Social Play in the Domestic Cat. Department of Psychology, Cornell University, Ithaca, New York 14850

Transcription

1 AMER. ZOOI.., 14: (1974). Social Play in the Domestic Cat MEREDITH WEST Department of Psychology, Cornell University, Ithaca, New York SYNOPSIS. The social play of domestic cats in the laboratory and at home was examined. Categories and sequences of motor patterns were identified and analyzed. The developmental period during which social play was most frequent was established to be from 4 weeks to 4 months. The results suggest that social play functions to provide specific forms of exercise and as a means of developing and maintaining social relations among littermates. The decline in social play appears to be related to dispersal of the young and to a decrease in preference for social contact. INTRODUCTION The topic of animal play has a curious history. For many years, play received extensive theoretical attention, but little experimental study (Muller-Schwarze, 1971). Theories abounded with respect to play's important contributions in aiding the young animal to learn about itself and its environment (Loizos, 1967). Although animal play is now beginning to be studied in a variety of mammalian species (Chepko, 1971; Farentinos, 1971; Steiner, 1971; Bekoff, 1972), the importance of play seems to be suffering a theoretical reversal. Even before the data are in or analyzed, recent writings have now dismissed play as an invalid or scientifically useless categorization of animal behavior (Berlyne, 1969; Welker, 1971). It was the purpose of this study to examine social play activity in the domestic cat and to explore its usefulness as an explanatory concept. This examination included identification of patterns of motor behavior associated with social play, description of developmental changes in play and analysis of some aspects of the structure of social play. The results indicate several ways in which social play may be beneficial to the young kitten. This research was supported by N.I.M.H. predoctoral fellowship award 1 FO1 MH I thank Dr. Robert E. Johnston for his invaluable assistance in the preparation of this manuscript. 427 METHODS The subjects in this study were 28 domestic cats that were studied for durations ranging from 12 weeks to 7 months over a period of 2 years. All were observed from birth. These animals constitute six litters obtained from four mothers, two of whom produced two litters. The mothers were kept with their young for the duration of the study. Due to the fact that the domestic cat, by virtue of its relationship with man, has been subjected to a broad range of living conditions, it was felt necessary to utilize several locations for this study. Fourteen kittens (3 litters) and their mothers were studied in the laboratory, where their movements were restricted to a 30 ft by 20 ft room and an indoor exercise area. Fourteen kittens (3 litters) and their mothers were housed in the observer's home in a rural area where they were permitted unlimited access to the outdoors by means of a special entrance which the cats could control. The mothers of the home-reared litters all "chose" to give birth indoors and to keep their young indoors for the first 6 weeks. The methods of care and observation were essentially the same for all litters.all the cats were fed twice a day at 8 AM and 4 PM. Their diet included milk, cottage cheese, Purina canned cat food, and Purina Cat Chow. The Purina Cat Chow and water were always available. Observations were also made on three

2 428 MEREDITH WEST feral mothers and their seven young. These cats lived exclusively outdoors, had no permanent indoor residence or formal attachment to humans. Due to the difficulties associated with observing these individuals, systematic data were hard to gather. Observations on these cats were made primarily to obtain information on the duration of the parent-young relationship and the use of space by the mothers and kittens. Observations were made daily on each litter for a period of 90 min, beginning on day 14 after birth. This observation period was always preceded by a variable period in which the animals were fed, handled, and their quarters cleaned. Observations on any one litter were made at the same time every day. This time, however, was varied across litters. During the observation period, written records were made on the following aspects of play behavior: description of the types of motor patterns used during social play, records of the content of social play encounters between kittens (what motor patterns were seen), records of when play motor patterns first appeared, measurement of the frequency of play within the observation period, and description of play activity between kittens and inanimate objects. An attempt was made to sample behavior of each kitten within the litter equally during each session. RESULTS Preliminary observations revealed that it was possible to identify eight categories of motor patterns associated with social play. These patterns constitute the building blocks for play. The structure of social play is created through the combination of these patterns into sequences involving several patterns and involving two or more kittens. Before discussing these sequences, a brief description of the patterns will be given. The first pattern defies any simple designation. The kitten lies on its back, its belly up, with all four limbs held in a semivertical position. The tail is straight back and may be moved back and forth. Typical paw movements associated with the bellyup (BU) posture are to move the back legs in a treading motion and to make reaching or pawing movements with the front legs. The mouth is held open and the teeth are exposed. In a social encounter, one kitten assumes the belly-up position and another kitten stands over it. Thus, the treading and pawing movements bring the kitten into contact with parts of the body of the standing kitten. Usually, these areas are the head, neck and ventral area. The second pattern in the stand-up (SU) position just mentioned. The kitten stands near or over another kitten with its head oriented toward the head and neck region of the other kitten. The stand-up kitten's mouth is open and it may direct "bites" toward the other kitten. The kitten may also raise one of its front paws and paw at the other kitten. The belly-up and stand-up position are often exchanged by two interacting kittens, so that they move as one tumbling unit as the exchanges are made. The third pattern is a face-off (FO). A kitten sits near another kitten and hunches its body forward, moving its tail back and forth, and lifts a front paw and moves it in the direction of the other kitten. The kitten's head and eyes are also oriented toward the other kitten. Two kittens may face-off simultaneously and direct their front paw movements at one another's face. The fourth pattern is a vertical stance (VS). A kitten assumes a sitting position and then rocks back on its hindquarters, lifts its front paws off the ground and stretches them out perpendicular to its body. The kitten also extends its back legs so that it is in a stationary bipedal position. The fifth pattern is a horizontal leap (HL). The kitten assumes a lateral position, with respect to another kitten, arches back slightly and curves its tail upwards and towards its body then leaps off the ground. The sixth position is somewhat similar. The kitten arches its back, curls its tail upwards and walks sideways toward or around another kitten or object. This is called a side-step (SS). The seventh pattern is a pounce (P). The kitten crouches with its head held low or

3 SOCIAL PLAY IN THE DOMESTIC CAT 429 touching the ground and its back legs tucked in and its tail straight back. The tail may be moved back and forth. The kitten moves its hindquarters back and forth and moves forward, the thrust coming from the extension of its back legs. 45 a o 30 o hi U a S 15 ui u a K oo o z 111 u a K 13 uo o z Id OK U FIRST POSITION POUNCE SIDE-STEP BELLY-UP STAND-UP FACE-OFF VERTICAL HORIZONTAL CHASE STANCE LEAP- MIDDLE POSITION POUNCE SIDE-STEP BELLY-UP STAND-UP FACE-OFF VERTICAL HORIZONTAL CHASE STANCE LEAP LAST POSITION POUNCE SIDE-STEP BELLY-UP STAND-UP FACE-OFF VERTICAL HORIZONTAL CHASE STANCE LEAP FIG the 1. Per cent occurrence of play patterns in (a) last position at six weeks of age. first position, (b) the middle position, or (c) the

4 430 MEREDITH WEST The eighth pattern is a chase (CH). Of all the patterns, this one moves the animal through the most area. A chase involves a kitten running after or from another kitten. It could, perhaps, be differentiated into pursuit and flight. The categories of play behavior just described typically appeared in sequences. In order to gain information about the structure of these sequences, shorthand records were made on 2500 sequences involving two kittens. The records were compiled from observations of 22 kittens, 13 males and 11 females from five litters, three reared in the lab, two at home. The ages ranged from 4y 2 weeks to 16 weeks. A sample of such a record is given below: Kitten One: P BU SU FO SU VS CH. Kitten Two: VS BU FO B BU HL. These records provided information on the number of patterns within a sequence, the frequency of particular patterns and positions of the patterns within sequences. With respect to the number of patterns within a sequence, there was an increase from a mean of 4.8 patterns per sequence at 6 weeks to 8.1 patterns/sequence at 12 weeks. Differences in ordering of patterns within sequences were also apparent (Fig. 1). At 6 weeks, 90% of the sequences were initiated with one of four patterns, belly-up (13%), stand-up (15%), side-step (20%) and pounce (42%). At 12 weeks, four patterns also accounted for 94% of play initiation, but there were differences with respect to which patterns were used. Side-steps were used only 3% and vertical stances, which had been used only 8%, were now used 24%. The proportions of pounces (37%), bellyup (16%), and stand-up (17%) were similar to those at 6 weeks. Although all patterns appeared during play sequences (as contrasted with the first or last position), several patterns were used more frequently. Belly-up and stand-up patterns accounted for 57% of the patterns used during a sequence at 6 weeks and belly-up, stand-up, and face-offs accounted for 63% at 12 weeks. Pounces, which occurred with high frequency as a first pattern, appeared only 7% during a sequence and were never used as the last pattern. Chases and horizontal leaps appeared as the last pattern for 70% at 12 weeks. Chases and face-offs were not used to initiate play. The side-step also declined to a very low percentage in any position at 12 weeks. Since social play typically involved two kittens, an attempt was made to see how the TABLE 1. Proportion of play patterns used in response to play initiation. Play pattern of initiator BU SU Belly-up Stand-up Pounce Side-step Vertical stance Chase Play pattern used in response P SS FO VS HL CH Other* The category "other" refers to any non-playful response C weeks Belly-up Stand-up Pounce Side-step Vertical stance Chase BU SU SS FO VS HL CH Other weeks

5 SOCIAL PLAY IN THE DOMESTIC CAT 431 performance of one kitten affected the other. It was only possible to look at this effect with respect to the initiation of social play (Table 1). The response of a kitten to play initiation appeared to be affected by the pattern used by the initiator. If a kitten was pounced upon, for example, it responded most often with a belly-up or vertical stance. Initiations involving either a belly-up or stand-up usually resulted in the performance of one of these same two patterns, e.g., stand-ups occurred in response to a belly-up 62% of the time at 6 weeks and 55% at 12 weeks. Other patterns appeared to be used infrequently regardless of the initiating pattern; for example, pounces never occurred as a response to play initiation, while chases, horizontal leaps and face-offs occurred only a small proportion of the time. (Table 1). During play sequences, there was rapid alternation of patterns. The most striking regularities were that belly-up and stand-up patterns were found to occur frequently in combination with one another. If a bellyup occurred, a stand-up followed it 67% of the time that any playful response occurred. A stand-up was followed by a belly-up 63% of the time. A face-off was seen in combination with a belly-up or stand-up 47% of the time. The different play patterns first appeared at different ages (Table 2). The first to be observed was the belly-up which appeared at the beginning of the fourth week, around day 21. AH play patterns, with the exception of the stand-up, were first observed in a solitary context, i.e., the movements were not directed toward another kitten. Upon emergence, the patterns were frequently repated over and over again. The TABLE 2. Ages at which the play patterns were first observed. Play patterns 1. Belly-up 2. Stand-up 3. Vertical stance 4. Horizontal leap 5. Pounce 6. Face-off 7. Side-step 8. Chasing Age first seen days days days days days days days days first sequences of patterns thus contained only repeated performances of the same pattern, such as twelve side-steps. Sequences involving more than one pattern and more than one kitten first occured at 5 weeks. The content of these sequences was made up of 50% belly-up and stand-up patterns and 30% pounces and 20% chases. They contained an average of four patterns per kitten. By 6 weeks, the sequences contained all the patterns, but face-offs and vertical leaps occurred infrequently (Fig. 1). Social play sequences often involved more than two kittens during the fifth through the eighth week. At 5 weeks, 28% of the social sequences involved three or more kittens. At 6 weeks, the percentage was 39%, at 7 weeks, 33%, at 8 weeks, 15%, and by 9 weeks, the percentage dropped to 3% of the sequences were of this type. The number of social sequences observed during a 45-min interval also increased with age. At 4l/2 weeks, each kitten participated in an average of 2.1 sequences, at 5 weeks, 3.5 sequences, at 6 weeks, 6.5 sequences, and at 7 weeks, 9.2 sequences. After this point, the average number of sequences was 9.4 sequences per kitten from 8 to 16 weeks. Beyond this age, systematic data were not collected. The over-all frequency of play also changed with age (Fig. 2).This figure shows the per cent of time spent in play during observation periods averaged for each week from 4 to 18 weeks. Beyond this age, the amount of play seen continued to decline but systematic observations were not made. Cats at this age no longer played daily and thus measures of per cent of time were more difficult to obtain. The data for Figure 2 were obtained during the observation period but the kittens also played during non-observation periods. For all the kittens in this study, there were three or four daily play periods. For the three litters studied in the laboratory, these play periods occurred at approximately 7:30 AM, 1 PM, 6 PM and 10:30 PM. The length of these periods increased with age from an average of 18 min at 5 weeks to 46 min at 9 weeks. There were several events which seemed to coincide with the decline in social play

6 432 MEREDITH WEST AGE IN WEEKS FIG. 2. Per cent o time during observation period spent in social play from 4 weeks to 19 weeks. activity. In both the laboratory and in the home, kittens began to display changes in daily activity around 4 months of age. The amount of time spent in sleep during the observation period increased from an average of 8 min for 8 to 16-week-old kittens to an average of 24 min at 16 weeks and 31 min at 20 weeks. The amount of time spent in a "quiet, alert" state, i.e., sitting and visually attending to events in the environment but not moving, increased from an average of 14 min for 8-to-16-week-old kittens to an average of 32 min at 20 weeks. Social grooming was a frequent activity in kittens from 5 weeks of age onward and it also began to decline at 4 months. Kittens, in general, spent more time away from their littermates, and the activity of the kittens was less synchronised. If permitted to do so (as in the three home-reared litters), the kittens spent more time outdoors. At 12 weeks, such kittens were outdoors 2 hr a day, by 16 weeks they spent 4i/ hr outdoors in nice weather. During these outdoor periods, play occurred but in smaller proportion to the time devoted to exploration and hunting. These two activities accounted for 68% of the activity of 12 kittens observed daily outdoors from 15 to 18 weeks of age. In case of the feral cats observed, it was noted that each of the three mothers left her kittens at around 4 months of age. They appeared to move to another part of their range and were not observed near their kittens after this time. Thus, by 4 months of age, the feral kittens had to find all of their own food. Prior to this time, it was observed that the kittens remained close to the shelter or area where their mother had reared them. Several times they were observed accompanying their mothers when the mothers went to find food, but most of the time, the mothers did not permit their young to travel with them. At 4i/ months of age, sexual activities of males began to appear. Males began to display mounting, neck grasping and sniffing of the female's genital area. The female showed no corresponding interest and the increasingly frequent attempts by the males to engage in such activity appeared to make the female less tolerant of any close contact, including play. Such avoidance continued until the females began to reach sexual maturity, around 5y 2 months. Males did not direct any sexual behavior toward other males. Although this paper deals only with social play, it is worth pointing out that play with objects in the environment and other solitary play activities, such as tail chasing, also declined at 4 months of age, but the decline was more gradual. When the adults observed in this study played, they tended to play with objects by themselves. Ten of the adults studied developed

7 SOCIAL PLAY IN THE DOMESTIC CAT 433 idiosyncatic forms of play. These consisted of play with a particular type of object, such as a piece of string or a ball, play in a special location, such as under a "favorite" chair or repeated performances of a particular play pattern such as a vertical stance. DISCUSSION The results of this study indicate that the social play of domestic cats can be divided into eight categories of motor patterns. The structure of social play comes from the organization of these patterns into sequences. Analysis of the sequences revealed that differences existed with respect to the ordering of the patterns suggesting that the patterns may serve different roles. Pounces, for example, occurred almost exclusively as opening patterns and thus appear to be used to solicit play, i.e., to decrease social distance and initiate physical contact. Chases and horizontal leaps to be used to terminate play. The high frequency of belly-up and stand-up patterns anywhere in a sequence and the fact that they were used in combination with one anotther suggests that the goal of social play for kittens is to engage in a particular form of physical contact. The developmental changes in play also indicate this. Play progressed from a stage of solitary performance of individual patterns to solitary and then social sequences containing several patterns. The decline in group play and increase in pair play might be viewed as further progress in this direction in that pair play allows more direct interaction between individuals. In play involving three or more individuals the nature of the play patterns allows for only peripheral interaction by the "third" member. The play patterns are quite constant in form and could be described as performed at a "typical intensity" as described by Morris (1957). The appearance of typical intensity is thought to "enhance the signal value of movements by making them more distinctive" (Hinde, 1970, p. 671). The need for such signalling during play may be due to the fact that such close and vigorous physical contact could result in injury if the signals for playful intent were not clear. One might also view the use of tail movements, which accompany the performance of some patterns, as a more potentially variable indicator of playfulness, as the tail can be moved at different rates. Social play in the cat can thus be described as sieieulyped form of physical interaction between individuals as has been noted by other observers and has led to the use of the "play-fighting" to characterize cat's social play (Ewer, 1968). The results of this study suggest that the use of such semi-functional terms should be avoided until more information is gathered on the goals of play because it implies a greater similarity between the two activities than appears to exist. For example, although several motor patterns are common to the two activities (belly-up, stand-up and chase), these patterns occur in different combinations with other patterns which are not seen in both situations. During play, bellyup patterns are often preceded by pounces. During fights, this pattern is never preceded by pounces but by the "defensive archedback, posture" which never occurs during play (Leyhausen, 1956). The decline in the use of the side-step may possibly be related to its partial resemblance to the defensive arched back posture in that its use could cause confusion in interpreting playful or aggressive intent. In a fight, the two animals may also display differing motivational states, such as fear or threat, whereas during play, both animals appear to be in the same motivational state. The outcome of a fight may involve injury to the opponents, flight of one of the participants or the gaining of a disputed object, such as food. None of these factors appear to be relevant to play suggesting again that the term "playfighting" may be misleading with respect to the function of play. The search for the functions of animal play has a long history, but there is little experimental support for any of the functional interpretations (Loizos, 1966). The results of this study are not a sufficient basis upon which to identify all of the possible functions of play in cats, but they do sug-

8 434 MEREDITH WEST gest several ways in which social play may function during the period from birth to 4 months. This period can be characterized as one of social living, dependency upon a parent for food and protection and confinement of activities to a relatively small area. The movements of the young are first restricted by the kitten's inability to travel and later by the necessity that they be available to the female for nourishment and protection from environmental dangers. It is while the kitten is living under such conditions that social play occurs as frequent and regular activity. The fact that the kittens are living in a social environment means that they must develop modes of social interaction. Play represents one of several modes available to the young. Others include social investigation and grooming. While it is not possible to isolate the effects of only play with respect to how it may function to facilitate or maintain social relations among littermates, the high frequency of play suggests that it does help to keep the littermates on friendly terms and thus helps to keep the litter unit together which is important for their survival at this time. Grooming, which also appears to be a sign of friendly relations first occurs at the same time that social play does and although it does decline after 4 months, it remains a more frequent activity among familiar adults than does social play. Young kittens require physical exercise for the proper development of their growing bodies. In animals dependent upon a parent for food and protection, this need for exercise must be adapted to fit the particular environmental circumstances, such as a restricted living area. The physical nature of the social play patterns provides exercise for many parts of the kitten's body. Although social play moves the animal through space, the patterns are also adapted to promoting physical contact, thus keeping the kittens together, which is valuable for the kitten in that their mother can find them more easily. The decline of social play at 4 months appears to coincide with changes in the environment and in the kitten which may affect its needs for exercise and social inter action. At 4 months of age, the litter unit becomes independent from the mother and members of the litter begin to travel from the living area. Kittens, at this age, may also have to obtain their own food. These factors suggest then that play may decline because opportunities for social contact may decline, i.e., the kittens may no longer live together. The fact, however, that social play in cats living socially in the laboratory declines at the same time suggests that other mechanisms may be important. Koepke and Pribram (1971) have reported a similar decline at this age in cats reared in the laboratory from birth to 6 months. One of the mechanisms responsible for this decline in play may be the devlopment of sexual behavior which appears to alter formerly friendly relations among males and females. Thus play may decline not only due the tendency of kittens to disperse at this age thus decreasing opportunity, but also due to changing preferences of individuals for social contact or particular forms of social contact. It may also be that alternate sources of exercise affect the decline in play, for example, increased exertion related to finding food. The fact that kittens have reached much of their adult size may also affect their needs for exercise. Although all of these factors help to suggest why play may be useful for the young kitten, they do not explain the significance of the play patterns themselves. Why have such patterns evolved? The patterns used seem to be "suited" to a carnivore in that they emphasize behaviors of pursuit and capture of objects or peers. Although there is little comparative data on the patterns used by other species of cats, there is one instance where the patterns appear to reflect differences in hunting adaptations. In the cheetah, young cubs have incorporated into their play a "new slap" which is used to knock over peers. This pattern is also used by the adult to knock over prey (Ewer, 1973). This pattern has not been observed in either prey-catching or play in other felids. Play in canids also seems to reflect food-getting adaptions. The presence of various oral manipulatory patterns, such

9 SOCIAL PLAY IN THE DOMESTIC CAT 435 as the "scruff bite" or "face bite" (Bekoff, 1972) may be related to some canids' tendencies to bring down prey with their mouths (Ewer, 1973). Bekoff (1972) also reports differences in the frequency of social play among canids which appear to be related to whether they are social or non-social species, a factor which may also reflect the influence of food-getting specialisations on social play. The results of this study indicate two ways in which social play may function in kittens. First, social play may provide a means of developing and maintaining friendly social relations among members of a litter. Second, social play appears to provide ample opportunities for physical exercise. The formation and maintenance of friendly social relations may not only help to keep the kittens together and reduce potential social tension but may also provide opportunities for the acquisition of social communication skills which will be used in other social situations. For example, the ability to discriminate between playful and non-playful approaches would allow the animal to interpret the behavior of individuals in future social encounters. The opportunity for exercise is equally important for the growing kittens. The fact that the type of exercise gained from the performance of the play patterns stresses movement patterns used in future food-getting or aggressive situations may make play of even more value. Social play is one of several types of play activity found in the domestic cat. Whether all of these types can be discussed in terms of similar function remains to be seen. The fact that preferences for different types of play with age, such as a decrease in preference for social play and increase in solitary play with objects, suggests that the functions of play may differ. One element that does appear to be common to all types of play in the cat is that cats seem to enjoy playing. Although the essentially private nature of this element may prevent it from being studied objectively, its potential importance must not be ignored. The fact that play is a pleasurable activity may explain not only the high frequency of play but also changes in preference for types of play activity. Social play in kittens may be accompanied by a form of pleasurable or positive feedback which would result in its frequent repetition. When, however, the kitten's interest or "desire" for social contact decreases, social play may provide less positive feedback. Thus, the functions of social play may not only become inappropriate, but also the performance of social play may become less rewarding. The need for engaging in a pleasurable activity may still exist however, and thus, other types of play activity would continue to be performed by the adult. The results of this study suggest that knowledge concerning the developmental period during which play occurs is important in order to explain the content and function of social play in the domestic cat. As more information is gained on developmental differences among species of cats and other carnivores, it may be possible to relate differences in their social play to developmental factors, such as variation in periods of dependency upon the parent or the social bond among littermates. REFERENCES Bekoff, M The development of social interaction, play and metacommunication in mammals: an ethological perspective. Quart. Rev. Biol. 47: Berlyne, D. E Laughter, humor and play, p In G. Lindzey and E. Aronson [ed.], The Handbook of Social Psychology. Vol. 3. Addison- Wesley, Mass. Chepko, B. D A preliminary study of the effects of play deprivation on young goats. Z. Tierpsychol. 28: Ewer, R Ethology of mammals. Plenum Press, New York. Ewer, R The carnivores. Cornell Univ. Press, Ithaca, N.Y. Farentinos, R. C Some observations on the play behavior of the Steller sea lion (Eumetopias jabata). Z. Tierpsychol. 28: Hinde, R Animal behavior: a synthesis of ethology and comparative psychology. 2nd ed. McGraw-Hill, New York. Koepke, J. E., and Pribram, K. H Effect of milk on the maintenance of sucking behavior in kittens from birth to six months. J. Comp. Physiol. Psychol. 30: Loizos, C Play in mammals. Symp. Zool. Soc. Lond. 18:1-9.

10 436 MEREDITH WEST Loizos, C Play behavior in the higher primates: a review, p In D. Morris [ed.], Primate ethology. Aldine, Chicago. Leyhausen, P Verhaltcnsstudien an Katzcn. Z. Tierpsychol. Beiheft 2, Morris, D "Typical intensity" and its relation to the problem o ritualization. Behaviour 11: Muller-Schwarze, D Ludic behavior in young mammals, p In M. B. Sterman, D. J. McGinty, and A. M. Adinolfi [ed.], Brain development and behavior. Academic Press, New York. Stcincr, A. L Play activity o Columbian ground squincls. Z. Tierpsychol. 29: Welker, W Ontogeny of play and exploratory behaviors: a definition of problems and search for new conceptual solutions, p In H. Moltz [ed.], The ontogeny of vertebrate behavior. Academic Press, New York.