Journal Club: Cockatoos are smarter than human toddlers
SUMMARY: How do you know that something you cannot see is still present? A new study shows that "object permanence" cognitive abilities in a cockatoo rivals that of the great apes and four-year-old human children
How do you know that something you cannot see is still present? Full object permanence, including the ability to track invisible trajectories of objects through time and space, takes years to develop in human children.
Image: Alice Auersperg/University of Vienna.
I live with a couple cockatoos so I am always looking for new tasks for them to solve, just to keep them (and me!) mentally challenged. Whilst I was traveling last month, I ran across an interesting little study that examined object permanence and spatial tracking abilities in Goffin's cockatoos. Object permanence is the idea that, like a cashew nut hidden in a pocket, an object exists even when it is not visible to the observer. Further, it also includes spatial tracking so when that concealed cashew nut is removed to a new location, say, a parrot puzzle toy, and hidden there, the observer (or cockatoo) then knows to seek it in that new location, even though the nut could not be observed whilst being relocated.
This level of cognitive development does not occur in human children until they reach four years of age. But according to a newly published study by an international team of scientific researchers and a flock of cockatoos based at the University of Vienna, object permanence abilities in young cockatoos rival those of four-year-old human children. Which of course means that hiding your stash of cashew nuts whilst your pet cockatoo is watching may be a bad idea.
The basic idea
Swiss developmental psychologist and philosopher, Jean Piaget, was the first person to study object permanence in human babies and toddlers, and he identified and described six distinct stages of visible and invisible object permanence. Basically, visible object permanence is where an object is recognised as having its own separate existence even after it is moved from one location to another whilst in the subject's plain view. Invisible object permanence is the idea that objects continue to exist even when they are not visible to the observer.
For example, by the time a human infant is approximately four months of age, she or he will reach for an object that is partially hidden from view (this is Piagetian stage 3b). After reaching eighth months of age, toddlers typically retrieve an object after observing it being hidden (Piagetian stage 4a). By the time a human reaches one year of age, he or she will search a second location if they see the object being moved from one hiding place to another (Piagetian stages 5a & 5b).
Piaget found that only after a human toddler is somewhere between 18 and 24 months of age are they able to solve invisible displacement tasks (Piagetian stages 6I and 6II). This is where the desired object is concealed in or under a container that is moved behind one or more opaque screens where the object is then deposited. Afterwards, the empty container is shown to the child, which reinforces the idea that the object was hidden behind the last screen visited by the container. These are classic tests that are used to assess cognitive development in humans.
But so-called shell games (transposition experiments or non-Piagetian invisible displacement tests), which are popular gambling games at carnivals, go further than Piaget's cognitive development tests for children. In transposition experiments, several identical containers -- only one of which hides the desired object -- are moved around each other a number of times in plain view. There are several variations on this theme, such as rotation tests -- where the containers are set on a solid base that is rotated in plain view of the subject -- and translocation tests -- where the observer is moved around the containers.
These variations are more even cognitively demanding tasks. For example, research shows that human children cannot reliably solve rotation tasks until they are four years old -- and some people never manage it.
The experimental design
How do other animals perform on these tests? Findings from previous studies in great apes are somewhat controversial, but they seem to suggest that these animals perform reasonably well in invisible displacement tasks (Piagetian stage 6; doi:10.1037/0097-7403.32.3.239). However, similar studies with other primates and with domestic pets like cats and dogs are inconclusive (doi:10.1002/ajp.22118).
But what about birds? Unlike most other mammals, birds are very similar to humans in the way they perceive the world around them -- being visually and auditorily oriented, as well as intelligent and highly spatial. Thus, it makes sense that birds would perform quite well on these tests, reflecting similarities to human children in cognitive development.
So far, the performances turned in by a number of corvid and parrot species do rival those of human toddlers. But these studies had very few subjects -- only one or two generally -- and none were tested on all three non-Piagetian invisible displacement tasks (transposition, rotation, and translocation) that are also commonly used in human cognitive development tasks. What might we learn about how birds think and perceive the world if a larger number were tested, and if they were also tested on the three benchmark invisible displacement tasks?
To answer these questions, we once again visit that flock of Goffin's cockatoos, Cacatua goffiniana, who have been in the news fairly often during the past few months (for example, read this and this). This flock of 14 captive-bred cockatoos resides in a large indoor-outdoor aviary at the University of Vienna in Austria. They regularly participate in a variety of non-invasive behavioural experiments as test subjects (meet the flock, larger view):
The Goffin Lab's flock of cockatoos. The group consists mainly of parrots hatched in 2010 and 2011. Four individuals are young adults (2006 and 2007 hatch).
Eight of these cockatoos were subjects in a series of tasks where they were asked to identify and track containers that concealed a favourite food reward (a cashew nut). All parrots (except one individual) were younger than one year old when they were tested.
The containers were magnetic so they could be "invisibly" moved manually from under the table using a magnet (see figure at right; larger view). Of course, when the parrot correctly identified which container concealed the food reward, he or she was allowed to retrieve the treat (larger view):
A Goffin's cockatoo retrieves the food reward that was concealed under the magnetic container.
Image: Alice Auersperg/University of Vienna [doi:10.1037/a0033272]
To observe how these cockatoos perform in Piaget's object permanence tests and in the three non-Piagetian invisible displacement tests, the team of researchers designed a number of tasks using the magnetic containers (experimental design; larger view):
Experimental design. [doi:10.1037/a0033272]
The above cartoon illustrates the experimental design as follows;
- Piagetian visible displacement tasks 3b and 4a (panel I, where in Piagetian 3b, the food reward is partially hidden and in 4a it is fully hidden after subject starts moving towards it)
- Piagetian visible displacement tasks 5a and 5b (panel II, where the food reward is fully hidden before the subject is released)
- Piagetian invisible displacement tasks (panel III, where a small container is moved behind one of three large screens where the food reward is hidden)
- Transposition tasks a and b (panel IV, where two of three containers are simultaneously exchanged, one of which conceals the food reward)
- Rotation tasks (panel V, where the containers are sat in a row upon a rotatable platform attached to a stick that is inserted through a hole in the table, thereby allowing it to be manually rotated by one of four different angles [90°, 180°, 270°, 360°])
- Translocation tasks (panel VI, is similar to the rotation task except the subject is moved rather than the experimental setup)
Of all these tasks, human children find translocation and rotation tasks to be most challenging: they solve translocation tasks usually between two and three years of age and they typically solve rotation tasks at four years of age.
The parrots were given 12 trials to correctly solve the tasks. All of the subject's responses were rated as the parrot performed and, because all the tests were videotaped, 20 percent of the trials were additionally scored later by an independent observer:
"The majority of our eight birds readily and spontaneously solved transposition, rotation and translocation tasks", said co-author Birgit Szabo, a cognitive biologist at the University of Vienna (figure 4; larger view):
Figure 4. Number of correct choices within the first 12 trials of the non-Piagetian invisible displacement tasks; * = Significantly above chance; brackets indicate significant paired tests between conditions. [doi:10.1037/a0033272]
They performed quite well on the transposition tasks.
"Transpositions are highly demanding in terms of attention since two occluding objects are moved simultaneously. Nevertheless, in contrast to [great] apes, which find single swaps easier than double [swaps], the cockatoos perform equally in both conditions", said lead author Dr Auersperg.
The cockatoos performed reasonably well the rotation and translocation tasks. Additionally -- and in contrast to human children who find translocation tasks easier than rotation tasks -- the cockatoos showed no significant differences between the two.
However, the cockatoos didn't perform as well on the Piagetian invisible displacement tasks (figure 3; larger view):
Figure 3. Number of correct choices within the first 12 trials of tasks 3b–6 II (6 II1 = hidden in first location; 6 II2 = hidden in second location); * = Significantly above chance on group level. [doi:10.1037/a0033272]
"[O]nly two out of eight choose immediately and reliably the correct location in the original Piagetian invisible displacement task in which a smaller cup is visiting two of three bigger screens", said Dr Szabo.
This surprised the researchers.
"Interestingly -- and just opposite to human toddlers -- our cockatoos had more problems solving the Piagetian invisible displacements [stages 6I and 6II] than the transposition task with which children struggle until the age of four", said Dr Auersperg.
The meaning of it all
This study found that the spatial reasoning abilities of Goffin's cockatoos rival those of corvids, great apes and even human children. This is something that many pet parrot owners could tell you just by living with these birds. But cockatoo intelligence didn't arise so they could be interesting and amusing companions for humans, so why did they evolve such advanced cognitive and spatial talents?
Well, the parrots' ecological role has probably provided a strong driving force in the evolution of these birds' mental strengths.
"[Survival] may be a candidate trait influencing the animals' performance in rotation and translocation tasks", said co-author Auguste von Bayern, a research associate in the Zoology Department at the University of Oxford.
Similar spatial abilities have been found in other bird species studied, which lends credence to the hypothesis that their cognitive abilities evolved to deal with the spatial and tracking demands associated with both flight and predator-prey interactions.
"We assume that the ability to fly and to prey upon or [to avoid] being preyed upon from the air is likely to require pronounced spatial-rotation abilities", said Dr von Bayern.
"Finding that Goffins [cockatoos] solve transposition, rotation and translocation tasks, which are likely to pose a large cognitive load on working memory, was surprising", agreed co-author Thomas Bugnyar, a Professor of Cognitive Ethology at the University of Vienna.
"[This] calls for more comparative data in order to better understand the relevance of such accurate tracking abilities in terms of ecology and sociality".
My guess is that such studies will find that all flighted birds have similar spatial and object tracking abilities.
Auersperg A.M.I., Szabo B., von Bayern A.M.P. & Bugnyar T. (2013). Object Permanence in the Goffin Cockatoo (Cacatua goffini), Journal of Comparative Psychology, [advance online publication] doi:10.1037/a0033272
Supplemental Material for Object Permanence in the Goffin Cockatoo (Cacatua goffini), Journal of Comparative Psychology, [advance online publication] doi:10.1037/a0033272.supp
University of Vienna press release
Previously published research from this group:
Polly gets his own cracker: clever cockatoo manufactures, uses tools -- the comments on this piece are also quite interesting.
Mallavarapu S., Perdue B.M., Stoinski T.S. & Maple T.L. (2013). Can Black-and-White Ruffed Lemurs (Varecia variegata) Solve Object Permanence Tasks? American Journal of Primatology, 75 (4) 376-386. doi:10.1002/ajp.22118
Barth J. & Call J. (2006). Tracking the displacement of objects: A series of tasks with great apes (Pan troglodytes, Pan paniscus, Gorilla gorilla, and Pongo pygmaeus) and young children (Homo sapiens), Journal of Experimental Psychology: Animal Behavior Processes, 32 (3) 239-252. doi:10.1037/0097-7403.32.3.239 [OA]
.. .. .. .. .. .. .. .. .. .. ..
.. .. .. .. .. .. .. .. .. .. ..