Jean Piaget was the first to note infants younger than 8 months engaging in perseverative reaching during the A-not-B task (Bornstein and Lamb, 1999). In this task, an experimenter hides a toy under box A, within reach of the infant. The infant then searches for the toy under box A. This sequence of events is repeated several times. Finally, in the experimental trial, the experimenter hides the toy under box B. After a short delay, the infant is allowed to reach for the box with the toy. Although the infant has seen the experimenter hide the toy under box B, and box B is within reach, the infant will continue to search for the toy under box A.
In Piaget’s theory of cognitive development, stage IV is critical. During stage IV of development, the infant appears to have some sense of object permanence, since she does initially search for the hidden toy. The object concept, however, has a limitation; to the child, the object is bound to its location (Gratch and Landers, 1971). Piaget asserted that this error (“stage IV error”, “perseverative error”, or “A-not-B error”) reveals an incomplete schema of object permanence. Recent researchers, however, found that the perseverative error can indicate deficits in other cognitive abilities, as well (Marcovitch and Zelazo, 1999). In this paper, I will discuss the full range of interpretations that can be derived from various results of Piaget’s A-not-B task, many of which have little, if anything, to do with object permanence. I will end with a study proposal to examine a newer version of the A-not-B task.
Best services for writing your paper according to Trustpilot
THE PROBLEM OF OBJECT PERMANENCE
In his work on the sensorimotor development of infants, Piaget asserted that infants at ages 8-10 months show a perseverative error during search tasks due to incomplete object permanence, and that the error indicates a failure to assimilate the new hiding place into the infant’s object concept (Marcovitch and Zelazo, 1999). When other researchers examined Piaget’s hypothesis, their studies yielded mixed results. Wellman, Cross and Bartsch (1986) conducted a meta-analysis on all studies conducted on the perseverative error before 1986 and found that infants ages 8-10 months show significant perseveration, reaching for location A much more often than they reached for location B. Length of delay, child’s age, and number of locations all had significant effects, while the number of trials reinforcing location A was unrelated to infants’ performance (Wellman, Cross and Bartsch, 1986).
Among the studies examined in the meta-analysis, Gratch, Appel, Evans, LeCompte and Wright (1974) offered some opposition to Piaget’s hypothesis. After testing 9-month-old infants with the A-not-B task at delay intervals of 0, 1, 3, and 7 seconds, the researchers found that infants showed the perseverative error in all interval conditions except the 0-second condition (i.e. no delay), with children just as likely to show the error at 1, 3, and 7 seconds of delay. These results contradict Piaget’s hypothesis of failure to assimilate the new hiding place; had this been the case, the infants should have shown the perseverative error even with no delay before search (Gratch et al., 1974). The researchers suggest that after the infant’s gaze is drawn to B, and without a delay to search, the infant is guided to the correct location (B) by virtue of already looking in that direction. With any amount of delay, however, infants revert back to their original scheme of the toy existing at location A (Gratch et al., 1974). Even incorporating the results, this theory is compatible with Piaget’s hypothesis of incomplete object permanence and lack of assimilation of the new hiding place into the infant’s schema.
Harris (1973) initially argued that the 10-month-old infants in his study presented the perseverative error due to proactive interference, evident by errors that increased when length of delay is increased. In a follow-up study, Harris (1974) discounts his prior ideas, asserting that a memory problem is not the likely culprit behind the perseverative error. In this study, Harris notes that 1-year-old infants, well beyond Piaget’s stage IV of sensorimotor development, approach both locations A and B when an object moves; that is, the infants do not treat object locations as mutually exclusive. In the A-not-B task, infants search behind both occluders to find the object (Harris, 1974). If infants are searching both locations, it is unlikely that proactive interference is the reason behind the perseverative error (Harris, 1974).
By 1975, researchers had still not found a satisfying cause of the perseverative error. Butterworth (1977), in an attempt to test Piaget’s hypothesis of incomplete object permanence, found that hiding the object is not a necessary condition for the perseverative error. In testing infants, Butterworth found that infants would sometimes look away from B before searching, usually indicating that proactive interference was a possibility; however, infants showed the perseverative error even while continuously looking at B, or continuously tracking the object (Butterworth, 1977). As a post-hoc theory, Butterworth suggests that the problem may not be that of incomplete object permanence, but that infants during stage IV simply have not acquired the skill to identify and retrieve objects in a coordinated manner.
THE SPATIAL LOCALIZATION HYPOTHESIS
Butterworth (1975) examined the perseverative error from a different point of view: does the location of the object change the rate of error? Under various conditions, errors were, indeed, found to be a function of the object’s location. Results from this study show that infants made the error when the object was moved away from the midline (Butterworth, 1975). Butterworth hypothesizes that errors were produced due to prior experience of the object at the midline (A), which made it difficult to assimilate the object at a location away from the midline (B). In this experiment, it is especially clear that infants are not simply repeating previously successful responses; that is, they make errors after a change in the relationship between locations, without prompting from the experimenter (Butterworth, 1975).
In a similar study, Bremner and Bryant (1977) set out to differentiate between Piaget’s egocentric response perseveration hypothesis, and, as an alternate hypothesis, perseveration to a certain location in space. There is a clear distinction between the hypotheses, perhaps best characterized by relative and absolute location. For example, if a child reaches for an object to her left, but changes the way she reaches if she is moved to the opposite side, then she is perseverating to a certain location in space. If the child continues to reach to her left even if she is moved to the opposite side, then she is showing egocentric response perseveration. Both hypotheses predict different results and different reasons for the perseveration error. Interestingly, Bremner and Bryant (1977) were able to confirm Piaget’s hypothesis, finding that infants were repeating past successful actions, even when they were re-located and facing a new direction. Perseveration does seem to be of responses, not places. To explain why infants persist in their responses in spite of location changes, the researchers hypothesize that stage IV infants, unlikely to be crawling yet, may ignore visual information when an object needs to be relocated. After initially finding an object (at location A), the infant, in most cases, simply needs to repeat the motion to find the object again (Bremner and Bryant, 1977).
In another study, Schuberth, Werner and Lipsitt (1978) created a condition in which half of the infants found a new toy at location B rather than the same toy they saw at location A. These infants were less likely to show the perseverative error than their same-toy counter-group, theoretically because they mapped the location as part of the toy’s scheme, consistent with Piaget’s object concept hypothesis. However, the researchers found that although most infants in the toy-change condition showed less perseveration, one-third of them continued to err and search at location A. It is possible that these infants failed to identify the new toy as different from the old toy, but other than this speculation, the researchers have no explanation for the contrasting behaviour. Their results support Piaget’s hypothesis over the spatial localization hypothesis: it appears that infants during stage IV do not differentiate between concept of object and concept of place (Schuberth et al., 1978).
MEANS-ENDS ABILITIES
After researchers failed to come up with tested alternative explanations for the stage IV error, they began to test infants’ object permanence itself, perhaps to reaffirm Piaget’s original hypothesis. In an infamous study by Baillargeon, Spelke and Wasserman (1985), five-month-old infants were habituated to a clear-screened drawbridge, moving from an upright position to a flat position. Infants were then exposed to a possible event and an impossible event: in the possible event, the screen stops when it reaches an occluding box; in the impossible event, the screen goes straight through the box. The results showed that infants looked significantly longer at the impossible event, indicating that the infants expected the screen to stop when it hit the box, and were surprised when it did not. This experiment served two purposes: (1) it disproves Piaget’s original claims, showing that five-month-old infants appear to have a schema of object permanence; and, (2) as object permanence is “an inseparable aspect of how objects behave in time and space” (Baillargeon et al., 1985, p. 206), the results raise even more questions about the perseverative error. If five-month-old infants understand the permanence of objects, their ability to exist when occluded, and that objects move on undeviating paths, why do 8- to 11-month-old infants commit the perseverative error during the A-not-B task?
Baillargeon, Graber, Devos and Black (1990) claim that infants do poorly on the A-not-B task because it requires them to produce means-ends sequences. Their general hypothesis is that these younger infants have trouble planning such sequences due to limited problem-solving ability. The researchers believe that the infants’ representations of the initial and goal states are unlikely to be the problem. Instead, it is likely that the infants are unable to reason about the actions required to transform the initial state to the goal state (Baillargeon et al., 1990). In this study, the researchers showed that infants do understand whether certain actions are sufficient to retrieve the object; consequently, the researchers hypothesize that infants are unable to select and sequence these actions adequately, even when they understand the actions. It seems, then, that it is a problem with planning.
Matthews, Ellis and Nelson (1996) partially agree with Baillargeon et al.’s (1990) hypothesis. After examining infants engage in a means-ends task several times during the longitudinal study, the researchers found the ratio of means-ends errors to total means-ends trials within a session to be non-significant. Means-ends errors dropped for infants across testing sessions, but no group differences between ages were found. Despite these results, infants continued to perseverate during the A-not-B task, leading the researchers to question what might be behind the perseverative error. They concede that despite the results of the means-ends portion of the study, the performance during the A-not-B tasks may, indeed, be due to some limitation of means-ends problem-solving ability. However, Matthews et al. (1996) also note that the function that mediates performance on the A-not-B task may be memory-related rather than one of perseveration.
Similarly, Munakata, McClelland and Johnson (1997) found that 7-month-old infants who were trained on means-ends behaviours still showed different behaviours in tasks that required the same means-ends abilities. For instance, after being trained, infants completed more toy retrievals when the toy was hidden behind a transparent occluder than when the toy was hidden behind an opaque occluder. In terms of means-ends abilities, both the transparent and opaque conditions required the same actions and the same effort. As the infants appeared to have more difficulty in the opaque condition, means-ends deficits were likely not the culprit. In a latter experiment, Munakata et al. (1997) found similar results, with infants showing greater toy retrieval in the transparent occluder condition. The researchers concluded that their difficulties with the opaque occluder could not be attributed to means-ends abilities alone. Essentially, deficits in means-ends abilities seem to not cause problems when the goal object is not hidden.
PERSPECTIVES FROM NEUROPSYCHOLOGY
Neuropsychology views the Piagetian A-not-B task as testing the immaturity of the frontal lobe. At stage IV of sensorimotor development, the frontal lobe, still underdeveloped, is unable to support cognitive skills such as working memory, inhibition, and attention. Bell and Adams (1999) examined 8-month-old infants and their performance on both the looking and the reaching versions of the A-not-B task. From a neuropsychological perspective, the aforementioned cognitive skills are used to search in both looking and reaching tasks. If both versions test the same skills, it is possible that the reaching task is simply more motorically complicated for infants to complete properly. Indeed, Bell and Adams (1999) showed, in a within-subjects design, that there is no difference in performance by infants in either the looking or the reaching version of the A-not-B task.
Clearfield, Diedrich, Smith and Thelen (2006) discuss successfully completing the A-not-B task as requiring a combination of two sets of cognitive processes: fast processes tied to the present, and slower processes tied to the past. Very young infants are often able to complete the task without error; Clearfield et al. (2006) attribute this to the use of the fast processes alone. These processes decay quickly, and the researchers argue that the perseverative error is how infants learn how to balance the fast and the slow processes, making it an important developmental milestone. In fact, the results of their experiments show that the infant must achieve some level of stability before perseveration occurs (Clearfield et al., 2006). With eight-month-old infants’ smooth reaches and five-month-old infants’ poorly-controlled reaches, the eight-month-olds are more likely to form strong motor memories, possibly leading to perseveration at location A. These motor memories, formed during stage IV, are also only likely to be used during stage IV (Clearfield et al., 2006).
Cuevas and Bell (2010) presented results consistent with the neuropsychological view, asserting that the cognitive skills required for looking and for reaching are very similar, and likely determined, in part, by the development of the prefrontal cortex. Infants initially appear to exhibit better performance on the looking version of the A-not-B task; later in age, they exhibit comparable performance on both looking and reaching versions. The researchers assert that this difference is due to differences in brain circuitry: infants show looking responses very early, while reaching responses are not consistently shown until 3-4 months, which still remain poorly-controlled until 8-9 months. Additionally, reaching creates more demands on cognition, requiring “memory of the hidden location, planning of a means-ends action sequence, and a reaching response” (Cuevas and Bell, 2010, p. 1369). It is possible that infants exhibit the perseverative error due to cognitive overload caused by reaching.
FUTURE DIRECTIONS – STUDY PROPOSAL
An interesting way to look at infant cognition during the A-not-B task is to use a gaze-tracking procedure. Once the eight-month-old infant is sitting on her parent’s lap, the researcher will begin the training trials, during which he will hide a toy under an opaque box (A), and then allow the infant to search for the toy. At this point, the infant’s gaze should remain on the researcher, the toy, and box A. During the test trials, the researcher will hide the toy under another opaque box (B), and, again, allow the infant to search for it. Before reaching (and presumably reaching for box A), where does the infant look? Does she look at box A immediately, or does her gaze linger on box B? Does she look at the researcher for clues, or does her gaze move back and forth between boxes? As the aforementioned research says that she is likely to perseverate at box A at eight months, when looking measures are unlikely to be used, it would be interesting to see if the infant has a more immediate response to the task before engaging in the reaching behaviour. Because looking develops before reaching, I predict that infants will not engage in perseverative behaviour through eye gaze, even if they ultimately reach for box A; that is, they will look at box B, but reach for box A.
CONCLUSION
Currently, there are few true answers to the problem of the perseverative error. Although it is a robust and well-researched area of developmental psychology, researchers are still unsure why infants in stage IV of sensorimotor development are unable to reach for the correct location during the A-not-B task. Future directions, as mentioned above, involve finding another way to test infants’ object concepts in order to circumvent the perseverative error.
