Tools from the past in the modern brain (Commentary on Stout et al.)


The human capacity for using and generating tools, from spoons to cars and computers, is far greater than that of any other species. Neuropsychological and neuroimaging research points to specific regions of the human brain which encode knowledge about tool use (Johnson-Frey, 2004). While many of these studies discuss possible evolutionary changes which might permit an explosion of tool use in the ancestors of modern humans, far fewer have attempted to examine the potential brain systems involved.

A paper in this issue of EJN adopts an expertise approach to this complex problem. The study by Stout et al. (2011) focuses on the toolmaking transition from the Oldowan method (2.5 million years ago) to the more advanced Acheulean method (0.5 million years ago). In both cases, the toolmaker shapes a core stone to use as a tool, but the methods differ in the complexity of the action planning and sequencing. In the Oldowan method, the toolmaker performs repeated targeted strikes of the core, each aiming to bring the tool shape closer to the desired shape. In the Acheulean method, the toolmaker also sometimes turns the core over and prepares the edge with small strikes before removing a larger flake from the initial surface. Thus, the Acheulean method involves a planned hierarchically structured sequence of actions, unlike the Oldowan method.

In Stout’s study, participants who were either naïve to stone-toolmaking, or trained to a moderate level of expertise, or highly proficient, watched movies of both types of toolmaking during fMRI scanning. While it is almost impossible to precisely control the components and timing of action in these naturalistic movies, the comparison of different types of tool use provides some insights into brain systems for understanding hierarchical actions and for tool-use expertise. The results show that observation of the more complex Acheulean toolmaking resulted in greater engagement of the action observation network (Grafton & Hamilton, 2007). One possible interpretation is that these regions have a specific role in processing the more complex hierarchical structure embedded in the Acheulean action sequences. A more mundane possibility is that the greater variety of actions in the Acheulean sequences leads to less repetition suppression and thus greater signal in regions encoding the individual action components. These two interpretations highlight the difficulty in finding ecologically valid ways to examine brain systems processing hierarchically structured actions.

Furthermore, participants who had training in stone toolmaking showed greater engagement of premotor regions when watching the movies. This is consistent with previous studies of expertise acquisition, in which premotor cortex is engaged when watching trained dance sequences (Cross et al., 2006). Curiously, the highly expert participants did not show premotor engagement, but there was a switch from left aIPS in naïve participants to right aIPS in experts which is consistent with the idea that right parietal cortex encodes more complex sequences than the equivalent region on the left (Grafton & Hamilton, 2007).

Overall, Stout’s study provides a new way to think about the human capacity for understanding and performing structured toolmaking actions, in relation to the evolution of these abilities millions of years ago. Further study of the comprehension and production of hierarchical action sequences will be crucial in understanding the evolutionary changes that enabled modern toolmaking sophistication.