Logic Is Far From Enough

Neural Nets and Philosophy.

Don’t get me wrong, logic is one of our best tools of discerning and understanding truth. Aside from being a software developer dependent on it, I’m pretty good at it too, but I can explain a number of reasons to go beyond reason or it actually goes before reason.

This essay is meant to push boundaries, challenge a lot of conventional thinking, and bring you where you have not been. More than to create understandings, as is the usual intent of my writings, this is meant to illuminate a new area to explore for your own understandings. It ranges widely, but then what a human is and how they think is very broad. This is meant to be advanced but then again it is about thought processes that were primary before philosophy. This may not seem so accessible but all of this is quite common, perhaps more so than logic and reason. It is just that we rarely notice it.

Basically, the point of this is that for all the power of the logic and reason of critical thinking that is the signature of philosophy, our best thoughts originate in the pattern recognition functionality of the neural net of the brain that does not use logic that is accessible to us. Not only that but our neural nets (yes plural) have problem-solving capability that is basically never considered and may not be exercised much. If you know of them though, they are easy enough to use and quite useful for problem-solving and for providing understanding.

Some premises are in order… (1) There are schools of philosophical thought that claim that truth can only be achieved through logic and reason because the senses can be fooled. I don’t follow that premise for a few reasons including that as a software developer I know how quickly the logic of any sophisticated problem becomes unmanageably complicated. Also, when wearing my evolutionary biologist hat, I’m acutely aware of this huge element of chance that is part of evolution and that will frustrate useful logical analysis. (2) A common definition of philosophy is “how to live a good life”. That is a reference to living a moral life. Of course what moral means is another issue that may need to be touched upon. (3) I believe that a primary purpose of philosophy is to give understanding. That is what the human mind is for and philosophy is a tool of the mind.

If this is about neural nets then we need to start with a good description of neural nets. The most obvious one is the human mind, composed of “white and gray matter”. I hope that explained a lot… Maybe not. That is one problem. It’s almost impossible to describe a neural net usefully, though recent work on artificial neural nets does give us some more information. What gives us a lot more information though is there is one neural net that is fairly easy to describe and measure. That is our vision. The term visual cortex describes a fairly discrete neural net, if there is such a thing in neurology. Most of this commentary is based upon what I learned during and after a lecture I watched decades ago that was basically the celebratory lecture tour of the researchers who did this work and had just completed the long process that admits what started as a theory into the body of knowledge that is known as science. They wanted to know how vision worked. Of course, there were theories, and work that had been done, but they took another path. What they did was to show slides with shapes on them to people, for measured amounts of time, mostly fractions of seconds. Doing so allowed them to record the timing and sequence of the operation of vision. When a slide with an “X” on it is shown to the viewer for .15 seconds, they know that they saw the “X”, but don’t know its location. Its location is known after about .3 seconds. Where it is in relation to perhaps an “O” is known at another very measurable time period. The point they learned though that is needed for this essay is that after the parts of something are identified and located visually, there is a critical point of recognition. The outline must be recognized and a gestalt must be generated. That is the pattern recognition of the visual neural net. A person recognizes what they have seen. That sequence allows a more general description of how all neural nets work, parts lead to recognition of the whole. A study of the neurology shows that those elements like the “X” and “O” are recognized by individual neurons (pretty much). Stay with me. It gets interesting soon.

So that is how vision works. Discrete elements of an image are visually detected. Then their location is determined, then an outline is formed to make an image. At some point, the pattern is recognized or not. You can reset the system by blinking, as when you are traveling in fog or darkness, trying to recognize what you can barely see. Your vision is biased by genetics and learning. You can prime the system to look for something. Most traffic on the optic nerve is actually from the brain to the eyes (80%, your eyes contain “pre-processor” neural nets). If the neural net cannot identify a pattern, it often creates one. If a hunter is looking for a bird, they are likely to see many birds that are not really there but some clue triggers a recognition of the pattern of a bird. No accessible logic is applied to this process.

There are a few useful points to this. It can better describe the neural net we think of as the mind and describe something about how to use it. It can also lead to the knowledge of other neural nets humans have available for problem solving and understanding, that few know about but that can be very useful.

In my book “When Barbara Explained Genius” is a long explanation of how the neural net of intelligence works and how to use it. It is very like vision in that it detects parts and identifies a pattern that can lead to understanding. Instead of sight though, intelligence produces understandings that are insights. A few points are relevant to mention. Intelligence evolved before cultural tools like language or logic so there can be a challenge to convert understanding into something that can be retained in memory let alone communicated. You may need to have an insight repeatedly to retain it. Insights build with repetition. While humans need to be trained to use logic consciously, the “mind” can trigger sophisticated logical processes to evaluate elements of an insight. You can retrigger insights like you can retrigger your vision by blinking… if you can. It goes on. One interesting point is the question of how to increase your intelligence. AI research discusses “filters”. Add filters together and you get more accurate AI intelligence. AIs create filters when forced to find a new way to solve a problem that they have already solved. How can humans create filters? By making mistakes. That is how intelligence can be increased. Another cool point about insight is that it is the ultimate drug. I think every seeker of knowledge knows the ecstasy of insight.

Okay, here we are now, at the rabbit hole. Neural nets are a tough problem, partly because they are not evolved to manage language but also because they do not work based upon our main conscious tools of understanding which are logic and reason. They are extremely subtle. Are there though other neural nets available to us for problem-solving? There are and if you know about them, you can use them. You already do use them but probably don’t know it. They can be remarkably powerful for problem-solving and producing understandings. As is characteristic of neural nets though, they require training. Neural nets do interact some naturally but the person trained to use them can use outputs from one as input data for another. While education and reading develop the neural nets, generally assume that “trained” means “self-trained”.

Let’s start with a fairly simple neural net that you may recognize. That is your speech center, another neural net but one that can understand and manage language. Here too you can see how trainable it is. Language is a huge thing. You can learn great eloquence and even power with it. It is the cultural tool that is the basic vehicle for manipulating the logic and reason of critical thinking. It can even become a problem if you develop your spoken or written skills enough that most others cannot follow them. A good reader does not read a word or even a phrase at a time, They recognize whole sentences at a time. If you don’t use it you won’t develop it and by the way, due to the wonderful plasticity of the neurology of the human mind, other intellectual functions will poach the neurons used for speech for their own work. This poaching is to some degree common to all neurological processes. (One theory holds that the visual component of dreams is to prevent this poaching from happening to your vision centers.) Here’s the fascinating part. Sure, language ability is many things and offers many potentials including using the cultural tools of logic and even Maths, but have you noticed a peculiarity? After an insight becomes understanding, it often continues to develop. Have you ever noticed that when your understanding of something really peaks you notice a poetic characteristic to it? It has a form and balance. The loose ends are now part of the whole. Michael Polanyi talked about this in Tacit Knowing. It’s hard to describe but the point of this is that when you really understand something with your neural net, it is given a natural shape. When it reaches that shape, that integrity, then it has a beauty and poetic form that makes you feel that it is truth. (It also may illustrate a bias and limitation of the nature of human comprehension, AI research suggests that but that’s way beyond this scope.)

Now for the good stuff. There are two other neural nets available that are very useful and interesting. They have also given fits to philosophers through the ages because they are so important to human thought and yet often don’t conform to the logic and reason that are the main tools philosophers use. These are older. This is in the realm of instinct. These are your moral and survival instincts. Both are powerful, trainable, deep, can solve problems and both can lead to understandings. These are what animals use mostly and were what humans almost exclusively used before language was available. They work. They don’t think the way “you” do and can give you novel answers for your conscious or reasoning mind to examine and use. Also, they are adapted to a different time and circumstances so it’s best to know about them so that you can avoid when they will trip you up. They are very very “Darwinian” and so often “think” in terms of blind competition. This is meant less to describe those instincts, which would take reams of words, as much as it is to point you to them so that you can recognize them in yourselves and then in others. It is especially to show their importance so that you will explore them and hopefully so that you might train them to be valuable tools of problem solving and understanding. We live in an easy time where we don’t use and develop them like we used to.

Morality is how you decide right and wrong. Morality is the combination of instinct and learned knowledge. Humans are programmed by nature (moral instinct) to seek out moral systems, use them, not question them… and to defend it to the death. To a biologist, a moral system is a survival strategy. To religions and philosophy they are constructs, often based on reason. Good luck with that. To nature, they are organic rules created by the trial and error of evolution, expressed as behavioral balances. Instincts are the outcomes of neural nets. A bird does not think “last night was cold, I should migrate”. Instead, many signals from the environment are recognized, combined, and evaluated over time until the neural net recognizes a pattern. Then a balance is shifted and the migration behavior starts. Instincts shift the balance between behaviors. In that moral behaviors have a large learned component, the moral instincts behind them may be obscure. Whenever survival is at play though, they come to attention. It is almost impossible to do something your moral instincts consider bad and are set against. You may very well even avoid doing something that your moral instincts just aren’t sure of. Again, they are trainable. Their importance and utility are shown in that traditionally the moral training of religions are lifelong processes. Do not misunderstand religion. Its greatest power has come from its function of husbanding and teaching morality… or used to. Pretty much every institution from Elks club to high school band to corporations to dance or martial arts have strong moral standards and often teachings. It is that important. Morality is all-pervasive.

Some interesting points come from this. A neural net is a trainable pattern recognition mechanism, with evolutionary biases built-in. A neural net can and does solve problems without the conscious mind. I had neighbors with very strong religious training. I knew that I could ask them a moral question and that it would be processed by their “moral neural net”. They could give me an answer they didn’t fully understand in response to a question they didn’t fully understand. It could be useful when trying to solve problems of how people think in survival terms. Another fascinating point is that you can talk to your moral and survival instincts. They aren’t from a time that would allow them to talk back, but they will listen. Your instincts are evolved for another time and circumstance. They may not agree with what your conscious mind knows and decides. They may strongly resist. This is likely to be in cases effecting reproduction or conflict. Your instincts may say “fight or run” and your conscious mind may say, “let’s not”. Your conscious mind may ask for patience to use another strategy that isn’t really natural to your instincts. In those cases, actually talking to your instincts may lead them to quit arguing. Again though, the neural net evolved for a particular purpose can answer questions on unrelated topics because all it is doing is recognizing patterns. It can be a useful tool and is especially well tuned to recognizing dangers that your conscious mind might never notice.

The most interesting and most important instinct or neural net to know about though is your survival instinct. It offers so much. It’s not as smart as your moral instincts but just knowing about it explains so many things and offers great understanding. Not surprisingly, more than understanding though, it can offer great strength. You have to figure that your survival instinct is going to be pretty basic. You just have to know about it, but it has been hidden from you. This is partly because it is quiet but also because in history the Catholic Church taught that humans have no instincts because that would mean we are animals rather than divine. That belief still exists as an important part of Western culture. You are unlikely to even notice it though because it is just part of the cultural background. You need to know the name of your survival instinct to even recognize it. You need to know about it to take the advantage of it and it is your greatest strength. It is like an emotion that can grow slowly or flower suddenly. You need to learn to recognize it in yourself and then you can recognize it in others. Its effect is so important and pervasive. It is one of the main answers or perhaps the main answer to the question “why”. The problem is that it is culturally hidden by hiding its name. The name given to Human Survival Instinct is “Faith” and religions have claimed it is an unsupported belief in a God but more importantly it is really an unsupported belief in oneself and the value of your own survival. Like other instincts, your survival instinct can be trained and strengthened by exercise, so maybe in the past it worked better as a part of religion focused around a God. Now, religion seems to lead to too many internal conflicts so we need to recognize faith as a belief in the importance of ourselves, our society, our civilization and humanity, the parts of survival. Morality is about how we choose what is right or wrong. Faith, our survival instinct, is why we make that choice. Why do you strive to do what is right? It is your instinctive struggle to survive, driven by an instinct. That survival instinct is known as faith. Look for faith in yourself and then you will see it in others. Look at what they do and you will recognize when the choices made were because of a decision to do the right thing. You can better see the beauty and rightness of what you and they have done. It too can solve problems by pattern recognition. It saved your ancestor's lives innumerable times by recognizing danger and opportunity. It is human survival instinct that drives the moral decisions that make the “good life” described as the goal of philosophy.