The scientific method, intelligent behavior and how the brain investigates how the world works (Part 4): The animal brain is living proof that scientific models are the most powerful tools to finding out why impossibly complex stuff comes about.

  The following is part of Chapter IV of "A Scientific Model Of The Brain: From Instinct to Reason, How Does The Mind Work". 

TABLE OF CONTENTS

1. The scientific method and how intelligent behavior is brought about in the brain.

1.1. The scientific method is - as a matter of fact - a very basic, all-purpose investigation method and, as such - contrary to general belief - not only is its application domain not limited to the natural sciences, but it is actually especially well-suited for impossibly complex, inscrutable questions such as the origin of Civilization, the State, power, love, intelligence, consciousness or human nature.

1.2. Predictions are the key to intelligent behavior: If we can predict what is going to happen, all there is to intelligent choices and behaviors is to choose the plan of action, predicted to yield the sought-after stuff.

1.3. The scientific method best describes how intelligent behavior is brought about in the brain: The strategy is to learn a model which replicates the investigated thing's behavior. Thus, as we try to replicate how things work, or simply experiment how to bring certain stuff about, we, little by little, discover what are the causes, reasons and intervening factors leading to stuff, and, as we little by little, find out why stuff comes about, we come to learn how things work and stuff comes about.

1.4. If our model replicates the investigated thing's behavior, we only need to run it from any given situation, in order to anticipate how the actual investigated thing will behave.

2. Physical things and abstract concepts

2.1 The brain's genius strategy to investigate how the world works, consists in building a model for each of the countless functionally distinct things in this world. Each of these individual models represent each of the concepts we form in our minds of said functionally distinct things.

2.2. A good definition of a physical thing will be made up of a detailed description, by which to recognize the thing, as well as a clear explanation of how it works. In contrast, when it comes to abstract concepts, since the point is not to recognize them, what the definition actually needs to explain is how the thing comes about. For instance, more than what is intelligence, we want to know how intelligent behavior comes about.

2.3. The brain learns models of the things in this world, because - once it knows how things work and what leads to stuff - all there is to intelligent behavior is to choose the plan of action predicted to yield the sought-after stuff.

3. The animal brain is living proof that scientific models are the most powerful tools to finding out why impossibly complex stuff comes about.

3.1. Especially when it comes to impossibly complex, inscrutable problems, rather than asking how the thing works, we will achieve better results if we investigate, what leads to the stuff we sense. If we find out what leads to certain stuff, we will basically know everything there is to it: namely, we will be able to predict when it is going to come about or we will know how to bring it about ourselves.

3.2. the scientific method is nothing but a formidably versatile and powerful investigation scheme, a set of steps and rules, that we can follow to find out what ingredients, factors and rules lead to certain stuff. It can be applied not only to research how things work or certain stuff comes about, but - fascinatingly enough - also to investigate how to attain some desirable stuff. Indeed, As we try to replicate how things work, or simply experiment how to bring certain stuff about, we, little by little, discover what are the causes, reasons and intervening factors leading to stuff, and, as we little by little, find out why stuff comes about, we come to learn how things work, stuff comes about or, plain simply, how to attain some desired stuff.

3.3. if the model replicates the investigated phenomenon's output, then - for all intends and purposes - the causes, reasons and intervening factors for the investigated phenomenon's output should be the same as the sources employed by the model to produce such same output. Indeed, it really does not matter if the model's underpinnings are actually not exactly the same as the investigated phenomenon's; so long the model replicates the phenomenon's output, we will only need to run the model, in order to anticipate which plan of action will yield the best outcome. 

3.4. When it comes to physical things, since there is something material, that we can observe through the senses, it is possible to play and experiment with the thing. Thus, - as compared to abstract concepts - the scientific investigation is far more straightforward, since we can investigate how the thing behaves in different scenarios.

3.5. A scientific model is particularly helpful to investigate abstract concepts, since it enables us to test, whether or not the hypothesized reasons reproduce the observed data.

FULL TEXT

1.1. The scientific method is - as a matter of fact - a very basic, all-purpose investigation method and, as such - contrary to general belief - not only is its application domain not limited to the natural sciences, but it is actually especially well-suited for impossibly complex, inscrutable questions such as the origin of Civilization, the State, power, love, intelligence, consciousness or human nature.

 See Part 1.

1.2. Predictions are the key to intelligent behavior: If we can predict what is going to happen, all there is to intelligent behavior is to choose the plan of action, predicted to yield the sought-after stuff.

See Part 2.

2. Physical things and abstract concepts

See Part 3.

3. The animal brain is living proof that scientific models are the most powerful tools to finding out why impossibly complex stuff comes about. 

3.1. Especially when it comes to impossibly complex, inscrutable problems, rather than asking how the thing works, we will achieve better results if we investigate, what leads to the stuff we sense. If we find out what leads to certain stuff, we will basically know everything there is to it: namely, we will be able to predict when it is going to come about or we will know how to bring it about ourselves.

The example of intelligence shows that we come up with abstract concepts, but generally have very little clue or insight of how they work, we only "sens" something. Clearly, when it comes to non-physical things, rather than to as how the thing works, we will achieve better results if we investigate, how it is that the stuff we sense comes about. For instance, rather than asking what is intelligence or how does intelligence work, - to the extent that we do not sense anything such as "intelligence" (whatever that means), but something more like "intelligent behavior" - we will get much better insight if we investigate how does intelligent behavior come about or, better yet, what leads to intelligent behavior? Similarly, rather than what is consciousness, we should consider what leads to our sense of consciousness (or, simply, what leads to conscious thought or our train of thoughts); rather than what is human nature, we will get better insight if we investigate what leads to good, selfless behavior and what leads to bad, selfish behavior; finally, rather than what is a French, a German, a Hispanic, an Indian, a Chinese, etc., we should focus our investigation on what leads to behaving like a French, a German, a Hispanic, an Indian, a Chinese, etc.. Yes, what we should really care about are the causes, reasons and intervening factors leading to whatever stuff is relevant to us. If we understand what are the ingredients and intervening factors for something, we basically know everything there is to it. Indeed, we will know how to bring it about or predict how is it going to come about. After all, the causes for something represent nothing but its sheer essence. For instance, if we understand what leads a human being to behaving like a French, we will know if French people are inherently different and represent an essentially distinct nature from other folks, or it is just that they go through different experiences. Now, if the former turns out to be the case, we will come to realize that it would have only been a waste of resources to build a specific model to understand French people; but a global model of human nature is a far more efficient and effective strategy. 

        

3.2. the scientific method is nothing but a formidably versatile and powerful investigation scheme, a set of steps and rules, that we can follow to find out what ingredients, factors and rules lead to certain stuff. It can be applied not only to research how things work or certain stuff comes about, but - fascinatingly enough - also to investigate how to attain some desirable stuff. Indeed, As we try to replicate how things work, or simply experiment how to bring certain stuff about, we, little by little, discover what are the causes, reasons and intervening factors leading to stuff, and, as we little by little, find out why stuff comes about, we come to learn how things work, stuff comes about or, plain simply, how to attain some desired stuff.

Now, crucially, if we have ever wanted to discover the reasons for stuff, that is exactly what the scientific method is all about. Indeed, the scientific method is nothing but a formidably versatile and powerful investigation scheme, a set of steps and rules, that we can follow to find out what ingredients, factors and rules lead to certain stuff. It can be applied not only to research how things work or certain stuff comes about, but - fascinatingly enough - also to investigate how to attain some desirable stuff. As discussed earlier, the key element of this general investigation method is the model. The basic, genius idea is to build some sort of physical or virtual artifact (the model), which seeks to replicate the data or simply generate some desired stuff. Building such a model is certainly easier said than done; but, assuming we manage to put together an artifact, which faithfully replicates the observed data or simply generates the sought-after stuff, the question of how the investigated thing works or how such stuff comes about, at last, receives a straightforward answer: namely, if the model's output is the same as the investigated phenomenon's output, then the investigated phenomenon's underpinnings - for all intends and purposes - should be the same as the model's underpinnings. In other words, the causes, reasons and intervening factors for the investigated phenomenon's output - for all intends and purposes - should be the same as the sources employed by the model to produce such same output. Consequently, as stated before, as we try to replicate how things work, or simply experiment how to bring certain stuff about, we, little by little, find out why stuff comes about, and, as we little by little, discover what are the causes, reasons and all sort of intervening factors leading to stuff, we come to learn how things work, stuff comes about or, plain simply, how to attain some desired stuff.

3.3. if the model replicates the investigated phenomenon's output, then - for all intends and purposes - the causes, reasons and intervening factors for the investigated phenomenon's output should be the same as the sources employed by the model to produce such same output. Indeed, it really does not matter if the model's underpinnings are actually not exactly the same as the investigated phenomenon's; so long the model replicates the phenomenon's output, we will only need to run the model, in order to anticipate which plan of action will yield the best outcome. 

It is thus easy to see why the model is such a formidably powerful device: Since the model works like the investigated phenomenon, if we want to predict what will be the phenomenon's behavior (or, simply, what kind of stuff is going to come about), not unlike weather models are employed to forecast the weather, all what we need to do is run the model from the same starting point, and observe what output it produces. In other words, it really does not matter if the model's underpinnings are actually not exactly the same as the investigated phenomenon's; so long the model replicates the phenomenon's behavior, we will be able to draw predictions and therefore use it to our benefit or, at least, anticipate and prepare for what is about to happen. Fact of the matter is, since it is impossible to demonstrate that something will always be correct, we will never be able to achieve absolute certainty on what are the exact causes, reasons and intervening factors for the investigated phenomenon's behavior. Consequently, so long the model's predictions are correct, - for all intends and purposes - the model's underpinnings are the same as the phenomenon's underpinnings. This is obviously not meant to contradict, however, that the day that one of the model's predictions turns out to be incorrect, it will be only intelligent to learn from the mistake by adjusting the model accordingly.     

3.4. When it comes to physical things, since there is something material, that we can observe through the senses, it is possible to play and experiment with the thing. Thus, - as compared to abstract concepts - the scientific investigation is far more straightforward, since we can investigate how the thing behaves in different scenarios.

Thus, considering what kind of powerful tool the model represents, it likewise becomes evident why it is far less unfathomable to investigate physical things. Since there is something material, that we can observe through the senses, it is possible to play and experiment with the thing. Now, we will not be able to mess around with the thing's inner-workings - which is the main benefit of a model -; but, at least, we can investigate how the thing behaves in different scenarios. Moreover, should we feel the need of building a model, our observations of the thing's behaviors will make this task far less impenetrable and daunting than, for example, a model of love, power or conscious thought.     

Medical research is an excellent example of this kind of mode of investigation: namely, as we explore the causes for disease and the corresponding therapies of cure, we, little by little, learn how the human body works. For obvious reasons medical doctors cannot just go about freely messing around with internal organs and body tissues, but careful observation already provides very helpful insights. As a matter of fact, many medical breakthroughs nowadays do not require any doctor to inspect any human body, but result of a cold statistical analysis. For instance, if heavy tobacco smokers exhibited alarming high rates of lung cancer, it did not take the brightest light bulb in the room to guess that tobacco smoking may be a cause or, at least, a significant factor in lung cancer. Now, as much as this may feel a very strong hypothesis, before we accept it, the scientific method instructs us to put it to the test. If we had a model of the human body, we would just put those artificial bodies to smoke like chimineys. If the hypothesis is correct, we prdict a significant number of them will develop lung cancer. However, such a model of the human body is not available and it would be only abhorrent to run an experiment of that sort on real human beings; thus instead we just select another sample of people and observe if they likewise confirm the prediction: namely, do we find in the new greater sample the same high correlation between heavy smoking and lung cancer?

 Another rather trickier example is the question of human nature. Again, instead of asking what is human nature, we will get better insights, if we consider what leads to good behavior and what leads to bad behavior. Now, needless to say, human nature is the kind of impossibly complex concept that it is generally believed to be beyond the application domain of the scientific method; to begin with, we would first have to reach an agreement on what constitutes good behavior and what constitutes bad behavior. Still, at least within the context of children parenting, it does not seem that unrealistic to conduct a scientific investigation on what parenting practices are more successful at raising respectful and considered individuals. Definitely, nowadays the most learned in the society insist children should at all times be wrapped in love, hugs and kisses in order to flourish. According to these modern parenting theories, when children feel loved, happy and safe, their brains are more open to absorbing new information, solving problems and developing critical thinking skills. However, an alternative, more traditional view on this question would argue that a child is not likely to learn how to behave, if mom hugs it after each and every mischief; whereas the message will be self-explanatory, if what follows is a gentle pat on the butt. Now, assuming both of these hypotheses are reasonable and worth considering, it is only sensible to follow the scientific method and put them to a test. As a matter of fact, for this kind of experiment, not only do we not need to build a model, but - to the extent that it would be really awkward to parent an artificial child - a model would probably be more of an impediment than an aid. Rather, it would suffice to form two groups of families: one made up with enthusiasts of the modern parenting theories and the other with more old-school type of parents. It would certainly take quite some years to get the results, but eventually we would know which of the two hypotheses is likely to yield better outcomes. Crucially, at that point, there would not be any space left for respectable opinions, but the data would have spoken, and Nature's survival-of-the-fittest will be the only authority qualified to interpret the veredict and deliver a final sentence..

3.5. A scientific model is particularly helpful to investigate abstract concepts, since it enables us to test, whether or not the hypothesized reasons reproduce the observed data.

It may not always be necessary to build a model, bbut that does not change the fact that a model is nevertheless always a fantastically helpful tool in the investigation of any question, in that it empower us to try and test all sort of possible answers to any given question: why did we lose the battle?, what leads to intelligent behavior?, is it going to rain tomorrow?, what leads to lung cancer?, is a true democracy ever possible if public opinion is controlled by the most wealthy in the society?, are children more likely to learn how to be a good person if they are kept wrapped in love, hugs and kisses, or does it work better if they are corrected whenever they misbehave? 

It is true that many scientific investigators do not put any effort in building a model of the investigated system. However this is not because it would not be useful, but because for most of these systems, not only it is extremely complicated to put together one such model, but often it is also unnecessary. This is particularly generally the case in the investigation of living things. As in the tobacco-smoking and children-parenting examples above, if we want to find out what strategies or practices are responsible for certain consequences, it is just easier to analyze what strategies and practices are followed by the various instances of the living thing we find in the real world, and what are the outcomes yielded by each of them. As discussed earlier, the same reasoning can be identically applied in the social sciences: if, for instance, we want to find out what policies, strategies and practices lead to certain consequences, it is only intelligent to look in the past for examples of such policies, strategies and practices, and analyze what were the outcomes. True, we will never be able to determine with absolute certainty, say, what were the exact causes for the defeat in a certain battle. Since the specific situation will never exactly repeat itself, we will never be able to reliably test, what would have really happened, if certain other choices had been made. However, there is no good reason why we could not explore the repercussions of some other more structural concepts. For instance, under what circumstances and in what shape is a strong central authority likely to be a positive or negative factor in the good government of a society? Separation of powers and checks and balances may be beneficial to prevent a single ruler from abusing and taking advantage of his unchecked power; however, in the context of a democratically elected authority, is it such a good idea to limit the authority of the government of the people?

 

Yes, regardless of whether a modle is actually built or not, the basic scheme is always the same: time and time again we try out and test all sort of possible answers to a given question, until eventually we reach a satisfactory resolution. But, what still needs to be clarified is what exactly represents an answer to the investigated question, as well as how will we be able to test them. In this sense, since we have come to realize, we will gain better insights if we craft our questions to interrogate how things work or how certain stuff comes about, our hypothesized answers will generally represent the causes and reasons for the stuff we observe or seek to produce. Thus, we basically want to find out what leads to certain stuff. Finally, particularly if a model is going to be built, still left to be resolved is how is that we can test our hypotheses; but, as crucial as it is, since such logistical issue is specific to the system under investigation, we better do not get diverted into that rabbit hole.