Individual orientation toward either systemizing or mentalizing (often measured as "empathizing") can be tested using short questionaires developed by the Autism Research Centre at Cambridge University. Such tests of cognitive orientation do not test for ability in either area; however, a correlation between orientation and ability is widely assumed. Most people are moderately skilled in both systemizing and mentalizing; extreme skills in either domain are rare. Severe deficits in mentalizing are symptomatic of autism spectrum disorders; the consequences of severe deficits in systemizing have received little attention and are not well characterized.

Systemizing reflects an ability to understand how the structures of systems enable the actions of particular forces and how those forces produce particular movements, i.e. particular overt system behaviors. The use of simple tools provides a domain in which to investigate the representation of structures, forces and motions. A wide variety of birds and mammals use tools, so methods of comparative functional neuroanatomy can be brought to bear. Drawing on such comparative data, I have shown that visual and premotor processes involved in tool use can be interpreted as implementations of structure mapping, an algorithm developed in the 1980s to explain human abilities to perform analogical reasoning (see "Implementation of structure mapping by event-file binding and action planning: A model of tool-improvisation analogies"). This result is interesting in part because it challenges a widely-held belief that structure mapping inferences require human-like language abilities. By showing that tool use abilities with a broad phylogenetic range are instances of structure mapping, the way is opened to investigate other areas where analogical inferences may be carried out by neurocognitive systems that do not rely on language. Based on a review of both neurocognitive and behavioral data, including pedagogical and linguistic data, I have hypothesized that even the abstract analogies employed in theoretical reasoning in physics may be implemented by the premotor system in a language-independent way (see "Motion as manipulation: Implementation of motion and force analogies by event-file binding and action planning"); I suspect that the same is true of much mathematical reasoning. This possibility is interesting in part because analogy formation is often involved in the unconscious thinking that results in insight or "Aha!" experiences.

It is a common observation that although people do not always enjoy working on problems, they almost always enjoy solving them. This is especially true of problem solving that involves systemizing - while understanding the intentions that made an agent act in one way or another often induces emotions such as fear or disgust, understanding how a mechanical system works nearly always induces at least mild pleasure. The association of positive affect with successful systemizing suggests that affect-driven feedback loops may be significantly involved in the development of both a cognitive orientation towards systemizing and systemizing ability. In "From "Oh, OK" to "Ah, yes" to "Aha!": Hyper-systemizing and the rewards of insight" I develop a neurofunctional model of the implementation of such an affective feedback loop, and show that suppression of activity in the default network of the brain, which is highly biased in most people towards mentalizing, is a natural outcome of the model. Interestingly, deactivation of the default network is also a common neurofunctional correlate of meditation, an activity that, like systemizing, requires intense attentional focus on objects or phenomena other than one's own beliefs, desires or other intentional or emotional states.

My current work on the system-brain project addresses two complementary questions. The first concerns the role of imagination, especially visualization, in systemizing and the structure and evolution of the neural mechanisms that enable switching of attention between perceptions and imaginations. Accurately distinguishing perceptions from imaginations, sometimes called "reality monitoring", is essential for maintaining appropriate overt, perceptually-guided behavior. One would, therefore, expect systemizing abilities that require sustained attention to imaginations to co-evolve with systems capable of detecting situations in which attention to the external world can be relaxed and control mechanisms capable of rapidly re-orienting attention from imagination to perception when the external situation becomes threatening or otherwise critically informative. Such systems are partially characterized in humans but as yet uncharacterized in any other animals.

The second question concerns the co-evolution of systemizing and mentalizing, selective pressures favoring one or the other cognitive style and hence neurofunctional network, and developmental and experiential triggers that shift problem solving between systemizing and mentalizing methods. Both uncertainty and emotions such as fear are known to increase the likelihood of mentalizing. Activation of the mirror neuron system, a neurofunctional network that enables imitation, may also be an indicator of agency and hence trigger mentalizing; however, the mirror system is also known to be involved in the implementation of systemizing. Although infants less than one year old appear to regard some sequences of motions as indicative of mechanistic as opposed to intentional causation, specific triggers of systemizing and their neurofunctional implementation remain poorly characterized.