives in a family in the making (newlyweds ; couple with children ; etc.) ? What about "foster families" for animals awaiting adoption or destined to become guide dogs or assistance dogs (Mouret, 2015) ?

49Animals can also be a source of conflict between family members who may have to negotiate roles, rules, boundaries, relationships and problem-solving strategies (Walsh, 2009b). What about these conflicting relationships resulting from the integration of animals into families ? Are these difficulties negotiated differently based on individual social environments and the normative or communicative, fusional or autonomous styles of interaction (Widmer et al., 2004) that characterize them ?

50Animals may also participate in the processes of socialization and the construction of family identity vis-à-vis the outside world. They can contribute to the social integration of the family into its environment by facilitating certain neighbourhood relations or obstruct it by complicating others, and also by making certain character traits and types of family functioning visible to outsiders. Animals can also reveal and play a role in social distinctions in families : the sociological profile of dog owners differs from that of cat owners (Héran, 1988) ; the horse has long been associated with "equestrian social distinction" (Roche, 2011), etc. The choice of a "purebred" or "mongrel" animal and breeding to enhance specific animal characteristics (morphological, ethological, biological, etc.) are also dimensions that should be further studied, both in the case of companion animals and in agricultural worlds where the identity of a family farm can be strongly linked to the characteristics of the animals in their herd. These are all legitimate avenues of investigation in the sociology of socialization, both interactionist and dispositionist.

Line 4 – How are descriptions of animal families used in/for the organization of human families ?

51Descriptions of animal families are many and varied. They illustrate the intimacy of family life when they are used, for example, to help a child determine which of the animals before them is the father, the mother or the baby. They also underpin naturalist studies on animal societies. From ordinary statements to scholarly analyses, from naturalistic diagnostics to metaphorical narratives, descriptions of animal families always convey, more or less explicitly, a normative concept of human families. These descriptions circulate in different forms and in different worlds. References to animals in children's literature or in educational institutions are part of a child’s sexual and moral education. Zoos, menageries and wildlife parks also provide illustrations of family socialization, in part because they offer staged encounters between human and animal families (Estebanez, 2011).

52All secular statements about and presentation mechanisms of animal families convey moral values that are relative, for example, to what appears similar to a "natural" model of the family. Conversely, the scientific literature on wild animals such as wolves, elephants and cetaceans is full of references to family values. How do biologists, ecologists and ethologists transform our conceptions of the human family by depicting animal families ? How are their analyses of animal families influenced by our ways of understanding human families ? Maternal instinct, intra- or interspecies adoption, gendered division of labour, fraternal behaviour between siblings, homosexuality, solidarity within families, single parenthood, etc. : The reciprocal influences between human and animal family theories and models are numerous but little studied. One example is Marga Vicedo's research, which demonstrates the decisive importance of the eventful history of the relationships between psychiatry, psychoanalysis, developmental psychology and ethology in establishing the theory of attachment that transformed the concept of maternal love into one of biological need and that "introduced a new justification for the prescriptive role of biology in human affairs and had profound - and negative -consequences for mothers and for the valuation of mother love" (Vicedo, 2013 : fourth of the coverage). But a study of the multiple ways of describing and organizing animal and human families and an analysis of the political and moral significance of the reciprocal influences that characterize them remains to our knowledge a largely neglected field of research. The task of pursuing this research between scientific discourses on the family and on animals could be taken up by science studies.

Line 5 - Integration of animals into families and the sociological (re)definition of the family

53Lastly, and logically, it is the sociology of the family that should address the question of the integration of animals into families ; its heuristic scope for the study of the family and its contemporary transformations seems significant to us. Since its inception, the sociology of the family has relied heavily on two main components : conjugality and parenting (Parsons and Bales, 1955). Since the 1970s, the construction of the family bond seems to be less and less linked to affinity and consanguinity. More recent research has shown that parental roles and recruitment by affinity are predominating (Dandurand, 1990 ; De Singly, 2016 ; Déchaux, 2009). The family is losing its institutional significance and becoming more "uncertain" (Roussel, 1989). In this context of uncertainty, affect is the glue that holds family relationships together : As a "secular religion" in the modern world, love promises that all single individuals will regain a sense of community (Beck, 2008 ; Beck and Beck-Gernsheim, 1995). The contemporary family is then redefined as an environment where each member (children and adults) must give and receive recognition of a particular type, a personal concern, support from the significant others (De Singly, 2016, 2017). From this perspective, the article by Émilie Morand and François De Singly in this thematic issue analyzes the bond that unites people to their dog or cat as a significant other, a significant-other animal that brings "comfort, security, validation of self and of one’s world" (Berger and Kellner, 1964 ; Morand and De Singly in this issue). In doing so, the authors explore a new concept of family that values the quality of relationships more than their status. These different concepts of the family have important implications on how to consider the integration of animals into the family. Conversely, the different ways in which animals are integrated into families in turn involve ways of understanding the family. When an animal is given "claimant" status because it does not possess the essential characteristics of humanness, the family remains an entity composed exclusively of human beings (Hickrod and Schmitt, 1982). In other cases where the species barrier is more labile, some authors refer to "more than human" (Irvine and Cilia, 2017) families, or to "post-human" homes (Smith, 2003). After a careful investigation, Morand and De Singly share this conclusion, thus showing the fertility of the crossover between family sociology and the study of human-animal relations.

Conclusion

54"Really ? You're offering me a dog ?" Somewhere between disappointment, disbelief and contained anger, Yolande, one of the characters in Family Resemblances comments on the gift she has just received from her in-laws. On a card, a photo of a puppy and an invitation to pick him up at the "Not Lost Kennel." As Martine Guyot-Bender (2003) suggests, this gift marks Yolande's definitive integration into her husband's family, and at the same time the assurance of experiencing a lonely, boring and unloved life, as her mother-in-law did. "A dog will never disappoint you," she explains, implying that the same cannot be said of humans, especially those close to us. "No one loved me, no one understood me like Freddy," she adds, referring to the dog that was her companion for 18 years, and who helped to fill the void in her family life. In this scene, which could have been written by supporters of the substitution thesis mentioned above, one perceives to what extent the integration of animals into the family – far from being anecdotal – is freighted with identity, and psychological and sociological issues. The puppy, offered as a gift, illustrates the mechanisms of transmission and transgenerational reproduction of a family model, of a definition of the family and the relationships between its members, that appear in plain sight. Looking at the photo on the gift card, Yolande says : "It looks like Henri's dog" (her brother-in-law). "It's the same breed !" comments the mother-in-law. "Since Freddy, I've always got the same breed." And Henri adds about Caruso, his paralyzed Labrador : "Yes, it was also my mother who gave me Caruso, the dog that doesn’t work anymore !" Henri's reply illustrates the functionalist prism through which the integration of animals into the family is often viewed. As Caruso can no longer walk, the dog no longer fulfills its functions as a substitute for human presence and interaction. "It's like a carpet, but alive," says Denis, the bartender in another scene. The objectification and "commodification" of animals are also presented in this scene, as well as the repeated evocation of the materiality and concreteness of what it means to integrate animals into the family. Still in shock over her unexpected gift, Yolande asks about the puppy : "How do you care for it ?" ("You water it once every morning, I think" replies Betty sarcastically). As an answer, a leash is handed to Yolande, materializing the link that now unites her, for better or worse, with the animal. The misunderstanding that follows the opening of the last birthday gift (an expensive piece of jewellery) is as cruel as it is hilarious : "Oh, another leash," says Yolande. "No honey, it's a necklace," says her husband. "But it's way too luxurious for a dog !" exclaims Yolande, causing her angry husband to explode at his wife's destroying his efforts to make this birthday a happy family reunion : "It's for you ! It's not for the dog !" A reminder of the rules of kinship (Wright Mills & Gerth, 1953), as well as a warning about respecting the anthropozoological boundary of a man who fails to match an idealized definition of family with the sad reality of his own family ties.

55These few microsequences from a fictional but realistic family life contain almost all the themes already explored by the literature on the integration of animals into families, as well as the various directions we have identified for future work. We can only imagine with pleasure what would result from solidly built research, anchored in an increasingly consequential state of the art, supported by innovative and reflective investigational approaches. The goal would be to develop a genuine sociology of family/animal relationships, one which more candidly takes on the implicit biases that have guided research to date (a clear positioning in relation to the three methods of integrating animals into families), and at the same time, seeks to denaturalize the categories it uses, and questions the regimes of knowledge production within which it is embedded. Let us hope that this path, inaugurated by this issue of Enfances Familles Générations, will be followed by other researchers in the future.