I stumbled upon Embodied Cognition while doing research for a sound study on music videos. In chapter nine (page 185, to be exact) of Carol Vernallis’s book Experiencing Music Video she, while discussing the use of metaphors in music videos, questions whether our ability to experience phenomenon metaphorically is a function of biology or culture. She goes on to say that two cognitive scientists, George Lakoff and Mark Johnson argue that these modes of thinking are intrinsic to us and are partially a function of genetics (this is acknowledging that culture’s role in constructing metaphorical thinking cannot be ruled out). Upon further research I discovered that the concept of metaphorical thinking being genetic is part of a larger, new theory, Embodied Cognition.
Embodied cognition is a very large and complex theory that I, personally, don’t quite fully understand; but it got me thinking a lot about how I function and the way my brain works, and so I wanted to pass the information along.
Samuel McNerney puts this large concept into one, easy to comprehend sentence – “our cognition is [limited to] our experiences in the physical world”. Until recently, cognitive scientist believed in the theory of dualism, that the brain and body are separate entities and that the body doesn’t have an influence on cognition. The mind was likened to a computing device in which the brain services as hardware that would analyze and send signals to the body, serving as the functional or practical element of these computations.
To explain this a little better, I will use an analogy created by Jeff Thompson:
Let’s say that you’re an outfielder in a baseball game. The ball is hit and is soaring in your direction, and it is up to you to catch the flying ball. The disembodied cognition theory mandates that you predict where the ball will land. In order to make this prediction you will need information about the ball, such as speed and direction, along with a model of the projectile motion of the ball. Your perception gives you the input your brain needs to construct a representational model of the projection. Using this information your brain will then command the body to move to the proper location.
Embodied cognition argues that this is incorrect. It is the theory that “our behavior emerges from the real-time interaction between a nervous system in a body with particular capabilities and an environment that offers opportunities for our behavior and information about those opportunities” (Thompson). This changes the brain’s job from simply receiving knowledge and outputting signals to the body to being a part of system that perception and action equally. First the brain analyses what resources are available and from there deducts if computation is necessary, or if the perceptual information is enough to guide action.
So how does this play out in our baseball game?
Well there are two possible scenarios that can play out in this situation. If you are in direct line with the arc of the ball, you begin to run as to make the ball appear to move at a constant velocity, then you will end up at the proper place. The second scenario is if you aren’t in the direct line of the ball, in which case you will move as to make the curved trajectory of the ball appear linear. This doesn’t involve any predication; instead, you just move in a particular way to solve your own problem.
You will notice that, during a game or practice, baseball players do not stop to think about where the ball might land – they just start moving towards it. This is because as they are running their mind is taking in information from their movement and environment and is adjusting the signals it is sending to the body (in regards to where it should go and how fast it should move).
To better demonstrate how these two types of cognition work, and why it is very obvious that we do in fact have an embodied cognition as opposed to disembodied cognition, take a look at these two robots.
Honda’s ASIMO functions with disembodied cognition. If you watch the video, you will see that ASIMO needs to compute before taking action. Even though ASIMO is very high functioning, it cannot compute fast enough to block a soccer ball from going into the net it’s guarding (a task that humans perform without error very frequently).
Then there is Boston Dynamic’s Big Dog which is robot ‘dog’ that uses embodied cognition. In fact, the robot has very little ‘brain’ at all, and instead produces movement based on the interaction between sensors on its legs and the surface it is on and any forces acting on it. Unlike ASIMO, Big Dog can very quickly readjust. It will not fall over when kicked nor will it fall on ice or get stuck in mud. It can also navigate itself away from and around tricky terrain (ASIMO can do this too, but will get stuck in the time it takes to compute and readjust). You will see that Big Dogs movement look almost identical to animal movement.
I would say that Big Dog is far closer to animal movement (and human movement) than ASIMO. For me, this proves that we do function on an embodied cognition – using our body and the environment to constantly readjust our behavior and movement – opposed to the disembodied cognition that has been thought of as hard fact for a very long time.
Which robot do you think more closely relates to humans?
I challenge you to think about a few of your actionsre that happen throughout the day. Are you noticing yourself calculating your actions before performing them? Or are you constantly readjusting your actions based on stimuli from your environment and body?
If you are interested more in this topic, George Lakoff and Mark Johnson wrote Metaphors We Live By together in 2003, about embodied cognition.