Samskaras in Neural Networks: Built-In But Not the Final Word
defined is that they are like ruts in the brain. Of course, it’s just an analogy, not a physical reality – thank goodness! But, nonetheless, it’s a disheartening image.
As yoga practitioners, we are committed to growth and evolution, so we don’t want to think of our thoughts stagnating. We want to be flexible, not trapped, in our behavior. Getting stagnant is not an inevitable consequence but, as we all know, sometimes choosing positive growth takes effort.
Computational neuroscientists use mathematics to visualize how neurons are networked together to produce a certain cognitive ability, like pattern recognition and storing memories, or a certain behavior, such as eye-hand coordination or navigating across a floor cluttered with obstacles. Such computational models can be used to study many different phenomena and sometimes can be used in place of animal research.
One of the earliest models of a neural network involved Hebbian Learning, a process where neurons that fire together wire together. In other words, the more we repeat a thought or behavior, the stronger the connections grow between the neurons that are responsible for that thought or behavior. The neural pathway is structurally reinforced, the chemical communicators (neurotransmitters, neuropeptides and hormones) used in the pathway are reinforced, and the behavior is reinforced – they’re all tightly linked together and are more probable to co-occur in the future.
Consequently, these following statements are all true… for better or worse…
Activity generated by computer simulations of neural networks sometimes produce a samskara-like pattern. That is, the neurons get stuck in producing the same output even though, by design, they weren’t explicitly programmed to act this way. They perseverate in a repetitive activity; the samskara is an emergent property.
have seen one in a shopping mall or amusement park. You place a coin at the top of the well and, from centripetal force, your quarter circles and circles until it draws into the center and disappears. No matter where you place your coin along the outer rim of the wishing well, the end result will be the same: the coin will circle its way to the center and disappear to the bottom.
This is karma in action. Around and around we go! Next stop… same place as before!
This attractive force is what a behavioral samskara is like: when the starting point is similar, any x-y point that falls along the surface of what computational neuroscientists call the attractor basin, the end result tends to be the same. The end result is getting caught in that attractor state, the valley or rut where all the x-y points eventually take you. It is a sort of limbo state; behavior runs on a loop. It can be a computer- programming inconvenience for computational neuroscientist, but if you’re the person living her life caught in a loop, the stakes are much higher.
This is the tendency of neurons that are linked together and are acting without a sense of higher purpose. But being only a tendency, it can be overcome. It is not an inevitability. Our brains can change. Our behavior can shift. Computer models don’t have strategic thinking, love and aspirations for improvement to help them break out of the attractor state. We do! We can be positively disruptive!