THE SEVERING OF A PET ATTACHMENT: RELATIONSHIP …
RUNNING HEAD: ATTACHMENT STYLE AND PET LOSS
Attachment Style and Pet Loss
In the Context of Attachment Theory
Lisa M. Orsini
Pacific Graduate School of Psychology
Table of Contents
Introduction 4
Background to the Problem 6
What aspects of Attachment are found in Grief? 6
The Companion Animal Bonding Scale 9
Defining Attachment Bonds 10
Re-Evaluating the CAB Scale 12
Attachment and Attachment Behaviors 14
Limits to Attachment Theory 17
Literature Review 20
Exploring the human-pet bond 20
Literature Review 25
Exploring the Broken Bond 26
Individual Differences 34
Bereavement Theory 37
Summary of Problem 41
Bowlby’s Attachment Theory 43
Question of Instinct 44
Attachment Behavior 45
Cybernetic Theory 46
Adaptive Value of Attachment Behavior 47
An Attachment Bond 48
Proximity Seeking 49
Control Theory 50
Pre-programmed Behavior 51
Goal Correction and Termination 52
Responding to Security Threats 53
Planning For Conditions 54
Set Goal of Proximity 54
Hierarchical Organization 55
Mediating Influences 56
Secure Base 57
Attachment Styles 57
Attachment Bond Versus Affectional Tie 59
Ethology 60
Felt Security 61
Studying Pet Loss 62
Attachment Theory 63
Methodology 75
Research Participants 76
Instrumentation 76
Procedures of Data Collection 76
Statistical Analysis 76
Apparatus 76
REFERENCES 84
Introduction
The relationship between a human and a pet has the potential to develop the qualities of an attachment relationship (Lagoni, Butler, & Hetts, 1994). Humans and their pets develop a unique relationship based on the quality of a care-giving and care-receiving environment that is similar to the attachment bond [1] most recognizable in many human relationships (e.g., infant-mother; romantic partners). According to attachment theorists, the experience of the attachment bond is precipitated by evolved instincts that have survival value and are common among the animal species. The survival value of these instincts is embedded in behaviors that establish and determine the quality of relationships between (human-pet) and within (human-human) species of animals (Bowlby, 1969/1982). The propensity for relatedness in the social animal species is discussed in studies of the evolution of behavioral traits (Hinde, 1979) and motivations of different animal species (Baerends, 1976). This study highlights the evolutionary and ethological foundations of attachment behavior, focusing on the phenomenon of attachment bonding and its effect in the grief response to losing a pet. Attachment theory as postulated by John Bowlby (1980) is utilized as a framework to guide an understanding of the origins of creating and breaking bonds, as is Ainsworth’s (1989) study of bonding and relationships similar to it.
According to Bowlby (1980), Parkes (1972) and Ainsworth (1989), understanding the quality of an attachment relationship is indispensable for understanding characteristic behavioral responses observed in grief reactions to loss. This project intends to test this theory as it relates to pet loss and attachment, and explores some related issues. Specifically, this study will utilize attachment style as a predictive measure of the human grief response to support the theory that attachment bonding occurs in human-pet relationships and variability in grief is dependent on the nature of an attachment relationship. This is a retrospective analysis following a similar path to that taken by Bowlby (1980) who observed very young children’s responses to the loss of the mothering person when refining his theory of attachment.
Background to the Problem
Fogel (1983) notes that the study of heterospecific attachment bonding is a challenging task, perhaps because, as Bolin stated (1987), “it does not lend itself easily to compartmentalization” (p.26). One of the more difficult challenges involved in such a study is finding an easily agreed upon method of measuring attachment bonding. According to Ainsworth (1989) a grief response to loss of an attachment relationship “demonstrates that the attachment bond has endured” (Ainsworth, 1989, p. 711; see also Archer, 2001; Bowlby, 1969/1982). Thus, studying grief following the death of a pet is one way to measure the existence and quality of a previous attachment bond.
What aspects of Attachment are found in Grief?
In a recent study (Planchon et al., 1996), a measure of death depression (Templer, Lavoie, Chalgujian, & Thomas-Dobson, 1990) and a measure of general attitudes toward pets (Templer, Salter, Dickey, & Baldwin, 1981) that overlaps with common descriptors of pet attachment (e.g., love and interaction) were utilized to determine factors related to pet grief. The study focused on human variables such as living alone and human-pet variables such as time spent with a pet, venturing beyond clinical impression to isolate and correlate a variety of measurable variables associated with pet grief. Not considered in this study, however, are qualities inherent to the relationship between an owner and his/her particular pet. Measurement of the relationship as a factor in grief has received relatively minor attention in the pet attachment literature.
Planchon & Templer (1996) summarized their goals in a review of the literature that includes Cowles’ (1980) contention that grief resulting from loss of a pet is a function of the owners’ attachment relationship with the pet, Thomas’s (1982) argument that grief is correlated with multiple variables including the relationship between the owner and the pet, and Keddie’s (1977) claim that grief after pet loss is “pathological.” The outstanding and common denominator of these arguments, attachment, was however, not addressed directly in the study. A measure of the relationship between owner and pet on a deeper psychological level could help clarify the significance of the fact that Planchon & Templer (1996) found no correlation between several human-pet variables, and a high correlation between human-human variables, with the amount of grief displayed.
The present study proposes that studies of grief following pet loss include a measure of pet attachment and a measure of the attachment style of the pet owner to improve the quality of grief findings, since attachment variables may overlap with human variables such as depression related to a death experience (e.g., measured by Death Depression Scale) and are not adequately accounted for by a scale that measures general positive attitudes toward pets (e.g., measured by Pet Attitude Scale).
The latter appears to be a pervasive problem in pet research in that measures of attachment to pets often utilize independent variables that do not fully explain the nature of the attachment bond. Variables such as positive attitudes toward pets (e.g., Pet Attitude Scale, Planchon et al., 1996), used in the previous study, are not equivalent to attachment bonding, and may even be more indicative of factors related to owning a pet (Poresky et al., 1988). Much of the newer research conducted on theories of grief[2] acknowledges attachment as an antecedent and perhaps cause of grief, validating the need for an adequate attachment measure to be used when researching pet grief. Without an objective measure that operationalizes and then quantifies the difference between variables externally similar to attachment (e.g., either ownership, affection, or positive attitude toward a pet) and companion animal [attachment] bonding (Poresky et al., 1988), important nuances in the grief reaction to losing an attachment relationship can be missed.
The Companion Animal Bonding Scale
Poresky et al. (1988) operationalized companion animal bonding in order to distinguish it from “pet ownership” (p.420). They utilized The Companion Animal Bonding Scale (CAB) (Poresky, Hendrix, Mosier, & Samuelson, 1987) devised specifically for this purpose to provide a new model for looking at pet ownership. Results from their study and others (Garrity et al., 1989) indicate the degree to which an analysis of attachment (Melson, 1990) was missing in earlier reports such as those documenting the health benefits of owning a pet (Friedmann et al., 1980; Mugford & M’Comisky, 1975; NIH 1988 as cited in Poresky & Daniels, 1998). The issue of using only subjective, non-quantifiable measurement has been noted as a problem in studies of pet support (Garrity & Stallones, 1997), and has been referred to as an issue in outcome measures of pet attachment (Wilson, 1994). In studies that have utilized an objective attachment measure, factor analysis reveals the attachment variable to be a multi-dimensional construct (Triebenbacher, 1999). Are the constructs identified effective for explaining the effect of the attachment relationship on some dependent variable (e.g., grief)?
Defining Attachment Bonds
In an effort to describe an attachment bond and differentiate it from an affectional tie or bond, Ainsworth (1989) points out that an attachment involves seeking “security and comfort” from the attachment figure (p.711; see also Cassidy, 1999).
Researchers such as the developers of the Companion Animal Bonding scale made an effort to find a “better operational definition of ‘bonding’ than pet ownership” (Poresky et al., 1988, p.420, having discovered a need for a reliable construct for measuring attachment bonding. Their study of the effect of early bonding experiences on later adult attitudes toward pets found that the strength of the age of first pet variable and the Companion Animal Bonding scale variable each resulted in a stronger correlation with contemporary attitudes toward a pet than the pet ownership variable alone (Poresky et al., 1988).
The results of this study are no doubt exciting, in terms of advancing the state of the art in measurement of attachment bonding. However, there appear to be some problems in regard the application of the theory of attachment when designing their measurement tool, the CAB scale.
A precise operational definition rests on the underlying theory of the construct, yet the literature is inconsistent with regard to defining attachment bonds (Poresky et al., 1997, 1998, 1999 as cited in Triebenbacher, 1999). There are theoretical and practical problems with using the Companion Animal Bonding scale to operationally define attachment bonding as understood by Ainsworth (1989) and Bowlby (1969/1982). In the CAB scale, Poresky et al. (1988) simplified the term bond, and appear to have taken it out of the larger context of attachment theory (p. 420).
Speaking on the difference between attachment and relationships similar to it (e.g., affectionate, “sociable” or “affiliative” behaviors[3]) Ainsworth (1967), then Bowlby (1969/1982) stressed the importance of secure base as a set goal of the child’s seeking proximity to his mother, the attachment figure (Bowlby, 1969/1982, p.373). Ainsworth (1982/1989) expanded on this concept stating that seeking security and comfort in the relationship with the attachment figure is essential to its property as a bond and for distinguishing it from an “affectional tie” (Ainsworth, 1973, p. 1 as cited in Poresky et al, 1988, p.420), “emotional tie” (Fogel & Melson, 1988, p190 as cited in Melson, 1990), or “affectional bond” ((Ainsworth, 1989, p. 711). Poresky et al (1988) overlooked this feature, defining “bonding” as simply “the establishment of a relationship which parallels, but may not be as strong as, Ainsworth’s concept of interpersonal attachment” (p.420). Their reference to a bond as something weaker than Ainsworth’s (1973) definition implies that a bond is a casual affectionate connection rather than a powerful element that ties one individual to another.
Re-Evaluating the CAB Scale
Triebenbacher (1999) re-evaluated the Companion Animal Bonding scale in an effort to extend the generalizability of the results of Poresky et al.’s (1988) study to a larger sample with a broader range of developmental levels, and to more precisely assess which factors of attachment are measured by the scale. To do this she revisited the construct of attachment as understood in attachment theory. The results of her study indicate that the original scales tap into the following attachment dimensions: emotional bonding, proximity, and caretaking. These results coincided with those reported by Poresky et al. (1987), but not with those reported by Poresky (1997); the latter study identified only two (e.g., involvement/relationship factor and sleeping arrangements/travel patterns factor) of the three of factors cited by Triebenbacher (1999). The study by Triebenbacher (1999) supports the view that the human-animal attachment bond is a multi-dimensional construct since several factors of attachment were identified.
Information related to the quality of human-companion animal attachment was not addressed in the latter study, however has been noted to significantly improve understanding of the construct of attachment. Ainsworth’s (1972) observation of the difference between the strength and quality of an attachment relationship clarifies the defining feature of an attachment bond - secure base. This feature does not appear to have been captured in the results derived from use of the CAB scale in Triebenbacher’s (1999) study or in studies (Poresky et al., 1987, Poresky, 1997) similar to it.
Ainsworth (1972) refers to Hinde’s (1979) concept of “penetration” to elucidate the conceptual difference between the strength and quality of a bond. According to Hinde (1979), the degree to which the individual (e.g., significant attachment) emotionally “penetrates a variety of aspects of another person’s life” (as cited Shaver & Cassidy, 1999, p.13) is significant. Indeed, attachment theory recognizes that relationships evolve over time and behaviors evidenced in relation to an attachment figure change as a function of the degree of the attachment bond. Hinde’s (1979) use of the word penetration is exemplified in the scenario of a child who eventually leaves home, but remains attached to his/her parents. On a different level, an attachment may, broadly speaking, be either secure or insecure. For example, a child who clings to mother is not necessarily securely attached, but rather may be displaying signs of a lack of an ability to use mother as a secure base (Cassidy & Shaver, 1999). This is an important distinction to be made in a study concerned with the effect of attachment on a particular outcome (e.g., health, grief etc.) since the quality of an attachment has been shown to differentially effect a variety of outcomes such as a child’s response to separation from his/her mother (Ainsworth, 1979) and in the area of pet research even the well-being of an individual (Wilson & Turner, 1998).
Attachment and Attachment Behaviors
Melson (1990) studied children’s attachment to their pets using a scale derived from the Companion Animal Bonding Scale (Poresky et al., 1987). In a similar manner to Poresky et al. (1988) Ainsworth’s (1989) notion of attachment bonding appears to have been taken out of the larger context of attachment theory. Melson (1990) extracts an interpretation that appears to rely more on a definition of attachment behavior than attachment bonding (Bowlby’s (1969/1982)[4]. He refers to the following interpretation of attachment: “a lasting emotional tie between people such that the individual strives to maintain closeness to the object of attachment and acts to ensure the relationship continues” (Fogel & Melson, 1988, p. 190 as cited in Melson, 1999, p.92). Thus, proximity seeking behavior defines attachment in Melson’s study. Proximity seeking is no doubt an attachment behavior, yet it alone is insufficient for qualifying a relationship an attachment in the sense of a bond (see Cassidy, 1999 in Cassidy and Shaver, 1999).
Ainsworth (1963), for instance, notes that attachment behaviors are mediated at times by exploratory behaviors[5], interacting dynamically to balance the infant’s need for a secure base. Approaching behaviors such as those that aim to seek proximity to a secure base (e.g., mother) are not at all times the “set-goal” (Bowlby, 1969/1982, p.373) of the infant, who seeks “not an object (e.g., mother), but rather a state – a maintenance of the desired distance from the mother, depending on the circumstances” (Bowlby, 1969/1982 as cited by Cassidy, 1999, p.6). Thus, Melson (1990) utilized the concept of proximity seeking to measure a component of the attachment system, but not attachment per se. He makes use of an important subtlety in the study of attachment, which is that it is not easily inferred from “presence or absence of attachment behavior” (Cassidy, 1999, p.12). However, he errs in his reference to proximity seeking as sufficient for defining a relationship an attachment. The theory underlying his definition is incomplete, since the concept of attachment bonding involves six criteria (e.g., to make an attachment bond), not just that one.
Despite the shortcomings in these studies, Poresky et al. (1987/1988), Triebenbacher (1999), Fogel & Melson (1988), Melson (1990) and pointed out that attachment is an important element in human-pet relationships. These researchers made important strides in developing a more fitting operational definition to use in cross study research. Respectively, they showed that there is a difference between owning a pet and having an affectional tie (not a bond) to a pet (Poresky et al, 1988), that several important factors of attachment are measured by the CAB scale (Triebenbacher, 1999), and there is difference between attachment behaviors and the attachment behavioral system (Melson, 1990; see also Ainsworth, 1989).
Limits to Attachment Theory
Post-modern researchers interested in theories of attachment have begun to debate the limits to using Bowlby’s (1969/1982) theory of attachment to explain the nature of the child’s ties to the attachment figure[6] when the relationship is viewed as dyadic (Ainsworth, 1989 as cited in Cassidy et al., 1999, p.12). Ainsworth (1982) reflects:
That there is a “relationship” between mother and child, in Hinde’s (1979) sense, from the time of the infant’s birth onward, and that the nature of this relationship stems from the interaction between them, is not to be gainsaid, but neither the mother-to-infant bond nor the emergent infant-to-mother attachment seems to me to comprehend all the important aspects of this relationship. (p. 24 as cited in Cassidy, 1999, p. 17).
Bretherton (1985) specifies:
A representational view of relationships...underscores that the two partners have, in another sense, two relationships: the relationship as mentally represented by the attached person and by the attachment figure (p. 34 as cited in Cassidy, 1999, p. 17).
Looking at attachment from a unidirectional (as opposed to a bi-directional) perspective and continuing to develop more precise methods for measuring the attachment bond proves useful for understanding how grief is experienced by the survivor of a lost attachment relationship. In noting the importance of the principle attachment figure, as a means to support, not censure (Lamb, Thompson, Gardner, & Charnov, 1985), Bowlby’s position on multiple attachment figures, Cassidy (1999) refers to an undoubtedly memorable and moving passage written by Bowlby (1980):
About four weeks after mother had died, [four-year-old] Wendy complained that no one loved her. In an attempt to reassure her, father named a long list of people who did (naming those who cared for her). On this Wendy complained aptly, “But when mommy wasn’t dead I didn’t need so many people - I needed just one” (Bowlby, 1980, p. 280 as cited in Cassidy, 1999, p.18).
This poignant passage delicately captures how powerfully the attachment figure is felt in the attachment relationship. Indeed, the relationship as implied in the passage appears to have been experienced by the child as something more powerful than care (as in terms of care-giving), since as noted, many people care for the child.
It would be useful to consider the reciprocal qualities of pet attachment (e.g., the perception of care received). However, similar to an infant or child, a subjective state such as feeling attached is difficult to elicit (e.g., Melson, 1999; Melson, 1988), and more so to quantify (e.g., Poresky et al., 1987) in a pet. To avoid this dilemma, the present study retrospectively examines human-pet bonding from the point of view of the bereaved. The need for a clearer understanding of bereaved pet owners provides an opportunity to find better ways to normalize or validate the experience of losing a pet - a concern which should not be dismissed given the continued stigmatization of those who grieve for their lost pets (Weismann, 1999; Boyce, 1998). Hence, the goal of the present study is to add to our understanding of attachment and grief as a natural part of the process of forming and losing a significant attachment relationship (Cassidy, 1999).
Literature Review
Companion animals have played a valuable role in human life for thousands of years (Boyce, 1998; Sable, 1995), dating as far back as prehistoric times (Serpell, 1989). Narrative literature (e.g., Carmack, 2003; Sharkin & Bahrick, 1990; Sorenson & Ross, 1998), self-report analyses (e.g., Albert & Bulcroft, 1987) and case studies (Voith, 1985) on this subject suggest that pets function as much more than objects or attendants (Singer, 1990 as cited in Boyce, 1998) for people. The following literature review provides a strong foundation for a theory of human-pet attachment bonding that is similar to theories of human-human bonding relationships.
Exploring the human-pet bond
Scales/Surveys
Loughlin & Dowrick (1993) undertook a descriptive exploratory study utilizing a survey design to determine the psychological benefits of owning birds. The results of their study found that (look for study).
In a survey of pet owners and non-pet owners in Providence Rhode Island, Albert & Bulcroft (1987) found that individuals who are strongly attached to their pets anthropomorphize their pets. Individuals least likely to anthropomorphize their pets were in first marriages or had two or more children. Adults in first marriages also scored lower on pet attachment (as measured by a scale they devised) than did single, divorced or widowed people. The authors concluded that lack of a present spouse or romantic partner may increase an individual’s need for emotional fulfillment. Higher subjective attachment to pets was found among people without children and among parents with children who live outside the home.
In Horn & Meer’s (1984) survey, a higher rate of pet owners than non-pet owners reported personal qualities such as being interesting and friendly and feeling a subjective sense of safety. In contrast to non-pet owners, pet owners reported feeling less lonely and remote from other people and believed that the presence of a pet increased relaxation. Pet owners also viewed the pet as a very important member of the household. Failure to control for demographic variables, however, has limited the usefulness of these results (see Poresky et al., 1988).
Stevens (1989) looked at attachment among urban and rural eighth grade students using the Pet Attachment Survey (PAS), an instrument he designed to assess human animal attachment levels for conventional pets. The results of this study indicated that urban children were more attached to their pets than were rural children.
Katcher (1980) surveyed children and adults from an East Coast sample and adult magazine respondents. From the combined total, 80.9 percent reported that they considered their horse(s) a member of the family. This number is contrasted with 48 percent of owners of other types of animal. In addition, 89.8 percent of her East Coast sample reported that they talk to their horse(s) as persons, as did 88 percent of dog owners. Many owners of both types of animals (dog or horse) reported confiding in their pets: 28 percent of the dog owners, 53.8 percent of horse owners under twenty in the East Coast sample and 57.7 percent of the adults in the overall sample. The percentage of owners who granted the animal status as a family member (76.9) was not much higher than the percentage who called their animals simply, pet (62.2), which suggests that the use of the term pet is not synonymous with family member. The difference in how animals were named was indicative of the amount of structured activity engaged in with the animal(s). The more activity engaged with an animal the more likely an owner considered an animal a family member, as opposed to just a pet.
Stallones et al. (1990) conducted a national survey in order to find a more representative sample of people who report on the beneficial value of attachment to pets, and to examine the relationship of attachment to ownership (Marx, 1984). Their study found that attachment to pets, but not pet ownership, was inversely associated to human social support, with stronger attachment to pets being reported by those who had less human social support and higher emotional distress.
Only one other national survey (Garrity et al., 1989), had been carried out, a fact which prompted the authors to conduct this survey. They stated: “descriptive studies using representative samples...were needed in order to proceed to test specific hypotheses” (p.100). Indeed, more representative samples would expand our understanding of the benefits of companion animals throughout the human life cycle, a problem that should not be overlooked given the following example. Although many survey studies have found positive correlations between ownership, attachment, and good health variables in the elderly (Lago, Connell, & Knight, 1983; Mugford & M’Comisky, 1975; Ory and Goldberg 1983) researchers using nationally representative samples (e.g., Garrity et al., 1989) found that only elderly individuals with recent negative life stressors and less confidants had more depressive symptoms (as cited in Stallones et al., 1990).
Staats, Miller, Carnot, Rada & Turnes (1996) utilized the Miller-Rada Commitment To Pets scale (MRCPS) and the Pet Relationship Scale (PRS) to assess their hypothesis that commitment to pets is a different, but overlapping construct of attachment to pets. This study found positive, moderate correlations between owner commitment variables and attachment variables, such as pets contributing to quality of life. However, commitment to pets was differentially associated with age of owner, gender, marital status and owner’s health status (e.g., younger, single persons in good health were more committed, but less attached to their pets). Attachment, on the other hand, was more strongly associated with female gender. Lack of a significant correlation between income and commitment paralleled survey results found by Albert and Bulcroft (1987), which suggests that ownership, but not attachment is affected by income.
Bodsworth & Coleman (2001) addressed the issue of whether children who have reduced access to parents have a stronger attachment bond with their pets. Using the Companion Animal Bonding Scale (CAB), this study found that children in single-parent families had higher levels of attachment to dogs that those in two-parent families and that children in early childhood stage (age range was 3-to-11) in single-parent families showed higher levels of attachment than those in two-parent households.
Case Studies
Voith (1985), a veterinarian, reported four case studies documenting the potential for strong attachment relationships between humans and companion animals. Pet owners in her study appeared to demonstrate a strong desire to keep their pets despite incurred monetary costs (commitment construct) and/or the risk of danger that keeping the pet potentially posed to children or themselves. Voith suggests that certain types of attachments, like those found in her four case studies, have the quality of a parent-child relationship.
Literature Review
The scientific study of pet loss grief has only recently begun to proliferate in psychological literature concerning the topic of attachment, and grief (Archer, 1994). Relying on relatively new validated instruments, some studies have focused on examining correlates of grief associated with type of pet (Planchon et al. (1996), antecedents of grief (Gosse & Barnes, 1999), or compared the experience of loss of a human to loss of a pet (Archer, 1994; Gerwolls & Labott, 1994). These studies provide insight into understanding grief within the context of attachment theory.
Exploring the Broken Bond
Surveys
Quackenbush et al. (1984) examined emotional responses of pet owners and found that pet death is experienced in a manner similar to human death in terms of sleep difficulty, days missed from work and other psychological and social difficulties. Planchon et al. (1996) utilized a retrospective design to determine correlates of pet grief. This study found that human variables (as opposed to human-pet variables) are more associated with pet grief. The authors conclude that psychologically well-adjusted individuals may fare better than less adjusted individuals with regard to amount and kind of grieving following pet death. Fogel et al. (1998) surveyed veterinarians in an attempt to increase awareness of the pervasiveness of severe grief among professionals practicing euthanasia. Their survey looked at the association between sex, age, years in practice and feelings after euthanizing a pet. The results of their study indicate that a high percentage of female (28% out of 82) and male (72% out of 82) veterinarians have cried (86% females, 41% males), felt the need to be alone (52% females, 34% males), and felt depressed (66% females, 44% males) after euthanizing a pet. This study provides information on symptoms of grief, but does not include attachment as a variable; it is included in this section because symptoms of grief might have been precipitated by an attachment relationship. The results of their study highlight the significance of asking the question: What determines an attachment relationship?
Scales
Using two validated instruments [the Pet Attachment Survey (PAS) and the Grief Experiences Inventory (GEI)], Gerwolls & Labott (1994) assessed the difference between loss of a human and loss (e.g., parent, spouse or child) of a pet. Their study found that loss of a pet after two weeks and 8 weeks revealed grief scores that are similar to those reported after losing a human relationship (Sanders et al., 1985). Also, pet loss subjects did not score significantly lower than individuals suffering a human loss after 8 weeks (except on a measure of Despair and Somatization). Also, 6 months following pet loss subjects did not score significantly lower than individuals who lost a human relationship (Sanders et al., 1985) on any clinical bereavement scales. The authors concluded that losing a pet is very similar to losing a human relationship.
Gosse (1989) also used the GEI and PAS. His study found degree of attachment to a pet to be a strong predictor of psychological adjustment to the death of a pet. He found a significant positive relationship between attachment and the grief response.
Gosse & Barnes (1994) investigated the antecedent of the grief response to losing a pet. This study utilized three clinical bereavement scales and three validity scales from the GEI, the PAS, and a Schedule of Recent Experience (SRE). The results of their study found partial support for a hypothesis that pet owners without children evidence more intense grief. A significant difference was found in the predicted direction only for a Social Isolation index. This finding appears to support survey results from Albert & Bulcroft (1987), which indicated that pet owners without children living home believe pets encourage them to interact more socially with others. Another hypothesis predicting that living alone would evidence higher levels grief did not reach statistical significance. With regard to this latter finding, however, once owners living alone or with one other adult were compared to owners with more than one other adult and/or living with children, a higher score for Social Isolation was obtained.
Archer (1994) designed a 40-item questionnaire based on some of the published descriptive data reported in the grief literature (e.g., Parkes, 1972,1985; Shuchter, 1986) in order to examine the psychometric properties of bereavement. Attachment variables included were age and circumstances of pet owners, the type of pet, and length of ownership. An emotional attachment variable was included and measured by asking subjects questions about the importance of their pet. Those who described their pets with descriptors such as “baby” or “loved” were given higher ratings than “friend” or “protector,” respectively). Results from this study support the hypothesis that grief following pet death is similar to grief following loss of a human. In this study, over half the owners reported feeling numbness and/or disbelief after hearing of the death of their pet and reported being drawn toward reminders of the pet. Half to four fifths reported being preoccupied with thoughts about the pet, and obsessive rumination about events before the loss. In addition, a minority (one quarter) reported having an urge to search for the lost pet, tried to avoid thinking about the pet or imagined the pet still alive and nearby as a means to assuage level of grief. A quarter of the respondents also reported feelings of anxiety, depression, anger and self-reproach. Finally, grief loss scores were not associated with attachment variables such as time since the loss, length of time owning pet, and type of pet, but instead with attachment affective variables. Especially high grief scores were associated with the variable living alone. The authors pointed out that a level of affect familiar to those who grieve the loss of a human was missing in this study, which may have been due to the measurements used. Additional caveats of this study are that the sample size was not nationally derived, limiting the generalizability of the findings, and over half of respondents were representative of pet owners who regarded their companion animal as simply a pet (51/88), which presents a bias in favor of those who are less emotionally involved with their pets. In addition to these problems, it is worth noting that Archer (1994) gave the descriptor “protector” a lower rating than friendship. The author of the present study conjectures that, due to the fundamental security and comfort-seeking component of an attachment, a descriptor such as protector may be more indicative of a bond than friendship (see also Ainsworth, 1989).
Counseling
Carmack (1985) found that bereaved individuals seek counseling after pet loss and report grief symptoms that are similar to those reported by parent’s grieving the death of a child (as cited in Jarolmen, 1998). Weisman (1990-91) found several common themes among bereaved owners self-referred for grief counseling following pet loss. These include anthropomorphism, empathy, regret and flashbacks as well as a tendency to view the deceased animal as a primary attachment figure. From his perspective, the extent of grief in bereaved pet owners may approach clinical proportions, especially in individuals who “valued their pets more than friends or relatives” (p.246). More severe grief symptoms included feeling preoccupied with the pet, mistaking shadows and sounds for the lost pet, and not disposing of pet’s belongings. He concluded that animal bonding is a unique relationship unlike other human relationships, and suggests individual and support group counseling as an effective method for reducing grief symptoms. Ross et al. (1998) share individuals’ grief reactions to loss of a pet and offer advice to practitioners on how to handle a wide variety of pet loss experiences (e.g., lost, stolen, accidental death, illness, euthanasia) and how to understand the impact of breaking an attachment bond. The researchers also discuss individual differences in pet grief experiences throughout the life cycle. Carmack (1985) also reported that some clients told their veterinarian surgeons that they would rather lose their spouse than their pet. Ross et al. (1988) reported that in many cases clients who lose a pet after a veterinarian procedure develop a great deal of anger and it is often directed toward the veterinarian doctors (see also Carmack, 1985). Instances such as this have also been reported in grief after experiencing the loss of a human attachment relationship (Archer, 1994).
Case Studies
Keddie (1977) described pet ownership as “prophylactic” and pet grief as “pathological,” to emphasize the significance of pet companions on the emotional well being of the owner (p.21). In a case study format, he reports negative mental health consequences resulting from loss of a “special” relationship with a pet. He describes a special relationship, as one in which there is evidence of “overdependence” to point out the effect of the shrinking family unit (e.g., single parents, single or widowed adults living alone) on mental health and perhaps to emphasize his personal understanding of the role pets play in people’s lives. Rynearson (1978) also referred to the bond observed between human and pets as an “overdetermined” misplaced need for attachment in humans. He pointed out that a biological theory is necessary for understanding the role that animals play in human lives and to understand reciprocal attachment needs between humans and pets. He theorizes that early frustration in attachment relationships leads to pathological levels of need for nurturance that manifest in relationships with pets. He provides case examples of grief reactions to the loss of a pet to support his argument that prolonged and intense grieving responses to the loss of pets are the consequence of a “distrust” of human attachment relationships (p. 550-551).
Interviews
Cowles (1985) gave in depth interviews to adults grieving the loss of a pet. Her subjects appeared to lack commonly experienced guilt reactions (Harris, 1984; Keddie, 1977), but did report memories of their deceased pet (i.e., stimulated by activities normally engaged in with the pet), searching behaviors (Bowlby, 1980; Parkes, 1998), obsessive rumination of the events leading up to the pet’s death, thoughts of previous significant losses both animal and human (Harris, 1984) and fears of losing of control or “going crazy” (Parkes, 1998).
Individual Differences
As has been noted in studies of attachment style (e.g., Hazen & Shaver, 1987; Main, Kaplan & Cassidy, 1985) and separation and loss or grief (e.g., Hazen & Shaver, 1992), individuals differ in their experience of the attachment relationship and the principle figure in that relationship. Such differences have been implied in studies of pet attachment. Poresky et al.’s (1988) study found that childhood bonding with a pet influences later pet bonding relationships. This finding might also be an indicator of continuity in attachment styles since early bonding was held as an indicator of later positive attitudes toward pets. Evidence of continuity in attachment related behaviors from childhood to adulthood has been found (e.g., see Belsky & Pensky, 1988; Bowlby, 1973, 1980; Ricks, 1985; Rutter, 1988), however it is unknown how it plays a role in attachment to pets.
According to Main (1990), flexibility is a biological tendency that parallels attachment and is useful for managing positive and negative aspects of the caregiving environment. Flexibility is an indication of the quality of the attachment bond (Bowlby, 1956; Ainsworth, 1972). It is a function of an individual’s ability to anticipate the consistency of an attachment figure in a given situation (Cassidy, 1999). The concept of a “working model” or “representational model” was utilized by Bowlby (1971,1975) to describe the means by which an individual will benefit from the security of an absent attachment figure. An infant who flexibly benefits from using mother as a secure base in times of need (e.g., clinging when introduced to a new situation or people) while maintaining his independence (e.g., willing to explore in less dangerously perceived environments) is one who has achieved a sense of safety; such a child would be described as securely attached (Bowlby, 1988 as cited in Cassidy, 1999, p.12). Other major patterns of attachment emerged in Ainsworth, Blehar, Waters, & Wall (1978) strange situation experiment. These are: anxious-resistant, and avoidant (as cited in Fraley & Shaver, 1999, p.735). Later, Main & Solomon (1986) categorized a disorganized type of attachment and new studies are looking at subtypes of avoidant attachment[7] (Bartholomew & Horowitz, 1991; Fraley & Bonanno, 2003). Bartholomew & Horowitz (1991) developed and tested a Four-Category Model of attachment. The results of their study confirm Bowlby’s (1973) conceptualization that working models differ in terms of self and other, in that there is one internal model that refers to the self and a second internal model that refers to one’s perception of how he/she is perceived by others (as cited in Bartholomew & Horowitz, 1991).
In lieu of the results of Ainsworth et al.’s, (1978) strange situation experiment[8], and other studies that examine the role of individual differences in attachment style (Hazen & Shaver, 1987) and loss (Hazen & Shaver, 1992), the present study proposes that the degree to which an individual aims to seek security and comfort from a relationship with a pet will be an important and necessary factor for determining the existence of an attachment bond. Also, the degree to which an individual reports having been flexible in his/her relationship with a pet will be looked at as a function of individual differences in attachment style and a means to identify variability in the grief response. The present study will look at the human-pet attachment relationship from the perspective of the bereaved party as a means to measure the bond in the relationship. A unidirectional theory of attachment underlies this approach, as grief is assumed to be a reflection of the degree of attachment experienced by the bereaved in the previous relationship.
Bereavement Theory
Bowlby (1977, 1981), Ainsworth (1979, 1989) and Parkes (1975), laid groundwork for a theory of bereavement based in attachment theory. They acknowledged an evolutionary basis of attachment that sprang from the Darwinian theory survival of the fittest. Although Bowlby (1969/1982) mostly referred to survival as the goal of the attachment behavioral system, his theory makes sense in light of modern day evolutionary paradigms that regard reproductive fitness as the definitive adaptive goal of human social behavior (Simpson, 1999).
With regard to a theory of grief, Bowlby (1981) and Parkes (1975) pointed out that grief is an adaptive consequence of separation from or loss of an attachment bond (as cited in Archer, 2001). Remarking on the adaptive benefit of grief proposed by these authors, Archer (2001) concedes that social bonding would be “impossible if attachments ceased whenever the attachment figure was out of sight” (p.555). In order to account for the stability of attachments, Bowlby (1969/1982) proposes that a mental representation of the attachment figure is formed that maintains (during separation) the sense of stability and security felt with regard to the attachment figure. But, Archer (2001) notes, there is something futile about a grieving mechanism that lacks a check-and-balance system for distinguishing separations from losses. The incongruity of an adaptive grieving mechanism that fails to make such a distinction is better understood if it is looked at from the perspective of a deficit reaction that serves to enlist those familiar behaviors helpful for creating and maintaining the attachment bond (Archer, 2001, p.555). According to attachment theory, the usefulness of the grieving mechanism lies in the fact that it allows an individual to maintain bonds that are threatened by loss (e.g., separation), with the set goal of preventing loss from ever occurring (Bowlby, 1969/1982). Then, from the point of view of modern evolutionary theory, Archer (2001) states that grief could be looked at as a side effect or a “ by-product of maintaining the attachment bond, which itself is necessary for maintaining relationships important for fitness” (p.555). Bowlby (1969/1982) argued that a mental representation of an attachment figure that persists after separation is useful for maintaining bonds and preventing them from breaking. In Making and Breaking of Affectional Bonds, Bowlby also (1977) stated:
What for convenience I am terming attachment theory is a way of conceptualizing the propensity of human beings to make strong affectional bonds to particular others and of explaining the many forms of emotional distress and personality disturbance, including anxiety, anger, depression and emotional detachment, to which unwilling separation and loss give rise (p. 201).
Thus, the deficit reaction is the severe emotional reaction (e.g., protest, anger, searching) that occurs in response to dissimilarity between the individual’s mental representation of the existence of the attachment figure and the reality of its absence. Archer (2001) states: “a mental model of the other...is continually compared with the input from the outside world... a discrepancy between the input and the mental model...generates a simple deficit-activated reaction” (Archer, 2001, p.555).
According to Parkes (1975, 19850, grief is tantamount to change perceived in an individual’s assumptive world. Parkes (1985) acknowledged the difficulty of making psychosocial transitions during bereavement in that they require adapting previous assumptions about reality to a new unfamiliar, perhaps unsafe reality. He affirmed that such changes are resisted by the bereaved, despite the fact that the old assumptions do not make sense in the context of new circumstances (p.12). This theory is consistent with a trauma perspective of grief in that it involves holding on to assumptions that concern one’s “safety and stability in the world” (Archer, 2001, p.556; see also Horowitz, 1988; Parkes, 1988).
Ainsworth’s (1979) study of the relationship between attachment styles and grief also utilizes the foundation of attachment theory to understand variability in the grief response. Results of studies on the strange situation led her to conclude that infants respond to separation differently depending on the quality of their attachment relationship with the attachment figure. She distinguished between securely attached infants and anxiously attached infants, noting that their behaviors varied on a continuum of anxiety.
These theorists recognize grief after loss of a significant attachment as a natural adaptive process that has benefit for human survival and fitness. In light of attachment theory, the symptoms to which separation and loss give rise only appear maladaptive or “immature” but are in fact reflective of a natural response to maintain attachment bonds (Bowlby, 1980 as cited in Cassidy, 1999, p.736). Attachment theory proposes utilizing an evolutionary perspective to understand the grief response (Archer, 2001).
Grief resulting from death of a pet makes sense within the paradigm of attachment theory. However, the extent to which pets function as attachment figures in their relationships with humans has yet to be conclusively decided. Since most studies on pet attachment are based on qualitative, descriptive and subjective data, it is mostly only inferred that grief following pet death is the result of an attachment bond, as it as yet to be empirically determined that the nature of the relationship is based on an attachment bond. If some pets do function as attachment figures in their relationships with humans then, in terms of attachment theory, it would follow that the quality of pet attachment and the grief response will be mediated by individual differences.
Summary of Problem
The pet attachment literature reviewed has mostly relied on descriptive and subjective (Quackenbush and Glickman, 1985) data to support a theory of human-pet attachment (e.g., Albert & Bulcroft, 1987) and attachment and loss (e.g., Rynearson, 1978).
The pet loss literature provides a more in-depth look at grief following pet loss in that they many utilized empirical analysis (e.g., Planchon et al, 1996; Gosse, 1989). However none appear to have been designed with the intention of comprehensively measuring attachment bonding as a precipitant to grief within the framework of attachment theory (Bowlby, 1969/1989) or modern evolutionary theory (Archer, 2001; Bowlby, 1969/1982; Parkes, 1975, 1998).
Regarding attachment measures, the Pet Attitude Scale (PAS) is a measure of feelings toward a pet (Herbert & Wilson, 1996) and the Companion Animal Bonding (CAB) scale only partially measures attachment bonding. These studies lack a systematic analysis of the grief reaction following pet loss and do not look at individual differences in attachment style as a measure of the grief response. The studies reviewed also lack a measurement tool designed to distinguish between attachment and simple pet affection.
A study that aims to accomplish these tasks would advance understanding of the emotional and behavioral responses of grief, including more seemingly irrational aspects such as searching, calling or clinging, imagining sights, sounds and smells, keeping mementos (e.g., toys or pet remains) of the pet’s existence (e.g., pet’s dish and objects in their usual place in the home)(Fraley & Shaver, 1999). Hence, the present study aims to scientifically validate the experience of human-pet (heterospecific) attachment bonding and loss by examining its similarity to human-human (conspecific) attachment bonding and loss and as it relates to an individual’s attachment style. The present study also aims to dispel notions of pathology as a prerequisite of pet attachment and as a necessary consequence in grief resulting from pet loss (Keddie, 1977; Rynearson, 1978).
Bowlby’s (1969/1982) theory of attachment will be utilized to facilitate the process. The following review of attachment theory intends to examine the range of criteria necessary for qualifying a bond as an attachment bond, with the aim of relating it to a theory of pet attachment bonding while exploring the difference between affectional bonds, attachment bonds and attachment behaviors.
Bowlby’s Attachment Theory
The initial stages of Bowlby’s (1980) attachment theory described qualities inherent to the mother-child bond. For Bowlby such qualities predicted the intensity of the emotional connection between mother and child both at the time of separation and reunion. Attachment theory is based on a deep reserve of clinical observations recorded as early as the 1940s. This era represents the first time that the manifested behaviors of infants and young children upon separation from the mothering person were given clinical relevance (for review see Bowlby, 1980). Bowlby later described these observations as vivid and typical accounts of the grief reaction and the observation methods used to organize such data as complementary to traditional retrospective methods for acquiring a knowledge base of personality characteristics- useful for making predictions about the effects of loss (Bowlby, 1969/1980, p. 24).
Question of Instinct
The development of the attachment theory became the focus of Bowlby’s work as such a theory most reliably explained qualities of stability or predictability in human behavior. According to Bowlby’s (1969/1982) theory man exhibits observable functions that indicate possession of a species characteristic (p. 39). Attachment behavior is an example of this characteristic in that it is a stable and/or instinctive behavior[9] similar to the reciprocal mother-young behavior observed in many lower order species (e.g., feeding nestlings). Observations of what Bowlby described as the idiosyncratic performance of individuals under certain environmental conditions contributed to the development of this theory. Behaviors such as mating and caring for young babies show a clear pattern of goal directed behavior with predictable “benefit [for the] individual or species” (Bowlby, 1969/1982, p. 39). Accordingly, instinctive or attachment behaviors in man are environmentally driven, goal-directed and have clear survival value (Bowlby, 1969/1982).
Attachment Behavior
Attachment behaviors are not inherited, but rather conceptualized by Bowlby (1980) as an aspect of an inherited potential. That is, the potential to adapt to a given set of environmental conditions and ensure the survival of the individual or species. Attachment behaviors transpire between a mother (or caretaker) and her young and are relatively stable across the species. They are recognized as a pattern of distinct behaviors brought about under particular environmental conditions. Bowlby (1969/1982) also described them as functions in a biologically driven system and ultimately, as the consequence of man’s instinctive evolutionary adaptedness[10] (p. 125). For example, mothering behaviors such as holding, feeding, watching young are functions designed to promote the safety or survival of the young and are also a consequence of a mothering person’s inherited potential to ensure the survival of her family genes. Ainsworth (1979) states: “one major aspect of the environment of evolutionary adaptedness for infant attachment behavior is not merely a mother figure but one who is sensitively responsive to infant behavioral cues” (p.5).
Cybernetic Theory
Bowlby (1980) distinguished biological functions from the system’s cause of activation. Rather, the cause is driven by “causal factors” or conditions presented by an individual’s internal (e.g., hormones) and external environment (e.g., lack of food). These factors are relatively cooperative, working in unison as “a product of gene action and environment,” and influencing one another in a constant feedback loop (Bowlby, 1969/1982, p.85). Elaborating on the attachment behaviors described above, a mother’s reaction to protect her child from falling is activated by an environmental condition (e.g., presence of stairs along child’s pathway), which in turn stimulates a perhaps, already active hormone level (e.g., designed for staying alert) into producing more efficient hormones (e.g., fight or flight hormones or adrenaline rush) necessary for accomplishing the task of running to protect a child.
Adaptive Value of Attachment Behavior
Attachment behaviors are a part of a complex system of human development that is of absolute value to each stage of an individual’s development. Behaviors that entail caring for the young describe a type of instinctive behavior designed to promote species survival, but are not sufficient for explaining the kinds of behaviors observed most clearly in young children and adolescents such as clinging to and searching for a lost attachment figure. These behaviors also function to promote species survival and only appear more “egotistic” as they are complementary to those of caring and are an example of the collaborative element necessary for the evolution and propagation of a species (Bowlby, 1969/1982, p.131; Cassidy & Shaver, 1999).
An Attachment Bond
Bowlby (1969/1982) stated that to understand the response of a child to separation or loss of his mother-figure one must also understand the nature of the attachment bond between the child and the mother-figure. The attachment bond is recognized in this dissertation as originally conceptualized by Bowlby (1969/1982), that is, as a consequence of a physiological system that functions in an environment of evolutionary adaptedness to promote the survival of the human species. The attachment bond is the product of the social relationship that develops between two individuals, but more importantly and for the purpose of this dissertation project, it is a measure of the investment in survival and security reported by and observed in the attached member.
In Bowlby’s (1969/1982) work, which focuses specifically on the relationship between a mother and child, the attachment bond is described as a “strong tie” between a mother-figure and an infant or child (p. 177). This bond describes the kind of relationship infants have with their mothers, and is recognized as “the foundation stone” of the child’s developing personality (Bowlby, 1969/1982). It is differentiated from attachment behavior in that the former qualifies the propensity of the attached to seek contact with a specific figure under certain environmental circumstances on the presumption that that person has value as a secure base[11] (Ainsworth, 1969; see also Bowlby, 1988).
Proximity Seeking
A wealth of data collected by researchers (e.g., see Bowlby, 1969; 1982) interested in understanding the developmental process of attachment behavior in infants[12] indicates that the reciprocal tie between mothers and infants is reinforced with proximity maintenance behavior. The term attachment behavior is derived from observations of proximity seeking behavior, which is the result of the infant’s interaction with the environment and the principal figure in that environment (Bowlby, 1969, 1982, p. 179-180). This theory is an alternative to a drive or need theory, in that it places greater emphasis on primary social evolutionary aspects of behavior rather than on secondary instinctive goals of behavior, such as eating, to explain the infant’s interaction with the environment (Bowlby, 1969, 1982, p.179; see also Freud, 1940). Attachment behavior is viewed as a consequence sufficient unto itself with distinct survival advantage, rather than as a secondary consequence to a behavior[13] (Bowlby, 1969; 1982). Thus, unlike the behavior of eating, which is done to satisfy hunger, proximity seeking is done to satisfy an innate need to seek proximity. In drive theory, socialization is posited as a learned consequence of behaviors such as those aimed at satisfying hunger (Freud, 1926, 1940; see also Dollard & Miller, 1950).
Control Theory
According to Bowlby, the goal-corrected function of attachment behaviors such as proximity seeking is better explained using the foundation of control theory. The motivation behind the decision to substitute control theory for drive theory was generated by Bowlby’s interest in ethology and Lorenz’s ethnological studies of the imprinting behavior of young goslings or ducks (Lorenz, 1935), Scott’s (1963) studies of the clinging and exploratory behavior of monkeys, puppies and lambs and Shipley’s (1963) observations of the contact seeking behaviors of guinea-pigs (Lorenz, 1935; Scott, 1963; Shipley, 1963 as cited in Bowlby, 1969/1982, p. 211-12; for a review of other animal behavior studies see also Carins, 1966a). These studies showed that proximity-seeking behaviors are not dependent on secondary benefits (e.g., food, warmth), as the animals observed exhibited seeking, clinging or following behaviors in spite of the lack of a predictable reward (Bowlby, 1969/1982).
Pre-programmed Behavior
Thus, proximity seeking behavior is considered a component of the infant’s attachment behavioral system, a larger system with goal-corrected functions that does not depend on learning in response to primary rewards such as food or warmth as derived from the secondary social relationship established with the attachment figure (e.g., mothering person) (Bowlby, 1988, p.24). The goal-corrected aspects of proximity seeking are viewed as an integral part of the reciprocal relationship between mother and child and as pre-programmed[14], which basically means that they are present at birth and only elicited in response to certain activating conditions. For instance, an analysis of the reciprocal nature of the behaviors between an infant and mother shows that changes in distance between the infant and the mothering person are corrected by approaching behaviors (Bowlby, 1988, 3-5).
Goal Correction and Termination
Bowlby (1969/1982) distinguishes between approaching behaviors such as calling and locomotion from other attachment behaviors such as crying or babbling, noting the exclusively goal-corrected organization of the former (p.248). Thus, unlike an attachment behavior such as crying, which also has proximity to the mothering person as a goal, goal-corrected approaching behaviors such as crawling are said to mediate attachment in that they are a start of a chain of relevant active behaviors aimed at restoring proximity and terminate when proximity is restored (Bowlby, 1969/1982; Ainsworth, 1979). For example, the behavior is terminated once a mothering person picks up a crawling child (e.g., in the process of seeking proximity), the environmental stimulus (e.g., picking up child) terminates further seeking. The relevance of approaching behavior and other types of goal-corrected behaviors to control theory lies in the shifts of responsibility between mother and child, with mothering offering the secure base or goal of the infant’s proximity seeking. These shifts change over time, as the child grows accustomed to the security available.
Responding to Security Threats
The process of maintaining and restoring proximity is brought about by a host of behaviors engaged in by both the infant and his mother. Calling, crying, and protesting would be included in a cursory list of infant behaviors. These behaviors increase and decrease in frequency and level of intensity as dictated by changes in environmental circumstances and the biological (including cognitive) development of the infant. Bowlby (1969, 1982) notes that already at 6 months-of-age an infant possesses an ability to recognize his own mother and henceforth continues to develop an uncanny ability to anticipate her “impending departure” [original emphasis](p. 204). According to Bowlby, the infant’s protest reactions are a communication of perceived threat to his sense of security, which was maintained earlier by a secure awareness of mother’s availability and was disrupted by her subsequent departure (Kobak, 1999; see also Bowlby, 1973).
Planning For Conditions
Ongoing experiences with the mothering figure contribute to the development of a keener capacity to appraise her availability. This capacity is reinforced by numerous “attractive” behaviors, such as smiling and babbling, which are often undertaken by an infant to foster a shared investment in maintaining his sense of security (Ainsworth, 1990, p. 474). Ainsworth (1990) notes that these behaviors have the appearance of “planning for conditions” that will increase the infant’s sense of “felt security” (Ainsworth, 1990, p.474 as cited in Kobak, 1999, p.31; Stroufe and Waters, 1977 as cited in Kobak, 1999, p.30). It is this ability to plan for conditions that has survival advantage and influences the infant’s personality development.
Set Goal of Proximity
In Bowlby’s later work greater emphasis was placed on the infant’s cognitive appraisals of his mother’s availability to explain the infant’s “set-goal” of proximity maintenance (Kobak, 1999, p.31; see also Ainsworth, 1990). According to Ainsworth (1979) the setting of the infant’s set goal to seek proximity to the mothering person is continuously activated, but shifts from low to high (or high to low) in response to his perception of his safety in the environment. For example, in times of high stress (e.g., strange environment or people) an infant may shift his set-goal from mere proximity maintenance (e.g., keeping eye contact with mother or holding her hand) to close bodily contact (e.g., clinging or screaming so that she may come to him). Which attachment behavior is chosen is determined by the child’s degree of perceived confidence and his ability to flexibly manage his environment.
Hierarchical Organization
It has been shown that the organization of attachment behaviors becomes increasingly more sophisticated as an individual develops. Once a child is around two or three these behaviors begin to be organized around plans (Ainsworth, 1979, p.4). The significance of these plans lies in the fact that attachment behaviors are beginning to be used to achieve multiple ends. For example, a child may engage in an approaching behavior such as crawling both to be closer to mother and to explore the environment (Ainsworth, 1979).
The strange situation provides a typical example of how locomotor approach can serve equally a set-goal of proximity maintenance and exploratory play. In this experiment, the attachment behavioral system was balanced by the exploratory system, since the attachment behavioral system was either not activated or less activated when infants were engaged in exploratory play. Ainsworth (1979) states that when an infant manifests a capacity to shift from a priority for safety to a priority for independence, he is exhibiting his capacity for “use of mother as a secure base” (Ainsworth, 1979, p.4). As stated by Main (1990) such behavior would be described as flexible (see Cassidy, 1999). She also states that individual differences in the organization of attachment behaviors (e.g., capacity to deploy behaviors for use to achieve multiple ends) are dependent on individual personal experiences (e.g., in terms of feedback and the consequences of engaging in certain behaviors).
Mediating Influences
Following the strange situation procedure it was concluded that the attachment system is subject to the mediating influence of other behavioral systems. Cassidy (1999) discusses these four systems: the exploratory system, the fear system, the care-giving system, and the sociable system. They are identified according to their potential to activate or deactivate attachment or proximity seeking behavior (i.e., proximity maintenance behaviors can be analyzed from the point of view of each of these systems). Each of the four systems varies in level of intensity of activation and is influenced by the age (cognitive and developmental) of the child.
Secure Base
Ainsworth (1967) first used the term secure base to describe the nature of the attachment bond and the impact of that bond on the infant’s fitness (e.g., in terms of flexibility and survival) for exploratory behavior (Ainsworth, 1963 as cited in Cassidy, 1999, p.12; see also Ainsworth et al., 1977,1978; Bowlby, 1962, 1989). How the attachment figure is relied upon as a secure base depends, however upon the mother’s responsiveness to the infant’s proximity seeking and call for security.
Attachment Styles
In the strange situation, it appeared that each child’s appraisal of his mother’s availability at separation was defined by a repertoire of individually familiar stressful experiences. In other words, each child appeared predisposed to respond in a particular manner to his mother’s availability, which in turn appeared to determine his adopted strategy for seeking proximity and exploratory play. This experiment led to a more refined explanation of proximity seeking, beyond explaining its goal literally in terms of physical distance toward embracing a cognitive appraisal of emotional distance, and later to an acknowledgement of other interacting behavioral systems.
Bowlby (1958) postulated that individual differences in attachment style are the result of the mother’s developed pattern of responsiveness to the infant’s attachment behaviors (e.g., calling, crying)[15].
The finding of individual differences in response to separation from an attachment figure that resulted from the strange situation, led to the classification of several different patterns of attachment, some of which continue to serve as markers to determine the quality of an attachment bond in other pair bonds (e.g., romantic partners). They have been useful in prediction studies of the adult grief response to loss or separation (Hindy & Schwartz, 1994; for a review see Bartholomew & Horowitz, 1991).
Attachment Bond Versus Affectional Tie
Five criteria for identifying an attachment bond were noted by Ainsworth (1989) and Bowlby (1977) as similar to those used to describe an affectional bond, with one notable exception, which makes up a total of six criteria that are presently used by attachment theorist to define the attachment bond (Cassidy, 1999, p. 12). Cassidy (1999) refers to Ainsworth’s (1989) criteria for identifying the existence of an attachment bond, which is distinguished from strong affection in a relationship. These include: 1) the existence of a non-interchangeable individual (e.g., the mothering person), 2) emotional investment in that individual, 3) persistent as opposed to transitory affection for that individual, 4) proximity seeking and maintaining behaviors in relation to the individual, 5) a manifested capacity for distress upon separation or imminent separation or after a mere threat of separation from the individual. Yet, in order to be truly defined as an attachment bond, Ainsworth (1989) states that an individual must also strive to 6) seek “security and comfort in the relationship” with the other person (Cassidy, 1999, p.12).
Ethology
Bowlby’s theory of attachment developed naturally from his work as an ethologist, which makes it particularly relevant to the topic of cross-species attachment. Many of his early hypotheses of attachment behavior are derived from observations of the survival behavior of animals, most particularly observations of strong attachment bonds in newborn animals - from puppies to monkeys - and their mothers. These behaviors were later recognized as comparable to the attachment behaviors of newborn children.
The notion that a human-animal bond can develop is based on the parallel individual observations of infant humans and infant animals. Observations such as the protest response of a human infant to the departure of his mother and observations of the phenomenon known as contact comfort in rhesus monkeys (Harlow, 1940) have been explained developmentally and behaviorally (Bowlby, 1980; King, 1993; Parkes, 1998). These behaviors are the result of an interaction between genetic, environmental and neurophysiological factors.
Inter-species attachment may be described as part of the repertoire of predictable behaviors characteristic of individual species – “not just in man” – to pursue evolutionary fitness (Bowlby, 1969/1982, p. 39; Archer, 1996). Bowlby (1869/1982) ascertains that the higher order function of human behaviors is best viewed as a modification - perhaps an elaborate modification in that functioning is augmented by a capacity to cognitively appraise the environment, but nevertheless derivative - of the prototypical “anatomical and physiological equipment” of other species of animals (p. 39).
Felt Security
The notion of continuity between humans and pets forms the basis for the argument that animals and humans have the potential to develop reciprocal attachment behaviors (e.g., respond to feedback from the environment, seeking ultimately to find felt security) (Bowlby, 1962/1989, p.40). Many theories have sought to explain the experience of felt security acquired in an attachment relationship (Ainsworth, 1963). Control theory, for example, relies on physical phenomenon to explain the process of attachment: sequences of learned behavior, guided by physiological changes in the organism, that respond to environmental cues and culminate in the form of mutual care-giving and care-receiving (Langer, 1967 as cited in Bowlby, 1969/1982). Such a proposition allows similarities between humans and pets that are based in anatomy and physiology to transcend their cognitive differences. They will be used to help advance the theory of human-pet attachment.
Studying Pet Loss
The attachment relationship between a pet and a human is considered an example of the instinctive response of the social animal species to form relationships of mutual care to promote survival (Gosse, 1998, p.7). The evolution of human and pet qualities is viewed in parallel as a function of adaptation and to increase the potential of either species to form an attachment bond (Brickle, 1982; Gosse, 1998). The attachment bond is noted as a caregiving and care-receiving relationship and as a necessary precursor and measure of the mourning experience (Bowlby, 1980; Parkes, 1998; Rynearson, 1978; Gosse, 1988). Thus, the bereavement response to the loss of the attachment bond is viewed in the present study as a measure of the intensity of the broken bond.
A Commonality
Rynearson (1978) regards biology as the starting point from which the study of inter-species attachment should begin. In particular, he makes note of the neurophysiological basis of the predictable behaviors of caretaking and relational proximity involved in attachment relationships. He recognizes a commonality in the behaviors of humans and pets (Rynearson, 1978, p.550). The recognition of the commonality between humans and pets has led to a greater general understanding of inter-species attachment. It has also given us the opportunity to notice other shared symptoms and traits in the behaviors of the social animal species. For example, the similarities found in the human mourning responses of humans for humans and humans for their pets, which is a focus of the present study. Information about different attachment styles in humans should help further this process and increase our knowledge of the role attachment behaviors play in creating and sustaining human-pet bonds, in particular, those they have been severed by loss.
Attachment Theory
Detaching versus Decathecting
An understanding of the mourning response of a human for another human may facilitate this task. In Mourning and Melancholia, Freud (1917) explained that successful recovery from loss hinges on an individual’s ability to detach or decathect from a lost attachment figure (Freud, 1917). Freud used the term decathect to describe the ideal end-point of “the work of mourning,” a process that culminates in freeing the ego from a hyper-cathected position resulting from an individual’s resistance to letting-go (Freud, 1917 as cited in Reiff, 1963 p.166; Fraley and Shaver, 1999). This theory was rejected by Bowlby (1969/1982). Bowlby (1980) disagreed with the idea that detachment is necessary (Fraley and Shaver, 1999). Although he observed a process of detachment in his early work with children who were briefly separated from their mothers, he did not view it as the key to recovery or as a final stage. Instead, Bowlby (1980) used the term detachment to describe the behaviors of children upon return of an attachment figure, defining it as an expression of discomfort (e.g., “defensive reaction”) with the separation just endured (Fraley and Shaver, 1999, p.749).
Is Grief Irrational and Immature?
To understand Bowlby’s (1980) theory of loss and the grief experience a revisit to his 1980 volume Loss is inevitable. Some of Bowlby’s (1980) most important contributions to the subject of loss are found in this first volume. Fraley and Shaver (1999) summarize these contributions in the form of “goals” that Bowlby intended to achieve in writing Loss. According to these authors, one of Bowlby’s goals was to restore dignity to “seemingly irrational or ‘immature’ reactions to loss” by finding evidence of the naturalness of those behaviors in the wild. Bowlby’s (1969/1982) background as an ethologist provided him with a fresh perspective on attachment behaviors in humans. He used this knowledge to better understand why human adult responses to loss often involve a shutdown of a higher order capacity to rationally process the permanence of loss and to legitimize the significance of attachment[16]
Bowlby found an explanation for this phenomenon in his research and observations of the characteristics, qualities, and behaviors of infants of many different species (Bowlby, 1977). Research supports Bowlby’s 1977 theory regarding a motivational system, and is framed to include the reciprocal nature of a need for proximity between caregivers and care receivers (Lorenz, 1950; Gardner & Wallach, 1965; Fuller & Reiling, 1976; Sternglanz et al., 1977 as cited in Hinde, find date). An aspect of the motivational system includes the quality of certain physical features of infants, such as large eyes and forehead, and engagement in behaviors such as crying, which serve as lures for achieving and maintaining proximity with caregivers (Fraley and Shaver, 1999, p.736; see also Archer, 1997; Hinde, find date). According to Bowlby, the motivational system functions as a predetermined mechanism for ensuring the creation of an attachment relationship[17]. This aspect of the attachment system, perhaps most appropriately explains the “seemingly irrational” behaviors observed in grieving. According to this line of reasoning, if the characteristics and behaviors that make up the motivational system are necessary precursors to the creation of an attachment relationship then these same characteristic and behaviors must play a role in preventing its severance (Parkes, 1972).
Adaptive Benefit of Grief
Grieving is a natural response to the uncontrollable situation of loss. It is defense against the incomprehensible and is set in motion by both awareness of a separation (i.e., impending or final) and a biologically determined need to prevent its occurrence. Bowlby (1980) notes that “protest” reactions like “disbelief, anger, searching, and sensing the continued presence of a lost attachment figure” are adaptable and make sense in terms of attachment theory. Evidence of aggravated medical conditions such as cardiac difficulties and psychological symptoms such as depression, restlessness, anxiety and insomnia as well as increased rates of suicide after bereavement lend support a “broken heart” metaphor (Parkes, 1998). Fittingly, this metaphor references the biological substrate of the attachment phenomenon.
Biological theory of Grief
Biological theories of grief sprang from Darwin’s (1872) early observations documented in The Expressions of the Emotions of Man and Animals. Before this work an acknowledgement of the commonality of the grief response between animals, young children and adult human was conspicuously lacking in the literature (Parkes, 1998). Bowlby’s work on the effects of maternal deprivation appeared in a World Health monograph in 1951 and quite independently, Parkes (1998) submitted a dissertation containing his “preliminary formulation” of a biological theory of grief. The two researchers subsequently joined at the Tavistock Institute of Human Relations in 1962 to begin a life long collaboration on a biological theory of grief.
Fight or Flight System
It appears that the basis of the interrelationship between biology and loss is a phylogenetic link inherited from lower-order species known as the “flight or fight” or “alarm” system (Parkes, 1998, p.31). According to the research, the alarm system is a stress signal in animals and human beings that produces physiological changes in the body necessary to support an appropriate response to real or imagined threats to survival (Cannon, 1929 as cited in Parkes, 1998). In support of attachment theory, the chemical substances produced in response to stress linger in the body well after the cessation of a threat of harm real or imagined (Cannon, 1929; Parkes, 1998).[18] With regard to the permanent loss of an attachment figure, the body’s stress signal may continue despite eventual cognitive restructuring and the passage of time (Rando, 1986; Sanders, 1989). This phenomenon explains the biological root of irrational behaviors in the grief response. However, the outward manifestation of behaviors such as protesting, searching and calling and the continued engagement in intense feeling states that are often experienced after loss are supported by other critical aspects of the attachment relationship; that is, in addition to having a biological root (Bowlby, 1977; Parkes, 1998).
Alarm Reaction to Loss
Parkes (1998) states that the alarm system is activated in situations of “absence of an escape route, lack of a haven or a home in which the individual can feel secure, or in social animals, absence of fellow members of the species who normally provide or share defense (e.g., parents or mate)...To this must be added one other category or situation capable of producing a high-state of alarm, namely the loss of a child or child-substitute” (p. 35). These observations are consistent with the concept secure base, and support an additional hypothesis in this study, which is that attachments that have a strong caregiving or care-receiving element should lead to a more intense grief experience following pet loss.
Grief and the establishment of relationship
According to Parkes (1972), intensity and resolution of grief depend on three main variables: (1) the existence of an attachment relationship, (2) the strength or quality, of the attachment relationship (3) the reason for its establishment. With regard to the above hypothesis, the present study postulates that when the reason for the establishment of a relationship with a pet is based on a desire to substitute pet(s) for human social support and to increase felt sense of safety and security a stronger attachment to the pet will be experienced. In addition, when the reason for the establishment of a relationship with a pet is based on a desire to substitute a pet for lack of children, a stronger felt attachment to the pet will be experienced. These hypotheses are derived from survey analyses conducted by Albert & Bulcroft (1987) and from studies conducted by Archer (1994) and Gosse (1994). As was indicated in these studies, stronger attachment relationships and grief reactions are often found in owners who do not have children at home, who live alone or with only one other adult in the home. Thus, the present study expects to find that when (1) the point of view of a caregiver (e.g., mother) is taken in the relationship with the pet, or (2) the point of view of a care recipient is taken in the relationship with the pet, a more intense grief reaction will result. Human-pet relationships established for these two reasons have also been supported in Bowlby’s (1969/1982) discussion on the nature of the child’s ties and in theories advanced by Ainsworth (1988) and Cassidy (1999) on nature of the parent’s tie. Further, in the study conducted by Gerwolls & Labott (1994), it was demonstrated that grief following loss of a pet is very similar to grief following the loss of a child.
Bonding and Nurturance
A human propensity toward creating attachment relationships was recognized by Bowlby in his theory of the evolution of attachment behaviors. Interest in nurturance, defined by Fogel and Melson (1986) as “the provision of guidance, protection and care for the purpose of fostering developmental change” appears to be evident by early childhood (Melson & Fogel, 1996). The development of an attachment relationship to pets appears to be precipitated by an interest in nurturance and perhaps being nurtured. This theory provides further support for considering variables such as care-given (Stern, 1985) and care-received (Bolin, 1987) in a measure of the attachment bond.
Making and Breaking Bonds
In order to narrow the focus of the vast amount of data, analysis and literature on theories of attachment, Bowlby (1969/1982) drew up a concise list of criteria of patterns of attachment. One criterion, the protest reaction, had hitherto been used to define the strength of an attachment (Schaffer & Emerson, 1964a as cited in Bowlby, 1969/1982, p.333). However, it has since been discovered that some children fail to exhibit protest upon separation from their mothers, with some exhibiting a kind of “stoic” reaction [19] (Ainsworth, 1963). As a result of this discovery Ainsworth (1972) and Bowlby (1969/1982) proposed that while the protest reaction is indicative of a type of attachment relationship, it is not a necessary criteria. Other attachment behaviors included in Bowlby’s (1969/1982) list are summarized as follows:
(a) Behavior that initiates interaction
(b) Behavior that maintains interaction
(c) Behavior aimed to avoid separations
(d) Exploratory Behavior
(e) Fear behavior with oriented reference to
attachment figure.
(f) Behaviors characteristic of greeting and
types of rejecting behaviors upon reunion after
stressful separation (p. 334).
Bonding Beyond Infancy
In a preliminary theory of other attachment behaviors beyond infancy, Ainsworth (1989) noted the key features of an attachment bond. In her discussion she also points to several issues that are important for considering attachment bonds with pets. First, noting the stability and variability of social relationships, she states that affectional bonds are represented throughout the lifespan and probably in all cultures, as they are species-characteristic [20] behaviors. These types of relationships are noted, however, as different from affectional bonds in three ways: (1) affectional bonds endure in time while relationships may be transient; (2) affectional bonds are characteristic of the individual (they utilize an internal representation of the attachment figure to achieve sense of security) while relationships are dyadic; (3) relationships are made up of varied components some of which are irrelevant to establishing an affectional bond or tie.
In a summary definition of an affectional bond, Ainsworth (1989) notes the necessary criteria: (1) enduring tie; (2) a non-interchangeable individual; (3) proximity seeking and maintenance behavior; (4) emotionally significant relationship; [21] (5) affective response upon reunion or after a stressful or long separation[22] (p. 711).
In contrast to the above criteria, Ainsworth (1989) states that in addition to the above criteria an attachment bond involves “the experience of security and comfort obtained from the relationship with the partner, and yet the ability to move off from the secure base provided by the partner, with confidence to engage in other activities” (p.711). In addition, and most importantly, for the purpose of this dissertation project, Ainsworth underscores that since all attachments are not classified as secure the latter criterion should be modified as “seeking of the closeness that if found, would result in feeling secure and comfortable in relation to the partner” (p.711).
Ainsworth’s (1989) discussion on attachments beyond infancy provides inspiration for the present study, as pet attachment bonds and grief following loss shows similarity to attachments and loses of significant human attachment figures. This similarity is apparent from her discussion on the topic of other attachment figures, such as child-parent bonds in adulthood. Ainsworth (1989) states that the response of an adult child bereaving the loss of a parent, “...usually demonstrates that the attachment bond has endured” (p.711). That pet loss is responded to by grief is alone indicative of an attachment relationship. How that relationship is qualified is best explored by a close examination of individual differences in the grief experience.
Methodology
The present study utilizes a correlational cross-sectional survey design to assess the relationship between pet grief and attachment style in a sample of adult individuals who lost a pet in death or separation within 12 months of the date of data collection. This study employs multiple regression techniques to examine the correlational relationship between the dependent or criterion variables, 1) severity of the grief response, 2) expression of continuing bonds, and the independent or predictor variables 1) pet attachment, 2) attachment style (e.g., secure, avoidant, anxious-ambivalent). The easy availability of the data encouraged the use of additional predictor variables such as level of perceived social support and human-pet variables such as age, gender, residential status (e.g., living alone), marital status, presence of additional pets in the household, and number of years the pet was owned.
Research Participants
One hundred study participants will be solicited by means of a flyer placed in locations frequented by pet owners and pet loss survivors. These locations include various pet loss support groups, pet crematoriums and cemeteries, humane societies (one in San Francisco and one in Santa Clara), the San Francisco Society for prevention of cruelty to animals and the Animal Rescue Foundation. Personal and colleague referrals will also be accepted.
Criteria for selecting research participants: 1) must be 18 years-of-age or older, 2) must have lost a pet in death or separation (e.g., missing, stolen) within 12 months from the date of data collection.
The rationale for stipulating these criteria is to facilitate generalization to other studies utilizing similar participant qualifications and to confine the grief experience to a time frame appropriate for collecting grief data that could be revealed across a range of pet attachment experiences and attachment styles.
Participants will not incur any direct benefits from participation in this study, will not be compensated monetarily, but should not incur any grave risks for their participation. The only risks anticipated are evoked sadness or uncomfortable and upsetting thoughts or feelings. Participants will not be identified by the data.
Once participants make the decision to participate in the research they will be provided with an informed consent form stipulating the purpose of the study, which summaries the limits to confidentiality, benefits and risk of harm for participating and their rights as study participants. This form is to be signed in the presence of a witness, one of several designated research assistant to be contacted at one of the locations listed above.
Instrumentation
Companion Animal Bonding Scale (CABS)
Poresky, Hendrix, Mosier, and Samuelson (1987) developed an eight-item self-report behavioral scale to assess the relationship between engaging in different human-animal activities such as sleeping next to a pet and attachment bonding. The behavioral items measured by the scale characterize different aspects of an attachment bond established between an owner and his/her pet. The CAB scale was the first of its kind to provide a better operational definition of human-animal bonding than ownership of a pet.
The first SPSS X Reliability analysis of the internal reliability of the scale yielded a Cronbach alpha of 0.77 for the scale when it was worded in the past tense (e.g., to capture the childhood relationship between an owner and pet) and 0.82 when worded in the present tense (e.g., to capture the contemporary relationship between owner and pet).
Scale Characteristics
One hundred thirteen sets of responses were produced from the childhood scale. A mean score of 26.7 (SD =5.7), with a range of 9 through 45 resulted from the childhood scale and a mean score of 28.6 (N = 99, SD = 6.3) resulted from the adult or contemporary scale. The difference between the groups was explained by a Pearson correlational statistic (r96 = .35; p = .001) as statistically significant (t95 = 2.26, p < 0.05). The “behavioral nature of the questions” (Poresky et al., 1987) is understood to have affected the resulting difference between the two arrays since a greater level of responsibility is assumed taken in the care of an animal as a child ages.
Factor analysis of the original scale yielded three principle factors. The first factor, bonding or involvement, accounted for 41.1% of the total variance. The second factor, size of the animal (e.g., determined sleeping arrangement), accounted for 14.5%. And, the third factor, autonomy and responsiveness of pet to owner, accounted for 9.1%.
To assess the scale’s construct validity, the CABS was correlated with the Pet Attitude Scale (PAS). This correlation yielded statistics of .42 and .38, respectively in the child-pet scale and the adult or contemporary owner-pet scale.
Multidimensional Scale of Perceived Social Support (MSPSS). Zimet, Dahlem, and Farley (1988) designed this self-report measure to assess an individuals’ perception of the amount of social support available. The measure is divided into three distinct social support constructs: Family members (Fam), Friends (Fri), Significant Others (SO). The different domains are divided into 4 items each, which results in a total of 12 items on the scale. Responses to these items are presented in a Likert format ranging from 1-7 starting with “very strongly disagree” (scale range 1) and ending with “Very strongly agree” (scale range 7).
Scale Characteristics
Clara, Cox, Enns, Murray & Torgrude (2003) conducted a factor analysis of this scale using a 3-factor model in two samples (college students, N=549 and psychiatric outpatient, N=156). The MSPSS was supported in both samples. In this study Family and Friends revealed the more consistent and stronger associations with symptomology. The results of this study lend support for the use of the MSPSS as a “brief instrument for assessing the hierarchical nature of perceived social support in a variety of samples.”
Although this study does not seek to assess the hierarchical nature of perceived social support, the results of the above study indicate that the instrument does tap into the construct intended to assess, that is degree of social support available and presence of symptomology as a function of social support.
King (1993) assessed the relationship between social support and depression (measured using the Hopkins Symptom Checklist) on a sample of 275 grieving pet owners. The results of her study found moderately high intercorrelations for the SO and Fri subscales (r=. 63) and internal reliability correlations of .91 (SO), .87 (Fam), and .85 (Fri). The full-scale reliability was .88.
King retested on 69 of the 275 subjects originally tested. Test-retest reliability yielded .72 (SO), .85 (Fam), and .75 (Fri). Test-retest for the full scale was yielded .85, which indicated good internal reliability over time as well.
King’s study also found a negative correlation between MSPSS and anxiety and depression (measured using the Hopkins Symptoms Checklist (HSCL). The three subscales all correlated negatively with depression (“in the r= 24 range”). King’s results support her assumption that perceived social support acts as a “buffer” to mitigate depression levels. Thus, King’s study supports the theory of a “moderating effect of social support between stressful life events and depression” (p.58).
The CENSURE Pet Attachment Survey (PAS) was developed by Holcomb, Williams & Richards (1985) at the Center for the Study of Human-Animal Relationships and Environments (CENSURE). The PAS is a 27-item self-report survey designed to measure attachment to a conventional pet (e.g., dog or a cat).
It is a four-item response measure presented in a Likert-type format that begins with “almost never” (scale range 1) and ends with “almost always” (scale range 4).
The PAS consists of two subscales, a Relationship Maintenance Scale and Intimacy scale. The Relationship Maintenance subscale is comprised of 16 items related to both time and financial investment, communication sensual and physical interactions with a pet. The Intimacy subscale is comprised of 11 items related to attitudes and feelings surrounding the emotional importance of the pet and desire to maintain physical proximity to the pet.
The University of Minnesota Veterinary Teaching Hospital (UMVTH)(1984) and Kerlinger (1986) were the first to validate the CENSURE PAS (as cited in Gosse, 1988). At the UMVTH it was administered to a variety of populations that included veterinary euthanasia clients (Daly, Anderson, and Kelly, 1985; Holcomb, Williams and Richards, 1985).
Scale Characteristics
Kerlinger’s (1986) study found a Cronbach Alpha Coefficient of .74 for intimacy. Other data indicate a Cronbach Alpha coefficient of 0.83 for relationship maintenance. According to Holcomb et al. (1985) full-scale reliability scores have not been reported, but construct validity data confirms use of the survey to measure attachment along the above two dimensions.
Gosse (1988) validated an adapted version of the survey. (He changed the tense of the items from present to past tense and retained only the 11 items belonging to the Intimacy Subscale). Gosse’s rational for retaining these 11 items only pertains to the results of the two studies cited above. In both of those studies, according to Gosse (1988), dog owners scored significantly higher than cat owners on the Relationship Maintenance Scale.
Since the present study does not limit ownership of a pet to either a cat or a dog, but extends it to different species of animals including wild animals such as rabbits and exotic animals such as parrots, the same rational does not fit. For example, it is uncertain if these other species of animals will also score lower than dogs on the Relationship Maintenance subscale and it is precisely a difference in the attachment relationship with different species of animals that the present study hypothesizes to discover. Nevertheless, like Gosse (1988) this study also modified the CENSURE PAS (e.g., used the past tense) to appropriately accommodate the circumstances of pet owners who have lost a pet.
Experiences In Close Relationships-Revised (ECR_R)
Fraley, Waller & Brennan (2000) developed a 36-item self-report measure of adult attachment. According to Brennan, Clark & Shaver (1998), this measure is multi-item measure derived from the full spectrum of multi-item attachment measures to date. Item-response theory (IRT) analysis was employed to narrow the original pool of 323 items down to 36.
Scale Characteristics
The scale’s response categories are presented in a Likert type format that ranges in dimension from 1 to 7, beginning with “Disagree strongly” (scale range 1) to “Agree Strongly” (scale range 7).The total item Cronbach alpha correlation yielded .94 on the Avoidance subscale and .91 on the Anxiety subscale.
The scale yields scores on two subscales, an Avoidance scale and an Anxiety scale. The first subscale, Avoidance, taps into the construct of discomfort with closeness and discomfort with depending on others, and the second subscale, Anxiety taps into fear or rejection and abandonment. These two subscales were found to be uncorrelated (r= .11) in the original study for the design of the ECR-R (Brennan et al, 1998).
There are four options for responding. They include Secure (e.g., “It is easy for me to become emotionally close to others”), Fearful (e.g., “I am uncomfortable getting close to others.”), Preoccupied (e.g., “I want to be completely emotionally intimate with others, but I often find that others are reluctant to get as close as I would like.”), Dismissing (e.g., “I am comfortable without close emotional relationships.”).
Pet Grief Measure
Archer et al. (1994) assessed pet grief measure using a 40-item questionnaire presented in a three-point Likert-type format. Responses range from the statement “Definitely applies to me” (scale range 1) to “Does not apply” (scale range 3). According to Archer et al. the rationale for restricting the number of possible responses to three was due to the length of the measure (e.g., 40-items) and because the same was limited to members of the local community.
Scale Characteristics
The 40-item scale was examined using Cronbach alpha (.94, after reversal of negative items) split-half reliabilities, and item-to-whole correlations (data for the latter two were not provided). 11 items were discarded from the original design due to high inter-item correlations and redundant wording of these items. For example, according to Archer et al. (1994), item 17 (“I flared up more easily after the death of my pet”) was highly correlated (r= 86) with item 27 (“After the death of my pet I quarreled more than usual with family and friends). These items (17 and 27) and 9 additional items were omitted (10, 15, 21, 22, 30, 31, 34, 37 and 40) to eliminate this problem.
The following items from the original scale were omitted due to their redundancy with items from the Continuing Bonds measure used in the present study. The first set of omitted items is: 21, 30, 4, 19, 33, 23 (These items dealt with “Preoccupation with the loss or the lost pet”). The second set is: 8, 36, 14, 32 (These items dealt with “An urge to search for what was lost”). The final set is: 26, 7 (These items dealt with Mitigating and avoiding grief). Within these three sets a few items were retained since they appeared to be concerned with emotions that were not included in the continuing bonds measure.
Procedures for Data Collection
Statistical Analysis
Apparatus
REFERENCES
Albert, A., Bulcroft, K. (1987). Pets and urban life. Anthrozoös, 1(1), 9-25.
Archer, J. (2001). Evolving theories of behavior. Psychologist, 14(8), 414-.
Archer, J., Winchester, G. (1994). Bereavement following death of a pet. British Journal of Psychology, 85, 259-271.
Archer, J. (1997). Why do people love their pets? Evolution and Human Behavior, 18(4), 237-259.
Baerends, G. P. (1976) The functional organization of behavior. Animal Behavior, 24, 726-738.
Bodsworth, W., Coleman, G.,J. (2001). Child-companion animal attachment bonds in single and two-parent families. Anthrozoös, 14(4), 216-223.
Bolin, S. E. (1987). The effects of companion animals during conjugal bereavement. Anthrozoös, 1(1), 26-35.
Bowlby, J (1988). A secure base: parent-child attachment and healthy human development. New York: Basic Books.
Bowlby, J (1969/1982). Attachment and loss: Vol 1. Attachment. London: Basic Books.
Bowlby, J. (1977). The making and breaking of affectional bonds. British Journal of Psychiatry, 130, 201-210.
Bowlby, J. (1961). Processes of mourning. International Journal of Psycho-analysis, XLII(4-5), 317-331.
Brown, B. H., Richards, H. C., Wilson, C. A. (1996). Pet bonding and pet bereavement among adolescents. Journal of Counseling and Development, 74(5), p. 505.
Carmack, B. J. (2003). Grieving the death of a pet. Minneapolis: Augsburg Books.
Cassidy, J. (1999). The nature of the child’s ties. In J.Cassidy, P. R. Shaver, Handbook of Attachment: Theory, Research and Clinical Applications (pp. 3-20). New York: The Guliford Press.
Cowles, K. V. (1985). The death of a pet: human responses to the breaking of the bond. Pets and the Family, 135-161.
Fogel, B., Abramson, D. (1989). Pet loss: a survey of the attitudes and feelings of practicing veterinarians. Anthrozoös, III(3), 143-150.
Fraley, C. R., Shaver, P. R. (1999). Loss and bereavement: attachment theory and recent controversies concerning “grief work” and the nature of detachment. In J.Cassidy, P. R. Shaver, Handbook of Attachment: Theory, Research and Clinical Applications (pp. 735-759). New York: The Guliford Press.
Garrity, T. F., Stallones, L., Marx, M. B., Johnson, T. P.(1989). Pet ownership and attachment as supportive factors in the health of the elderly. Anthrozoös, III(1), 35-44.
Geary, D. C. (2000). Attachment, caregiving and parental involvement. Psychological Inquiry, II(2), 84-123.
Gerwolls, M. K., Labott, S. M. (1994). Adjustments to the death of a companion animal. Anthrozoös, VII(3), 172-187.
Gosse, G. H., Barnes, M. J. (1994). Human grief resulting from the death of a pet. Anthrozoös, VII(2), 103-112.
Hinde, R. A., Barden, L. A. (1985). The evolution of the teddy bear. Animal Behavior, 33(4), 1371-1373.
Horn, J., Meer, J., (1984). The pleasure of their company: a report on Psychology Today’s survey on pets and people. Psychology Today, August, 52-57.
Jarolmen, J., (1998). A comparison of the grief reaction of children and adults: focusing on pet loss and bereavement. Omega, 37(2), 133-150.
Katcher, A. K., Beck, A. M. (1987). Health and caring for living things. Anthrozoös, 1(3), 175-183.
Keddie, K. M. G. (1977). Pathological mourning after the death of a domestic pet. British Journal of Psychiatry, 131, 21-25.
Kidd, A. H., Kidd, R. M. (1980). Personality characteristics and preferences in pet ownership. Psychological Reports, 46, 939-949.
Kogure, N., Yamazaki, K. (1989). Attitudes to animal euthanasia in japan: a brief review of cultural influences. Anthrozoös, III(3), 151-154.
Lagoni, L., Butler, C., Hetts, S. (1994). The human-animal bond and grief. New York: W.B. Saunders Company.
Lasher, M. (1998). A relational approach to the human-animal bond. Anthrozoös, II(3), 130-133.
Lawrence, E. A. (1985). The death of a faithful dog: impact and meaning. Journal of Psychoanalytic Anthropology, 8(4), 283-286.
Lloyd-Evans, M. (1987). Human-animal interactions - A. Anthrozoös, I(1), 52-53.
McGuire, M. T., Fairbanks, L. A. (1977) Ethological psychiatry: psychopathology in the context of evolutionary biology. New York: Grune & Stratton, Inc.
Melson, G. F. (1990). Studying children’s attachment to their pets: a conceptual and methodological review. Anthrozoös, IV(2), 91-99.
Melson, G. F., Fogel, A. (1996). Parental perceptions of their children’s involvement with household pets: a test of a specificity model of nurturance. Anthrozoös, IX(2/3), 95-106.
Miles, M. S., Demi, A. S. (1983-84). Toward the development of a theory of bereavement guilt: sources of guilt in bereaved parents. Omega, 14(4), 299-314.
Miller, M., Lago, D. (1990). Observed pet-owner in-home interactions: species differences and association with the pet relationship scale. Anthrozoös, IV(1), 49-54.
Parkes, C. M. (1964). Recent bereavement as a cause of mental illness. British Journal of Psychiatry, 110, 198-204.
Parkes, C. M. (1985). Bereavement. British Journal of Psychiatry, 146, 11-17.
Parkes, C. M. (1998). Bereavement: studies of grief in adult life (3rd edition). Connetticut: International Universities Press, Inc.
Planchon, L. A., Templer, D. I. (1996). The correlates of grief after death of pet. Anthrozoös, IX(2/3), 107-113.
Poresky, R. H., Daniels, A. M. (1998). Demographics of pet presence and attachment. Anthrozoös, 11(4), 236-241.
Poresky, R. H., Hendrix, C., Mosier, J. E., Samuelson, M. L. (1987). The companion animal bonding scale: internal reliability and construct validity. Psychological Reports, 60, 743-746.
Poresky, R. H., Hendrix, C., Mosier, J. E., Samuelson, M. L. (1988). Young childrenÕs companion animal bonding and adult’s pet attitudes: a retrospective study. Psychological Reports, 62, 419-425.
Quackenbush, J. (1985). The death of a pet - how it can affect owners. Veterinary Clinics of North America: Small Animal Practice, 15(2), 395-402.
Rajaram, S. S., Garrity, T. F., Stallones, L., Marx, M. B. (1990). Bereavement: loss of a pet and loss of a human. Anthrozoös, VII(1), 8-16.
Ross, C. B., Baron-Sorenson, J. (1998) Pet loss and human emotion. Philedelphia: Taylor & Francis Group.
Rynearson, E. K. (1978). Humans and pets and attachment. British Journal of Psychiatry, 133, 550-555.
Simpson, J. A. (1999). Attachment theory in modern evolutionary perspective. In J.Cassidy, P. R. Shaver, Handbook of Attachment: Theory, Research and Clinical Applications (pp. 115-140). New York: The Guliford Press.
Staats, S., Miller, D., Carnot, M. J., Rada, K., Turnes, J. (1996). The Miller-Rada commitment to pets scale. Anthrozoös. IX(2/3), 88-94.
Stallones, L. (1994). Pet loss and mental health. Anthrozoös, VII(1), 43-54.
Wilson, C. C., Turner, D. C. (1998) Companion animals in human health. London, Sage Publications
-----------------------
[1] Ainsworth (1989) makes a distinction between an “attachment bond” and an “affectional bond.” She notes that the former is the investment of one individual in another preferred individual (e.g., mother-child bond) (Ainsworth, 1989 as cited in Cassidy, 1999, p.12).
[2] Archer (2001) examines the role of evolution and attachment in the grieving process to highlight the significance of adaptive versus maladaptive mechanisms of grief.
[3] See Bowlby (1969/1982, p.229; see also Murry, 1938) for a discussion on “affiliation” and Greenberg & Marvin, 1982; Ainsworth, 1989 for a discussion on the difference between “sociable system” and “affiliative system.
[4] See Bowlby’s (1969/1982) review of attachment behaviors necessary for establishing an attachment (p.334), and “Attachment as an organizational concept” (p.371).
[5] See other Mediating Influences in this document. For a review of systems that mediate the attachment system see Cassidy, 1999.
[6] In Attachment and Loss (2nd ed.), Bowlby (1969/1982 stated: “the parent-child relationship is a complementary one…. Since a bond is a property of two parties, the bond with which we are concerned should be designated as one of attachment-caregiving” (p.377; see also Hinde, 1979 as cited in Bowlby, 1969/1982).
[7] In response to the results found in a study by Hazen & Shaver (1987) and Kobak & Sceery (1988) Bartholomew & Horowitz stated “a single avoidant-detached category may obscure conceptually separable patterns of avoidance in adulthood” (p.227). Fraley & Shaver (2003) examined the difference between fearfully avoidant, and dismissingly avoidant individuals.
[8] The Strange Situation investigated the interaction between exploratory play and children’s response to separation from the attachment figure (see Fraley and Shaver, 1999, p.739).
[9] Bowlby (1969/1982) refers to instinctive as a “rough and ready way to behaviour that in the environment of evolutionary adaptedness has consequences that are vital to the survival of the species and that is controlled by systems which in that environment are usually fairly stable” (p. 136). Bowlby confirmed, however, that there “can be no cut-off point between what is called instinctive and what is not,” since many environmental systems ranging from the most stable to the least stable (e.g., in terms of survival value) control instinctive behavior. In other words, there is a difference between a ‘need’ and a behavior that has survival value).
[10] The term ‘evolutionary adaptedness’ is described by Bowlby (1969/1982) as the quality of the environment allowing for adaptation. For example, the oxygen level in the air is a condition without which human survival would fail.
[11] Ainsworth (1969) introduced the concept of secure base.
[12] This applies primarily to the mammalian species, as the behavior of birds appears to be related to a different behavioral mechanism; see Bowlby, 1969, 1982, p. 183.
[13] Bowlby (1969, 1982) likened attachment behavior to a biological function, defined as a “consequence that in the course of evolution has led the behavior in question to become incorporated into the biological equipment of a species. Such incorporation occurs as the result of some advantage (in terms of survival and differential breeding success) that the behavior confers on those possessed of it” (p. 225).
[14] Bowlby (1969/1982, 1988) used the term pre-programmed according to cybernetic theory, which states that when a transformation in behavior occurs it is the result of the environments influence on the system (e.g. attachment behavioral system) or the influence of the system on the environment. Attachment behaviors follows the pattern of a closed or negative feedback loop by which goal-seeking behavior is maintained (see )
[15] Thus, does the mother respond to every child cry or sometimes wait until the child stops crying before picking him/her up. Bowlby’s (1958) theory regarding individual differences in attachment has received empirical support in studies conducted by the following researchers: Ainsworth, Blehar, Waters, & Wall, 1978; see also DeWolff & van Ijzendoorn 1997 (as cited in Cassidy, 1999, p.7).
[16] “Attachment behavior is conceived as a class of behavior distinct from feeding behavior and sexual behavior and of at least an equal significance in human life. There is nothing intrinsically childish or pathological about it” (Bowlby, 1977, p.204).
[17] This system is differentiated from learning theory and drive theory- see above (need references, Freud).
[18] This latter theory has been used to explain symptoms of Post Traumatic Stress Disorder (PTSD), and the chemical link between substance use and dependence and relapse (need reference).
[19] Reporting on her observations of Ganda infants, Ainsworth (1963) stated: some of the infants…who seem most solidly attached t their mothers displayed little protest behavior…but rather showed their strength…through their readiness to use her as a secure base from which they could explore” (as cited in Bowlby, 1969/1982, p.333).
[20] Species characteristic behaviors have survival advantage as noted in review of Bowlby’s attachment theory.
[21] Cassidy (1999) derived this criterion from Ainsworth’s (1989) theoretical analysis.
[22] In reference to unwilling or permanent losses she states: “Inexplicable separation tends to cause distress, and permanent loss would cause grief” (p.711). Also, as has been noted on the previous page the type of affective response may range from greeting behavior to a form of rejection, which will often be congruent with affect (Bowlby, 1969/1982).
................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.
Related searches
- the importance of a college education
- the importance of a name
- the importance of a teacher
- the importance of a will
- the purpose of a business plan
- what is the salary of a doctor
- the role of a finance manager
- derivative of the area of a triangle
- the represents the domain of a function
- example of the theme of a story
- a the abundance of a ground beetle species in a meadow b the zonation of seaweed
- the purpose of a chamber of commerce