Differences and Similarities in Humans' Perceptions of the Thinking and Feeling of a Dog and a Boy

Jeffrey Lee Rasmussen and D. W. Rajecki
Indiana University - Purdue University at Indianapolis

College students' perceptions of companion dog mentality were systematically compared with perceptions of human child mentality. Independent groups of respondents rated capacities of a dog or a boy on 12 categories of thinking and 30 items of remorseful feelings for misbehavior. The boy received superior ratings for so-called "complex" (but not "simple") thinking categories and "upper level" (but not "lower level") remorse items. Even so, there were strong associations between dog and boy means across all 12 thinking categories ( r = .74) and all 30 remorse items ( r = .72). Thus elements of thinking and feeling that were judged likely (or not) for the boy were also relatively likely (or not) for the dog. These several comparisons were taken to indicate that whereas the dog and boy were perceived by subjects as having mentalities that were quantitatively different , those respective mentalities were nevertheless viewed as qualitatively similar . Findings were discussed in terms of social-cognition theories of anthropomorphism.

Evidence exists that laypersons are often willing to credit some form of mindedness to many of their nonhuman companions. For example, in a survey at a veterinary clinic, Katcher (1981) found that fully 81% of owners thought their companion animals were sensitive to their human feelings. Ninety-nine percent of the respondents reported talking to companion animals, and 94% did so in the manner of talking with a person. Some 28% of the owners confided in their companion animals and 28% talked of events of the day (cf. Johnson, Garrity, & Stallones, 1992; Hirsch-Pasek & Treiman, 1982). Other analyses indicated that people feel they are loved by their companion animals (Serpell, 1986) or that their companion animals experience jealousy (Mathes & Deuger, 1982).

We feel that patterns such as these can be understood in terms of a view of popular anthropomorphism that is based on social cognition processes. Two such cognitive theories -- shared experience, and the social construction of humanness - will be taken up in our Discussion. However, before turning to a theory of lay impressions of companion animal mindedness, it is necessary to know just what phenomena to theorize about. A first concern is the establishment of quantitative or qualitative descriptions of people's perceptions of animal mentality. Apart from the examples sketched above, what manner of mind do humans sense in nonhuman animals?

Mentality Ascriptions in Humans' Perception of Animals

We would identify several researches as having to do with humans' ascriptions of certain psychological traits or capacities to nonhumans: Burghardt (1985), Eddy, Gallup, and Povinelli (1993), Hogan (1994), and Sanders (1993). Perhaps the most elaborate experiment in this vein was that of Rasmussen, Rajecki, and Craft (1993). Using bipolar rating scales, they asked college students to indicate whether it was reasonable to say (or not) that a companion dog, cat, bird, fish, and a school-age child had the capacity for 12 commonplace human mental capacities or experiences.

Their factor analysis of responses identified two levels of ascriptions that were labeled simple thinking (e.g., pleasure and displeasure; emotion; gratitude) and complex thinking (e.g., object permanence; enumeration and sorting; memory and foresight). The child and all animals were credited with simple thinking, but respondents were much more likely to restrict their ascriptions of complex thinking to the child. Because this paradigm appears to be fruitful for mapping people's perceptions of animal mentality, we attempted to broaden such findings.

Design Considerations

The above general approach (Rasmussen et al., 1993) was adopted for the current project. That earlier research showed the utility of comparison judgments about a child target, so the current work also included material about a hypothetical boy. However, this paper offers several important distinctions compared with the original. The essential design of the Rasmussen et al. study was within-subject; every respondent made judgments about each of four animal targets and a human child. This approach opens the possibility that human/nonhuman contrasts from a within-subject design might be due in part to the sheer opportunity to make comparisons.

We took a deliberate step to control this potential moderating variable. Here, a between- subject design was employed for actor contrasts. A subject made judgments only about a single target: a dog or a boy. Obviously, a comparison of our between-subject results with earlier within-subject results would be valuable.

Further, while human impressions of the mentality of companion cats, birds, fish, and other species are of theoretical and practical importance, for the sake of matching the behavior of an animal to that of a boy (below), we chose to focus the nonhuman comparison on a dog. The choice of the domestic dog as a representative companion for this study was defensible based on their numbers and highly social nature (Sanders, 1993).

Design Extensions

Rasmussen et al. (1993) limited their measures to questions about "thinking." The current study extended the range of investigation to include thinking and certain forms of "feeling." It happens that the most frequent dog behavioral problem presented at clinics is inappropriate aggression (Landsberg, 1991). There could be a tendency on the part of some to see such aggression as intentional , which opens the possibility that the animal itself might be held psychologically accountable to some extent for this presumed misbehavior. Some lay judges might expect that the dog could and should feel shame and guilt for its transgression. (Note that this is not a question of what should be done with the animal.) For quantitative measures of guilt and shame we turned to scales from the child development and social psychology literature.

Method

Versions of a standardized questionnaire were used to record respondents' estimates of whether it was reasonable to say that an actor (dog or boy) had certain mental capacities, and the capability to experience certain elements of guilt and shame. Conditions in a factorial experimental design were represented in particular versions of the instrument.

Mental Operations (MO) Categories

Labels for the 12 capacities or experiences researched by Rasmussen et al. (1993) - hereafter referred to as mental operations, or MO categories - are listed in Table 1, below. For the sake of multiple measurement two questionnaire scale items were used for each of the 12 categories. The two items per category were meant to be combined in the eventual analysis. For a list of all 24 items see Appendix A. In the instruments themselves, the questions about a given actor type occupied both sides of a single sheet. At the top of the first page of items a paragraph outlined a standard hypothetical situation for each of the actor types. In this example the actor is a boy:

Situation : The setting is a typical household. In the scene there is an elementary school-age boy. In front of the boy is a cookie, a treat. Please answer the following questions about the situation.

Question : Is it reasonable to say that the boy...

This paragraph was followed by the 24 MO items, in some order as specified below.

Guilt/Shame (G/S) Items

Because we focus here on an actor's own psychological accountability for a presumed transgression, we sought established measures of guilt and shame. Many such examples were listed in recent reports by Ferguson, Stegge, and Damhuis (1991) and Kugler and Jones (1992); we are indebted to those authors for suggesting scale items.

Our items were developed by partially modifying earlier lists on the basis of whether it might be expected that a boy and a dog could show the demeanor or emotion. For example, earlier items, "I really want to tell what happened," and "then I'd really get red in my face" were not used. Some additional items (e.g., "feel mad" and "whimper") were created for the present study as they appeared to be expectable reactions for the boy and dog.

At the appropriate location in the instrument - immediately following the transgression scenario (see below) - subjects were asked their opinion of how an actor felt or reacted. The question stem was: "Is it reasonable to say that [the] Dog (Boy) would...", followed by 30 terms. Those 30 terms are listed in Table 2, below, and are hereafter referred to as guilt/shame, or G/S items.

Dog and Boy Transgressions

Levels of aggressive biting and threatening were chosen to represent potential transgressions common to children (Abramovitch & Strayer, 1978; Matson, Manikam, & Ladatto, 1990; Rajecki & Flanery, 1981; Solomons & Elardo, 1989) and dogs (Crowell-Davis, 1991; Fisher, 1993; Oswald, 1991).

At the beginning of the G/S section of a questionnaire the items were preceded by a story about two neighboring families. The story concerned either a male dog or a boy who acted as an inappropriate aggressor. Initially, the degree of transgression was intended to be at one of four levels. However, results and hindsight now indicate that these levels did not interact in a significant way with the key dog/boy variable. Only one such transgression will be described here; others involved some level of intensity of biting or threatening.

A story about a dog designed to show what we took to be an extreme amount of inappropriate aggression was:

In a certain residential neighborhood, Family A and Family B lived next door to one another. A low picket fence divided the yards of the two houses. There was a backyard gate in the fence, allowing travel between the two properties. Because the two families were generally friendly, the gate was sometimes left open, and sometimes shut. Family A included a school-age boy called A. Family B included a young male dog called B.

One day Dog B was given a new toy, which he liked very much. That day the gate between the two yards happened to be open. Boy A saw the new toy on the ground in the Family B yard. Curious, Boy A entered Family B's yard and picked up the toy. Unexpectedly, Dog B reacted immediately.

Dog B flung himself on Boy A and bit him hard, causing a serious tear in the skin. Boy A became terrified, dropped the toy, and ran screaming back into his own yard.

Dog B looked at the window of Family B's house and saw his adult owners staring at him in a disapproving manner.

In stories with the boy as aggressor, "Dog B" was replaced by "Boy B" and appropriate terminology. Following the presentation of a scenario, a subject was asked to indicate how mild or severe he or she believed Dog/Boy B's wrongdoing was, using a 7-point VERY SEVERE (1) to VERY MILD (7) scale.

Instrument Construction

The questionnaire consisted of three main sections. The first section gave general information about the study, presented safeguard information, provided spaces for subjects to indicate gender and age, and gave information on how to use the rating scales. The second and third sections contained the items for the MO and G/S scales. Half the instruments presented the MO questions first, followed by the G/S questions; half presented the G/S questions first, followed by the MO questions. A marker sheet separated the first set of questions from the second, and another marker sheet was placed at the end of the questionnaire. These sheets indicated the end of a section and requested that the subject check that she or he had answered every question.

Mental Operations (MO) Items

The items for each generic category for the MO variable separated into two lists of 12 each. The order of items in both lists were then independently randomized and recombined into a list of 24 items. A mirror-image list of the 24 items was also generated and used in about half of the eventually usable sample of questionnaires. Each item was accompanied by a 7-point scale with bipolar endpoints of UNREASONABLE (1) and REASONABLE (7), with a mathematical midpoint of 4. Half the items had UNREASONABLE-REASONABLE for the endpoint ordering; half had REASONABLE-UNREASONABLE.

Guilt/Shame (G/S) Items

For the 30 G/S items, one list was randomly generated and its mirror image was also generated. The G/S items used the same scale format as the MO items.

Subjects

This project was assigned "exempt" status by the local university Institutional Review Board. Convenience samples of volunteer subjects were tested in introductory psychology classes from late spring through summer, 1993. Of the questionnaires that were returned, six were discarded because subjects failed to follow instructions. The remaining 189 usable questionnaires came from 130 women and 59 men. (This gender split could be expected from a convenience sample due to the normal demographics of the classes tested.) Ninety-one subjects filled out the dog questionnaire version, and 98 filled out the boy version. Of the available instruments, 181 were complete and 8 had one piece of missing data.

Review of Design

Between-Subject Variable

The key between-subject factor was determined by the actor presented in a given questionnaire: dog or boy. Many analyses in the results section (below) hinge on this distinction.

Other between-subject factors were level of transgression and sex of subject. However, we have already noted that level of transgression did not interact with the dog/boy factor, and neither did subject gender produce statistical interactions. Therefore, these particular variables do not inform the current issues. Although included in all formal statistical tests as potential sources of error variance, the variables of transgression level and gender are dropped from further discussion.

Further, various item orders, and MO versus G/S section orders in instruments (as noted) were systematically shuffled. Different combinations of given orders were used for the sake of counterbalancing the presentation of material, and are dropped from consideration as variables.

Within-Subject Variables

The individual MO items were combined appropriately to yield 12 levels of an ascribed mental operations variable, and the individual G/S items can be considered as 30 levels of a perceived guilt and shame variable. Where appropriate, these within-subject factors were analyzed along with the between-subject factor in mixed designs (e.g., Actor (dog/boy) Type X G/S Items).

Results

Mental Operations (MO) Categories and Analyses

As reported, ratings on the two items per MO category were combined in the analysis. Table 1 presents the means for the resulting 12 MO categories for the dog and boy. According to subjects'judgments it was most reasonable to say that the actors were capable of Emotion and Sensation/perception, and least reasonable to say they were capable of Conservation. The general patterns of category means in Table 1 are similar to those from Rasmussen et al. (1993, Table 1), keeping methodological differences in mind.

Previous research by Rasmussen et al. (1993) indicated that responses to the MO item categories were factor analyzable into two factors. Thus an orthoganal varimax analysis was used on the present data (the BMDP 4M program; see Dixon, 1990). Three factor analyses were run: the first (common) included the dog/boy questionnaire responses together, and the remaining two were carried out for the dog and boy types separately.

Results of the common and separate factor analyses showed the same factor structure. Further, the current common factor structure was virtually identical to that reported by Rasmussen et al. (1993, Table 2). Both researches found that ratings of the categories of Emotion, Sensation/perception, Pleasure/displeasure, Gratitude, and Play/imagine loaded on Factor 1 (called "simple thinking") and Object permanence, Memory/foresight, Enumeration/sorting, Dream, Morality, Schemata, and Conservation loaded on Factor 2 (called "complex thinking"). These factor affiliations are also shown in Table 1.

To reduce the 12 MO dimensions to two indices, the average of the unweighted linear composite of the five Factor 1 categories listed in Table 1 was formed into one index, and that from the remaining seven categories (Factor 2) was formed into another. Taken together these make up two levels of a repeated measures variable, which will be referred to as "factor." As noted, following Rasmussen et al. (1993) these two levels will be labeled respectively as simple thinking and complex thinking.

A 2 x 2 (Actor Type X Factor) mixed-design ANOVA was carried out. There was a main effect for Actor Type, F (1,188) = 40.45, p < .001. Subjects reported that overall it was more reasonable to say that the boy ( M = 5.45) possessed the two levels of the mental operations rather than the dog ( M = 4.79). Further, there was a significant main effect for Factor, F (1,188) = 426.42, p < .001, which indicated that subjects believed it more reasonable for actors to possess simple thinking ( M = 5.75) than complex thinking ( M = 4.49). Finally, there was a significant Actor Type X Factor interaction, F (1,188) = 127.08, p < .001, with respective means presented in the lower part of Table 1. It can be seen that there was a small difference between ratings of simple thinking for the dog and the boy, and a large difference between the ratings of complex thinking for the two actor types. In other words, both dog and boy were credited with simple thinking, but ascriptions of complex thinking were more likely for the boy.

Correlation Analysis of Dog/Boy MO Means

Note in Table 1 that whereas the boy was generally superior to the dog on complex thinking (Factor 2) the canine did receive ratings that were above the scale midpoint (4.00) on three of those categories. Thus complex thinking was not seen as altogether alien to the dog. The degree of judged commonality between the dog and boy is further suggested by an examination of the outlier pairs in Table 1. If one group of subjects thought it was reasonable to say that the boy had the capacity for Emotion ( M = 6.39), the other group also thought it reasonable of the dog ( M = 6.12). As a parallel, if it was much less reasonable to say that the boy was capable of Conservation ( M = 3.96), it was also less reasonable to say this of the dog ( M = 2.34). Taking all data points into account, there was a strong positive correlation, r = .74, p < .01, between dog and boy means for the 12 MO categories in Table 1. A scatterplot of this relationship (in triangles) is shown in Figure 1.

Analyses of G/S Items

In an effort to reduce the total G/S items to a smaller number of indexes, a factor analysis of these measures was also carried out. However, on the basis of our brief literature search there was no strong reason to expect the procedure to yield a clear factor structure. Indeed the results of the factor analysis gave no interpretable results, so all 30 items were retained.

An ANOVA, including an Actor Type X G/S Item contrast, was carried out. The results indicated a significant effect for Actor Type, F (1,172) = 21.26, p < .01, which meant that it was less reasonable to say that the dog ( M = 4.06) than the boy ( M = 4.81) would show the various G/S reactions. (The only and nonsignificant exception was that subjects believed that the dog was somewhat more likely to want to hide compared with the boy.)

Importantly, there was also a significant Actor Type X G/S Item interaction, F (29,4988) = 3.32, p < .001, which is shown in Table 2. The table gives the averages for the G/S items for the dog and boy actor types, rank ordered in terms of F ratios contrasting individual pairs of means. The F ratios, which were calculated using separate error variance for each item, ranged from F (1,186) = 39.14 to 8.03 for the significant ratios ( p < .01), and F (1,186) = 7.36 to 0.00 for the nonsignificant ratios ( p > .01). (Given the number of comparisons carried out, it was deemed more appropriate to use the p = .01 than the p = .05 level to establish significance.)

By convention, entries in the top panel of Table 2 are termed "upper level" items, and those in the bottom panel are termed "lower level" items. Given the unproductive factor analysis of G/S material (above), this upper/lower distinction provides at least one meaningful way to sort out subsets of these items. Of course, by choosing the terms "upper" and "lower" we mean to connote a judgmental distinction between the two levels, just as in the case of simple and complex thinking from Rasmussen et al. (1993).

To put the upper/lower differences another way, as seen at the bottom of Table 2, both dog and boy were credited with lower-level remorse, but ascriptions of the upper-level type were more likely for the boy.

Correlation Analysis of Dog/Boy G/S Means

Upper level G/S items are thus by definition those on which the boy target received considerably higher ratings. However, this does not imply that raters thought the dog target was utterly devoid of these experiences. On six of the 16 items in the top panel of Table 2 the means for the dog are above the scale mathematical midpoint (4.00).

Further, the mean G/S upper and lower level item scores in Table 2 are ordered by F ratios. This method of display partly obscures a marked similarity between values for the boy and the dog at each level. Using the outlier pairs in Table 2 as examples, if one group of subjects thought it was reasonable to say that the boy would "be afraid of being punished" ( M = 5.96), the other group of subjects thought it was also reasonable of the dog ( M = 5.57). As a parallel, if it was much less reasonable to say that the boy would "experience physical discomfort" ( M = 3.71), it was also less reasonable to say this of the dog ( M = 2.81). Taking all data points into account, there was a strong positive correlation, r = .72, p < .001, between the dog and the boy mean scores for the 30 G/S items in Table 2. A scatterplot of this relationship (in circles) is shown in Figure 1.

For the record, when analyzed separately the degree of dog X boy associations between means proved even stronger for the 16 G/S upper-level items, r = .87, p < .001; and the 14 G/S lower-level items, r = .90, p < .001.

Exploratory Analyses of MO and G/S Covariation

Because the current guilt/shame data did not show a factor structure it was not possible to arrive at a straightforward correlational analysis based on composite factor scores. Nevertheless, an exploratory analysis was carried out that compared subjects' responses on the simple and complex thinking dimensions, and their responses on the upper level and lower level feeling items. Individuals' scores on both the simple and complex MO composites were moderately predictive ( r s = .28 and .36) of their positions on the lower level G/S index. That is, respondents who assigned relatively high simple or complex thinking scores were somewhat more likely to perceive the transgressing actors as experiencing guilt or shame of the lower-level variety.

In contrast, respondents who assigned relatively high complex thinking scores were quite likely ( r = .56) to also assign relatively high scores on the measures that fell in the upper level G/S category. Of perhaps equal import, position on the simple MO composite was effectively unpredictive ( r = .13) of standing on the upper level index. That is, for the combined dog/boy analysis, perceptions of upper level guilt and shame were related only to mental ascriptions of the complex sort.

Finally, for readers versed in the technique, a canonical correlation analysis of the 12 thinking categories and the 30 feeling items was undertaken. In brief, the squared canonical correlation, which indicates the variance accounted for in the guilt/shame canonical factor by the mental operations canonical factor (or vice versa) was a large .61. Such covariation patterns warrant refined and extended future research.

Discussion

Mental Operation Categories

Recall that the Rasmussen et al. (1993) comparative study of ascriptions of mental operations employed a within-subject approach, whereas the current cross-species data are based on a between-subject design. Any parallels in the results of the two projects would therefore strengthen faith in the lay perceptions as revealed.

Accordingly, it is impressive that the factor structure for the MO categories found in the current study is highly similar to the factor structure for those same categories reported by Rasmussen et al. (1993, Table 2). Further, the 2 X 2 table of means for thinking factor X actor type at the bottom of the current Table 1 are a close match to a comparable set of means from that earlier study (1993, Table 3).

Apparently the MO categories employed in these separate studies tapped about the same lay impressions of human (child) and animal thinking, regardless of the differences in within- versus between-subject procedures. Subjects in our study may or may not have used a template of human mind to make estimates for the dog, but at least the current procedure did not force them to make direct cross-species comparisons.

Clearly, subjects in this study considered a human child more likely to have a capacity for thinking, compared with a dog. But the dog was surely not judged as a being without mental operations. As seen in Table 1, dog mean ratings of all Factor 1 (simple thinking) categories (100%) exceeded the scale mathematical midpoint of 4.00, and several Factor 2 (complex thinking) ratings (43%) did so as well.

Of course we realize that all the current conclusions are based on a sample of college students, whose judgments may or may not generalize to other sectors of society. But it is useful to note that the earlier project by Rasmussen et al. (1993) employed college students in their Experiment 1, and academic veterinarians and psychologists in their Experiment 2. The findings of those two experiments were quite similar.

Guilt/Shame Items

Clearly, also, subjects in this study considered a human child more likely to have a capacity for remorse, compared with a dog. But the dog was not seen as an actor who entirely lacked a conscience. As seen in Table 2, dog mean ratings of many lower level G/S items (86%) exceeded the scale mathematical midpoint of 4.00, and several upper level G/S ratings (38%) did so, too.

Quantitative Differences and Qualitative Similarities

Accordingly, a key to impressions of canine mentality is found in Tables 1 and 2. Not surprisingly, perhaps, independent groups of raters judged that the child and dog were not equal regarding mental operations or guilt/shame reactions. An important first point, however, is that differences in those measures did not appear "across the board." That is, the dog was not seen to have a mentality that was less than the child's by some constant over all dimensions. Ratings revealed subtle distinctions between both types of mental activities. Subjects were willing to grant the dog the benefit of the doubt where simple thinking and lower level guilt and shame were at issue, but they frequently drew the line at complex thinking and upper level remorse.

Even so, a second important point is that while unequal in critical ways, the dog and the child turned out to still have a great deal in common . The intra-measure, cross-species correlations (all r s greater than .70) for the MO categories and the G/S items are very interesting. They tell us that where specific scale items were under consideration, independent samples of subjects sensed that what was reasonable to say about the child was also relatively reasonable to say about the dog, as seen in Figure 1. The correlational picture tells us that even though the dog and the boy were seen as clearly unequal in expectable or interpretable ways, at base subjects employed the same "model" of boy and dog mentality. To put it another way, the dog and boy were seen as quantitatively different , but qualitatively similar .

Theory: The Social Cognition of Anthropomorphism

The current research has revealed an essentially anthropomorphic picture of students' views of dog mindedness. What is the basis for the tendency to view animal mentality in human terms? To the extent that people already use anthropomorphic models to understand one another, it seems no less legitimate to use anthropomorphism to predict or explain certain animal behavior. In our view, the tendency to anthropomorphize companion animal mentality can be understood in terms of concepts compatible with those found in research on person perception and social cognition. Briefly, we wish to sketch two social-cognitive explanations for the phenomenon: (i) shared experience, and (ii) the social construction of humanness.

This perspective is based on the assumption that human beings have minds. Mind is defined as the "ability to monitor your own mental states, and the corresponding capacity to use your experience to infer the experience of others" (Gallup, 1985. p. 633). Mind allows us to make note of our own experiences in situations. Then, when we observe others in similar situations, we are able to use memory to model their private experiences. In other words, because a person monitors her or his own mental states and their connection with external events, this provides a means of achieving intuitive knowledge of other peoples' mental states.

The anthropomorphization of companion animal mentality could stem in part from this basis. In domestic settings people and their companion animals frequently face similar situations. To the extent that a companion animal's reaction to a situation has something in common with that of the human observer, the tendency would be to interpret the animal behavior in human subjective terms. For example, when the doorbell rings the dog barks and the person jumps up from the chair. Why? Apparently - to the person - they both "wonder who is at the door."

Canines are not the mental equals of people, but companion dogs are often viewed as members of the family or as close friends. As it happens, not all persons are the mental equals of others, but many families and other social units meaningfully function with members who are profoundly retarded. These severely handicapped, alingual individuals enjoy considerable social status by virtue of the social construction of humanness (Bogdan & Taylor, 1989).

Four basic factors are involved in this social construction: (i) the nondisabled attribute thinking to the disabled and interpret their sounds, postures and expressions as reflections of intelligence, (ii) the disabled are seen as individuals having distinct personalities and unique histories, (iii) the disabled are seen to reciprocate in relationships as true companions and objects of caring, and (iv) the disabled are incorporated into a social place by having a role in daily patterns of events and routines (after Sanders, 1993).

In an ethnographic investigation of caretakers' attributions of mindedness to their dogs, Sanders (1993) found that all four of the above factors could be documented in the interactions of people with their companion animals. To the extent that social construction of humanness can occur for retarded, alingual persons in families, the same thing could happen for simple-thinking, alingual animals that live in our families.

Recommendations

The shared experience and social construction theories noted above are not mutually exclusive, but are different enough to inspire particular lines of inquiry. We think they can be gainfully applied to the question of anthropomorphism in important areas of human-animal social relations. Further, the current questionnaire approach to mapping humans' perceptions of nonhuman mindedness and experience has again proved fruitful. This paradigm allows a combination of survey and experimental methodology through the use of structured scale items, and the representation of designed conditions in versions of booklets.

An integration of social cognition theory with appropriate methodology could therefore usefully expand knowledge of person/companion animal relationships in particular, and human/animal relations in general. Various human populations, animal species, and mental traits, capacities or processes are open to such study.

References

Abramovitch, R., & Strayer, F. (1978). Preschool social organization: Agonistic, spacing, and attentional behaviors. In L. Krames, P. Pliner, & T. Alloway (Eds.), Aggression, dominance, and individual spacing . New York: Plenum Press.

Bogdan, R., & Taylor, S. J. (1989). Relationships with severely disabled people: The social construction of humanness. Social Problems , 36 , 135-148.

Burghardt, G. M. (1985). Animal awareness. American Psychologist , 40 , 905-919.

Crowell-Davis, S. L. (1991). Identifying and correcting human-directed dominance aggression of dogs. Veterinary Medicine , 86 , 990-998.

Dixon, W. J. (1990). BMDP manual [Computer program manual]. Berkeley, CA: University of California Press.

Eddy, T. J., Gallup, G. G., Jr., & Povinelli, D. J. (1993). Attribution of cognitive states to animals: Anthropomorphism in comparative perspective. Journal of Social Issues , 49 , 87-101.

Ferguson, T. J., Stegge, H., & Damhuis, I. (1991). Children's understanding of guilt and shame. Child Development , 62 , 827-839.

Fisher, J. (1993, June). Rage syndrome - Possesive aggression or normal behavior? International Society for Anthrozoology Newsletter , pp. 5-6.

Gallup, G. G., Jr. (1985). Do minds exist in species other than our own? Neuroscience and Biobehavioral Reviews , 9 , 631-641.

Hirsch-Pasek, K., & Treiman, R. (1982). Doggerel: Motherese in a new context. Journal of Child Language , 9 , 229-237.

Hogan, W. (1994, Fall). Can a cockroach have a nice day? The Latham Letter , pp. 1, 14-15.

Johnson, T. P., Garrity, T. F., & Stallones. L. (1992). Psychometric evaluation of the Lexington Attachment to Pet Scale (LAPS). Anthrozoos , 5 , 160-175.

Katcher, A. H. (1981). Interactions between people and their companion animals: Form and function. In B. Fogel (Ed.), Interrelations between people and companion animals (pp. 41-67). Springfield, IL: Charles C. Thomas.

Kugler, K., & Jones, W. H. (1992). On Conceptualizing and assessing guilt. Journal of Personality and Social Psychology , 62 , 318-327.

Landsberg, G. M. (1991). The distribution of canine behavior cases at three behavior referral practices. Veterinary Medicine , 86 , 1011-1018.

Mathes, E. W., & Deuger, D. J. (1982). Jealousy, a creation of human culture? Psychological Reports , 51 , 351-354.

Matson, J. L., Manikam, R., & Ladatto, J. (1990). A long-term follow-up of a recreate the scene, DRO, overcorrection and lemon juice therapy program for severe aggressive biting. Scandanavian Journal of Behaviour Therapy , 19 , 33-38.

Oswald, M. (1991). Report on the potentially dangerous dog program: Multnomah County, Oregon. Anthrozoos , 4 , 247-254.

Rajecki, D. W. & Flanery, R. C. (1981). Social conflict and dominance in children: A case for a primate homology. In M. E. Lamb & A. L. Brown (Eds.), Advances in developmental psychology (Vol. 1, pp. 87-129). Hillsdale, NJ: Lawrence Erlbaum Associates.

Rasmussen, J. L., Rajecki, D. W., & Craft, H. D. (1993). Humans' perceptions of animal mentality: Ascriptions of thinking. Journal of Comparative Psychology , 107 , 283-290.

Sanders, C. R. (1993). Understanding dogs: Caretakers' attributions of mindedness in canine-human relationships. Journal of Contemporary Ethnography , 22 , 205-226.

Serpell, J. (1986). In the company of animals . New York: Basil Blackwell. Solomons, H. C., & Elardo, R. (1989). Bite injuries at a day care center. Early Childhood Research Quarterly , 4 , 89-96.

For abstracts of all issues, including the most current, click Article Abstracts

To order Society & Animals Journal, go to our secure online ordering page

You can Search the online issues of Society & Animals, as well as the entire Society & Animals Forum (formerly PSYETA) website,
for topics and keywords of your interest: