The “evolution” of interest in individual differences in the intelligence of nonhuman animals has followed a circuitous route. Individual differences in intelligence were a central focus of early psychologists (note the inclusion of chapters in our first textbooks; e.g., Seashore Reference Seashore1923), as well as, some decades later, of the first animal learning theorists (e.g., Thorndike's studies in the 1930s). However, with the increasing fixation on the “experimental approach” and reductionism, interest in individual differences waned overall, and systematic studies of variations in intelligence within animal species were virtually abandoned between 1940 and 2000. This trend has shifted dramatically in recent times, with increasing interest in between-species comparisons of intelligence (G), and more dramatically, in within-species variations in intelligence (g). In this spirit, Burkart et al. have done commendable work summarizing the advances, insights, and limitations of animal research on individual differences in intelligence, and have placed this work in the important context of contemporary evolution theory.
Although we agree with many of Burkart et al.'s conclusions, we are skeptical of their inference that the evolution of intelligence, as well as individual differences in intelligence, is inextricably tied to brain size. Brain size does appear to explain differences in the cognitive capacities of closely related species, although the relationship begins to break down across families and higher taxonomic groups. It is similarly problematic that while Neanderthal brain size ranged from 1,300–1,600 grams, their human counterparts had brain sizes of 1,200–1,500 grams. Current theory suggests that competition between the cognitively superior humans and cognitively inferior Neanderthal accounted for the latter's rapid extinction (Banks et al. Reference Banks, d'Errico, Peterson, Kageyama, Sima and Sanchez-Goni2008; Gilpin et al. Reference Gilpin, Feldman and Aoki2016). Relatedly, the size of the human brain has decreased during the last 100,000 years (Aiello & Dean Reference Aiello and Dean1990), a time during which we underwent unusually rapid cognitive gains.
Although brain size does have some value in explaining the cognitive capacities of closely related species (i.e., G), it is less successful when applied to individual differences within a species. Early estimates suggested a weak relationship between brain size and intelligence (r
2 = 0.02–0.07; reviewed in Van Valen Reference Van Valen1974), and meta-analyses based on modern imaging techniques find only a marginal increase in this estimate (r
2 = 0.08; reviewed in McDaniel Reference McDaniel2005). Furthermore, the strength of correlations between brain size and intelligence vary across specialized abilities, and in the case of some abilities, no correlation is observed (van Leeuwen et al. Reference van Leeuwen, Peper and van den Berg2009; Wickett et al. Reference Wickett, Verbnon and Lee2000), suggesting that variations in brain size may instantiate differences in specific abilities, but not variations in general intelligence. So why might any correlation exist between brain size and intelligence? A possibility that is widely ignored is that more intelligent individuals interact more extensively with their environments (e.g., they explore more, they learn more; Light et al. Reference Light, Grossman, Kolata, Wass and Matzel2011; Matzel et al. Reference Matzel, Townsend, Grossman, Han, Hale, Zappulla, Light and Kolata2006), and this “environmental enrichment” promotes brain growth (Rosenzweig & Bennett Reference Rosenzweig and Bennett1996). Simply stated, brain size might be influenced by intelligence, but might not itself cause differences in intelligence. This possibility has received wide support outside of the field of intelligence (Clayton Reference Clayton2001; Maguire et al. Reference Maguire, Gadian, Johnsrude, Good, Ashburner, Frackowiak and Frith2000; van Praag et al. Reference van Praag, Kempermann and Gage2000; Will et al. Reference Will, Galani, Kelche and Rosenzweig2004), and can explain the paradoxical observation that the correlation between IQ and brain size only emerges after age 7 (by which time differential experiences will have begun to accumulate; McDaniel Reference McDaniel2005).
The role of brain size in intelligence may matter less than we intuit. It is important to be reminded that brain size is only a very indirect measure of how general intelligence is instantiated. Higher cognition is highly complex, and the circuitry, neurochemistry, and intracellular components of the brain all contribute to its computational capacity. For example, as noted by Burkart et al., we have reported that general intelligence in mice is correlated with the expression in the prefrontal cortex (PFC) of a dopaminergic gene cluster( Kolata et al. Reference Kolata, Light, Wass, Colas-Zelin, Roy and Matzel2010), and smarter mice express higher dopamine-induced activity in the prefrontal cortex (Wass et al. Reference Wass, Pizzo, Sauce, Kawasumi, Sturzoiu, Ree, Otto and Matzel2013). In humans, the dopaminergic system in the PFC seems also to be closely linked with executive functions and intelligence (McNab et al. Reference McNab, Varrone, Farde, Jucaite, Bystritsky, Forssberg and Klingberg2009; Miller & Cohen Reference Miller and Cohen2001). And whereas the brain of birds differs strikingly from the mammalian brain (e.g., it lacks the 6 layers of lamination in the neocortex), the avian nidopallium caudolaterale (NCL) is remarkably similar to the mammalian PFC. Like the PFC, the NCL is a hub of multimodal integration connecting the higher-order sensory input to limbic and motor structures (Gunturkun & Kroner Reference Gunturkun and Kroner1999) , and dopamine in the avian NCL seems to play a similar functional role in higher cognition as it does in the mammalian PFC (Karakuyu et al. Reference Karakuyu, Herold, Gunturkun and Diekamp2007; Veit et al. Reference Veit, Hartmann and Nieder2014). This confluence of evidence across taxonomic groups (humans, mice, and birds) is compelling, and at least as parsimonious as the descriptions of intelligence based on variations in brain size.
Burkart et al. imply in their current article and state explicitly elsewhere (van Schaik et al. Reference van Schaik, Isler and Burkart2012) that “general intelligence is not a uniquely derived human trait but instead a phylogenetically old phenomenon, found among primates, rodents and birds” (p. 280). However, the PFC and NCL are on opposite ends of the cerebrum and possess distinct genetic expression patterns, leading some to claim that these regions are not homologous but, rather, represent a case of evolutionary convergence (Gunturkun Reference Gunturkun2012). Thus, non-homologous fields converged over the course of 300 million years into mammalian and avian prefrontal areas that generate the same cognitive functions (e.g., working memory capacity; Diekamp et al. Reference Diekamp, Kalt and Gunturkun2002; Matzel et al. Reference Matzel, Sauce-Silva and Wass2013) that contribute to the establishment of general intelligence. In other words, general intelligence could have evolved multiple times in different taxonomic groups. Of course this is a matter of considerable controversy (Karten Reference Karten2015), and the question is far from resolved. Nonetheless, this type of solution is more parsimonious than one based solely on brain size, and mitigates the extant problem of the “cost” of bigger brains. We hope that the “evolution” of interest in the variation in general intelligence follows this route for the next decade.
The “evolution” of interest in individual differences in the intelligence of nonhuman animals has followed a circuitous route. Individual differences in intelligence were a central focus of early psychologists (note the inclusion of chapters in our first textbooks; e.g., Seashore Reference Seashore1923), as well as, some decades later, of the first animal learning theorists (e.g., Thorndike's studies in the 1930s). However, with the increasing fixation on the “experimental approach” and reductionism, interest in individual differences waned overall, and systematic studies of variations in intelligence within animal species were virtually abandoned between 1940 and 2000. This trend has shifted dramatically in recent times, with increasing interest in between-species comparisons of intelligence (G), and more dramatically, in within-species variations in intelligence (g). In this spirit, Burkart et al. have done commendable work summarizing the advances, insights, and limitations of animal research on individual differences in intelligence, and have placed this work in the important context of contemporary evolution theory.
Although we agree with many of Burkart et al.'s conclusions, we are skeptical of their inference that the evolution of intelligence, as well as individual differences in intelligence, is inextricably tied to brain size. Brain size does appear to explain differences in the cognitive capacities of closely related species, although the relationship begins to break down across families and higher taxonomic groups. It is similarly problematic that while Neanderthal brain size ranged from 1,300–1,600 grams, their human counterparts had brain sizes of 1,200–1,500 grams. Current theory suggests that competition between the cognitively superior humans and cognitively inferior Neanderthal accounted for the latter's rapid extinction (Banks et al. Reference Banks, d'Errico, Peterson, Kageyama, Sima and Sanchez-Goni2008; Gilpin et al. Reference Gilpin, Feldman and Aoki2016). Relatedly, the size of the human brain has decreased during the last 100,000 years (Aiello & Dean Reference Aiello and Dean1990), a time during which we underwent unusually rapid cognitive gains.
Although brain size does have some value in explaining the cognitive capacities of closely related species (i.e., G), it is less successful when applied to individual differences within a species. Early estimates suggested a weak relationship between brain size and intelligence (r 2 = 0.02–0.07; reviewed in Van Valen Reference Van Valen1974), and meta-analyses based on modern imaging techniques find only a marginal increase in this estimate (r 2 = 0.08; reviewed in McDaniel Reference McDaniel2005). Furthermore, the strength of correlations between brain size and intelligence vary across specialized abilities, and in the case of some abilities, no correlation is observed (van Leeuwen et al. Reference van Leeuwen, Peper and van den Berg2009; Wickett et al. Reference Wickett, Verbnon and Lee2000), suggesting that variations in brain size may instantiate differences in specific abilities, but not variations in general intelligence. So why might any correlation exist between brain size and intelligence? A possibility that is widely ignored is that more intelligent individuals interact more extensively with their environments (e.g., they explore more, they learn more; Light et al. Reference Light, Grossman, Kolata, Wass and Matzel2011; Matzel et al. Reference Matzel, Townsend, Grossman, Han, Hale, Zappulla, Light and Kolata2006), and this “environmental enrichment” promotes brain growth (Rosenzweig & Bennett Reference Rosenzweig and Bennett1996). Simply stated, brain size might be influenced by intelligence, but might not itself cause differences in intelligence. This possibility has received wide support outside of the field of intelligence (Clayton Reference Clayton2001; Maguire et al. Reference Maguire, Gadian, Johnsrude, Good, Ashburner, Frackowiak and Frith2000; van Praag et al. Reference van Praag, Kempermann and Gage2000; Will et al. Reference Will, Galani, Kelche and Rosenzweig2004), and can explain the paradoxical observation that the correlation between IQ and brain size only emerges after age 7 (by which time differential experiences will have begun to accumulate; McDaniel Reference McDaniel2005).
The role of brain size in intelligence may matter less than we intuit. It is important to be reminded that brain size is only a very indirect measure of how general intelligence is instantiated. Higher cognition is highly complex, and the circuitry, neurochemistry, and intracellular components of the brain all contribute to its computational capacity. For example, as noted by Burkart et al., we have reported that general intelligence in mice is correlated with the expression in the prefrontal cortex (PFC) of a dopaminergic gene cluster( Kolata et al. Reference Kolata, Light, Wass, Colas-Zelin, Roy and Matzel2010), and smarter mice express higher dopamine-induced activity in the prefrontal cortex (Wass et al. Reference Wass, Pizzo, Sauce, Kawasumi, Sturzoiu, Ree, Otto and Matzel2013). In humans, the dopaminergic system in the PFC seems also to be closely linked with executive functions and intelligence (McNab et al. Reference McNab, Varrone, Farde, Jucaite, Bystritsky, Forssberg and Klingberg2009; Miller & Cohen Reference Miller and Cohen2001). And whereas the brain of birds differs strikingly from the mammalian brain (e.g., it lacks the 6 layers of lamination in the neocortex), the avian nidopallium caudolaterale (NCL) is remarkably similar to the mammalian PFC. Like the PFC, the NCL is a hub of multimodal integration connecting the higher-order sensory input to limbic and motor structures (Gunturkun & Kroner Reference Gunturkun and Kroner1999) , and dopamine in the avian NCL seems to play a similar functional role in higher cognition as it does in the mammalian PFC (Karakuyu et al. Reference Karakuyu, Herold, Gunturkun and Diekamp2007; Veit et al. Reference Veit, Hartmann and Nieder2014). This confluence of evidence across taxonomic groups (humans, mice, and birds) is compelling, and at least as parsimonious as the descriptions of intelligence based on variations in brain size.
Burkart et al. imply in their current article and state explicitly elsewhere (van Schaik et al. Reference van Schaik, Isler and Burkart2012) that “general intelligence is not a uniquely derived human trait but instead a phylogenetically old phenomenon, found among primates, rodents and birds” (p. 280). However, the PFC and NCL are on opposite ends of the cerebrum and possess distinct genetic expression patterns, leading some to claim that these regions are not homologous but, rather, represent a case of evolutionary convergence (Gunturkun Reference Gunturkun2012). Thus, non-homologous fields converged over the course of 300 million years into mammalian and avian prefrontal areas that generate the same cognitive functions (e.g., working memory capacity; Diekamp et al. Reference Diekamp, Kalt and Gunturkun2002; Matzel et al. Reference Matzel, Sauce-Silva and Wass2013) that contribute to the establishment of general intelligence. In other words, general intelligence could have evolved multiple times in different taxonomic groups. Of course this is a matter of considerable controversy (Karten Reference Karten2015), and the question is far from resolved. Nonetheless, this type of solution is more parsimonious than one based solely on brain size, and mitigates the extant problem of the “cost” of bigger brains. We hope that the “evolution” of interest in the variation in general intelligence follows this route for the next decade.