Verified by 
Neuroscience
How Does Your Brain Create Your Conscious Mind?
Discoveries through a study of how we learn.
Posted August 17, 2021 Reviewed by Lybi Ma
We need a unified and accessible explanation of the main mental processes that define the human condition, ranging from perception through cognition and emotion to action, in both health and disease. These mental processes are emergent, or interactive, properties that arise when multiple brain regions interact together. To explain how these emergent processes work, one needs to use neural models that show how brain evolution leads to behavioral success. Discovering and developing such models with over 100 gifted Ph.D. students, postdoctoral fellows, and faculty has been my life’s work for the past 64 years.
This is of particular interest to psychologists because most of our deepest insights were discovered through a study of how we learn, indeed how we learn so quickly throughout life about a changing world without new learning forcing catastrophic forgetting of what we learned before. I call this problem the stability-plasticity dilemma because it asks how our minds can be so plastic, and thereby learn quickly, without sacrificing the stability of already learned memories. To realize how well we have solved this dilemma, we have only to attend a party and learn the faces of a lot of new people within minutes. We do this without worrying that these new memories will force us to forget the faces of people we already know, including our mom’s face. This is an amazing accomplishment of our brains!
This learning problem has fascinated me since I took an introductory psychology course in 1957 when I was a freshman at Dartmouth College. Even in 1957, a lot was known about how people learn lists of events, whether the list is a story, a dance, or a route from home to the supermarket. I began to study these data with a passion until my head exploded with ideas about how list learning may occur. These ideas led me to introduce the field of neural networks, a field that is all over the news today, often about a neural network that is called Deep Learning. But it was back in 1957, as a college freshman, that I begin to explain how our brains make our minds, and to thereby start solving the classical mind-body problem.
At Dartmouth, using simple psychological ideas about learning as my hypotheses, I derived equations for the foundational processes whereby our brains work. These equations for short-term memory, or STM; medium-term memory, or MTM; and long-term memory, or LTM, still form the foundation of every biological model of how brains make minds. LTM is the kind of process whereby we learn and remember things about the world. I used these laws to explain the main psychological data about how humans learn lists.
To explain how learning that solves the stability-plasticity dilemma works, I later showed how learned expectations about what may happen next focus attention upon combinations of critical features in the world, while suppressing features that are irrelevant. These attended critical features are what we learn about and use to control and predict our thoughts and actions.
When an active expectation matches attended critical features well enough, excitatory bottom-up and top-down signals mutually reinforce each other between the features and the active recognition category that is reading out the expectation. The result is a state of resonance in which the matched signals are synchronized, amplified, and prolonged long enough for us to consciously recognize the attended object or event. Remarkably, such a feature-category resonance also solves the stability-plasticity dilemma.
These processes are modeled by the most advanced cognitive and neural theory of how our brains learn to attend, recognize, and predict objects and events in a changing world. I introduced this Adaptive Resonance Theory, or ART, in 1976 and it has been steadily developed to the present time.
In summary, the words “resonant” and “conscious” are important because resonance is needed to learn quickly without experiencing catastrophic forgetting, while also supporting “consciousness.”
With this foundation, I could also show how, where in our brains, and why evolution may have been driven to discover the conscious states of mind whereby our brains consciously see, hear, feel, and know things about the world. In brief, conscious states enable us to effectively act upon the world, whether by seeing to look and reach, hearing to communicate and speak, or feeling to acquire valued goals. Six kinds of adaptive resonances in different parts of our brains support these conscious experiences. When they all occur together, they support life-long learning as we
develop an ever-expanding sense of self.
Our conscious awareness pervades all aspects of our lives. We can use the example of art to illustrate how our brains consciously see. How do visual artists, including Banksy, da Vinci, Matisse, Monet, Seurat, and Stella, achieve the aesthetic effects in their paintings? And how do humans consciously see these paintings? These questions now have scientific answers.
After modeling and explaining lots of data about cognition and cognitive-emotional interactions, I was able to derive them from thought experiments about how we learn. A thought experiment uses simple facts that are familiar to everyone to derive profound insights about how the world works. Einstein introduced several famous thought experiments, one of which, called the elevator thought experiment, helped him to derive the laws of General Relativity Theory. My own first thought experiment was published in Psychological Review in 1980. It analyzed how any system can autonomously learn to correct predictive errors in a changing world.
The hypotheses on which this thought experiment is based are familiar facts that we all know about from daily life. They are familiar because they represent ubiquitous environmental constraints on the evolution of our brains. When a few such constraints act together, they lead to ART.
Nowhere during these thought experiments are the words mind or brain mentioned. Their results are thus a universal class of solutions to the problem of autonomous error correction in a changing world.
The three main words that encapsulate this type of universality are autonomous adaptive intelligence. Specializations of these brain designs will continue to be used during the coming century to design autonomous adaptive intelligent algorithms and mobile robots for engineering, technology, and artificial intelligence.
I also contrast our brains’ natural intelligence with problems of popular machine learning algorithms such as Deep Learning. Deep Learning is unreliable (because it can experience catastrophic forgetting) and untrustworthy (because it is not explainable). Use it at your own risk in any application with life-or-death consequences. Adaptive Resonance Theory overcomes these and other such problems.
Another thought experiment helped me to understand how cognition and emotion interact; that is, how thought and feeling interact, to acquire valued goals. Many facts about cognitive-emotional interactions follow, including how we learn which events are causal and which are accidental.
All processes in Nature can break down. If the cognitive-emotional processes break down in prescribed ways, then behavioral symptoms of mental disorders that afflict millions of people can be explained, including behavioral symptoms of Alzheimer's disease, autism, amnesia, schizophrenia, ADHD, PTSD, and disorders of slow-wave sleep. These explanations may help to discover more effective treatments for these disorders.
The universality of the models leads to biologically-based insights into the human condition. These include biological bases of creativity, morality, and religion, notably why our brains are biased towards the good so that values are not purely relative; why many beliefs and decisions are superstitious, irrational, and self-defeating despite evolution's selection of adaptive behaviors; and why some beliefs persistently resist disconfirming evidence, as many societal problems today tragically illustrate.
Insights about the neuronal cells in our brains generalize to conclusions about all cellular living organisms, neural or not. These shared laws embody a universal developmental code wherein variations of STM and LTM laws, albeit realized by distinct physical and biochemical substrates, occur in all living cellular tissues, from slime molds to humans.
Principles of complementarity, uncertainty, and resonance underlie the laws of theoretical physics, leading me to speculate how brain designs, which also embody principles of complementary, uncertainty, and resonance, reflect the laws of the physical world with which our brains ceaselessly interact. These interactions enable our brains to incrementally learn to understand those laws and to thereby understand the world scientifically.
