In the project, Understanding Textual Language, two words, Understanding and Language, carry most of the weight. Whether textual, spoken, or signed, the type of language involved seems tertiary to the questions: What, in fact, is understanding? And how is understanding linked to language? Consequently this article submerges the modifier “textual” and concentrates on the two most shadowy elements of the project: understanding and it’s linkage to language. And, of these two elements, understanding is certainly the darkest, even the meaning and etymology of the word is relatively opaque as you can see here, here and here. So, let us start with Understanding to see if we can forge a clear-enough meaning of the word and, if possible, of how, whatever it might be, it is represented in the brain.
Understanding is one of those terms that seems to be the focus of a collection of meanings which, taken in the aggregate, we all understand in a general way. But no one seems to have much knowledge of how understanding actually works, nor how it is organized in the brain, retrieved, manipulated and then formulated conceptually into thought, and thought into speech, this in spite of the fact that each of us understand new things effortlessly every day, and know it when we do it; it is, in a sense, like walking or riding a bicycle: after one learns it, it is quite easy and there is no need to think about just how to do it, but if one had to explain it one would sink very quickly, and deeply, into mystery.
Another difficulty is that one cannot really see anything happening as understanding takes place; the process is hidden from our eyes. Seeing complex mechanisms, and how their elements operate and interact together, is normally one of our basic tools of learning, but here we can observe only the product of thought in the form of an individual’s actions, it’s internal operations are opaque, quite hidden in the ultimate “black box”, our brain. Only recently are we beginning to peer inside this box with instruments that show us roughly where activity occurs, and to some degree the mechanics and chemistry of how it works.
To oversimplify only slightly, there are fundamentally two main schools of thought concerning understanding; they are what might be called the holistic school and the mechanistic school:
The holistic school argues that understanding, and in particular the phenomena of consciousness and self awareness, is unique to our species and fundamentally cannot be explained rationally or, at the very least, has not been. This argument is rooted in a metaphysical view of life, one that says, We humans have something unique that no other species has. Interesting, and supportive in this respect, is the peculiar fact that, of all animals, humans, so far as anyone can tell, are the only species to have religion, even in its weakest manifestations. This is not to imply that there is necessarily a tight linkage between religion and the holistic approach; many non-religious people also have significant doubts about the alternative, the mechanistic approach, or what might more accurately be called the process approach or the programmatic approach:
This second school argues that understanding is essentially mechanistic. That is to say that understanding, even at its most subtle, could theoretically be achieved by a machine, or some construction of sufficient complexity, even though such a mechanism has yet to be assembled. The first modern explanation of this approach was articulated by Alan Turing in 1950 when he devised the so-called Turing Test under which, if a machine such as a computer could pass this test, it might legitimately be called “artificially intelligent”, which means that it has understanding approaching, equivalent to, or even, conceivably, greater than, humans. Since that time a number of others have attempted to validate that approach and still others to refute it. Here we will look in some depth at the arguments that have been made on each side of this divide and develop a few arguments of our own as well.
Since hazy issues are often best understood by observing them at their margins, we will start by inspecting some of these margins of understanding and, since it has often been remarked pejoratively of some people that they are as “dumb as a stone,” which certainly seems a margin of some sort, let us start with a stone. Can a stone be said to understand? At first glance it seems unlikely. If we suppose that a stone might “understand,” it seems an extravagantly skewed use of the word. Yet let us have a second glance. But first a small segway:
When we say that a child misbehaves it is sometimes said, usually forgivingly, that “he just doesn’t understand”. When that phrase is uttered, what is meant? It is either that the child doesn’t understand the normal rules of behavior or that he chooses willfully not to follow them. By rules here is meant the rules of society, the rules our tribe has evolved over millennia for its survival and increase. So clearly we can say that the term understanding, at least in this sense, has to do with the knowledge of rules, of our culture, which is to say humans’ systems of rules. Now, with this notion of rules in mind, let us return once more to the stone.
Certainly a stone obeys certain rules, those of physics, and of those, most prominently, those of gravity and solidity. And objects of even larger mass: mountains, oceans, planets, suns and galaxies follow a clear, if complex, and as yet only partially understood, set of rules. Yet we would not ordinarily speak of objects such as these as having understanding. So one aspect of this issue to keep in mind is the extent to which the term understanding is merely a question of semantics, a question of commonly understood meaning. In our way of thinking, understanding is not something possessed by inanimate objects or, if it is—endowing our sun with thoughtful qualities, for example—it would certainly be a manifestation of metaphysics. Here we reach only for simple physics.
If we think of the range of objects on our planet that we know something at all about, it is easy to follow a long spectrum. We can think of this spectrum as though it extends upwards like a ladder, from inanimate objects of various toward the bottom, beginning perhaps with a stone and the mineral class in general and then we can proceed step by step up through the various classes of flora and then of fauna and finally, at the very top of that animate class, we generally classify humans, the natural objects that—from our viewpoint at least—manifest the most obvious characteristics of what we think of as understanding.
In search of margins, we move back down now. Just one rung below humans on the understanding ladder we might then place the primates, and below them, in a more or less orderly regression: the other classes of mammals, general vertebrates, invertebrates, flora, the classes of inanimate objects and thus back down eventually to the stone, perhaps the archetype of the inanimate class. It will be helpful to keep this entire spectrum in mind if we wish to make sense of our terminology and explore the various boundaries within this range.
We can begin with humans at the top step of the ladder, and inspect the intellectual margins of this segment of the spectrum. There are two aspects of this segment that will be considered here: the first aspect has these two margins: the very youngest children and how they learn in comparison with the very oldest people and how, with age, they forget; each of these of course can be compared with the midrange, the normal understanding of the mass of people in their prime. The other marginal aspect here is to look at the various kinds of disabilities or abnormalities in this class, considering these because some interesting experiments have been conducted concerning this aspect of the spectrum that may tell us something significant.
Very young children seem preprogrammed to learn using all of their senses. Once a child learns to grasp things, she will explore her then limited world by picking up a wide range of objects, looking at them, feeling them and, if they make sounds to hear them; she may even attempt to taste and thus smell them too, very gradually answering for herself—without words—the elemental yet very important questions: Is it solid? Is it squishy? Is it powdery? Can I move it? Is it heavy? Is it bright, shiny? Does it make a noise? Does it feel good? Does it taste good? Do I like it? And occasionally she learns that it is too hot, too cold, too prickly, that it is granular or viscous, that it can’t be picked up, that it’s unpleasant or even that it causes pain. It is important to note that all this information is gathered long before the child begins to learn the terminology for such sensations, she learns the sensations themselves. It is a revelation to see the energy and determination brought to this elemental class of learning by the very young.
At the other margin of this aspect of the human segment are very old people. By definition these are people who have survived a relatively long time which, in the domain we inspect here, is our particular interest in them. The understanding of this class margin, like the class margin of the very young, seems also, in a peculiar way, dominated by their senses: as their mobility, sight, hearing and even taste and smell gradually deteriorate, these same people, once at their most productive, now, at this period of their lives, find their intellect changing as well, and this among people whose brains may physically be untouched by disease—we are not writing of Alzheimer’s disease or other cerebral malfunctions here. It is perhaps more accurate to speculate that their understanding now tracks the diminished physical abilities of their senses, their interests gradually reflecting their now decreased capabilities, so that the verve that had once enchanted them in their prime can no longer be reconstituted. Whether the cause is a diminished database of deep sensual information, or lack of desire to access it, it is difficult to know. But above all it seems clear that even here, at this margin, understanding is highly linked to sensation, though in this case it is, tellingly, a negative linkage.
Moving now to that aspect of the human portion of our spectrum that deals with abnormalities, there have been some interesting experiments done with people who have lost limbs and with people whose brains or motor-sensory function have in some way been damaged. In one experiment on a patient, reported in the book, Mirrors in the Brain, by Giacomo Rizzolatti & Corrado Sinigaglia and translated by Frances Anderson, reviewed by Vilayanur S. Ramachandran in Nature Magazine:
… If a mirror is propped up vertically on a table in front of a patient with, for example, a paralysed left hand (so that one edge of the mirror is against his chest), the patient gets the illusion that the left hand is moving when he moves his right hand. We and others have found that this causes recovery from paralysis, perhaps by visually reviving dormant mirror neurons.
A similar experiment was conducted in a different venue with a person who had lost a hand and a part of the lower arm and was suffering “pain in the lost hand”. Similar treatment with a mirror temporarily relieved this pain, though in this case there seemed no permanent relief.
Regardless of the more general implications of so-called mirror neurons, or of their cultural implications—there has been considerable recent discussion, pro and con, of both their existence and, if so, what could be the implication of this phenomenon? These experiments seem to confirm that at least some element of what we would call understanding in the brain, the understanding of manipulation, is primarily driven by, or correlated with, sensation.
Numerous experiments and observations have been recorded concerning people with certain portions of their brain damaged either by accident or through genetic disorders, too many to report here.
The weight of these observations and experiments imply that at a minimum a very large part of what we choose to call understanding is driven by human sensory capabilities, or, in the abnormal case, the lack of it. It is especially worth noting that in many instances language has not at all been involved. So the next obvious question is, To what extent does understanding exist independent of, separate from, sensibility? Or is all, of what we think of as understanding, sensory-based? Now we will look at this issue in more depth, moving increasingly down the understanding spectrum, in search of meaningful margins.
Anyone who has spent time with dogs, cats or horses, or other domesticated animals will say unequivocally that they have understanding. Certainly they do not have human understanding, but often their understanding seems in many respects to resemble that of humans. This is significant in part because of terminology: When we say that an animal understands we are in a sense saying that their cognition is, at least in some things, in some ways, similar to ours. When we see an animal react to pain we understand this reaction completely. When we see a cat plan carefully to capture prey, whether for food or for “sport”, or perhaps even for “practice”, we understand this “gaming” completely. When we see a house pet come "home" at night, eat, and settle down comfortably on a small rug in the corner, we understand this contentment perfectly. Most significantly perhaps, when we see a domesticated animal acting contrary to what we perceive as its “nature”, such as a very large and potentially dangerous dog interacting playfully with very small, young children, even perhaps to the dogs occasional distress, we are impressed with, well… their “humanity”, in this case their restraint, their understanding of the young as immature beings, in the same way in which they might understand their own young.
Today we think of the domesticated animals as “pets”, but they became domesticated in the first place through their symbiosis with mankind, a relationship that benefited each. Think of the skill of sheepdogs, and of quarter horses; imagine the benefit of short-legged dogs bred to flush pests from holes; consider “barn” cats, unfed by humans, who’s every instinct and skill is devoted to the planning and execution of the control of barn rodents and, not incidentally, the gaining of their supper. And all of this knowledge and skill has been learned essentially with minimal use of “language”.
It seems likely that the semantic term understanding is not completely determined by some rigorous and objective measure. This supposition will be reinforced as we proceed with our brief survey.
Outside of the domesticated animals, with whom our intimacy in some sense warps us toward anthropomorphism (as evidenced here by the initially unconscious (and incorrect) use of “whom” in this sentence!). The animals we are most likely to think of as having “understanding” are those closest to us genetically: the primates; they share with us a distinct resemblance in structure, in the collaboration of tribalism—and tribalism’s need for resolution of conflict—in facial and bodily expression and in their general predilection for comforts similar to those which we enjoy.
Most neurobiological experiments are done using chimpanzees and orangutans for the very reason that their brains seem to be “wired” most closely to our own—not to mention that we are often reluctant to experiment on our own brains, a somewhat chancy undertaking.
In several seminal experiments performed recently, a robotic arm was wired to the brain of a macaque monkey and the monkey manipulated a mechanical arm To eat a marshmallow using only its thought; a brief summary is reported in this article in the Massachusetts Institute of Technology’s, Technology Review. An excerpt:
In a dramatic display of the potential of prosthetic arms, a monkey at the University of Pittsburgh was able to use his brain to directly control a robotic arm and feed himself a marshmallow. The research, published today in the journal Nature, is the first to show that an interface that converts brain signals directly into action is sophisticated enough to perform a practical function: eating. Researchers who led the work have just begun human tests of a related technology.
Certainly in this experiment it can be said that the monkey is exercising volition, certainly a necessary element of understanding. Again, significantly, language is not involved here—except of course for the silent language of reward.
It is probably true that we “grant” understanding in some degree to virtually all mammals and, further, that they have it when we think they have it, irrespective of precisely what it might be objectively. This bias is most apparent when reflecting on cetaceans, most particularly whales, but also to some considerable extent with dolphins and porpoises. For some reason these animals’ size seems to be significant to us; perhaps it is the novelty of seeing such large beasts acting in response to cues that we might also understand; perhaps it is because although they appear in most ways very like fish, in other ways they act like mammals, some would say recognizing us as their Cousins, and surprising us all the more for that. We humans commonly anthropomorphize these large fishlike and most un-mammalian-seeming of the mammals, awarding them a rather high level of understanding on very limited objective bases, Moreover, this tendency of ours toward seeing understanding in cetaceans is a very old one; as Jonah himself might attest.
For a précis of the extent to which undomesticated animals can apparently learn read this article published in the Washington Post by Christine Kenneally.
Next, consider these undomesticated animals: the large cats, wolves, jackals and others in which understanding exhibits itself in the form of reflection, categorization, planning and execution, and that not only individually but also, in a highly sophisticated, almost tribal, life of cooperation, competition to establish leadership, and finally in the execution of the plans thus developed and implemented as a group. And again, all of this understanding is acquired using only the yips and yaps, growls and whines, and the mute and brute language of aggression and submission, of success and failure, reward and punishment. Yet who in a position to observe these activities could deny that in these acts is revealed a strong aspect of understanding?
As we further descend the ladder spectrum we have constructed, we should ask whether fish and birds and reptiles understand? We are perhaps more niggardly in awarding understanding to these animals than we are with the primates, cetaceans and the other mammals, though of course they each perform roughly the same functionality, though with different natural skills and more or less reliance on each of their senses, particularly in the activities of foraging and hunting. But it is generally acknowledged that in some dimmer sense than humans they have understanding, though here, it seems to be increasingly displayed through what we often term merely instinct. Instinct is, in a semantic sense, a denial of understanding, a view that where volition and planning and feedback is absent, understanding is not precisely the correct term.
Continuing down: Do insects understand? Communal insects, such as bees, ants and termites, seem to have a limited sort of understanding. But the understanding here is subtly different; it is an understanding that, from the standpoint of an individual animal, turns out on close inspection to be quite minimal, yet when aggregated into a community their understanding as a whole seems surprisingly complex. This is significant because it is a relatively condensed view of what might be called communal understanding, as opposed to individual understanding, an important phenomenon that we can observe in pack-hunting animals, and even of course, and especially, within human society. Think of what has been termed Internet 2.0, and one might especially consider Wikipedia as a sort of hive of activity that after a not inconsiderable period of time has, in a sense, become greater than the sum of its parts.
Can flora be said to understand? Certainly they obey a complex system of rules. Yet here volition seems to be lacking: Even the ability of flora to face the sun, or other light, is known as simply a physical reaction: phototropism. And though we sometimes speak metaphorically of a "wise old tree” and speculate on what it must have “seen”, having sometimes survived for centuries, this way of speaking of understanding smacks of poesy more than of objectivity. Except in this lyrical sense, few would think of flora as having understanding, and if they did it would certainly be considered mystical, in the same way as sensate qualities are sometimes ascribed to our planet as a whole: Survey here what has come to be called Gaia.
And here we are, having arrived back at the inanimate rungs of the ladder, and the stone that we began with.
So what have we learned from our brief perusal of the understanding spectrum? Perhaps the most important is that the term understanding is itself arbitrary: instead of objective criteria we humans ascribe it only to animate objects and only metaphorically to inanimate objects such as the sun and flora. Further, we have learned that the term is self-referential; we tend to scale what we term understanding in other animals strongly in proportion to the extent to which they seem similar to us; which is to say that semantically we see ourselves as the gold standard of understanding.
We have also learned that understanding is achieved by us largely in relation to our senses and not especially, or solely, to our words: the way we “feel” about things, the way we understand things, is to a great extent sense-based, not word-based. Furthermore, and important for our project which attempts to reflect understanding via language, what we term understanding is exhibited long before any utterance of speech.
No matter that stones and planets and galaxies follow a complex system of rules, we differentiate these physical rules—which of course our bodies follow as well—from systems of thought that seem somehow especially human, particularly volitional thought. As we will see below, in moving toward defining understanding objectively, we continually come to a fork in the road: is it all about us, or is there an objective measure?
Having outlined the semantics of what might be called the natural spectrum of understanding, thus putting at least some boundaries on the term itself, let us now move to the mechanistic, or what might be called the programmatic, view of intellect. Here we will review the arguments that have been put forward in favor of artificial intelligence and the counterarguments that this view has provoked.
As a preamble to this discussion, we should realize that the very term artificial intelligence has, for a number of reasons which won’t concern us here, been used in the last few decades somewhat indiscriminately, so researchers have come to distinguish between “strong” artificial intelligence and “weak” AI, the weak versions are those that attempt something less than what we have chosen here to call understanding, those versions that might have some human resemblance to understanding in a particular domain but which lack the generality and flexibility that seem somehow the essence of human understanding. This overview of the field will concern itself solely with so-called strong artificial intelligence.
Alan Turing, though perhaps not the very first to consider the matter of machine intelligence—that distinction might belong to Charles Babbage and Lady (Ada) Lovelace, who made a valiant attempt to explain Babbage’s mechanical calculating machine—was probably the first to do so in the modern context of electronic computers and is certainly the most well recognized person to be linked with the issue we discuss here.
Well aware of the semantic fork in the road described above, Turing attempted to rationalize the potential of machine "intelligence" by developing a rather clever finesse of the semantic problem that did not resort to speculation or philosophy, which is now, famously, and succinctly, called simply The Turing Test:
In essence, Turing proposed to change the question from "Do machines think?" into "Can machines do what we (as thinking entities) can do?" The advantage of the new question, Turing argued, was that it "drew a fairly sharp line between the physical and intellectual capacities of a man.
From this point on, the question of understanding has been delicately refocused toward one of its aspects: intelligence or intellect. It is easy to overlook this subtle distinction, yet it is important to keep this change in mind as we proceed, for it has the effect of slightly refocusing the term understanding; language has been interjected, and in such a way as to now limit the discussion to humans.
The Turing test has something of a special relevance here for our project because of its emphasis on text as a communication device. This is the hypothetical experiment he posed in a 1950 paper, Computing Machinery and Intelligence:
[A] human judge engages in a natural language conversation with one human and one machine, each of which try to appear human; if the judge cannot reliably tell which is which, then the machine is said to pass the test. In order to test the machine's intelligence rather than its ability to render words into audio, the conversation is limited to a text-only channel such as a computer keyboard and screen…
In addition he attempted to rebut, rather loosely, what he saw as nine common objections to machine intelligence including one which, interestingly, and showing some continuity with the past, he termed Lady Lovelace’s Objection. They are worth reviewing because although his rebuttals are rather curt, they at least pretty much list the entire gamut of objections, a useful exercise in itself.
Perhaps the next significant discussion of machine intelligence, and one given in rebuttal to Turing, is a paper delivered in 1980 by John Searle: Minds, Brains, and Programs from The Behavioral and Brain Sciences, vol. 3, published by Cambridge University Press. It has become known simply as The Chinese Room Argument. In it Searle argues essentially that a machine can never have a mind or understanding regardless of how intelligently it may behave. Unfortunately, philosopher Searle’s description of this thought experiment is not nearly as concise as mathematician Turing’s. To the extent of the reader’s interest it may be read in full or in summary by using one of the links above, but here is a fragment of it that provides the essence of his basic argument:
Suppose that I’m locked in a room and given a large batch of Chinese writing. Suppose furthermore (as is indeed the case) that I know no Chinese, either written or spoken, and that I’m not even confident that I could recognize Chinese writing as Chinese writing distinct from, say, Japanese writing or meaningless squiggles. To me, Chinese writing is just so many meaningless squiggles. Now suppose further that after this first batch of Chinese writing I am given a second batch of Chinese script together with a set of rules for correlating the second batch with the first batch. The rules are in English, and I understand these rules as well as any other native speaker of English. They enable me to correlate one set of formal symbols with another set of formal symbols, and all that "formal" means here is that I can identify the symbols entirely by their shapes. Now suppose also that I am given a third batch of Chinese symbols together with some instructions, again in English, that enable me to correlate elements of this third batch with the first two batches, and these rules instruct me how to give back certain Chinese symbols with certain sorts of shapes in response to certain sorts of shapes given me in the third batch. Unknown to me, the people who are giving me all of these symbols call the first batch a "script," they call the second batch a "story," and they call the third batch "questions." Furthermore, they call the symbols I give them back in response to the third batch "answers to the questions," and the set of rules in English that they gave me, they call the "program." Now just to complicate the story a little, imagine that these people also give me stories in English, which I understand, and they then ask me questions in English about these stories, and I give them back answers in English. Suppose also that after a while I get so good at following the instructions for manipulating the Chinese symbols and the programmers get so good at writing the programs that from the external point of view—that is, from tile point of view of somebody outside the room in which I am locked—my answers to the questions are absolutely indistinguishable from those of native Chinese speakers. Nobody just looking at my answers can tell that I don’t speak a word of Chinese. Let us also suppose that my answers to the English questions are, as they no doubt would be, indistinguishable from those of other native English speakers, for the simple reason that I am a native English speaker. From the external point of view—from the point of view of someone reading my "answers"—the answers to the Chinese questions and the English questions are equally good. But in the Chinese case, unlike the English case, I produce the answers by manipulating uninterpreted formal symbols. As far as the Chinese is concerned, I simply behave like a computer; I perform computational operations on formally specified elements. For the purposes of the Chinese, I am simply an instantiation of the computer program. …
Like Turing’s Test itself, this argument has drawn counter arguments. The gist of these rebuttals support Turing and the mechanistic or programmatic view of the term understanding and say, in effect: Viewed objectively, if understanding looks like a duck, and quacks like a duck, and walks like a duck, it is a duck. These opinions can be found here. Two of these counterarguments seem especially persuasive: that of the system argument and that of the complexity and speed argument, so here we will outline these rebuttals of Searle:
The system argument maintains that the thing that has understanding is not the man in the Chinese room who, after all, still does not understand Chinese, Searle’s main objection. Rather it is the system of the man and his room full of appurtenances: his files of translations, his scribbles and, in effect, the room taken as a whole; nevermind that it might take hundreds of years or more for him to come up with a reply; here we consider the theory and overlook small practical problems such as that of time, which in the case of a computer could be shortened to the point of insignificance.
In support of the mechanistic approach it should be noted that the notion—of The Chinese Room as not simply a man, but as a system en toto, yet not excluding the man either—is, in a sense, mimicked, surprisingly and unknowingly, by man himself in his everyday construction of what could be termed biological machines equally as smart as man: the creation of new men. We as a species have in fact found this feat trivially easy—not to mention pleasant—over millennium, and the manufactory works unceasingly every day of the week. There are, after all, roughly 6.5 billion of our tribe as of this writing, and there can be no question that we are instrumental in our own construction. Yet, in this endeavor, we are sadly lacking, as is the man in the Chinese room who does not speak Chinese; like him, we have only a vague idea of how the construction is done, other than, of course, the initial genesis which we (and, for the religious, the Deity) provide.
The element of the system that is missing in this human example, the one that collaborates with us and enables the endeavor of man to replicate himself. It is the system of evolution. This system is a close counterpart, in a different domain, to all of the appurtenances, coding and decoding, pencil and scratch paper available to the man in The Chinese Room. It is a system of which we ourselves have only an inchoate knowledge of how the codes were developed and by what provenance.
The complexity and speed argument replies to Searle that he simply cannot visualize a machine doing anything as subtle as understanding because he has never seen a machine constructed that has anything approaching the sort of complexity and versatility that the “machinery" of humans exhibit. This rebuttal argues essentially that it is merely a logically inductive argument—and here a mistake—to think that because during the millennia of men’s constructive building efforts no one has ever fabricated such a machine, therefore the construction of such a machine is very likely impossible.
One crisp scenario—taking place, curiously, in yet another room—that clearly highlights the complexity and speed argument objection is that of Paul and Patricia Churchland who offer this clever rebuttal:
Suppose a philosopher finds it inconceivable that light is caused by waves of electromagnetism. He could go into a dark room and wave a magnet up down. He would see no light, of course, and he could claim that he had proved light is not a magnetic wave and that he has refuted Maxwell's equations. The problem is that he would have to wave the magnet up and down something like 450,000,000,000 times a second in order to see anything.
This is a rather neat way of restating the same argument: we, as humans, simply cannot visualize—for clearly it would not be possible—a human waving a magnet up and down at a frequency sufficient to “see light”, therefore, using an inductive argument, very often a good argument, light is impossible! Yet while this method seems, and is, impossible, we humans have in fact constructed light emitting electrical gear for over 200 years and, using fire which yields light, have used it for thousands of years before that, though in the this instance we had, at the time, no idea why it worked as it did.
In our quest to understand understanding we have not quite yet run out of “thoughtful” rooms. Another cogent argument is made by Frank Jackson, a philosopher who, poses this interesting, and nicely succinct, question, sometimes termed Mary’s Room:
Mary is a brilliant scientist who is, for whatever reason, forced to investigate the world from a black and white room via a black and white television monitor. She specializes in the neurophysiology of vision and acquires, let us suppose, all the physical information there is to obtain about what goes on when we see ripe tomatoes, or the sky, and use terms like ‘red’, ‘blue’, and so on. She discovers, for example, just which wavelength combinations from the sky stimulate the retina, and exactly how this produces via the central nervous system the contraction of the vocal cords and expulsion of air from the lungs that results in the uttering of the sentence ‘The sky is blue’. [...] What will happen when Mary is released from her black and white room or is given a color television monitor? Will she learn anything [new] or not?
Fundamentally, all these rooms return us to the sensory question: how much of what we term understanding is inherently sensation and is understanding “sensational” to such a degree that it is impossible to imagine that a machine, such as a computer, without these senses could in any way understand?
In each of our lives occasions present themselves in which we learn something fundamentally new, something at first unnatural, something that so alters our state of mind that our brains are forever after changed in some foundational way and cannot unlearn it. These incidents are transcendental, they transcend the normal boundaries that we had previously imagined were there. Most of these occasions are learned in some way but in this learning we are a aided unknowingly by predispositions in portions of our brain inherited from our ancestry, both human and pre-human:
Though few can remember it, learning to walk was certainly one such occasion. Once it is learned its benefits seem so powerful and so natural to some portion of our brain, that after only a year of age or so one adopts it automatically, and after only another short time one need not actively remember how to do it; Furthermore most of us cannot remember explicitly ever having learned it. This skill now verges on those of the autonomic systems of our nervous system, those that cause our heart to beat and, instinctively, remind us to breathe (which, incidentally, I can testify can be inhibited by the excruciating phenomenon called sleep-apnea). The automaticity of walking is also testified to by the great difficulty that some stroke victims have when, in their adult life, they must attempt to relearn the process.
Most of us probably remember the first time we rode a bicycle. Once one’s brain, through a brief period of practice and mistake, comprehend the notion of balance provided by the gyroscopic effect of forward speed on two wheels, it is unlike anything we have previously experienced. Not only that, but after only a short period we forget how we do it; we just do it. And no matter how long this skill may lie dormant for one reason or another it is a skill now so firmly embedded somewhere in our neural structure that it is essentially impossible to forget it.
Speed and the effect of time is an important, if poorly understood, aspect of understanding: in the nineteen thirties and forties some cheap paper books intended for youngsters were printed in such a way that in the upper right hand corner of each odd numbered page was a small picture that was incidental to the text itself. As one read the book which contained jejune stories one occasionally glanced up at the little picture which generally seemed pretty much the same page-to-page though it seemed to vary slightly as though from the effect of a slight misalignment in printing. But if one closed the book entirely and then, with ones thumb at the top of the book, flipped through the pages quickly, beginning to end, while watching only this small picture, it marvelously transformed itself into what today we would call a moving picture or, rather, a motion picture cartoon.
These books were made in such a way that the characters in the picture seemed to move only very, very slightly from page to page (we would now say frame-to-frame) as one was slowly reading the book itself, yet the rapid page-flipping motion—the speed—transformed these innocuous little pictures into small dancers or little moving scenes, something that was, at that speed, sensed altogether differently by the book’s reader; it then magically transformed itself so as to seem somehow alive! The experience was qualitatively different. In their heyday these books may have been the forerunners of motion pictures.
Think of the audiovisual motion pictures in color that we watch routinely today: some of us, when we are young, and some even when older, become swept up as though into another consciousness as we follow the actions and sounds of the movie that seems so vividly to mimic life. As we transfer our thoughts to those of the protagonists of the production, our own emotions are touched, sometimes deeply, even though we understand with one part of our brain that it is not real. And this human capability can be contrasted with that of most or all domestic animals which seem not to engage with this artificial drama, except perhaps as a somewhat annoying noise. Clearly, there is a higher level of understanding involved in paintings and graphics and music, which do not depend entirely upon language to trigger deeply emotional feelings.
Now, imagine a future when we are sensing a scene not merely with our eyes and our ears but with all of our senses including touch, in a sort of super-virtual reality; would we know if it were real or if it were a drama? The power of all this combined sensibility combined with the speed of change (rapidly, frame-to-frame, so to speak) would be such that it is not ridiculous to wonder whether perhaps that which we sense as consciousness is in fact simply the speed of events and our rapid and complete sensing of them impinging on our intellect in an essentially continuous fashion, on many portions of our brain simultaneously. In this scenario it is the depth and broadness of the sensation combined with the speed of events that would deliver the feeling of completeness and totality. This is not an argument that everything we sense is illusory. It is simply that the completeness of this experience might easily be sensed as self awareness or consciousness, the holy grail of thought as it were, the I, the me, the ego.
Until now, we have used the term understanding in a very general sense, one that we understand when we see it, but one that is still, after all this, remarkably ill-defined. Here we will try to deconstruct it in such a way that we better understand its elements. An awareness of the individual elements of the system may help us better understand the whole. There are several ways this could be done. The simplest of course is merely to enumerate the various qualities or properties of understanding, as best we know them.
Another way though, and the one which will be adopted here, is to think of understanding not as simply a list of properties, but as a process which, while it involves these properties, the order and schema of the process, more than merely enumerating its elements, is what will most clearly resemble the actions of our mind-body system, actions we will recognize intuitively as understanding. This process can be seen in the simple diagram at the right. Begin with process element 1.
This description of the process is, in its way, so simplistic (intentionally so), that it applies in large part whether the telltale actions, generated by process element 4, are those of humans or those of the lower animals. Although for the very lowest of animals, and let us choose the oyster for example, the execution of function in the processing of sensation is so simple that it may consist only in reflexive action of opening and closing its shell to obtain nutrients, so that the first three chart elements, characteristic of most animals, even those only a little higher on the scale, are entirely absent in these simple mollusks.
In humans the process is remarkably complex and the chart is highly abstracted from the complicated processes that actually go on, as well as from the geographic areas of the brain wherein such functionality takes place.
But at some level of abstraction, these same basic elements, however crudely, are exhibited by nearly all animals; one supposes that it is part of our genetic heritage developed slowly—no doubt perilously—over environmental time. It should be understood that in addition to being highly abstracted, the chart is misleading in one way and silent in another:
· Implicit in a flow chart is the notion of serial processing, yet the evidence is overwhelming that animals exhibit a high degree of what, in software engineering, would be termed parallel processing. Besides that, changes of state can occur not merely through electrical signals but, as well, through chemical processes, some of which occur quite rapidly and simultaneously over a wide range of neurons. One merely has to think of the chemical adrenaline, for example, to realize just how fast, and upon just how many of these elements of the control system that only this single chemical acts upon all at once, and then extrapolate from that to the effect of the myriad chemicals that our body creates routinely, those generated regularly by the pituitary gland for example or others, as with adrenaline, on demand.
· The chart fails to indicate quiescence in any way and thus implies continuous action, whereas humans and most other animals have periods of rest when at least some of these elements, for example the Execution and direction of function element, are at rest, or perhaps merely idling, or they are inhibited, if those are appropriate terms. Precisely what it is that goes on in a dream state is an interesting question but one that will not be addressed here.
Nevertheless, the diagram shown above, simplistic as it is, can serve as a reference for our discussion of some of the workings of these elements. Researchers have, for example, through various new techniques, isolated areas of the brain, where certain functions go on. In this interactive diagram, on the web site of the American Medical Association, the various areas of the brain are shown in some detail and their function is described at the link. The picture, just below, reproduced from that site, is an overview of the primary regions of the human brain which, like the functional flow chart is highly abstracted and shows only the major lobes of the human brain, the continents, so to speak, but not the countries or cities—though it does show the major rivers (of blood vessels in this case). The links provides considerable additional explanatory detail of each lobe:
1. Processing of sensation is dispersed among a number of areas. Autonomic control is handled largely in the brain stem, while other functions, vision, for example, is handled, counter-intuitively, in the visual cortex located near the rear of the brain, not near the eyes at all. Hearing is processed in the primary auditory cortex located rather more logically near the ears, while sensations of pain and touch are located in the parietal lobe and its neighbor, the primary somatic sensory cortex.
2. Abstraction, thought and reflection are not localized, however much of this function is thought to occur in the prefrontal cortex at the very front of the frontal lobe and, more generally, in the left cerebral hemisphere, particularly in Wernicke's area (interpretation of sensory data) and Broca’s area where it becomes intertwined with other functionality such as speech development.
3. Planning occurs most significantly in the prefrontal cortex and rehearsal of complex manipulations probably takes place in the premotor cortex toward the rear of the frontal lobe (actual motion is controlled by the motor cortex in the left hemisphere).
4. The conscious functions of Execution and direction is broadly dispersed through the left and right cerebral hemispheres.
Like the flow diagram itself, this brief outline of the location in the brain of the control of various tasks is certainly idealized; both descriptions attempt to roughly extricate and localize functionality that is in fact more highly dispersed than is implied by the descriptions. This is done not because of any desire or need to delve into neurophysiology here, but for two other reasons: first, to try to isolate functionality logically so that the operations of the basic control system can be more easily discussed, wherever and however it actually takes place; and also to emphasize that such functions are real and more or less distinct and that they have particular brain locations even though the precision with which these workings can be localized and understood still leaves a great deal to be desired.
A telling example is that the location of memory is still only vaguely understood, and this even though considerable information is known about the micro-workings of the process (not least because of the recent, increasing awareness of Alzheimer’s disease), yet even such concepts as short- and long-term memory, previously thought to be foundational concepts, still remain in shadow, as this quotation and its references taken from Wikipedia makes clear:
Hebb distinguished between short-term and long-term memory. He postulated that any memory that stayed in short-term storage for a long enough time would be consolidated into a long-term memory. Later research showed this to be false. Research has shown that direct injections of cortisol or epinephrine help the storage of recent experiences. This is also true for stimulation of the amygdala. This proves that excitement enhances memory by the stimulation of hormones that affect the amygdala.
This quotation is interesting on two levels: first, it indicates the uncertainty of even what had been the most basic of memory concepts; second, it mentions a conclusion that while not germane to the immediate topic here, is of considerable importance in other ways to this project, and that is the statement that “excitement enhances memory.” When we reach the discussion of language this is a topic that will interest us.
In spite of the crudeness of this analysis of understanding, the combined weight of the facts of the matter seem to exclude the notion that understanding as a whole is stored in the brain mainly in the form of words; considerable learning, memory, volition and direction is manifest in humans long before language is comprehended, and by other animals as well. And this manifestation of understanding is achieved even by those lower animals which will never develop, or will develop only minimal, “language” (in whatever form of expression). However, in spite of this they clearly do exhibit considerable understanding in the broad sense in which we have been using it here.
The recognition that for all organisms, including ourselves, much of understanding may be instinctive would, if true, seem to confirm, not invalidate, the observation that understanding is to a large degree some complex form of the encapsulation of sensation, one not yet clearly understood, and not a neat and simple network of language or words. No matter, this does not preclude the certainty that abstraction, categorization and reflection itself, most especially in humans, is performed in the brain, using some code, in some still-unknown form.
It is tempting to speculate that as we proceed upwards on the animate portion of the understanding ladder-spectrum, more and more of the categorization, planning and management elements of the process of understanding in animals is handled symbolically, thoughtfully. Less and less, as we ascend the ladder, is managed instinctively, just as the elaboration of the structure of the brain itself seems to have proceeded from the bottom-rear, the brain stem, and to have grown slowly over evolutionary time, upwards and to the front, adding layer after layer, node after node until reaching the point where any further evolutionary expansion of the human skull would preclude birth in the normal fashion.
In some of these overlays upon the original prehistoric organ, organisms increasingly developed conceptualization, planning and some form of language expression. Yet, these new facilities did not come sui generis but were piggybacked, so to speak, on the initial sensory apparatus already in use. For example, when we, for some reason, try to recall the name of Albert Einstein, it seems difficult or impossible not to picture this scientist with his characteristic expression, his electrified hair, and perhaps with his tongue sticking out, as we have seen him so in dozens of pictures, this conceptualization seeming to come unbidden from the sensory information from which the name Einstein itself has been abstracted or associated. And this leads us of course into the role of language.