The Word from Neuroscience

From the 1990s onward, with functional magnetic resonance imaging (fMRI), it became possible to measure the areas of the brain that are activated when we carry out various activities, thanks to the fact that magnetic resonance is sensitive to changes in blood flow and in its oxygenation inside the brain¹. Very recent studies in neuroscience have thus succeeded in determining the sectors of the brain called upon during work with computer code.

The results of these studies indicate that when a human “codes” – in an artistic framework or not – his brain activity is not comparable to that associated with natural language; on the contrary, it is the so-called multiple demand areas of the brain that are activated, as when the brain tackles mathematics, logical problems and/or executive control tasks². These areas are called upon bilaterally during coding³, marking yet another difference from the use of natural language, which mobilizes mainly the left hemisphere⁴.

Thus, even if computer code and graphical interfaces make use of words drawn from human language, with meanings that sometimes resemble those activated when these words are mobilized by natural language, human-machine interaction plays out, from the very start of the process, on a strictly logical, operational plane – as when doing a jigsaw puzzle, or else when tinkering (executive process).

Certain studies suggest that for the brain, thinking and acting “amount to the same thing⁵”, which reveals that the brain network that manages cognitive activities – such as calculation and the solving of spatio-temporal problems – was structured during the evolution of the species upon the brain network that manages motor function⁶. On the machine side, the link between words and operations is infinitely more direct, the symbolic level of code being anchored point by point, in a one-to-one manner, on a gestural base, on gestures schematized and crystallized in the very structure of the machine.

Having attempted to clarify the nature of the interaction that arises between the artist and the algorithm, which should be apprehended less as a linguistic dynamic than as an operative process dominated by gesture, we will now try to grasp what this artistic gesture, which must manifest itself and without which the work will not come into being, may consist of.

We will be led to consider, on the one hand, an artistic gesture that is expressed within digital creative processes, but outside the algorithmic core: it will concern the analog periphery and partially the protocol zone. This gesture will be associated with a sudden temporality, with a form of discharge that seems to mark, for an instant and through the different layers of mediation, a direct contact with the matter of the world. On the other hand, there will be the question of an algorithmic gesture, which manifests itself within the algorithmic core and unfolds in a discontinuous and indirect way, admitting delays and latency phenomena, for it must reconcile its temporality with that of the machine.

1. K K Kwong J W Belliveau D A Chesler et al., «Dynamic magnetic resonance imaging of human brain activity during primary sensory stimulation. ». Proceedings of the National Academy of Sciences [en ligne]. June 1992, vol. 89, n° 12, p. 5675–5679.↩︎

2. Anna A Ivanova Shashank Srikant Yotaro Sueoka et al., «Comprehension of computer code relies primarily on domain-general executive brain regions ». eLife [en ligne]. December 2020, vol. 9, p. e58906.↩︎

3. Ibid.↩︎

4. One study largely confirms the results found by Ivanova et al. while suggesting a certain degree of lateralization, varying between different subjects, during code reading: Yun-Fei Liu Judy Kim Colin Wilson et al., «Computer code comprehension shares neural resources with formal logical inference in the fronto-parietal network ». eLife [en ligne]. December 2020, vol. 9, p. e59340.↩︎

5. Radek Ptak Armin Schnider and Julia Fellrath, «The Dorsal Frontoparietal Network: A Core System for Emulated Action ». Trends in Cognitive Sciences [en ligne]. August 2017, vol. 21, n° 8, p. 589–599.↩︎

6. Ibid.↩︎