Kea show three signatures of domain-general statistical inference

Abstract

One key aspect of domain-general thought is the ability to integrate information across different cognitive domains. Here, we tested whether kea (Nestor notabilis) can use relative quantities when predicting sampling outcomes, and then integrate both physical information about the presence of a barrier, and social information about the biased sampling of an experimenter, into their predictions. Our results show that kea exhibit three signatures of statistical inference, and therefore can integrate knowledge across different cognitive domains to flexibly adjust their predictions of sampling events. This result provides evidence that true statistical inference is found outside of the great apes, and that aspects of domaingeneral thinking can convergently evolve in brains with a highly different structure from primates. This has important implications not only for our understanding of how intelligence evolves, but also for research focused on how to create artificial domain-general thought processes.

More than the sum of its parts: Exploring the development of ratio magnitude versus simple magnitude perception

“…F I G U R E 1 Examples of nonsymbolic ratio magnitudes. From left to right, line-ratios, circle-ratios, and dot-ratios animals such as monkeys (Drucker et al, 2016;Vallentin & Nieder, 2008), parrots (Bastos & Taylor, 2020), and chicks (Rugani et al, 2016) demonstrate this ability. For example, McCrink and Wynn (2007) found that infants who were habituated to multiple examples of a 2:1 ratio of different colored dots looked longer at novel 4:1 ratios, indicating that they could discriminate ratio magnitudes.…”

Macphail’s Null Hypothesis of Vertebrate Intelligence: Insights From Avian Cognition

“…The ability to reason about use probabilistic information to make predictions about uncertain events also appears to have evolved convergently in the great apes and parrots (Rakoczy et al, 2014;Bastos and Taylor, 2020). When choosing between two hidden samples taken from two mixed populations of rewarding and unrewarding objects, capuchin monkeys appear to use a heuristic strategy of simply avoiding the sample from the population with the greatest absolute number of unrewarding objects (Tecwyn et al, 2017).…”

“…When choosing between two hidden samples taken from two mixed populations of rewarding and unrewarding objects, capuchin monkeys appear to use a heuristic strategy of simply avoiding the sample from the population with the greatest absolute number of unrewarding objects ( Tecwyn et al, 2017 ). On the other hand, both the great apes and the New Zealand mountain parrot, the kea, make their choices by relying on probabilistic information, by comparing the ratios of objects within and between the two populations ( Rakoczy et al, 2014 ; Bastos and Taylor, 2020 ). This ability, known as true statistical inference, has so far not been conclusively shown outside of these two taxa, as other studies on primates and birds have not been able to exclude the absolute number heuristic as a potential strategy ( Clements et al, 2018 ; Roberts et al, 2018 ; De Petrillo and Rosati, 2019 ; Placì et al, 2019 ).…”

“…Initial evidence that birds might have more domain-general cognitive processes comes from studies examining the ability of non-tool users to solve tool problems ( Bird et al, 2009 ; Auersperg et al, 2010 , 2011a , b , 2012b , 2016 ; Laumer et al, 2016 ). More recently, a study in kea showed that they can not only make accurate probabilistic comparisons between the two sampling events, as described above, but also integrate information across different domains ( Bastos and Taylor, 2020 ). In one experiment, the two jars contained a physical barrier, and the otherwise identical populations of tokens were unevenly distributed above and below these barriers.…”

Do infants and preschoolers quantify probabilities based on proportions?

“…The findings of the infant study encouraged a series of comparative studies with other animals [ 8 – 13 ]. These studies are important to better apprehend the distribution of probabilistic reasoning skills and come to a better understanding of their evolution.…”