Donna Haraway. Investigating the therapeutic benefits of companion animals: Problems and challenges.

Donna Haraway, Adrian Franklin, Michael Emmison and Max Travers. "Investigating the therapeutic benefits of companion animals: Problems and challenges." in: Qualitative Sociology Review, Special Issue: 'Animals and People'. Vol. 3, No. 1, April 2007. p. 42-58. (English).

Abstract

To investigate the health benefits of companion animals in a way that goes beyond finding statistical patterns involves appreciating the philosophical debates about the nature of animal consciousness that engage an inter-disciplinary field of scholarship cutting across the Great Divide of the hard sciences and humanities. It also requires developing a methodology to conduct empirical research which is often viewed as of secondary importance by researchers wishing to make a philosophical case about human beings and modernity. This paper considers the achievements of qualitative sociologists, particularly in the field of post- Meadian symbolic interactionism who have addressed these issues, and discusses ways of extending and deepening this agenda through cross- fertilization with similar work in ethnomethodology, conversation analysis and post-humanist sociology in investigating the health benefits of dogs.

Keywords

Animal-human relationship; Health; Methodology; Qualitative research; Ethnography

There has been a massive increase in the last twenty years of empirical research concerned with the health benefits of the animal-human relationship, particularly those between companion animals and older people. The objective of this paper is to focus on the methodological issues that arise in conducting research in this area. A lot of qualitative research has been conducted, particularly in the field of post-Meadian symbolic interactionism. There have also, however, been studies with significantly different methodological assumptions, and employing different qualitative methods, by ethnomethodologists and conversation analysts, and post- humanist sociologists. The paper will argue that these approaches can usefully be cross-fertilised. It is also important to understand the underlying issues to avoid fragmentation in this inter-disciplinary field

One of the basic problems that qualitative approaches in this area need to address is the status of the animal as an agent, in other words, what is the nature of animal action and interaction as it become available to the qualitative researcher? Cartesian influences have dominated a view of animal agency, particularly in the sciences, that reduce it to instinct or pre-set behavioural responses. Since the statistical studies which show the health benefits of animal companions underline the importance of the relationship between an owner and a companion animal, much depends on how those relationships come into being. If they are merely the result of instinct or the transfer of pack behaviour on the part of the dog, there would seem to be very little point in social scientists working alone or in collaboration with colleagues from science in order to document and analyse how relations evolve. More recent evidence seems to suggest however that dogs and cats do evolve distinct relationships with human companions and that there is considerable scope for relationships to evolve in particular, dialectical and contingent ways (Haraway 2003; Bekoff et al. 2002). If this is the case, and we will argue that it is, then the qualitative researcher is in a unique position to explore ways of obtaining and analysing this data.

The first part of this paper sets out the significance of this issue by summarising the statistical basis of claims for the health benefits of companion animals and also those studies that have established the very significant financial savings that such relationships make on national health budgets. The second part discusses the philosophical debates that are relevant to human-animal relationships and pays particular attention to more recent ways in which Cartesian and behaviourist models have been challenged. This opens up new opportunities for qualitative research to engage in important pure-basic research but, clearly, it is a very new area which poses certain difficulties. The third section, therefore, discusses these difficulties and in the light of this discussion offers a new methodology that might be applied in a number of contexts. Attention is given both to the analysis of observational data and the technologies that might assist in gathering it.

We argue that a mix of ethnographic, ethnomethodological and conservation analysis is capable of revealing and examining relationships between humans and companion animals although the implication of our discussion points to the benefits of very large, time-consuming projects even where the number of cases examined is small. We also conclude that there is further scope using this methodology to form collaborations with human physiologists (and other medical researchers) in order to examine the physical and psychological basis for human health benefits.

The health benefits of companion animals

There is considerable evidence to show that companion animals can be highly beneficial to human wellbeing. In 1992, Anderson et al found that in a survey of those attending a cardiovascular screening service in Melbourne, pet owners reported significantly fewer visits to doctors and significantly less consumption of specified medications (for high blood pressure, high cholesterol, sleeping difficulties or heart problems). Pet owners had “significantly lower systolic blood pressure and plasma triglycerides than non-owners” but the two groups “did not differ in body mass index, socio-economic indicators, or smoking habits” (Jennings et al. 1998:163). Moreover, pet owners in the study ate more meat and take-out food. Since then numerous international follow-up studies have largely confirmed these findings (Headey 1999; Friedmann et al. 2000).

In 1998, Jennings et al estimated the health benefits of companion animals based on their 1992 survey. Using 1993-4 health costs in Australia, the total savings were estimated at $189. 992 millioni, comprising of savings from GP visits of $26.244 million; savings on pharmaceuticals of $18.856 million and savings on hospitalisation of $144.892 million. According to a later study based on nationally representative data, the actual figure was considerably higher. Headey and associates replicated Anderson’s survey and found similar results: pet owners made significantly fewer visits to doctors and used significantly less medicine. Using 1994-5 Medicareii expenditure and assuming that all recurrent health expenditure can be divided up proportionately to the number of doctor visits people make, Headey (1999) calculated the saving to be $988 million, representing 2.7% of the nation’s health expenditure. However, as we enter a new phase of more intensive and detailed study of this phenomenon, the benefits may be more significant as a result of being able to direct them more effectively in the population through training and supervision. For example, Jennings et al strongly suggest that the critical benefit may not be from ownership per se but from specific types of relationship. They found, for example, that “non-partnered people who reported feeling close to their dogs made significantly fewer doctor visits and took less medication than non-partnered people who were not close to their dogs” (Jennings et al. 1998:168). This suggests that we need to understand the relationship itself and its variation, how close relationships develop and why they deliver health benefits. With greater confidence in their administration and armed with Headey’s finding that the over 55s have the most to gain, countries such as Australia which are facing the increasing health cost burden of an aging population might elect to increase the proportion of pet owners in this age group or enact policies that promote their take-up.

While cardiovascular disease is a prominent problem for the nation, the benefit of understanding the relationship we have with companion species is considerably wider. It extends to general wellbeing (Garrity and Stallones 1998), treatment of depression, loneliness and anxiety (Wilson 1998:61), and Alzheimer’s disease (Batson et al. 1998). Knowledge of this relationship could be significant for a multitude of interventional therapies that use companion species in prisons, hospital care, homes for the elderly, in special needs schools and psychiatric hospitals. In an aging society, understanding relationships that have a positive bearing on aspects of health and wellbeing for the elderly becomes imperative.

In their conclusion to a review of all evidence on the therapeutic benefits of companion animals, Friedmann, Thomas and Eddy (2000) argue that it provides “intriguing evidence that animals can be beneficial, particularly for cardiovascular health”. They use the word “intriguing” because studies so far have only provided solid statistical proof of the benefit, not an explanation for it. They suggest that considerably more work needs to be done, but clearly statistical studies have run about as far as they can take us. Two major statistical studies in Australia and one each in the UK and USA (Anderson et al. 1992; Jennings et al 1998; Serpell 1991) give us confidence to proceed to a more intensive, qualitative interrogation of this question using other methodologies. To generalise, existing explanations of contemporary relationships between humans and dogs fall into one of two types that can be called substitutive and anthropomorphic theses. We consider both to be deficient precisely because what they take for granted about the human-dog relation (i.e. that it is a substitution or a representation) requires detailed and painstaking qualitative research that has never been done.

Substitutive theses tend to argue that companion animal species have become more significant to contemporary modern humans since they have been substituted for forms of sociability and solidarity previously provided by significant (and embodied) relationships with friends, community and family. Veevers (1985:19) identified “the surrogate function” as one of three functions that companion animals play in the social life of families and, clearly, she felt that there was something problematic with it: it is interaction that “too closely approximates interactions with humans”. By the end of the century such concerns had softened and dogs and cats were seen as appropriate partners for domestic lives and companionships, especially where those with significant human companions had become fugitive and problematic. Jonica Newby (2001:170-177) for example, places great emphasis on the growth and scourge of loneliness in contemporary society as an effect of near- complete urbanisation. Garrity and Stallone’s (1998:8-9) summary of research on the effects of pet contact on human wellbeing emphasises two ways in which social support from companion animals is effective. The first, the “direct effect" view simply portrays social support as having an unmediated, direct impact on aspects of human wellbeing. This is not, however, explained. The second they call the buffering model which “views social support as somehow intervening to protect the individual from damage to well-being when the person is under siege from stressful life events”.

This thesis argues that animals can become love objects, objects for affection, bond creation, provide togetherness and loyalty etc, but the critical point is that their effects are merely substitute effects for human forms of sociability. They replace normative human needs necessary for wellbeing but do not add anything of themselves as animals, nor is anything important created from the combination of human and animal relationships. For this reason there has been little need to study the specifically hybrid nature of relations between humans and animals since the effect is not assumed to arise from the uniquely human-animal dimension but the replacement of human social support. And since this is so, the significance of what animals do, their agency, is downplayed or ignored, as is the ontological choreography that describes the way any one relationship develops (Haraway 2003). It is as if the main form of agency in the effect is the very acquisition of a companion animal and subsequent human imagination and, since that is a given independent variable, it needs no more investigation. This thesis is, however, unhelpful in explaining the pattern of health effects. One of the key findings in the statistical studies is that companion animals deliver health effects irrespective of whether a person has a human partner or not.

Anthropomorphic theses on the other hand suggest that the efficacy of companion animals for human health relies on the willingness of humans to project human meanings and motives (love, care, affection, loyalty etc.) onto animal behaviour and actions that simply do not exist or cannot be shown to exist. The typical sociological explanation can be summarised like this: because we have lots of interactions with animals that are ritualised, predictable and involve a shared focus of attention (in other words that they are intense), we make the assumption that there is reciprocity of perspectives, emotion and intent and think that we are loved. The simultaneous denial of communicative competence to animals and the assertion of human tendencies toward anthropomorphism descends from Mead and has remained very influential in scientific and psychological discourse. As Sanders (1999) observes, he was also influential on those rare moments when sociologists considered the human-animal relationship:

Since animals were not fully fledged social actors from the Meadian point of view, their encounters with humans were one-way exchanges, lacking the intersubjectivity at the heart of true social interaction. People interacted with animals-as-objects. The dog owner babbling endearments to his or her canine companion is engaged in a form of happy self-delusion; he or she is simply taking the role of the animals and projecting human-like attributes into it. (pp. 118-119)iii

The asymmetrical nature of this exchange is mirrored in other follow-up studies that try to hone in on the cause of the effect. One of the more common is to measure blood pressure before and after a human does something (look at, stroke, be with) with an animal, as if only human agency and human thoughtfulness are at play and need to be understood. While we completely agree that human agency, thought and imagination are critical to understand and inevitably play an important role in explanation, we do not agree that this is all we need to attend to nor where the whole answer lies. This suggests that there are two other objects that demand to be investigated: the companion animals themselves and the relationship itself. Again, the statistical studies of companion species and human health suggest that human self-delusion may be less important than the type of relationship and the species in question. Cats are equally the objects of human projection but the statistical studies show that they offer less health benefit than dogs (Freidmann et al. 2000). Empirical research might show the relationship to be both a hybrid cultural form and one built on agency rather than imagination.

Humandog relationships: philosophical perspectives

We know that the beneficial health effects from companion dogs hold irrespective of whether the people concerned have human partners or not (Freidmann et al. 2000). This allows us to assume that the benefit is not solely from companionship or social support per se and, therefore, not merely from the substitution of animal for human contact. At the same time, although humans can never be perfectly sure what their companion animals are thinking or intending, the idea that they are purely deluding themselves by anthropomorphic projection is now widely doubted (see Sanders 1999:119-147 and Beckoff, Allen and Burghardt 2002:87-113 for a good discussion) while complex forms of mutual communication have been recorded extensively (Haraway 2003). There have been several recent publications, linked to comparative genomics as well as to comparative psychology, that speak to the consequences of the long association of these two species. Dogs read human indexical behaviour better than chimpanzees or wolves. Kaminski et al (2004) found that at least one dog (Rico) acquires/learns words for objects in a manner that used to be thought restricted to growing children--and does it fast and well. This evidence is important for the case we make for taking other animals, especially here dogs, more seriously as social partners in semiotically dense communication, and not just as objects for human meaning-making. This recent evidence, the result of collaborations between anthropologists, behaviourists, geographers, medical psychologists and others (who comprise a new discipline of human-animal studies) also suggests there are reasons to believe that dogs’ basic biobehavioral heritage pre-adapts them for work in contemporary pet and social therapeutic relationships. Pepper and Smuts’ (1999) work on evolutionary pathways of cooperation specific to dogs and humans supports such a view. In the field of ethology, cognitive ethologists and sociobiologists have also questioned the behaviourist view that it is inappropriate to talk of an animal mind (Bekoff et al. 2002).

Underpinning much of this literature is a philosophical project based on the view that not only are previous views of animals as instinctual, insensate creatures entirely different to humans scientifically incorrect, but that this also has far-reaching implications for how we should treat animals and understand our place in the natural world. This is nicely expressed at the end of an essay by Eileen Crist (2002) that is about far more than Darwin’s views on the capacity of earthworms to make choices in constructing their burrows:

Does it matter whether earthworms are intelligent or experience their world? I would submit that what matters is that scientists be allowed and encouraged to pose these questions about worms and other animals. It is hoped that following their cue, common-sense views that are flippantly dismissive of such forms of awareness in the world will be discarded. Why is this desirable? The most significant reason today is the need to awaken and deepen our sense of wonder about the living world. For the erosion of this wonder – encouraged, in part, by the dominance of overly mechanistic models of animal behaviour in the twentieth century – is internally connected to the gathering speed of the human onslaught on the natural world, and to its darkest corollary, the sixth extinction. (p. 8)

This philosophical and ethical interest in the animal-human relationship informs most recent contributions to this literature. Donna Haraway (2003; 2007) has recently advanced an eloquent and powerful argument about the need to re-think our relations with animals, which although not cited by post-Meadian symbolic interactions or cognitive ethologists approaches the same issues from a distinctive philosophical position. Haraway began her intellectual career as a socialist-feminist, but more recently has argued as a poststructuralist that we need to rethink our relationship to objects and animals. Her writings in the areas of cyborg studies, primatology and latterly, about companion species have been inspired by a mix of Alfred Whitehead, feminist theory and science and technology studies. For Haraway, the key to understanding what happens between dogs and humans (and consequences such as health, wellbeing and happiness) is their relating and co- constitution. For her “beings do not pre-exist their relatings” and it is “[t]hrough their reaching into each other, through their ‘prehensions’ or graspings, [that] beings constitute each other and themselves” (Haraway 2003: 6). Any empirical work on this topic must recognise that both dogs and their human companions matter and it is what they do, how they reach out to each other, how they grasp each other (and their prehensions) that constitutes whatever relationship they have. This is what we researchers have to be there for (their embodied relatings); this is what we have to find ways of describing and analysing.

Two other contemporary social theorists, who are not usually mentioned in the animal studies or ethology literatures also use the case of animals to make philosophical arguments about humanity and modernity. The philosophers, Deleuze and Guattari, and especially their concept becoming animal provides the means to develop an alternative to humanist approaches to human-animal relations. Their concept “becoming animal” offers a way of exploring what a relationship with a companion species might involve and how that relationship can be therapeutic or beneficial to modern humans. Certainly for Deleuze and Guattari, animals provide a unique source of absorption in the other (i.e. being attentive to other animal “being in the world”) but more positively and radically, a means of “sweeping away” fixed notions of what it is to be properly human, producing a self more open, flexible and attentive to the world around it but also a self that is becoming more than a fixed human identity. Rather than a mere object for human contemplation then, becoming animal provides experiences that take humans beyond themselves. So becoming animal is an ‘“experimental” state of identity suspension but it is more than just this, as the word “becoming” suggests. In Deleuze and Guattari’s view, what humans become is not evolutionary through its usual terms of filiation and descent but through alliance; “in the domain of symbioses that bring into play beings of totally different scales and kingdoms with no possible filiation” (Deleuze and Guattari 1999: 238). In sum, “becoming animal” allows us to explore human-animal relationships both in terms of how they undermine the (modern) human sense of aloneness in the world, trapped inside the modern idea of a detached and perfectible self and how, at the same time, the parallel worlds of animals offer the possibility of embodied personal extension beyond the confines of the human, the experience of connectivity and the production of a new “humananimal” life world.

Writing in the tradition of Science and Technology Studies, Bruno Latour (1993) and others have also shown how non-humans are active agents in the human social world to a degree hitherto considered impossible. Their break-through was to abandon the Great Divide between the social and natural sciences and to commence treating all non-human objects and beings in a symmetrical way and as inextricably intertwined with the human world. Pickering (1995) argues that we must position ourselves in medias res, in the “thick of things”, and study the choreography of agency as it unfolds. This approach warrants an ethnographic approach with its stress on symmetrical attention to both humans and animals and suggests the importance of studying new human-companion animal relations from their beginning. Their specific history and pattern of agency is critical.

All these writers and researchers provide a rich set of ideas enabling one to view animals as conscious agents. Although we have characterised them as driven by philosophical and ethical interests, many also use empirical examples.iv However, it would also be fair to say that as philosophers and social theorists, these writers are not interested in the different methods one might use in investigating the animal-human relationship or in practical questions such as the health benefits of companion animals. The next part of this paper will consider these issues.

Investigating the relationship

The investigation of human-nonhuman animal relations poses a number of challenges and a variety of methods have been suggested and employed to good effect by qualitative researchers. Post-Meadian symbolic interactionists have conducted ethnographies, auto-ethnographies and interview studies. A prominent concern of these studies has been to demonstrate that animals have agency (e.g. Alger and Alger 1997, 1999) and that the relationship between humans and animals must not be restricted by the linguicentric constraint bequeathed by Mead (e.g. Myers 2003; Sanders 2003). Others have sought to provide first-hand information about their beneficial effects (Flynn 2000; Irvine 2004). There are, however, also autoethnographies informed by different theoretical frameworks (e.g. Goode 2007; Shapiro 1990; Haraway 2003; Smuts 2001), and conversation analytic and ethnomethodological studies that investigate maturally occurring interaction using audio and video-recordings (e.g. Tannen 2004; Laurier, Maze and Lundin 2006). All these studies are interesting and informative, but they also invite critical discussion on how they have obtained and analysed their data, and how one might ideally (that is to say with generous funding and a large research team) address this topic more systematically. We are particularly interested in the problems relating to whether one adopts a human or animal-centred perspective, the importance of studying a relationship over time as against a narrower period of interaction, and the possibility of a variety of relationships (which is arguably the key to understanding possible health benefits).

A human or animal-centred perspective?

Inevitably most research on human-animal interaction (this use of the term “interaction” rather than “relationship” is philosophically significant) has been conducted from an human centred perspective. To give an example, there have been a couple of methodologically sophisticated studies by conversation analysts based on analysing many hours of recordings of how people talk to or about their companion dogs. The focus, however, is on what humans are doing and saying. For example, Tannen (2004) looks at how ‘talking to the dog’ can be seen as a device by which humans manage conflict or do indirect challenges or complaints, a standard CA argument about the way that we prefer to work through interaction using inferences or implicature. Similarly, Robert’s (2004) study of ’animal-directed talk’ in veterinary clinics focuses on the way such talk facilitates the delivery of professional judgements, or the sustaining look of an expert demeanour by the vet. In both cases, the dogs interacting with these humans do not get much attention. Nevertheless, we argue that the fine-grained attention to the details of the interaction evidenced in these studies is something that must be replicated in an inquiry into how any health benefits are derived.

Another common way of presenting animals as if they were humans can be found in the post-Meadian symbolic interactionist literature. Although a number of positions are advanced by different researchers, it has become common to argue that although animals are unable to express themselves through language, they can be understood as having selves, exhibiting agency and expressing emotions in the same way as humans. Irvine (2004: 68-77) in a thoughtful discussion of these issues acknowledges the many differences between humans and animals, but defends a “critical anthropomorphism” (see also Bekoff et al. 2002). This involves finding a middle-ground between behaviourist description and sentimentalised “projection” of our own human feelings, preferences and attributes onto animals. She argues that this can be achieved by “informed, systematic interaction with and observation of an animal”:

For example, given what I know of cats, or about a particular cat, I can make reliable statements about when a cat feels contented as opposed to fearful. Cats use explicit body language, and anyone who pays close attention over time will come to understand that dilated eyes and flattened- back ears signal fear. If I ground my statements in knowledge about normal behavior, I can safely use anthropomorphic language to label it. Indeed, I have no other choice. Although I cannot know whether the cat’s experience of fear is the same as mine, the label ‘fear’ is justified. (Irvine 2004: 69-70)

It seems implied in this passage that cats experience “fear” or “contentment” as if they were humans without the ability to communicate the range of their emotions or desires through language. Later in the book, Irvine (ibidem) as a cat-owner also finds it easy to attribute agency, through anthropomorphising, to actions that could equally well be explained in behaviourist terms, without needing to suggest that cats have selves or feel emotions like humans:

Cats will frequently impose themselves on people’s activities to make their desire and intentions known. My cat Pusskin regularly paws at my arm for attention when I am working at the computer. Another cat, Leo, watches my husband shave. He also supervises (sic) all food preparation...Anyone who lives with cats is familiar with how they sit on reading material, making themselves the center of attention....Evidence of agency among animals helps explain why our experience of them as subjective beings is not solely the result of sentimental anthropomorphism. (pp. 132-3)

The difficulty here is that, despite the qualifications, there is some degree of anthropomorphism, and also the presentation of cats as always acting nicely towards other cats and humans. There is no mention, for example, of the fact that some cats (even the tamed variety) like fighting or hunting birds and mice, or spraying their territory or what happens during the mating season. This is another side to the anthropomorphism: the use of selective examples to advance a philosophical or ethically-driven argument.

Interaction or a relationship?

To date many studies of humans and nonhuman animals, partly because of time and resource constraints, have been concerned with case studies or single episodes of interaction. Others and particularly auto-ethnographies are concerned with how a relationship has developed over time with an animal that has (one is tempted to use “who” for “that”) a particular personality. This is why the term “relationship” seems preferable to that of “interaction” in acknowledging how humans and companion animals develop a strong emotional bond based on intimacy and mutual discovery over a course of time. We would argue that these relationships must be studied as they are created and unfold over time, ideally from their inception, and also they must be studied reactively, observing how the relationship is built upon a long series of transactions between the partners as a result of which conventions, habits, practices and rituals become established. In this way, we can begin to understand whether dogs and humans shape each other in species specific ways. In other words these relationships have a biography, they have an unfolding or a becoming.

A variety of relationships

It is also important to recognise that there will be different types of relationships. Haraway (2003) recognises that relationships between humans and animals may not always be harmonious and fulfilling: they can be like other types of relationships that require patience, hard-work and the mutual tolerance of irritating habits. One should also recognise that there are many kinds of people who develop a relationship with an animal, and the trajectories may well be different. To give some examples, it would be interesting to learn how an elderly person who has recently suffered a stroke gets on with a dog bought at the suggestion of a doctor, how children relate to dogs, or even at the experiences of families or individuals from different class or ethnic backgrounds. Recent research indicates almost half of Australian households contain a dog, eighty-eight per cent of owners thought that their dog was a member of the family and eighty-two per cent said that they acquired their dog for company. Australians are also expressing the significance of dogs in their lives through naming strategies: in the past fifty years dog names have shifted almost completely from dog specific names to human names (Franklin 2000; 2005).

Provocatively, one might add that there are also different types of dogs, and this does not simply mean different breeds, with their expected characteristics. Perhaps one can learn something from spending time with different families about these types, and how this shapes the relationship with a human or group of humans.

A proposed methodology

How though would one ideally wish to investigate the relationship between human beings and dogs, given generous funding and an interdisciplinary research team? We propose a methodology that will innovate a synthesis of traditions from anthropology and ethnographic fieldwork (where a cultural milieu is translated into terms understandable to those who live outside it) and animal behaviour research (where one species’ behaviour is rendered explicable to another species albeit predominantly humans). We need to innovate an entirely new approach, which we have given the working title of trans-species methodology. And in order to innovate this methodology we will need to bring together specialists who would normally work on either side of the humanities-science divide. We need to train ethnographers to understand (and work appropriately alongside) dogs and to bring both types of expertise into practice. For this reason the ideal research team should include ethnographers, veterinarians and ethologists who will combine their knowledge and approaches during data collection and analysis stages.

Ideally, we would like to achieve a cross-fertilisation of approaches from cognitive ethology, social anthropology and ethnomethodology. Cognitive ethology is a diverse, multi-disciplinary subject that takes seriously the argument that animals have both agency and consciousness (Bekoff et al. 2002). From social anthropology comes the practice of maintaining an intensive fieldwork relationship over a long period of exposure. It is also predicated on flexibility and immersion in the fieldwork milieu and working with very different cultural milieux, usually mastering new languages as of course. These skills are useful in working with trans-species relationships, especially when combined with an ethnomethodological focus. From ethnomethodology comes the discipline of focussing only on what eventuates during and from interaction. The focus is on how people or “members” (and in this case two species in companionate relations) construct their world. For ethnomethodologists the world has an orderly, if not an ordered, quality and this orderliness is produced over time by people (and animals together) in everyday life.

An excellent example of the value of such an approach is offered by Laurier, Maze and Lundin (2006). Their analysis of the video-record of people walking dogs in a park illustrates that mind can be conceived as “embodied-in-action”. If we locate dogs (and their humans) in contexts where cognition can be “naturally” deployed (such as a park) then the observational record documents in fine detail how the practical activity of dog-walking is accomplished through mutual gaze, bodily comportment and the contextual clues afforded by the paths and other environmental features. Although people (and animals) have to work continuously at making their own actions make sense to others, the social world constructed attains a taken for granted quality. Part of this taken for grantedness is the indexical character of communication whereby members have to fill in background assumptions that are unique to most interactions. It is the everyday, taken for granted nature of relationships between humans and companion species that render them difficult to identify and research using survey or interview techniques. We would agree that video analysis must be central to any investigation, but the challenge lies in making sense of what the animal as well as the human is doing, and how the other party responds to or makes sense of this.

People may not be aware of the depth and complexity of their everyday communication and culture shared with dogs. Only those immersive, in-depth and long-term methods are likely to make sense of these relationships and demonstrate the complexity of the interaction. We know that people talk to animals constantly and that animals make responses to these utterances creating a trans-species conversation. In the past this was confused with anthropomorphism, but this confusion stemmed from the one-sided focus on humans alone and a rather limited understanding of the social and communicative capabilities of companion species. Since the mid-1980s, Meadian behaviourism has been largely discredited and researchers have demonstrated complex communicative interactions and the ability of companion animals to make sense of human spoken language and even to adopt greater use of vocalisation in communications with companion humans. However, as Sanders (1999:142) suggests, human conversation is but one in a range of kinaesthetic bases for mutual empathetic exchanges between people and dogs which gives rise to “a shared physical grammar”. Importantly, Sanders is guided by Shapiro’s observation that dogs spend most of their time in “concernful absorption”.

The objective of a multi-disciplinary research project would be to study these mutual worlds with as much attention to detail as the ethnographer normally pays to human social and cultural worlds. It has been well documented that humans who live with companion animals spend a considerable amount of time talking to – and with – the animal (e.g. Arluke and Sanders 1996; Tannen 2004; Roberts 2004). Such studies have typically – and not surprisingly - focused upon the human contribution to these encounters and the “function” that such verbalizations might have, for example in mediating relations between family members or reinforcing the family’s identity. Our proposed study will, however, take the investigation of such communication one step further by systematically incorporating the contribution of the companion animal into the research data. We envisage generating transcripts of naturally occurring interaction between our human and animal subjects in which both the human and the animal contributions can be identified.

To achieve this ambitious goal, we would need to draw upon the advice and expertise of ethologists who would work alongside the social scientists. They would need to generate and develop ethograms of dog behaviours that could be analysed as part of the sequences studied by conversation analysts. The objective would be to generate records of the interaction that in conversation analysis have the canonical form:

1. Human comment (typically verbal, but also likely to include gestures)
2. Animal “response” (to be identified through the ethogram)
3. Human comment on the response, or putative “repetition” of what animal meant etc.

Such a record would, we maintain, avoid the human-centred focus of previous interactional studies. The objective would be to obtain an objective and permanent record which can be used by an inter-disciplinary team to investigate how the relationship between the human and the animal is developed and sustained. We cannot specify in advance of conducting the research precisely what form these encounters will take or what “phenomena” we are likely to find. We can, however, be confident that it is only through such fine-grained attention to the details of the interaction that we will appreciate the intersubjectivity which characterizes the animal- human relationship.

The approach we envisage adopting for investigating this trans-species relationship will also be based upon the pioneering ethnomethodological studies of adult communication with alingual or disabled children (e.g. Goode 1994; Pollner and McDonald- Wickler 1985). Such studies have documented the myriad of ways in which parents, and other care-givers, have managed to construct complex and rewarding worlds of mutual intelligibility with children unable to communicate through conventional verbal means and who have typically been seen by professional outsiders as lacking any communicative competence. Crucially these relationships, and the methods which sustain and constitute them, are built through prolonged and close contact between the parties. The methods are typically tacit and defy explicit coding or measurement but they are nevertheless researchable and describable. Achieving intersubjectivity in “worlds without words” requires, as Goode shows, paying close and detailed attention to habitual routines, the spatialization of domestic life and tactile and embodied actions. We hypothesize that similar kinds of methods and communicative resources will be found in the relationships between humans and their companion animals. Indeed Goode (1994) makes this possibility explicit:

certainly one would not be surprised to see similar findings in studies of interaction with very young children, mentally retarded children, or, as long as we are careful about the juxtaposition of people with mental retardation and animals, in communicative interaction with other species. (pp. 89-90)

Finally, because we understand human-dog relations as emergent, neither given in biology or culture nor seen as systems or structures, and we are interested in the diversity of relationships, we advocate an approach that explores them biographically over their life course. The research programme we envisage would combine several types of data collection:

1. Regular observation of interaction in natural settings, especially the home and during walks. This includes rigorous and symmetrical attention to both human and dog actions and interactions and the development of an ethogram or method of notation specific to human-dog interaction.

2. Observation and analysis of video film sent from a video cameras set up in living areas of the home. Video data and analysis is important because it allows the researchers to capture fleeting actions and sequences of action in detail that can then be analysed using conversation analysis. The video film also makes it possible to check that observed interactions are not biased by the presence of the observer, although one can accept that ethnography and discourse analysis are each valuable in conducting qualitative research (Travers 2001: 105-6).

3. Interviews with the human partners to obtain their understanding of the relationship as it develops over time.

4. Diaries kept by human partners to record anything they think is of interest and new in the relationship with their dog.

Because companion animals figure differently and perhaps in compound ways through the life course, the research would involve sampling the life course at four points: childhood (age 8-12); stable adult partnerships prior to childbirth (say, 25-39); post-child-rearing period (50-60) and post-retirement (60 plus). We would also wish to make sure our cases are not all drawn from one type of place, but places that might in a general sense be typical or useful. It might, for example, be interesting in an Australian context to distribute one set of four case studies in each of two cities: a provincial city that is stable or in population decline (such as Hobart) and a major metropolitan city that is growing fast (an example would be Brisbane). Since commentators (e.g. Melson 1998) have argued that the degree of metropolitan growth and lifestyles have impacted both on human quality of life and companion animal recruitment, these distinctly different cities provide an important comparison and test.

Conclusion: companion animals and human well-being

This paper began by engaging with a problem raised by years of statistical research about the health benefits of companion animals. We know that animals have health benefits but we do not know how this is achieved. It has argued that there is no simple answer, because it is difficult to describe the animal-human relationship without being drawn into deep philosophical disagreements both about this and the problems raised by modernity. Matters are compounded by the fact that most qualitative studies have been conducted by researchers who love animals and are already committed to a philosophical, political or ethical viewpoint about the need to treat them humanely. For this reason, great care needs to be taken in taking the case studies and vignettes in ethnographies by, for example, Irvine and Haraway, which are convincing about the contribution animals have made to their well-being, as saying everything about the relationship or the mechanisms behind the health benefits. We have argued that a range of methodologies might be useful, including ethnomethodologically informed video-analysis. We have also argued against the intellectual fragmentation that appears to characterise the field of animals and society. It is unfortunate that post-Meadian symbolic interactionists, ethnomethodologists, conversation analysts and varieties of post-humanist philosophy do not engage with each other, given that the arguments are similar and it is possible to deepen an analysis through cross-fertilisation. It is also unfortunate that there is not more contact and collaboration between ethologists and sociologists. Ethograms that attempt to represent animal perspectives may have conceptual problems, but there should be more communication across the Great Divide.

Although ethology is a well-resourced scientific discipline, there seem few opportunities, outside the small-scale or autoethnographic study, for exploratory or pure research on the animal-human relationship employing qualitative methods in the manner we have described. However, it is worth concluding by again suggesting that understanding the animal-human relationship can have practical benefits as well as contributing to ethical and philosophical debate about the human condition. In fact, we see funding from government for research on the health benefits of animals as offering a good opportunity to bring together a team of inter-disciplinary researchers to conduct a large-scale, longitudinal project.

To date, explanations for the health benefits of companion animals have been conjectures based largely on statistical data and little additional data has been produced. Whereas previous studies of human-animal relations confined their attention to largely anthropomorphic constructions, interpretations and projections that humans might place on their relations with companion animals, which may in turn account for the health benefits, we hope that future research will also look closely at the nature of exchange and communication between humans and their animal companions. This will have to document how very specific relationships are established or created over time and investigate them as social spaces of interaction in which very tangible exchanges of communication, support and emotion take place. In other words, we are advocating an approach that in addition to asking what human-animal relationships mean (to humans) will also ask the more symmetrical and empirical question: what do they (i.e. both humans and animals) do and what ongoing partnerships are produced as a result. In the longer-term, it will explore whether there is anything about this activity and interaction that contains the bases for health and other benefits to humans.

We hypothesise that in such relationships people may experience two types of benefit. First, that companion animals and humans are capable of developing and have a propensity to develop a symbiotic relationship that entails both social support and mutual advantage. These relationships may develop and mature over time and cannot, therefore, be deemed to have a general effect based solely on ownership itself or co-presence per se. We want to discover how such relationships develop, how they are expressed as cultural repertoires of practice and how they coalesce into a social-spatial habitus. Second, we hypothesise that the social space of this interaction provides relief, or escape from, or perhaps an antidote to, concerns, tensions and anxieties that are produced in contemporary (human) society. Humans are drawn not only into relations with animals but into their world and a constructed parallel world of human-animal relations. On the face of it, dog owners spend considerable periods of time locked into this in-between world and they frequently report experiences of intense pleasure, but the temporal pattern and duration of this experience has not been recorded or analysed in a systematic or comparative framework. A well-resourced, inter-disciplinary project of this kind would be both valuable as a means of exploring philosophical debates and arguments, and might also contribute to improving human health.

Endnotes

i This figure (and all subsequent figures) is expressed in Australian dollars. Currently 1 Australian dollar equals 0.77 US dollars

ii Medicare is the name of the Australian Government’s publically funded health care system.

iii In recent years, this position has been challenged from a number of disciplinary perspectives. In particular, post-Meadian symbolic interactionists have argued that there are ways of communicating with animals in the absence of language and that animals have selves; see, for example, Myers (2003), Irvine (2004) and Sanders (2003). These studies suggest the need to look closely at the animal-human relationship. This is particularly relevant in studies about health effects where benefits are often seen as the fantasy creation of the human and what the dog is thinking or doing matters very little. We return to this issue, and these studies, in a later section of the paper.

iv As noted above, Crist (2002) considers Darwin’s research on earthworms, and Haraway writes about her own relationship with two dogs.

References

Alger, Janet. M. and Alger Steven F. (1997) “Beyond Mead: symbolic interaction between humans and felines.” Society and Animals 5(1): 65-81.

------. (1999) “Cat Culture, Human Culture: an ethnographic study of a cat shelter.” Society and Animals 7(3): 199-218.

Anderson, Warwick, Christopher Reid and Garry Jennings (1992) “Pet Ownership and Risk Factors for Cardiovascular Disease.” Medical Journal of Australia 157: 298-301.

Arluke, Arnold and Clinton R. Sanders (1996) Regarding Animals. Temple University Press Philadelphia:

Batson, Kathryn, Barbara McCabe, Mara Baun and Cindy Wilson (1998) “The effect of a therapy dog on socialization and physiological indicators of stress in persons diagnosed with Alzheimer’s disease.” Pp. 203-216 in Companion Animals in Human Health, edited by C. C. Wilson and D. C. Turner. Thousand Oaks: Sage.

Bekoff, Marc, Colin Allen and Gordon Burghardt (2002) The Cognitive Animal: Empirical and Theoretical Perspectives on Animal Cognition. Cambridge, Mass.: MIT Press.

Corson, S. A. and Corson E. O’Leary (1980) “Pet animals as nonverbal communication mediators in psychotherapy in institutional settings.” Pp. 87-101 in Ethology and Nonverbal Communication in Mental Health, edited by S.A. Corson and E. O’Leary Corson. Oxford: Pergamon Press.

Crist, Eileen (2002) “The inner life of earthworms: Darwin’s argument and its implications.” Pp. 4 - 8 in The Cognitive Animal: Empirical and Theoretical Perspectives on Animal Cognition edited by M. Bekoff, C. Allen and G. Burghardt. Cambridge, Mass.: MIT Press.

Deleuze, Giles and Felix Guattari (1999) A Thousand Plateaus. London: Athlone Press.

Flynn, Clifton P. (2000) “Battered Women and their Animal Companions: symbolic interaction between human and nonhuman animals.” Society and Animals 8(2): 101-127.

Franklin, Adrian S. (1999) Animals and Modern Cultures. London: Sage.

------. (2005) Animal Nation: The True Story of Animals and Australia. University of NSW Press.

Freidmann, Erika, Sue Thomas and Timothy Eddy (2000) “Companion animals and human health: physical and cardiovascular influences.” Pp. 125-142 in Companion Animals and Us – Exploring the relationships between people and pets edited by A. Podberscek, E. Paul and J. Serpell. Cambridge: Cambridge University Press.

Garrity, Thomas and Lorann Stallones (1998) “Effects of pet contact on human well- being: review of recent research.” Pp. 3-22 in Companion Animals in Human Health edited by C. C. Wilson and D. C. Turner. Thousand Oaks: Sage.

Goode, David (1994) A World Without Words: the social construction of children born deaf and blind. Philadelphia: Temple University Press

------. (2007) Playing with My Dog Katie: An Ethnomethodological Study of Canine- Human Interaction. Ashland, Ohio: Purdue University Press.

Haraway, Donna (2003) The Companion Species Manifesto. Paradigm Press.

------. (2007) When Species Meet. Minnesota: University of Minnesota Press,.

Headey, Bruce (1999) “Health Benefits and Health Cost Savings due to Pets: preliminary estimates from an Australian national survey.” Social Indicators Research 47: 233-243.

Irvine, Leslie (2004) If You Tame Me: Understanding our Connection with Animals. Philadelphia: Temple University Press.

Jennings, Garry, Christopher Reid, Irene Christy and Warwick Anderson (1998) “Animals and cardiovascular health.” Pp. 161-172 in Companion Animals in Human Health edited by C. C. Wilson and D. C. Turner. Thousand Oaks: Sage.

Kaminski, Julianne, Joseph Call and Julian Fisher (2004) “Word Learning in a Domestic Dog: evidence for fast mapping.” Science 304: 1682-83.

Latour, Bruno (1993) We Have Never Been Modern. Wheatsheaf

Laurier, Eric, Ramia Maze and Johan Lundin (2006) “Putting the Dog Back in the Park: animal and human mind-in-action.” Mind, Culture and Activity 13(1): 2-24.

Melson, Gail (1998) “The role of companion animals in human development.” Pp. 219-236 in Companion Animals in Human Health edited by C. C. Wilson and D. C. Turner. Thousand Oaks: Sage.

Myers, Olin E. (2003) “No Longer the Lonely Species: a post-Mead perspective on animals and sociology.” International Journal of Sociology and Social Policy 23(3): 46-68.

Newby, Jonica. (1999) The Animal Attraction: Humans and their Animal Companions. Sydney: ABC Books.

Pepper John and Barbara Smuts (1999) “The evolution of cooperation in an ecological context: an agent-based model.” Pp. 205-213 in Dynamics in Human and Primate Societies edited by T. Kohler and G. Gumerman. Oxford: Oxford University Press.

Pickering, Andrew (1995) The Mangle of Practice: Chicago: University of Chicago Press

Pollner, Melvin and Lynn McDonald-Wikler (1985) “The Social Construction of Unreality: a case study of a family’s attribution of competence to a severely retarded child.” Family Process 24: 241-254.

Roberts, Felicia (2004) “Speaking to and for Animals in a Veterinary Clinic: a practice for managing interpersonal interaction.” Research on Language and Social Interaction 37(4): 421-446.

Sanders, Clinton R. (1993) “Understanding Dogs: caretakers’ attributions of mindedness in canine-human relationships.” Journal of Contemporary Ethnography 22(2): 205-26.

------. (1999) Understanding Dogs. Philadelphia: Temple University Press.

------. (2003) “Actions Speak Louder than Words: close relationships between humans and nonhuman animals.” Symbolic Interaction 26(3): 405-426.

Serpell, James, editor (1999) The Domestic Dog. Cambridge: Cambridge University Press.

Shapiro, Kenneth (1990) “Understanding dogs through kinaesthetic empathy, social construction and history.” Anthrozoos 3 (3): 184-195.

Smuts, Barbara (2001) “Encounters with Animal Minds.” Journal of Consciousness 8(5-7): 87-113.

Tannen, Deborah (2004) “Talking the Dog: framing pets as interactional resources in family discourse.” Research on Language and Social Interaction 37(4) 399-420.

Travers, Max (2001) Qualitative Research Through Case Studies. London: Sage.

Veevers, Jean (1985) “The social meaning of pets: alternative roles for companion animals.” Pp. 11-29 in Pets and the Family edited by M. B. Sussman. New York: Haworth.

Wilson, Cindy C. (1998) “A conceptual framework for human-animal interaction research.” Pp. 61-90 in Companion Animals in Human Health edited by C. C. Wilson and D. C. Turner. Thousand Oaks: Sage.

Citation

Franklin, Adrian, Michael Emmison, Donna Haraway and Max Travers (2007) “Investigating the therapeutic benefits of companion animals: Problems and challenges.” Qualitative Sociology Review, Vol. III Issue 1. Retrieved Month, Year (http://www.qualitativesociologyreview.org /ENG/archive_eng.php)