¡Descarga Object Permanence in Animals: A Comparative Study y más Guías, Proyectos, Investigaciones en PDF de Programación Orientada a Objetos solo en Docsity!
Animal Learning & Behavior
Object permanence in four species of psittacine
birds: An African Grey parrot (Psittacus
erithacus), an Illiger mini macaw (Ara maracana),
a parakeet (Melopsittacus undulatus),
and a cockatiel tNymphicus hollandicus)
IRENE M. PEPPERBERG and MILDRED S. FUNK Northwestern University, Evanston, Illinois
Four psittacids-an African Grey parrot, an Illiger macaw, a cockatiel, and a parakeet-were tested on object-permanence tasks that are commonly used to assess levels of understanding in human infants during their first 2 years. These birds showed Stage 6 competence, demonstrat- ing that object permanence is not limited to mammals. The results for these birds were compar- able to those of an African Grey parrot that ·had been trained in interspecies communication prior to object-permanence testing. Our findings thus suggest that although language-like behavior provides a communication channel that facilitates testing, such language training is unlikely to affect the outcome of the tests.
Object pennanence-the notion that objects are separate entities that continue to exist when out of sight of the observer-would seem to be one of the more important cognitive concepts, and thus an appropriate topic for com- parative study. Object permanence was nevertheless largely ignored during almost a century of comparative research (note Burghardt, 1984; Macphail, 1987), pos- sibly because it was considered both innate and unitary-a concept so basic that researchers could not imagine how any organism could exist without it (see Flavell, 1985). Piaget was the first to suggest that this concept was not inborn, but was acquired and could be acquired to vary- ing degrees. Only after Piaget's writings became widely available (Piaget, 1971, 1978, 1980) did researchers rou- tinely test animals on the concept of object permanence (e.g., Chevalier-Skolnikoff, 1976, 1981, 1989). These studies, which involved nonhuman primates and a few nonprimate mammals, have demonstrated that the capac- ity of nonhuman mammals for solving some object- permanence problems is comparable to that of humans. Additional data suggest that nonmammals may also comprehend object permanence. Observation of the be- haviors of several avian species in seminatural surround- ings and in the wild (e.g., nutcrackers, Nucifraga colum-
We thank the Ravid family, Jean Benzies, Linda Rusin, and Ann Shanks for their patience and for allowing their birds to be used in our
study, Lois Sears for assistingwith the testing, and Vincent M. LoLordo
and two reviewers for their comments on an earlier version of the manuscript. I. M.P. thanks the NationalScience Foundation for support (BNS 84-14483). The authors' address is Departmentof Anthropology, Northwestern University, 1810 Hinman Avenue, Evanston, IL 60208-1310.
biana, Kamil & Balda, 1985; Vander Wall, 1982; European jays, Garrulus glandarius, Bossema, 1979; jackdaws, Corvus monedula, Etienne, 1976-1977; Lorenz, 1970; marsh tits, Parus palustris, Sherry, 1982; Shettleworth & Krebs, 1982; chickadees, Parus atricapillus, Sherry, 1984; Shettleworth & Krebs, 1986) provide evidence for behaviors such as recovery of cached food or cavity nesting that would appear to require a con- cept of object permanence. I And although laboratory studies initially provided evidence for only a limited con- cept of object permanence in birds such as domestic chickens (Etienne, 1973a, 1973b), a recent study (Pep- perberg & Kozak, 1986) has demonstrated the full ca- pacity in an African Grey parrot (Psittacus erithacus). The results of the Pepperberg and Kozak (1986) study, however, might not be representative of psittacine abili- ties. African Greys, for example, have scored better than other psittacids on tests of certain cognitive capacities (Ko- lar, 1972; Rensch, 1956). Moreover, the subject of the Pepperberg and Kozak study, Alex, had received con- siderable prior training on tasks involving animal-human communication (Pepperberg, 1979, 1981, 1983, 1987a, 1987b), and some investigators (Hall, Braggio, Buchanan, & Nadler, 1982; Hall et al., 1980; Premack, 1983, 1984) have suggested that "language-trained" subjects have spe- cial skills and abilities that enhance their capacities for problem-solving. The present study, therefore, had two objectives: (1) to assess the ability of a variety of nonlanguage-trained psit- tacids to solve standard object-permanence tasks, and (2) to compare the performance of these nonlanguage- trained psittacids with that of Alex. Comparisons among the nonlanguage-trainedpsittacids would enable usto learn
97 Copyright 1990 Psychonomic Society, Inc.
98 PEPPERBERG AND FUNK
whether object permanence, like certain more advanced cognitive capacities (e.g., numerical competency; Koeh- ler, 1943, 1950), might vary across species. Compari- sons between Alex and the nonlanguage-trainedbirds, par- ticularly an untrained African Grey, would provide information on how such training might influence a study of object permanence.
Background According to Piaget (1953, 1954), development of the object concept is the earliest step in intellectual develop- ment and is a prerequisite for successful interaction with the external world. A subject that comprehends the per- manence, separateness, and continuity of external objects can also comprehend its own identity as a separate entity among all others in its environment, and can thus situate itself in time and space. Piaget described the development of this concept in terms of a sensorimotor period of six stages, each stage showing progressive manipulatory skills and increasing awareness of self and surroundings. Although the Piagetian framework was designed for use with humans, its methodology is consistent with that of animal studies and thus facilitates cross-species compari- sons of observational and experimental data. The specific behavioral features that the Piagetian framework uses to categorize narrative data (observations and note-taking about the subject's activity in its natural habitat or in the laboratory; see Chevalier-Skolnikoff (^) & Poirier, 1977) are similar to those of the early ethologists (e.g., Lorenz, 1970). Moreover, experiments based on the Piagetian framework directly measure levels of performance on specific tasks that are often easily adaptable for use with many species (Chevalier-Skolnikoff, 1989). A series of tasks designed by Uzgiris and Hunt (1975; see Appen- dix) to compare capacities for object permanence in de- veloping human infants has, for example, also been used in animal studies (e.g., chimpanzees, Pan troglodytes: Hallock & Worobey, 1984; Wood, Moriarty, Gardner, & Gardner, 1980; lowland gorillas, Gorilla gorilla: Red- shaw, 1978; for a review of other animal studies, see Dore & Dumas, 1987). The Uzgiris and Hunt (1975) scale not only can be used to measure relative rates of development across subjects, but can also be administered in adulthood to assess an ul- timate level of acquired competence (Triana & Pasnak, 1981). The scale thus enabled us to compare adult sub- jects of varying ages. The level of competence is deter-
mined by performance on an increasingly difficult series of 15 tasks. Success on a later task presupposes ability on earlier ones. Successful completion of specific sets of tasks can be correlated with Piaget's six stages of under- standing of object permanence (e.g., Wadsworth, 1984; see below). The Uzgiris and Hunt scale of tasks thus not only facilitates cross-species comparisons of the concept of object permanence, but, by dividing the Piagetian stages more fmely, also provides additional within-stage data on
behavioral similarities and differences across species.
Recently, however, researchers have argued whether the Uzgiris and Hunt (1975) scale provides the best basis for comparison. The arguments center on two issues: (1) whether such standardized tests are sensitive enough to pinpoint the age at which many of these abilities emerge (e.g., Bower, 1982; Flavell, 1985; Harris, 1983), and (2) whether the tasks actually quantify the existent cog- nitive level of the subject or, instead, measure the level of proficiency reached with respect to training, learning, and memory (Bjork & Cummings, 1984; Dore & Dumas, 1987). The first issue is not relevant if, as in our study, the scale is used to determine the relative ultimate level of achievement rather than relative rates of development. The criticism contained in the second issue is more difficult to counter (see General Discussion); however, the invariance of the procedures ensures that the effects of learning, training, and memory would influence the results for all subject species equally. Furthermore, stan- dardized tasks such as those of Uzgiris and Hunt prevent the variations in experimental design between laborato- ries that often render ambiguous the results of other com- parative studies (see Wood et al., 1980; note also Bate- son, 1979; Kroodsma, 1982; Kroodsma, Baker, Baptista, & Petrinovich, 1984).
MEmOD
Subjects
We tested 1 individual from each of the four subject species. In object-permanence studies, single-subject designs are known as "power studies" (Triana & Pasnak, 1981). If the subject of a power study reliably demonstrates object permanence, the implication is that this aptitude is within the capacity of that species. Negative results from one subject in such a study, however, do not neces- sarily imply the lack of such an aptitude for the species as a whole
(see discussion in Triana & Pasnak, 1981).
The 4 subjects were all pets. They were: Wok, a 13-year-old Afri-
can Grey parrot (Psittacus erithacus); Fred, an adult llliger mini
macaw (Ara maracana); Bruce, an 18-month-old parakeet (Melop-
sittacus undulatusi; and Yellow Bird, an 8-month-old cockatiel
(Nymphicus hollandicus).
Apparatus: Search Objects and Covers
Objects for which the birds searched included food and nonfood items. Food items included seeds, grapes, crackers, peanuts, and
ice cream or water in a small cup for Fred; rolled oats, cream cheese,
banana, and egg yolk for Bruce; jelly beans, cookie pieces, choco- late matzoh, and cherries for Wok. Fred and Bruce were also tested with mirrors, yarn, and a hair clip. Wok was tested only with food because scheduling difficulties prevented replication with nonfood items. Yellow Bird was tested only with nonfood items-rings, clips, and a key chain. We chose these classes of objects on the basis of the extent to which they would motivate the subjects (see Discus- sion). For all subjects, these items were hidden under crumpled paper toweling or newspaper, paper plates, bags, or measuring cups.
Tasks
The Uzgiris and Hunt (1975) Scale 1 tasks used in this experi- ment are given in the Appendix. Tasks 1 and 2 test if a subject can visually follow a slowly moving and then disappearing object, and correspond, respectively, to Piaget's Stages 2 and 3. Tasks 3 and 4 test Stage 4, the ability to retrieve an object that is, respectively,
100 PEPPERBERG AND FUNK
Scoring Criterion
The performance criterion was similar to that used in previous studies: no more than one incorrect trial in two of three tests of the Uzgiris and Hunt (1975) tasks (see Pepperberg & Kozak, 1986; Wood et aI., 1980). A trial was done correctly if the subject un- covered the hidden item or performed one of the other acceptable actions within 30 sec of the hiding of the item. For each of the tasks, Uzgiris and Hunt list three to five alternative erroneous actions that subjects could perform, from pulling off the wrong cover to losing interest in the object or doing a haphazard search. Correct searches must be done consecutively three, four, or five times as noted for each task. Mistrials were recorded when our subjects (1) failed to go to any of the sites within 30 sec, (2) removed and immediately manipulated the covers without searching for the hidden object, or (3) went so fast that they interfered with the placement of the ob- ject. Errors were recorded when the subjects lifted and searched under the wrong cover.
RESULTS AND DISCUSSION
Except for the parakeet, Bruce, who died before test- ing could be completed, all subjects met the criteria for full Stage 6 object permanence according to the Uzgiris and Hunt (1975) scale. The results are summarized in Ta- bles 1 and 2. Because there were often individual differ- ences in how the subjects solved the tasks (note Fletcher, 1965), detailed descriptions of such behaviors are given below.
Table 1 Results of Single-Trial Sessions Per Task for Wok, Bruce, and Fred
Bird
Task Wok Bruce Fred
Visual Pursuit of Slowly Moving Objects I 3 3 3 2 3 3 3 Search for Simply Hidden Objects 3 3 3* 3 4 3 4 3 5 3 2 2
6 3 3 (I)^3 (I)
7 5 (1)^5 5
Search Following More Complex Findings 8 3 (1)*^3
Search Following an Invisible Displacement 10 3 3 3 11 2 3 (I)^ 2* 12 3 (1) 3 3
13 6 (1)*^ N/A^ 5 (1)
Search Following Successive Invisible Displacements 14 4* N/A 4 15 2* N/A 2 Note-Numbers represent the number of times the birds performed the action designated by Uzgiris and Hunt (1975) as critical for compre- hension. The number of correct repetitions designated as necessary for comprehension is given in the Appendix. Numbers in parentheses represent the number of errors; see text for details. Asterisks represent mistrials. In most mistrials, the subjects manipulated the covers and ig- nored the hidden objects.
Table 2 Results of Multiple-Trial Sessions Per Task for Yellow Bird Number of Trials Task Session 1 Session 2 Session 3 Visual Pursuit of Slowly Moving Objects 133 3 233 3 Search for Simply Hidden Objects 3 3 3 3 4 3 (2) 3 3 5 2 2 2
Search Following More Complex Findings
Search Following an Invisible Displacement 10 3 3 3 11 2 3(1) 2(1) 12 3 (3) 3 (2) 3 (5) 13 5* 5 (2) 5 Search Following Successive Invisible Displacements W 4 4 4 15 2 2 2 Note-Numbers represent the number of times the bird performed the action designated by Uzgiris and Hunt (1975) as critical for compre- hension. The number of correct repetitions designated as necessary for comprehension is given in the Appendix. Numbers in parentheses represent the number of errors; see text for details. Asterisks represent mistrials. In most mistrials, the subject manipulated the cover and ig- nored the hidden objects.
Reaction to the Tasks and Apparatus Wok, the African Grey parrot. Wok initially tried to avoid the investigators and the search objects. He was par- ticularly disturbed by novelty. When a novel cover (a paper towel) was introduced (Task 4), he did not remove it, but instead moved as far away as was possible, began to feather-pluck, and emitted distress calls. When we replaced the towel with a familiar cover (a paper bag), he responded appropriately. We therefore introduced other novel covers (e.g., plastic measuring cups) in ses- sions separate from the object-permanence trials. Tasks 1-4. Wok responded appropriately to all trials with familiar covers. Tasks 5-8. Wok responded appropriately on most trials of these tasks. He had several mistrials during Tasks 7 and 8: he was more interested in the plastic measuring- cup covers than in the hidden objects. He would ignore the actions of the investigators, then grab a cup and play with it for several minutes. He made one search error in- itially in each of Tasks 7 and 8 (searching the cup closest to him), but then correctly completed the requisite num- ber of trials. In another session, he responded appropri- ately to all trials of Task 8. Task 9. The first trial was interrupted by construction noises in an adjacent room, but Wok removed two of the
OBJECT PERMANENCE IN PSITTACINE BIRDS 101
three superimposed cups before becoming distracted. Ad-
ditional trials proceeded without disturbance, and Wok
responded correctly. Tasks 10-13. Wok erred once each on Tasks 12 and
- On Trial 2 of Task 12, he chose the cup that had been correct on the initial trial. He also had a mistrial on Task 13, when he ignored the task and played with the covers. Task 14. On the first trial, for which we placed all the items on a paper-lined plastic tray, Wok pulled the paper out from under the covers, thus causing a mistrial. His next two responses were correct: once he searched under all the screens in order of hiding and once he searched directly under the correct screen. Task 15. After one mistrial, in which he played with the cups used as covers, Wok-appropriately-searched under the cover where the object logically should have been hidden. Bruce, the parakeet. Bruce seemed unperturbed by testing. Even our initial use of a cover that was too large for him to remove-a quarter of a cocktail napkin-did not prevent him from obtaining the hidden object; he pulled it from under the cover. He removed the covers when we subsequently used eighths of napkins or propped pieces of paper plates between him and the objects. Bruce differed from Wok not only in his tractability. We did not often use food objects, because Bruce was most willing to do a task if it involved uncovering a hand mirror. Too, possibly because parakeets are ground feeders (Forshaw, 1969, 1973), he was more likely to work on the floor than on his cage or a perch. Bruce died before he could complete the series. About a month before his death, he was uncooperative and ap- peared ill: during the last scheduled session, he ignored the hidden objects and the researchers. These behaviors could have been caused by the increasing difficulty of the task or the onset of ill health. Tasks 1-4. Bruce watched and edged away from yam as it moved around him, and watched as it was dropped past his perch. In Task 3, he withdrew partially covered pieces of banana from under a cocktail napkin (see above). He occasionally shredded the napkins during and after the trials. For Task 4, we cut the napkin cover into eighths, and he responded appropriately to all trials. Tasks 5-12. Bruce made only two errors on all trials of these tasks. One error occurred on the first trial of Task 6, when he searched randomly under both covers. The other error occurred on Trial 2 of Task 11, when he searched under the screen where the item had previously been found. Tasks 13-15. Task 13 was the last task for which we obtained any data. During the last session, Bruce correctly completed three invisible displacements.
YeUow Bird, the cockatiel. Yellow Bird worked more rapidly than the other birds, but also spent relatively more
time playing with the covers or in displacement behaviors (e.g., feeding) at the outset of the sessions. In contrast to the parakeet, he could easily lift the covers. Because he would manipulate small metallic objects, he was tested entirely with nonfood items: rings, hair clips, and a key chain. Series 1: Tasks 1-3. Yellow Bird responded appropriately to all trials of these tasks. Task 4. Yellow Bird clearly erred once, by reacting to the disappearance of a ring without attempting its recov- ery. In another trial, he pulled at the cover but not strongly enough to get the ring; however, he then noticed, ap- proached, and pecked at the ring that the experimenter was wearing. Tasks 5-8. Yellow Bird made one error on the first trial of Task 6; as in Task 4, he reacted to loss but made no attempt at recovery. Task 9. On the first trial, Yellow Bird lifted only two of the three covers. He did one trial correctly, engaged in displacement behaviors for several minutes, then com- pleted the requisite number of trials. Tasks 10-11. Yellow Bird was initially more interested in the covers than in the hidden objects, but he did com- plete the appropriate number of trials for these tasks. Task 12. This task was presented in the same session as Tasks 10 and 11. He chose the wrong screen three times before responding appropriately three times; he therefore did not meet criterion, although he did eventually com- plete the requisite number of trials. He appeared fatigued, and we ended the session. Tasks 13-15. We advanced to the next tasks in the next session. His initial responses were not appropriate, although they were of considerable interest: he would carry the objects and covers to the edge of the table, drop them, watch the descent, and follow the experimenters' retrievals.' Only after he lost interest in these procedures did he respond appropriately to the trials for these tasks. Series 2: Tasks 1-10. Yellow Bird reached criterion on all tasks during one session. Tasks 11-13. Yellow Bird was uncooperative during most of the sessions devoted to these tasks. He again picked up, carried, and dropped the covers and the ob- jects, and watched the experimenters' retrievals. He bobbed his head, banged his beak, and fluffed his feathers, although, unlike Wok, he did not try to avoid the ex- perimenter or the apparatus. He would push a recently retrieved object under a cover or a nearby calendar. He made more errors on these tasks than he had on the first series. One session ended when he tried to bite the ex- perimenter. Tasks 14-15. In contrast to his behavior during the previous tasks, Yellow Bird-on a subsequent day- responded immediately and appropriately on all trials.
OBJECT PERMANENCE IN PSITTACINE BIRDS 103
like tasks facilitates conceptual thinking and would likely
affect the results of object-permanence studies. Our non- language-trained subjects, however, exhibited the same Stage 6 capacities as did Alex, the language-trained Afri- can Grey studied by Pepperberg and Kozak (1986). It is therefore possible that object-permanence tasks do not re- quire the level of conceptual ability that is affected by language-like training. Language training may, for ex- ample, help a subject form a precise representation of the hidden object (see Vygotsky, 1962), but a subject in the Uzgiris and Hunt (1975) tasks need not have a represen- tation of the specific hidden object-only that something has disappeared. Non-language-trained birds also may successfully complete even the more advanced sorting tasks of the Piagetian preoperational period (see Chardard- Segurel, 1984; Hall et al., 1980; Hall et al., 1982; cf. Braggio, Hall, Buchanan, & Nadler, 1982). The point at which language-like training affects object concept is not, therefore, easily established (Premack, 1983). Whatever the effect of language training on mental ca- pacity, the ability to communicate with a subject in a language-based code can affect the ease and speed of object-permanence testing. Vocal signals such as "Pay attention!" or "Go find the cork" would refocus the at- tention of the trained African Grey, Alex, when he be- gan to ignore the apparatus (e.g., by beginning to preen). He could thus be given 1-2 min to complete a search, and disinterested behaviors were not immediate grounds for a mistrial. In contrast, we could use vocal signals with the subjects of the present study only to direct them to begin a search, not to bring them back to the task: they either searched for the item within 30 sec or did not respond at all.
Testing a Constellation of Abilities and Individual Traits Tests of object permanence actually test a constellation of attributes and abilities: motor skills, motivation, atten- tion, memory, and socialization (for animals, tractabil- ity). Some of these abilities (motor skills) are likely to be independent of a concept of object permanence, whereas others (memory) are closely correlated with such a concept. Many researchers, therefore, have argued that the Piagetian framework in general-and the Uzgiris and Hunt (1975) tasks specifically-do not facilitate fine- grained comparative studies. The Uzgiris and Hunt (1975) tasks, for example, often do not distinguish between comprehension of object per- manence and the physical ability needed to perform the task (note Dore & Dumas, 1987; Flavell, 1985). The parakeet in the present study, for example, could not re- move the cocktail-napkin cover initially used in Task 3; he instead removed the object from under the napkin. His behavior indicated a sense of object permanence, but would have been ineffectual with Task 9. The Uzgiris and Hunt tasks, therefore, must be constructed and ad- ministered so that comprehension and physical compe- tence are not confounded.
A subject's motivation to search for the hidden objects can affect the results of object-permanence testing, although motivation is not a reflection of cognitive abili- ties (note Thinus-Blanc & Scardigli, 1981). Birds in par- ticular appear less motivated than children and nonhuman primates to search for a hidden toy, although several avian species do manipulate objects in apparently playful be- haviors (see Pepperberg, 1989), and our subjects would engage in a search if the objects were varied from trial to trial. Food items were not particularly potent rein- forcers, probably because our subjects were well-fed pets. Occasionally, our subjects would preferentially manipu- late the objects used as covers, or would ignore the ex- perimenters and all experimental materials. Such trials (counted as mistrials) clearly demonstrated a lack of moti- vation and not a lack of understanding of the task: the animals did not choose the wrong cover; rather, they chose not to participate. The extent to which a subject attends to the tasks also affects the results. Attention can be divided into two parts: (1) the attention span, or the length of time that a subject will devote to a task; and (2) how well the subject no- tices and processes the information being presented. At- tention span need not be related to cognitive capacity, but the quality of attention may be indicative of such capac- ity. Both forms can vary not just between but within sub- jects. Attention span varied considerably within the sub- jects in our study. Yellow Bird, for example, in a single l-h session, performed 10 tasks, but he completed no tasks in a subsequent session because he would not attend to the task. Similar variations occurred for the other birds. The quality of the attention that our subjects devoted to the tasks was less easy to document. The Uzgiris and Hunt (1975) tasks, for example, do not demonstrate whether the subjects are searching just for something of interest or for the particular object that had been hidden. A search for the latter would suggest that they had attended to and coded the specific attributes of the object as well as its spatial representation, and would demonstrate a complete understanding of object permanence: not just that an ob- ject maintains its existence when out of sight, but also that the found item is identical to the one hidden (note Flavell, 1985). This latter level of comprehension can be demon- strated by substituting different objects during invisible displacements and noting whether the subjects are sur- prised at finding something other than what had been hid- den (see leCompte & Gratch, 1972). Tinklepaugh (1928,
- showed that monkeys and chimpanzees continued to search for the expected hidden object if it had been replaced with a less favored item. Repetition of the present study with this "substitution" protocol would be neces- sary to provide information as to the expectancies and quality of attention of our subjects. Memory may affect the ability to respond correctly on object-permanence tasks. Bjork and Cummings (1984), for example, suggested that the Stage 4 "AB search er- ror" (Task 6) often observed in children is a function of memory. The child continues to search where the object
104 PEPPERBERG AND FUNK
has been hidden and found on previous occasions, even though placement under a different cover is clearly demonstrated. The subject thus seems not to remember information from the second hiding (see review in Flavell, 1985). Piaget does not consider this type of error to be related to memory. In his view, an AB error indicates that the object is seen as an integral part of the hiding place, rather than as something with an independent existence; that is, the error is based on an incomplete comprehen- sion of object permanence. But the error may cease to occur if an infant is allowed to search immediately after the object is hidden in the second location (i.e., without any delay), if the discriminability of the covers is in- creased, or if removal of the covers is simplified. Moreover, when children have more than two choices of cover, their errors are not all to the previous location, but instead are clustered around the correct choice (Bjork & Cummings, 1984). We used only two choices of cover and allowed immediate searches (as specified by Uzgiris and Hunt, 1975); Bruce, Yellow Bird, and Fred picked the wrong cover on initial trials of Task 6, but not in sub- sequent trials. Bruce and Wok did make AB errors, but on Tasks 11 and 12, respectively, and not on Task 6. Given our experimental design, our birds were unlikely to make their errors because of problems in remember- ing the location of the second hiding.! In our study, socialization (or tractability) often deter- mined whether a bird would be an appropriate subject and the ease with which it could be tested. Mistrials occurred when our subjects were frightened by external noises or novel covers (Wok), or when they chose to ignore the entire testing situation. Their patterns of behavior could be correlated to some degree with the extent to which they accepted strangers and novel situations-that is, their degree of socialization. It is unclear, however, whether there exists a correlation between the extent to which a subject can be socialized and the level of its cognitive achievement. Even if such a correlation exists, our con- clusions would be unaffected by our choice of subjects on the basis of their sociability: our study was designed to test if these conceptual levels are within the capacity of a species (Triana & Pasnak, 1981). In sum, it is not always possible to disassociate the in-
dividual abilities that are tested during a study of object permanence, particularly if the study uses the tasks
designed by Uzgiris and Hunt (1975). Despite the limita-
tions of that scale for individual comparisons of each of the relevant abilities (e.g., see Bower, 1982; Flavell,
1985), the tasks nevertheless provide a standardized means for assessing cross-species similarities and differences for
the constellation of capacities. As such, the tasks provide an important starting point for more detailed cross-species comparisons of cognitive capacities.
Effects of Training and Learning on Success in Object-Permanence Tasks In addition to the questions concerning which capaci- ties are being examined during testing for object perma-
nence, another question arises as to whether the proce-
dures assess existing capacities or, instead, actually teach the concept (see Cornell, 1978; Hediger, 1981; Thomas & Walden, 1985). It is, for example, possible that (1) subjects could learn the task from observing the ex- perimenters' behaviors, (2) the repetitions of the tasks and even the mistrials allow subjects to engage in trial-and- error learning of each task (note Dore & Dumas, 1987), or (3) the tasks provide a step-by-step program for teach- ing the concept. Analysis of our procedures and results suggests that our study did indeed assess existent abilities. We were particularly careful to ensure that our birds were not being shown the correct response to any task. When a bird did not respond within 30 sec, the cover was not-as in some studies (Uzgiris & Hunt, 1966)-lifted in the bird's presence. Instead, both the cover and the ob- ject were removed and only then was the object displayed for another trial. Furthermore, our subjects were not likely to have learned to respond to perceptual cues (e.g., to the location at which the object was hidden) because none of our subjects was shown an object being hidden more than once in any location on a trial in a given session. Although it was impossible to ensure that one-trial learn- ing was not taking place, the pattern of errors was gener- ally inconsistent with an interpretation based on trial-and- error learning for the individual tasks. Had our subjects' competence on object-permanence tasks been a product of learning, there would have been such evidence in the data from Yellow Bird, who was the one subject to per- form the tasks three times. He did not, however, improve from the first series to the next; in fact, he made some mistakes on previously correct tasks (Tasks 11 and 13). He did improve on the third series, but not dramatically so. Even on this series, however, his mistakes were not grouped at the beginning and then reduced as one would expect if he were learning a concept. For the other sub- jects as well, choices did not get better with the number of repetitions. Thus, although tasks were repeated several times to show that chance was not a factor, such repeti- tions did not necessarily improve performance. For our subjects, the tasks did not seem to provide a stepwise means of learning the concept of object perma- nence. Success on the later tasks did not appear to be de- pendent upon success on the earlier ones. Yellow Bird made more mistakes in the later, more difficult tasks in all series, but he and another subject also responded cor- rectly on more complex tasks prior to their completing simpler ones: Fred reached criterion on Tasks 8-10 be- fore Task 7, and Yellow Bird completed Tasks 13-15 be- fore he was retested on Task 12. Experience on the sim- pler tasks did, however, teach our subjects that the experimenters and the apparatus were nonthreatening. Such learning did not affect our assessment of existent abilities, but did facilitate our efforts to perform such an assessment.
Implications of Having a Concept of Object Permanence Most animals, including humans, are not born with a concept of object permanence but acquire some sense of
106 PEPPERBERG AND FUNK
HEDIGER, H. K. P. (1981). TheCleverHans phenomenon froman animal psychologist'spoint of view. In T. A. Sebeok& R. Rosenthal (Eds.),
The CleverHans phenomenon (pp. 1-17). New York: Annals of the
New York Academy of Sciences. KAMIL, A. C., &; BALDA, R. P. (1985). Cache recovery and spatial
memory in Clark's nutcrackers (Nucifraga columbiana). Journalof
Experimental Psychology: Animal Behavior Processes, 11, 95-111.
KOEHLER, O. (1943). "Zahl"-Versuche an einem Kolkraben und Ver-
gleichsversuche an Menschen. Zeitschrift fUr Tierpsychologie, 5,
KOEHLER, O. (1950). The abilityof birds to "count. " Bulletin of Animal
Behaviour, 9, 41-45.
KOLAR, K. (1972). Der Papageien.In Grzimeks Tierleben VIII, Munich:
Kindler-Verlag; English translation, van Nostrand Reinhold.
KRooDSMA, D. E. (1982). Learning and the ontogeny of sound signals
in birds. In D. E. Kroodsma & E. H. Miller (Eds.), Acousticcom-
munication in birds (Vol. 2, pp. 1-23). New York: Academic Press.
KRooDSMA, D. E., BAKER, M. C., BAPTISTA, L. F., &; PETRINOVICH, L.
(1984). Vocal "dialects" in Nuttall's white-erownedsparrows. Cur-
rent Ornithology, 2, 103-133.
LECOMYTE, G. K., &; GRATCH, G. (1972). Violation ofa rule as a method
of diagnosing infants' levels of object concept. Child Development,
loRENZ, K. (1970). Studies in animaland human behavior (Vol. 1).
Cambridge, MA: Harvard University Press.
MACPHAIL, E. (1982). Brain and intelligence in vertebrates. Oxford:
Oxford University Press.
MACPHAIL, E. (1987). The comparativepsychologyof intelligence. Be-
havioral &: Brain Sciences, 10, 645-695.
PARKER, S. T., &; GIBSON, K. R. (1977). Object manipulation, tool use and sensorimotor intelligence as feeding adaptations in Cebus mon-
keys and great apes. Journal of Human Evolution, 6, 623-641.
PEPPERBERG, I. M. (1979, June). Functionol word use in an African
Grey parrot. Paper presented at the annual meeting of the Animal
Behavior Society, New Orleans. PEPPERBERG, I. M. (1981). Functionalvocalizations by an African Grey
parrot (Psittacus erithacusi, Zeitschrift fUr Tierpsychologie, 55,
PEPPERBERG, I. M. (1983).Cognition in the AfricanGrey parrot: Prelimi-
nary evidence for auditoryIvocal comprehensionof the class concept.
Animal Learning &: Behavior, 11, 179-185.
PEPPERBERG, I. M. (1987a). Evidence for conceptual quantitative abil-
ities in the African Grey parrot: Labeling of cardinal sets. Ethology,
PEPPERBERG, I. M. (1987b). Interspecies communication: A tool for
assessing conceptual abilities in the African Grey parrot iPsittacus
erithacus). In G. Greenberg & E. Tobach (Eds.), Language, cogni-
tion, consciousness: Integrative levels (pp. 31-56). Hillsdale, NJ:
Erlbaum. PEPPERBERG, I. M. (1989). Tool use in birds: An avian monkeywrench?
Behavioral &: Brain Sciences, 12, 604-605.
PEPPERBERG, I. M., &; KOZAK, F. A. (1986). Object permanence in the
African Grey parrot tPsittacus erithacus). Animal Learning &: Be-
havior, 14, 322-330.
PIAGET, J. (1953). Origin of intelligence in the child. London: Rout-
ledge & Kegan Paul.
PIAGET, J. (1954). Theconstruction of realityin the child. New York:
Basic Books.
PIAGET, J. (1971). Biology and knowledge. Chicago: University of
Chicago Press.
PIAGET, J. (1978). Behaviorand evolution. New York: Pantheon.
PIAGET, J. (1980). Adaptation and intelligence: Organic selection and
phenocopy. Chicago: University of Chicago Press.
PREMACK, D. (1983). The codes of man and beasts. Behavioral &: Brain
Sciences, 6, 125-167.
PREMACK, D. (1984). Upgrading a mind. In T. G. Bever, J. M. Car-
roll, & G. A. Miller (Eds.), Talking minds (pp. 181-206). Cambridge,
MA: MIT Press.
PREMACK, D. (1986). Gavagai. Cambridge, MA: MIT Press.
REDSHAW, M. (1978). Cognitive development in human and gorilla in-
fants. Journal of Human Evolution, 7, 133-141.
RENSCH, B. (1956). Increaseoflearning capability with increasein brain
size. American Naturalist, 90, 81-95.
ROZIN, P. (1976). The evolution of intelligence and access to the cog-
nitiveunconscious. In J. M. Sprague& A. N. Epstein (Eds.), Progress
in psychobiology andphysiological psychology (Vol. 6, pp. 245-280).
New York: Academic Press. SCARR-SALAPATEK, S. (1976). An evolutionary perspective on infant intelligence: Species patterns and individual variations. In M. Lewis
(Ed.), Origins of intelligence (pp. 165-197). New York: Plenum.
SHERRY, D. F. (1982). Food storage, memory, and marsh tits. Animal
Behaviour, 30, 631-633.
SHERRY, D. F. (1984). Food storage by black-capped chickadees:
Memory for the location and contents of caches. Animal Behaviour,
SHETTLEWORTH, S., '" KREBS, J. (1982). How marsh tits find their
hoards: The roles of site preference and spatial memory. Journalof
Experimental Psychology: AnimalBehaviorProcesses, 8, 354-375.
SHETTLEWORTH, S., '" KREBS, J. (1986). Stored and encountered seeds: A comparison of two spatial memory tasks in marsh tits and chicka-
dees. Journal of Experimental Psychology: AnimalBehavior Processes,
STETTNER, L. J., &; MATYNIAK, K. (1968). The brain of birds. Scien-
tific American, 218, 64-76.
THINUS-BLANC, C., POUCET, B., &; CHAPUIS, N. (1982). Object per-
manence in cats: Analysisin locomotorspace. Behavioural Processes,
THINUS-BLANC, C.; &; ScARDIGU, P. (1981). Object permanencein the
golden hamster. Perceptual &: Motor Skills, 53, 1010.
THOMAS, R. K., &; WALDEN, E. L. (1985). The assessment of cogni-
tive development in humanand nonhuman primates. In E. Watts (Ed.),
Nonhuman primate models for human growth and development
(pp. 187-215). Atlanta: Alan R. Liss. TINKLEPAUGH, O. T. (1928). An experimental study of representative
factors in monkeys. Journal of Comparative Psychology, 8, 197-236.
T1NKLEPAUGH, O. T. (1932). Multiple delayed reaction with chimpan-
zees and monkeys. Journal of Comparative Psychology, 13, 207-243.
TRIANA, E., &; PASNAK, R. (1981). Object permanence in eats and dogs.
Animal Learning &: Behavior, 9, 135-139.
UZGIRlS, I. C., &; HUNT, J. McV. (1966). Ordinal scales of infantde-
velopment, No. I-Object permanence [Film]. Champaign, IL:
University of Dlinois Films.
UZGIRlS, I. C., &; HUNT, J. McV. (1975). Assessment in infancy: Or-
dinal scales of psychological development. Champaign-Urbana, IL:
University of illinois Press. V ANDER WALL, S. B. (1982).An experimental analysisof cache recovery
in Clark's nutcrackers. Animal Behaviour, 30, 84-94.
VAUCLAlR, J. (1982).Sensorimotor intelligence in human and non-human
primates. Journal of Human Evolution, 11, 257-264.
VYGOTSKY, L. (1962). Thought and language. Cambridge, MA: MIT
Press.
WADSWORTH, B. J. (1984). Piaget's theory of cognitive and affective
development. New York: Longman.
WELLMAN, H. M., CROSS, D., &; BARTSCH, K. (1986). Infant search
and object permanence: A meta-analysis of the A-not-Berror. Mono-
graphs of the Society for Research in ChildDevelopment, 51(3, Serial
No. 214). WOOD, S., MORIARTY, K. M., GARDNER, B.T., &; GARDNER, R. A.
(1980). Object permanence in child and chimpanzee. AnimalLearn-
ing &: Behavior, 8, 3-9.
ZENTALL, T. R., EDWARDS, C. A., MOORE, B. S., &; HOGAN, D. E.
(1981). Identity: The basis for both matching and oddity learning in
pigeons. Journal of Experimental Psychology: Animal Behavior
Processes, 7, 70-86.
ZENTALL, T. R., .t. HOGAN, D. E. (1974). AbstracI concept learning
in the pigeon. Journal of Experimental Psychology, 102, 393-398.
NOTES
I. The following behaviors demonstrate Stages 5 and 6 understand- ing of object permanence. Stage 5 (considers sequential displacements whilesearchingfor vanishedobjects): nutcrackers, preventedfrom each-
OBJECT PERMANENCE IN PSITTACINE BIRDS 107
ing but allowed to watch other birds cache and recache seeds, recovered
seeds at levels above chance (Vander WaH, 1982). Stage 6 (images of absent objects, representations of displacements): (1) interocular transfer tests show that marsh tits rely on information stored in their brain to recover caches (Sherry, 1982); (2) marsh tits avoid already emptied caches during a second recovery session, suggesting that their behavior is based on memory rather than on simple rules of movement; this be- havior also indicates a memory for displacement (Sherry, 1982); (3) chickadees that cache two different types of food recover their fa- vored food first, suggesting memory for the particular type of food cached, rather than for just "something of interest" (Sherry, 1984);
(4) nutcrackers persist in searching areas where they have cached seeds
even when the experimenter has removed the seeds to prevent odor cues
(a test comparable to Task 15 ofUzgiris & Hunt, 1975); such recovery is, however, disrupted when local landmarks are moved or removed (Balda, 1980; Vander Wall, 1982).
- Yellow Bird's behaviors, as well as his later repeated hiding of an object under a calendar after retrieving it, are "tertiary circular reac- tions" that indicate Stage 5 competence in another Uzgiris and Hunt (1975) series, "The Construction of Object Relations in Space. " Such behaviors are important to note because they were thought to be limited to hominids (Parker & Gibson, 1977).
- Wellman, Cross, and Bartsch (1986), however, proposed that memory plays only a small role in the AB error. They suggested that the error comes from an incompletely developed ability to process spatiotemporal information-that is, a problem in understanding object movement. The small number of AB errors in our study suggests that these birds could correctly process spatiotemporal information.
APPENDIX
Scale I: The Development of Visual Pursuit and
the Permanence of Objects, from Uzgiris and Hunt (1975) and
Wood et al. (1980)
A. Visual pursuit of slowly moving objects
I. Follows a slowly moving object through a 180^0 arc (3-4)
a. Does not follow object b. Follows jerkily through part of arc c. Follows smoothly through part of arc *d. Follows object smoothly through complete arc
- Noticing the disappearance of a slowly moving object (3-4)
a. Does not follow to point of disappearance b. Loses interest as soon as object disappears *c. Lingers with a glance on point of disappearance *d. Returns glance to starting point after several presentations e. Searches around point of disappearance
B. Search for simply hidden objects
- Finding an object that is partially covered (3)
a. Loses interest b. Reacts to the loss, but does not obtain object *c. Obtains the object
- Finding an object that is completely covered (3)
a. Loses interest b. Reacts to loss, but does not obtain object c. Pulls screen, but not enough to obtain object *d. Pulls screen off and obtains object
- Finding an object completely covered in two places (2)
a. Loses interest b. Searches for object where it was previously found (with respect to location) *c. Searches for object where it is last hidden
- Finding an object completely covered in two places alternately (3-5)
a. Becomes perplexed and loses interest b. Searches haphazardly under one or both screens *c. Searches correctly under each of the screens
- Finding an object completely covered in three places (5-7)
a. Loses interest b. Searches haphazardly under some or all screens *c. Searches directly under the correct screen
C. Search following more complex hidings
- Finding an object after successive visible displacements (3-5)
a. Does not follow successive hidings b. Searches only under the first screen c. Searches under screen where object was previously found d. Searches haphazardly under all screens e. Searches in order of hiding *f. Searches directly under the last screen in path
- Finding an object under three superimposed screens (2- 3)
a. Loses interest b. Lifts one or two screens, but fails to find object *c. Removes all screens and obtains object
D. Search following an invisible displacement
- Finding an object following one invisible displacement (3)
a. Loses interest b. Reacts to loss, does not search c. Searches only in the box *d. Checks the box and searches under the screen *e. Searches under the screen directly
- Finding an object following one invisible displacement with two screens (2)
a. Searches only in the box b. Searches under screen where object was previously found *c. Searches directly under correct screen
- Finding an object following one invisible displacement with two screens alternated (3)
a. Loses interest b. Searches haphazardly under all screens *c. Searches directly under correct screen d. Other-searches under screen where last found
- Finding an object following one invisible displacement with three screens (5-7)
a. Loses interest b. Searches haphazardly under all screens
