When I first read the famous paper “The Extended Mind” by Andy Clark and David Chalmers several years ago, I was quite fascinated by it. It describes some simple thought experiments to cleverly cast doubt on our common intuition that our cognitive processes take place only within our heads. Instead, it suggests that our mental processes can extend beyond our brains and into the environment. As this paper has had a central role in the development of my motivation to start this website, I’d like to discuss the basics of the Hypothesis of the Extended Mind in this first post.
Clark and Chalmers start their paper with a thought experiment. In this experiment the reader is asked to consider three cases of problem solving and to think about how much cognition is present in each case:
- A person sits in front of a computer screen which displays geometrical shapes and is asked whether the shapes fit in the displayed ‘sockets’. The person has to answer by rotating the shapes in his mind.
- The same situation as in (a), but now the person can choose either to mentally rotate the shape or physically rotate it by pressing a button, the latter having some speed advantage.
- The same situation in a possible future, where the person can choose between using his neural implant that does the rotation operation as fast as the computer does in (b), or using ‘old-fashioned’ mental rotation.
We are tempted to say that in case (2) less cognition is present than in (1) or (3); we have a strong intuition that our mental processes take place within our heads. Clark & Chalmers state that these three cases are actually similar, since there isn’t any real fundamental difference between case (2) and (3), other than the boundary of skin and skull.
Now, this thought experiment is actually not as abstract as it might seem, but describes playing the video game Tetris. The experiment is based on the findings of researchers Kirsh and Maglio. In their paper about cognitive performances while playing Tetris, Kirsh and Maglio found that the physical rotation in (b) is actually much faster than the mental rotation. They also found that players were not only physically rotating the shapes to fit the slot; the players were also trying to determine whether the shape fits in the slot. Kirsh and Maglio call this an epistemic action: an action to “alter the world so as to aid and augment cognitive processes such as recognition and search.” This is opposed to pragmatic actions, in which the environment is altered to come physically closer to a goal state.
It is this example of the human capacity to solve problems by using the environment, that Clark and Chalmers employ to introduce the Parity Principle:
“if a part of the world functions as a process which, were it done in the head, we would have no hesitation in recognizing as part of the cognitive process, then that part of the world is (so we claim) part of the cognitive process.” 1
To make it plausible that not only our cognitive processes, but also true mental events can extend in the environment, Clark and Chalmers present a thought experiment involving the fictional characters Otto and Inga, who are remembering how to get to the museum. Otto has Alzheimer’s disease and uses a paper notebook he always carries with him to serve the function of his memory, while Inga’s biological memory is functioning properly. Inga is thought to have a belief about the location of the museum, before she recalls this from her internal, biological memory. In the same manner, Clark and Chalmers argue, Otto can be said to have a belief about the location of the museum before he actually consults his notebook. According to the parity principle, under the right circumstances, the notebook can be seen as an extension of Otto’s memory. By showing how beliefs are not bound by the borders of the body, Clark and Chalmers show that true mental events can extend in the environment as well.
But if skin and skull do not form boundaries for the human mind, how should we then decide where our minds end? Clark and Chalmers suggest three criteria to decide whether a part of the world should (temporarily) be seen as part of the cognitive process:
- The information retrieved from the external source should be directly endorsed and trusted;
- The technological aid should always be available when needed;
- The external resource should be directly available without difficulty.
Now, these criteria give a good starting point and show why certain parts of the environment should not be seen as part of the cognitive process. The information in the dictionary in your bookcase for example is not always available when you need it. Moreover, looking up a lemma can be quite time consuming and difficult compared to our biological resources. The information is not readily and easily available and should therefore not be considered part of you own knowledge. 2
An example of a technological aid that does meet these criteria, is pen and paper for doing long multiplications. To calculate the product of two numbers, we use an algorithm that divides the process of multiplying arbitrary large numbers into very simple steps. By writing down figures in certain locations, we use the pen to manipulate the external memory source: the paper. While we are solving the problem, the information on the paper is directly and constantly available. As the writing utensils play a crucial role in this cognitive process, according to the parity principle, they should be considered as part of this process.
These examples of cognitive extension, and others that Clark gives in his books and papers, are not the typical futuristic technologies that come to mind when thinking about humans merging with technology. Although the possibilities of Brain-Machine Interfaces and neural implants offer very exciting new ways of communicating with technology, this direct interaction with brains is by no means necessary to become part of the extended mind. In fact, most of Clark’s examples are about technologies that have become ubiquitous and invisible, such as pen and paper.
In future posts I will discuss whether the digital information sources we interact with everyday could also be potential mind extensions.