What is an Explanation?
I closed the previous essay with a neologism ‘unsane’ as a portmanteau for my perception, my meta-story, about the interaction of human neural, perceptual, and behavioral systems within the context of civilization: “we are unsane.” I thought I should explain this in more detail, and I shall in a future essay, but then was attracted to the word ‘explain’ and got to wondering what is an explanation? I should explain explanation! Now that is an interesting problem contained within a circular phrase. I first thought that this might become an example of circular reasoning in which there is no point of entry that allows the circle to be broken and stretched out for examination.
I had no sooner mentalized the words ‘circular reasoning’ when I thought of the Definition Problem, where defining words can tend to be circular. If I were to try to define affection, I might refer to love, and to define love I might use the word affection. Then I thought of the subsidiary section of some definitions in the dictionary in which close synonyms are differentiated by examples of their usage. Ah ha! A definition is a set of rules that match a symbol to a perception – a pattern match – that is then used to think about and communicate that perception. Earlier I defined a story as the right words in the right order, the essential components of successful verbal communication. Accurate symbolization is necessary for successful transfer of information from one mind to another. Both minds must first operate with the same rules and symbols: a reason we have dictionaries.
The right words in the right order – that might be a useful beginning to the problem of explaining explanation. I recently completed an interesting electrical control project and the last thing I did was to present the owner of the machine with a manual of operation which included a schematic drawing of the motor control circuit and a written description of its operation. When the machine operator presses the Start button, a series of events happen and when the Stop button is pressed, a different series of events happens. This is a process, something happens, and there is a time order in the process that can be perceived and expressed. There are two interlocked ‘words in the right order’: each event element of the process is expressed in a sentence for which the words have to be in the right order, and each sentence has to be in the same order as the events being described.
I can point to every element in that controller, a wire, connection, contact, or coil and point to its symbol on the schematic. I can point to a symbol on the paper and then to the device on the controller. I can tell or write the name of each device or symbol using a word or words, another symbol conversion which can be applied to either the physical manifestation or to its symbol on the drawing. The words may be expressed invisibly in my mind or converted into a symbol in a communication channel, that is, spoken as vibrations in the air, or written and communicated as electromagnetic vibrations as light. Either the visual experience of the device or of the symbol on the schematic can evoke three different expressions, the thought, and the spoken or written symbol. At all of the intersections of physical manifestation and symbol, and symbol to symbol there is a one-to-one correspondence.
If someone were to look just at the controller, nothing much could be learned. As shown in the photograph it is incomplete because there are components that are out of sight. Our viewer could not treat it as a black box because the inputs and output are not accessible, connected or operable. In its completed installation our viewer could push the Start button and observe that there are some noises, a motor begins to run and a shaft rotates. When the Stop button is pressed there are some other noises, the motor stops, and the shaft no longer rotates. Push this button and this happens and push that button and something different happens. A one-to-one functionality is established between an operator input and the controller output: starting or stopping the motor and the shaft.
But perception is not explanation. Each device could be identified and a change in it observed, the ‘when,’ but not the how, why, and what it does. There are too many noises in too short a time to follow the process in time. The connections between the devices are hidden, and it is the connections that determines the process and the functionality. Each of the devices could be used in a very different circuit in a different connection pattern set and consequently with a different function.
“In the right order:” connections are interesting relationships and they establish the right order. It is no accident that the control circuit uses connections between devices to determine its functionality and process, that brains use connections between neurons for multiple functions and processes, and that words in their order in sentences and paragraphs express connections of various types.
Connections are expressed in the schematic diagram. (pic) Each device and wire has a symbol and it is the connection of the two ends of the wires that determines the process. It is a fully determined system with two-state input and output that does something useful by its operation. The process operates over time in a cause and effect chain. The schematic can be ‘operated’ in my mind by positing a change in an input and following its consequences in a concatenation of if-then’s. “If the Start button is pushed, then the control relay 1CR will energize.” This is a symbol to symbol exchange or conversion – from the schematic drawing to words in my mind. In a succession of symbol conversions, I can build the process, the operation of the controller in my mind going from one ‘if this happens then that will happen’ to another. I can also do another symbol conversion to write those words either on paper or a computer. Each of those symbol conversions has a one-to-one correspondence that can be verified by an observer. Likewise, each step in the operation and the overall function of the controller may be verified. Those symbols in my mind, in the air, or on paper are not arbitrary. They are constrained by reality and the accuracy of that constraint is openly verifiable by examining each symbol conversion.
This is an example of the discipline of reality. Each symbolization is directly convertible back to the symbol substrate – in this case a physical manifestation, the controller itself. And each style or type of symbolization is inter-convertible to another. Out of these tight correspondences arises verifiability and reliability and in turn generating confidence. There is absolutely no admixture of emotion, motivation or path dependency. That has been distilled out, or maybe better said, played no role in any of the symbolizations.
Some key concepts arise out of this discussion: physical to symbol conversion, symbol to symbol conversion, one-to-one correspondence of both types of symbol conversions from experience through perception to symbolic expressions, and the verifiability and reliability of these conversions. To these I can add the ‘process,’ ‘symbolization’ of a symbol substrate, ‘symbol exchange’ and ‘inter-convertibility.’ Connections all of them.
I find it quite interesting that in all my years of designing, building and troubleshooting these types of circuits, I have never before used the concepts of ‘true or false’ when thinking about them. It has always been right or wrong, as in ‘that is a wrong connection, move the wire from here to there.’ True and false arise somewhere else in some other context.
Where is the explanation? To a first level of understanding, the explanation occurs at the successful completion of a concatenation of if-then statements and their verification. If I push this button, then this relay will energize and when it changes its state this contact will close and if that contact is closed then this device will receive electrical power, etc. Once I have followed the process in my mind and verified it by applying power to the controller, I can say that I understand the circuit, its function, and the process by which the function is fulfilled.
But do I really understand the circuit? No. I do not have an electrical engineer’s understanding of the electrical and magnetic forces that transmute my finger pressing a button to the motor running. Further, they may not have a physicist’s understanding of electro-magnetism which is much more complex.
There are few things more fascinating to me than magnetism. When I was about 10 or 11, I had several cylindrical Alnico magnets removed from loudspeakers and I would play with them in a state of puzzlement. I knew about physical force, pulling or pushing by contact, but magnetic force exerted touchlessly and invisibly through space was astonishing. The magnets could lie on a table by themselves and nothing would happen. But push one closer to another and suddenly they would move of their own ‘volition,’ crashing together end to end. Sometimes one would twist around to another orientation while moving together. I would pick them up and they were held together by some invisible force. I could pull them apart but they resisted being separated. If I turned one of the magnets around, they resisted being forced together. They would pull together in a straight line, but the repulsive force had a shape. The magnets wanted to reverse their mutual orientation so that repulsion became attraction. I would pull them apart, turn them around and push them together repeatedly, fascinated by the invisible force. Eventually I learned about north and south poles and that unlike poles attract and like poles repel. I saw photographs of iron filings that illustrated the magnetic force field and thought that I understood magnetism. I was satisfied with that explanation.
The theoretical particle physicist, Richard Feynman, was asked if he could explain magnetism. He thought about for awhile and then answered “Yes, but it will take about 7 pages of high-level mathematics.” His explanation would revolve around the quantum mechanical effects of two motions of electrons around the nucleus of atoms of some metals. His equations could model with mathematics the intrinsic spin of the electron and its orbital motion extremely accurately, but could he tell me how that motion generated the force connecting those magnets? I don’t think so. He knew a lot more about that force and his explanation reveals the paucity of mine, but in the end its mystery remains. Magnetism just is. (And by the way, spin and orbital motion are analogies. Physicists can perform experiments that demonstrate something about the electron that is communicated by these analogies to more familiar perceptions.)
Feynman’s 7 pages of advanced mathematics are a functional explanation, that is, the mathematics functions precisely as do the measurements of an enormous number and variety of experiments. The math has a one-to-one correspondence to the observations and measurements of the experiments and are verifiable and reliable. It is not an essential explanation. The essence of magnetism remains unknown and perhaps unknowable to our neural systems. This reminds me of the story of the philosophy professor who entered the classroom for the final exam, then proceeded to write the single word ‘why’ on the blackboard and then sat down. Only one A was given to a student – the one who wrote ‘because’ in the blue book and turned it in. At the bottom, both Feynman and I had to say ‘because’ and admit our current inability to go father. He can just go a lot farther than I can. There are limits to our depth of explanation.
I was a ‘why’ child. I guess my essays reveal that I still am. My brain asks questions and questions ask for answers. It is as if there is a mental discomfort about not knowing that drives a ‘why question’ and a state of lower mental energy upon the satisfaction of curiosity in a ‘because story.’ I have used, and probably over-used the trope ‘patterns in my mind,’ but so far it is the best explanation for explanation. Now that is interesting, seeking an explanation for explanation, but pattern matching seems to work. Somehow, specific words connected together in a specific order – a pattern, a set of connections – does something in my mind, performs a function, that I find satisfying. The conclusion of that sentence then expands or morphs the original question ‘what is an explanation’ into ‘what is mental satisfaction and dissatisfaction?’ Feynman would have been quite unsatisfied with my basic explanation of magnetism by north and south poles. Was he any more satisfied with his extremely complex explanation than I with my very simple model?
Explaining explanation! How can that be done? So far, I have noted that essentialist explanations may or probably do not exist. Explanations are tentative and subject to further investigation yet they may be satisfactory. It is satisfaction that provides a clue to what the brain is doing. The brain, our complex neural system, is a functional element of our body. It does something like out kidneys filter blood. What does the brain do? To a first order of approximation, it takes sensory data, both internal and external, extracts useful and interesting data (the filter function), processes it in some useful manner (its survival rules) and outputs neural impulses which are converted into processes and behavior, internal or external. The brain and its neural system are then extended survival mechanisms, finding food, water and shelter, avoiding danger, finding mates and caring for offspring.
Out of this, I extract two related attributes: exploration and curiosity, each related to territory, to the location and extraction of energy and material resources from our environment. Both of these attributes are questions, what is out there, where is it, how can I use it, or is there any danger. We are a curious species; many species are curious and explore their territory for resources. To curiosity we add the pattern matching of symbols, words, to perception and then combine specific symbols in a specific manner to tell a story that communicates, that somehow affects another mind and neural system in a similar manner. The recipient knows without the time and effort to explore. This is the essence of socially shared knowledge.
There are two clues in this vignette: first, our minds try to find order and pattern in experience because of the possibility that an event might be repeated – our experience of order – and an event responded to and remembered might make it easier to respond appropriately to its repetition. And second, we are a social communicative species with shared knowledge, what I learn I communicate to you in shared knowledge making each of us more significant and useful to each other and to the group.
But there is something deeper to be grasped for our chimpanzee cousins do not tell stories about yesterday and they do not interpret those stories in meta-stories. They share knowledge, but in a different way. They can observe and imitate, to some extent they can operate as if they understand cause and effect, but they do not tell stories. They cannot explain. What do we do when we explain? Is it related to imitation? After successful communication your brain can imitate my brain and I can imitate your ideas. Your brain might function more like mine. Now that is an interesting and useful idea. I’ll try this conceptualization:
There is something interesting in ‘argument – tying another mind into such a tight knot that it has no choice but to acquiesce.’ What is happening? Dominance and submission, yes, but also possible improvement of the secondary mind. Argument is an error correction process, an artifact of shared partial, path and motivation dependent knowledge. It is a process of adding to and cutting out, of summing and averaging individual contributions, distilling out errors and distortions. This works in the milieu of our small-scale foraging hunter-gatherer societies with their limited number of individuals operating in the same ecology but becomes very difficult, even impossible and fails under the demographics of large-scale social groups characteristic of civilization with its high level of shared knowledge and multiplicity of stories and truths. Explanation becomes very much more difficult.
I wrote an essay on the epistemological problem. Here is a thought experiment using my circuit story to illuminate my understanding of this problem: how we can have knowledge of a world outside of our perception and outside of our minds – a prerequisite for making an explanation. Give me an observer, an epistemological philosopher for example, and set him down in front of that motor controller. It is highly probably that he has never seen such a device before and cannot discern its provenance, function or operation. The Start and Stop buttons are on the other side of the machine and out of his sight. I place a sealed envelop on top of the machine and tell him that it contains a statement of what I will do and a description of his exact perceptions. “Now watch and report.” Out of his sight I press the Start button and immediately he says “I heard in rapid succession some clicks and two big thumps, one a short time after the other.”
We open the envelope to reveal that I had written: I will press the Start button and do nothing else. You will hear some clicks and two big thumps spaced apart in time. You may see some metal part moving if your eye happens to be focused on them.” I ask this philosopher how he could have precisely the experience and perception that I predicted he would have, written on a piece of paper and sealed in the envelope. Nothing of this was communicated from my mind to his before or during the operation of the controller. I affected his internal reality in a specific manner and his symbol conversions or transformation were the same as mine. I could take any viewer and repeat this experiment as many times as desired with the same experiences, perceptions, and conceptual conversions or transformations. There are several symbol conversions involved and they are congruent. The only connection between my pushing the button and his report of his experience is an external reality of which he had an accurate, verifiable, and reliable perception.
At this level and for this example there is no epistemological problem. I can reliably and verifiably affect the external world in a predictable manner observable by another person. This may not be good enough for a philosopher who asks unanswerable questions, but it is good enough for me. I’ll conclude with this new conceptualization: fictional questions, lacking a clear symbol substrate, can only have fictional answers. Factual questions with clear chains back to a substrate get factual answers. My impression of philosophers is that they tend to think that if they create ever more exacting definitions of words and use them in ever more complex sentences and paragraphs, they can somehow arrive at truth. Here truth lies at the end of unanswerable questions, where answers intersect with faith to generate certitude. Hmmm, another essay lies in that sentence.
In my muse about magnetism I mentioned quantum mechanics. My knowledge of that subject approaches zero as a limit (a tongue in cheek reference to the fundamental theorem of differential calculus) but I have read that physicists admit to not really understanding it either. There is one school of thought that advises quantum physicists to ‘shut up and calculate.’ Mental and verbal models are inadequate or even fail to convey the complexities and subtleties. Only mathematical models are adequate to their current level of understanding. Quantum mechanics may not be ‘explainable,’ only modellable and then not very well in words.
I reread about accurate symbolization being required for accurate transmission of information from mind to mind, and got to thinking about what happens in the process of receiving information. For some reason I first thought about activation of mirror neurons in the recipient’s mind, then “speech conveys the power to rewire other brains,” and finally, using the concept set of this essay: communication establishes congruent neuron connections in separate brains that allow for similar emotions, motivations and behaviors. That is the social function of communication: shared knowledge, motivations and functionality.
This is my longest essay. So, why did I go to so much trouble to explain explanation? Because I will in succeeding essays attempt to make evolutionary explanations. Explanation may be difficult and evolution is an extraordinarily divisive thesis and the combination is, well ….