00:00

Bateson

The thing of it is that these are very difficult things to talk about because there are three aspects of the matter which people think are different problems, different concerns, which in fact boil down to being all one matter. And I put these three up on the board, but one of them is evolutionary theory, and that’s a matter you know which is dealt in one sort of book. Another is mind-body problems, and that is dealt with another sort of a book. And the third is epistemology: that is dealt with, again, in another sort of book. And what I want to get across to you is that these three apparently different matters are, in fact, all one subject of discourse, and that you cannot handle the one without simultaneously handling the others. And if we are going to talk about consciousness, I would like to aim that word specifically at an awareness of these three things and their interrelations.

You see, if we start from where we were last night, we had a lot of Cartesian diagrams in which time was horizontal and some thing or other—responsibility or narcissism or something or other—was in a vertical coordinate. And these were on the model of what Descartes thought was the way to think, and on a model which has been an extraordinarily profitable model in thinking about a lot of things—like planets, and temperatures, and even perhaps populations. I’m not so sure about populations. It’s [???] fashionable, still, among populations [???] study populations.

Now, it’s not an accident that the man who designed those graphs also is the man who formalized the dualism between mind and matter. And it’s very curious that this should be so. Now, the notions that I want to get across to you is that, when you do this—if you start arranging your words and explanations on that sort of a tautology, that sort of a basic set of notions about how things are related—you will, of necessity, end up with the sort of split between mind and body that Descartes ended up with.

Now, that split, you know, has been the battleground of, especially of biology, for a very long time. And the problem is how to get away from it. You see, the moment you go to the extreme materialistic end—which has these dimensions and quantities in it, and the nice curves, and all the rest of it—you specialize over there, the thing bubbles up over here with all sorts of mental, spiritual notions which you excluded from here. You’ve squeezed them out from here and they bubble up there. And the moment you do that they split wide open. Now, it may be, you know, that there are total splits in the universe. I prefer to believe that it is rather one universe than two. And the only real argument for that, you know, is Occam’s razor: that it’s less trouble to believe in one universe than to believe in two. It’s miraculous enough to believe in one, believe me!

Okay. There is also, you see, a consciousness of how it is to think, how it is to engage in trial and error, and so on. And that consciousness, as far as I can make out, is roughly called prajñā in Tibetan Buddhism. It’s a useful word if that’s what it means—always difficult to be sure. So what do we do?

There is a very curious theorem. You remember: at school you were taught that you should not add apples and miles. And that was a very useful thing to learn. It’s very useful in reading equations to remember that, unless you can sort the things out, to have E = mc² you have to remember that m is a dimension of mass—not matter, but mass—c is of the dimensions of length divided by time; it’s a velocity squared, so it’s length squared divided by time squared, and E, therefore, is of the dimensions of a mass times the square of a length divided by the square of a time. So that’s all E is, you know? It’s a quantity of those dimensions.

Now, there’s a very funny theorem, called Euler’s theorem in topology, which says that if you take a polygon—that’s a solid three-dimensional figure with edges and faces such as polygons have, and apices where the edges meet—that the number of faces plus the number of apices equals the number of edges plus 2. Let me draw it on the blackboard. Faces + Apices = Edges + 2. And there’s a horrid question, you see: what is the dimension of the 2? We have grossly been adding surfaces to meeting points, to points, and then equating them with edges. And then there’s this 2. We’ve already mixed our dimensions hopelessly. Now what is the solution to this difficulty? The theorem stands; is probably the fundamental theorem of topology. How is it right that these things should be added in this funny way? And what is the dimension of 2?

I made a crack on the side last night while Jonas was speaking. He was talking about dimensions and dimensions, and this dimensions and that dimensions. And I said, “Are you sure you’re you’re talking about dimensions at all? That this subject matter we are dealing with is a subject matter within which the concept of dimensions—and therefore quantities and graphs and all this stuff that all follows, you see—that this is an appropriate language for talking about this at all?” And I sort of let that pass as a wisecrack last night, but we now have to face it a little bit. You see, he thinks he’s going at a speed—that fellow on the water, there—he thinks he can measure the speed with a speedometer. But that’s really, you see, all bunk. The truth of the matter is: he’s having fun! And the relation of the fun to the speed is very obscured. Probably what he’s having fun with is the probability of disaster. You know? Dimensions. What in heck are we talking about? Because, you see, we’ve been pulling these analogies, metaphors, out of physics and then trying to map humanity, behavior, love, hate, beauty, ugliness onto those metaphors.

But let’s go back to Euler. This is [???]. Euler’s theorem. It’s proven in a very curious way. Imagine a polygon. We’ve got a football, you see, and now we’re going to slice it so that it’s all flat faces. [???] and the number edges between the [???]. But what you do is, you take away for the proof one window out of the football; one face. And now you put your sums in, pull the whole thing out flat. Which leaves, you see—it doesn’t do anything to the relations. You consider what happened when you took out one window, and now you take the edges away one by one, showing that, at each operation, you leave the same relation between the number of edges and the number of faces, et cetera. And you end up with nothing on the paper, which is now a flat piece of paper. Right.

The notion that I’m trying to get across to you is that these things called edges—no. God, language is such a lousy invention, isn’t it? Hm?

Bateson

Language? I’m sure it’s not a dimension. I’m sure it’s an epistemology, you see? It’s a pickled epistemology. But very pickled in a solution that hardens it terribly. A hardening solution. Alright.

We have names of edges. Names of faces. Names of apices. And that’s what we’re playing with. Not edges, apices, and faces. And because the whole thing is removed by a level in a curious way, it becomes legitimate to add them together, and subtract them, and all the rest of it. And the 2 is sort of in the Platonic universe, the universe of ideas, but not in the universe of dimensions.

Now, it may be that there are real dimensions, you know, out there in some sense. But we are dealing not with real dimensions, but with descriptions of dimensions. And the central difficulty and the big enlightenment comes when you suddenly realize that all this stuff is descriptions. And when you realize that, when you realize that it’s possible to be wrong in your descriptions… and it’s possible to be wrong for this reason: that the creatures we talk about—people, sea urchins, starfish, beetles, plants, cabbages, whatnot—that all those creatures that we talk about themselves contain description; that the DNA are descriptive prescriptions, injunctions, for how to make a bird or a man or whatever. And these injunctions themselves contain epistemology. They contain a theory of the nature of description implicitly. And you can never get away from theories of the nature of description wherever you have descriptions. All descriptions are based upon a theory of how to make descriptions. You cannot claim to have no epistemology. Those who so claim have nothing but a bad epistemology. And every description is based and contains implicitly a theory of how to describe. The Cartesian coordinates are a theory of how to describe, and for many purposes, I believe, an inappropriate and dangerous theory by making the end leads to various sorts of quantification of things which probably should be regarded as patterns and not quantities.

You see, you can be wrong in describing the anatomy of a human being when you say that he has ten banana-like objects at the end of each limb. No—five on the end of each limb! Because, you see, he might have not five fingers on the end of each limb, but four angles between fingers. The question is: what is there in the genetic injunctions which are the prescriptive descriptions for how to make a hand? Is there a number there at all? Five or four or whatever. Is there, conceivably, a rule of symmetry there? What sort of stuff is—this is almost a total gap in our genetic knowledge. There are few little spots in genetics where there are indications of what the epistemology, what the theory of prescription, might be upon which the DNA become the prescription for the hand, the body, whatever. This is a matter which geneticists have hardly touched.

Let me give you a couple of cases. All the vertebrates and chordates are bilaterally symmetrical in their ectoderm and mesoderm on the whole. Endoderm—profoundly asymmetrical. There are some cases of asymmetry, some fairly superficial asymmetry, in ectoderm and mesoderm. Why the endoderm is more asymmetrical than the rest, lord alone knows. Alright. Where does the bilateral symmetry come from? Not the genes, oh no! It’s doubtful whether the DNA genes are ever able to tell the embryology how to orient itself. A frog’s egg—and this has been known since the twenties now—an unfertilized frog’s egg (and presumably this goes for overall chordates) is, so far as we know, radially symmetrical. It is… you go around the equator—it has a north pole and a south pole. It’s approximately spherical, it’s prigmented down to rather below the equator. This end is fairly clear of fat, and this end is heavily fat; is yolk. It’s sort of yellow down here and sort of nearly black up here. But it’s the same all the way around, so far as we know. Nucleus is located somewhere up here.

How does it decide on this line? The line of bilateral symmetry; plane of bilateral symmetry. And the answer is that a spermatozoan will enter somewhere below the equator, and that now gives it three points: two poles, and that point of entry. And that line of longitude, that meridian, will be this line. You don’t have to have a spermatozoan, you can do it with a fiber of a camel’s hair brush. Just prick it and an egg will develop that will make a complete frog—which will be haploid, it will only have half the number of chromosomes it should have and it will be sterile. But it will catch flies and hop like any proper frog. All the information for catching flies and hopping is there. This is a very curious little fact right in the middle of the problem of what the genetic code looks like.

You see, the genetic code that is the unfertilized egg has sufficient information in it to pose a question, to set the egg to a readiness to receive a piece of information. The genetic code does not contain the answer to that question, it depends upon something outside the egg—a spermatozoan or a camel’s hair fiber—to fix it. This, you see, sets a whole stage for, how shall I say, for saying, what is the unit of embryology? And the unit is not just that egg, the unit is the egg plus the answer. And without the egg plus the answer you cannot go on to the next phase, and so on.

Well, let me now give you another piece of data about the nature of this whole business, and that is this piece: that we take a newt’s embryo—that’s a fairly early one. It’s facing in that direction. And we put an eye in which wouldn’t be there yet, but that’ll help you keep oriented. That is now a [???] of the right hand side of a newt’s embryo. He’s towards you as I am, standing. And, at a certain point, a low mountain starts to grow up that swelling. And that swelling is the limb bud of this right limb. And remember that this limb is different from this one. And that difference couldn’t be in the DNA because the same DNA are in the cells right here that are in cells here, you know? So where does that come from? Very awkward. However, that’s the limb bud for this one.

We cut it out, lift it. That’s now the front edge, right? That’s the posterior edge, that’s dorsal and that’s ventral. Right? We lift it out, turn it through 180 degrees, and put it back again. That front edge is now the back edge, that’s the old ventral edge, that’s the old dorsal edge, and this is the old posterior, yes? Embryo. And when he grows, what do you think? It grows into a left leg. Why? It grows into a left leg because it knows—I use the word “know” deliberately: it has received the information and is governed by the information there. I’m not talking about consciousness, I’m talking about being determined by information. I don’t know whether it’s conscious. I’m not a limb bud [???] anthropologist. An amphibian.

Alright. It has gradients fore and aft, which are determined before it even became a mountain, a bulge. And those gradients are informational gradients. Lord knows what they depend on. Could be clockwork for all I know. It doesn’t matter, you see? Information’s got to be in some form or other, provided it’s there. Electrical, chemical, what have you. That fixes this as anterior. But the dorsal/ventral information comes much later. And we did our operation before it knew the dorsal/ventral answers, which later it gets from its neighborhood. So now this one is told to be a dorsal by the neighbors. This one is told to be a ventral by the neighbors. And now we have that ground plan for our limb. It’s going to come out this way just like it would anyway. So the proximal distal dimension is unchanged. [???] the use of ground plan of a left leg. You have inverted one dimension but not the other. Okay? Between that one dimension of this, as in a mirror, you get that. This one faces that way, this one faces that way. Here’s the mirror in the middle. One-dimensional inversion gives you the mirror image. Two dimensions gives you this one again. Three dimensions gives you that one again.

And what I’m saying is that the world into which we are moving, the world in whose terms we have to think, is a world of patterns, is a world in which there are tautologies and logics which we can use for explaining things with, for building accurate language with, and some rigor with. But it ain’t like the language of quantities and such things. It’s a language of patterns and not an entirely simple business. We have a major problem in front of us to create the language in which we can talk about evolution, talk about—what are the other things? Epistemology. And talk about mind-body. We’re going to be able to do trial and error in these matters. This is like somewhat of a ship, approaching it from another angle. The old quantitative logics that we were brought up with are devoid of time. There were some nice pattern things in Euclid and elsewhere, but timeless. If straight lines are defined in this way, and points are defined in that way, and triangles are defined so-and-so, and if two triangles have three sides of the one equal to three sides of the other, then the two triangles are equal each to each. That’s the way I’ve learned it by heart, more or less. But look at the word “then.” There’s no time in that “then.” There’s nothing but logic in that “then.”

Now, if a frog’s egg receives a spermatozoan on a given meridian, then that meridian will define the plane of bilateral symmetry. That “then” has time in it. Sequential time, always with a delay. Alright. If Epimenides was right in saying that cretins always lie, and he was a cretin, was he a liar or not a liar? If he was a liar, then he was a liar. If he was not a liar, then it was untrue that cretins are always liars. A-ha! Now, if you look at the “then” in that paradox—if yes, then no; if no, then yes. You’re looking at a logic that has time in it. Which is essentially the same as the logic of the electric bell on the front door: if the circuit is complete, then a magnet is activated which will break the circuit. If the circuit is broken, then the magnet will not be activated and the circuit will be restored. If the circuit is restored, then the magnet will be activated and the circuit will be broken. And we get, you see, not this curve that we were talking about last night, but this one. It’s an off-on one for the magnet. It’s… a-ha.

But the paradox—if yes, then no; if no, then yes—contains, then, a real temporal “then,” and the nature of the system is such that you switch at the break (this middle point)—actually, when you closed systems, oscillating systems—this one can’t be said to have a middle point and this one, of course, the correctives are always starting to act all the time, really. But they change their magnitudes and you get a swinging system.

The moment you get a swinging system, you are in fact looking at something where the ideas which define it, as these exist within it and hopefully within our explanation, too—because it’s nice to have the explanation, you see, in step with the system of ideas within that you are explaining. That’s what I keep saying. If we’re going to say the thing has five fingers, you may be wrong because, really, it has four gaps between fingers: four relationships. Because this is governed by relationship, not by the absolutes.

But if you’re going to now face oscillating systems, you have a very curious world in which a certain degree of reality is imparted to the chunk of living matter that is a justification, of some sort, in drawing a line around it. That justification is based on the fact of that autonomy—a literal autonomy, in that the system names itself. You’ve got a system of injunctions in which the system of injunctions is auto- (“self”) -nomic: self-naming. Self-ruling, but essentially self-naming. And that is the only autonomy there is, as far as I know. It’s recursiveness. That recursiveness is crucial to a system with if-then, where the “then” is not a logical “then,” but a temporal “then.”

Now, all this—I seem to be going on too long. Ten—when did we start? Five or ten after ten. I’m not going on too long! Right. I’m now starting to build up slowly, to where we can begin to think.

We have if-then relations with “then” being purely temporal, being temporal, and therefore differing profoundly from classical logic. That doesn’t mean, you know, that it’s impossible to think! It means that logic is a poor simulation of cause. We used to say, “Can computers simulate logic?” But, you see, computers work on if-then, so they’re causal. If this transistor tickles that transistor, then such-and-such. But that’s a causal if-then with time in it. And the truth of the matter is that logic is a very poor simulation of computers and other causal systems, and still poorer, probably, of biological systems; classical logic. That does not mean there are not regularities, patterns, epistemologies—that is, ways of thinking—which shall, in some degree, be not too incorrect representations of how to think.

You see, we keep coming back—I keep coming back—to the assertion that what we deal with are descriptions. What we deal with are representations of how it is. How it is—we don’t know! We can’t get there. That’s the Ding an sich, which is always and inevitably out of reach. You have sense organs specially designed to keep the world out. It’s like the lining of your gut, which is specially designed to keep out foreign proteins: to break down the foreign protein before it enters—to its amino acids, or whatever. Only the amino acids are allowed through. And your sense organs—you see, which is another piece of this whole business: how you can know anything, how it is that anything knows anything, how it is that an egg can be organized—and you’re only eggs, and I’m only eggs, you know? We’re the tryout of eggs. The hen is the egg’s way of finding out whether it was a good egg. If the hen’s no good, the egg was lousy. It had the wrong genes or something. It’s a trial-and-error system. It’s not quite what Samuel Butler said, but pretty near it. He said the hen was the egg’s way of making another egg. It’s really the egg’s way of finding out whether it was any good—in evolutionary terms.

We’ve got trial and error, actually, coming in here, too. And we have—how you think—we have higher organized, as a physiological and behaving creature, the nature of adaptive action. And this is really the same sort of process as this one, as evolution. Evolution being: the same sort of thinking has got to be used to analyze evolution as the thinking you use to analyze thought. Not that they’re the same process. I mean, I don’t think what you think can alter your ova or spermatozoa. I’m not preaching a Lamarckian message at all; indeed, quite the contrary. But I am saying that the sort of trial-and-error, this sort of logic, this sort of way of thinking about things—which is a non-quantitative way, is a way immersed in time at every stage—that that is in common between the two. And therefore, epistemology and evolution go hand in hand. And mind-body is obviously the same sort of business. What you think about evolution is what’s going to be the reflection of what you think about mind-body relations, and all the rest of it. It’s all going to move along together.

Alright. Now then, let’s switch over and take a look at what matters. I mean, it’s nice to think straight, it’s nice to have all sorts of scopes and skills in one’s thinking. But we live in America, and in America things are supposed to matter according to purposes, certain adaptive sorts of coping and manipulation, and so forth. Pragmatism. Wonderful.

Let us suppose that, in biological evolution, there were a direct communicational bond between experience and the effects of environment upon the body—between somatic change, as it’s called, and the DNA injunctions to be passed on to the next generation. Let’s imagine, for the moment, a Lamarckian universe. That is, if I brown myself in the sun, tan myself in the sun, let us imagine that this were in some degree to be passed on as a increased browning, brownness, of the skin of my offspring. What my offspring would have lost by that would be a flexibility in terms of how much they would go brown under the sun. You see, I’m flexible. I go brown with the sun or I go bleach with no sun. The Lamarckian theory would presume a rigidity in my offspring, a reduction of their ability to bleach with no sun.

Well, if you begin to think about that, and to spread this to all sorts of variables through the body, as a general theory of evolution, you will see that, obviously, a Lamarckian theory will, in the end, be an increasing rigidity. A loss of the ability to adapt. It won’t do. Things are going to get tight. The body is certainly made of a very large number of variables—our description of the body, correction, is made with a very large number of variables which seem to be interlocking in all sorts of rings and loops. Now, if you start tightening any one of them—you see, they have that much flexibility, we’ll suppose; tolerance. Go up there, too little sugar in the blood. Go down there, too much sugar. Or up and down, whichever you want to do. Death ensues, you know? You can learn a little bit, and you can move this scale as well as moving the variable. But these things interlock quite a bit. It’s surprising what we can stand. But put yourself under a stress, push some of your variables up to that upper limit—which is what happens with disease, or a cold, or something. Why do we put people in bed and keep them warm when they have a cold? Because, essentially, they’ve lost a heck of a lot of flexibility by being stressed up to maxima or minima somewhere in their organization. And we, therefore, protect them during that period.

Alright. If you had a Lamarckian system of inheritance, certain things—this would always be done at a certain price. You would give up the ability to adjust certain variables. And a certain benefit. The benefit being that you wouldn’t have to adjust those variables; you’ve been born right, you see? Alright, that’s a problem for biological evolution and the barrier between somatic change and genetic change. What seemed to be quite important, if you verbalize it, it will be a rule that you shall not make an irreversible change until a long time has elapsed so that it is reasonably certain that the irreversible change will pay, and you won’t regret the irreversibility. This seems to be deeply characteristic of biological evolution.

Now we will take a look at social and cultural evolution. And note that there is no such barrier. That, within much less than a generation of the invention of carbon paper—which was adaptive, you know, in a sort of way in the first instance—it’s designed for a lawyer’s office where documents have to be copied critically, word for word. Contracts and stuff. And for this, carbon paper was an enormous advantage. And the sort of slavery in Dickens’s lawyer’s offices and so on where these miserable people are copying documents—now, that sort of thing, you see, was largely made obsolete by the invention of carbon paper. Alright. But within very few years of the invention of carbon paper, we start to use carbon paper for personal letters, even for love letters because, posthumously, we’d like our biographers to have access to our most romantic thoughts. That’s how it goes. That the invention of, say, carbon paper, and still more—automobiles, roads; what are these boxes, video tape, television—all this incredible stuff we have around becomes absolutely irreversible because it’s built deeply, irreversibly into the physiology of our society within—not minutes, but certainly very few years—of invention. There is no barrier between immediate adaptation and pickling the change into society.

It is, I think, for this reason more than any other that I distrust consciousness as a gimmick added to the evolutionary scene. Conscious cerebration is much too fast. It doesn’t give any time for growth into the new state of affairs, assimilation and—not assimilation, just—inevitable to assimilate—but tentative assimilation which will slowly flow, hesitate and flow, hesitate and flow, into new patterns. [???] within less than a human lifetime—I went to a bed with a candle when I was a boy. I went to school with a horse-drawn bus in the suburbs of London. At the foot of Wimbledon Hill two more horses were tied onto the bus to go up the hill. And so on. This is just outside London, where the championship tennis matches were already being played. It goes much too fast. You couldn’t get a horse up that hill anymore because the road surface is now committed to automobiles. And that goes all the way through the system. When you haven’t got a carbon, you’re now subject to all sorts of sanctions. So that the invention of the carbon—and later the xerox, I’m sorry to say—become irreversible parts of your physiology, almost.

And this goes because there’s no proper delay. We always used to joke between World War I and World War II about the American airplanes. That they’d spent World War I designing airplanes—fighter planes—and never really got around to making any because they kept changing them on the blueprint. Now, this was, you know, a deep biological wisdom in the Americans, really. If they had any sense, they’d be changing their entire culture on the blueprint for 100 years before they really manufactured the next version. [???]. World War II, they got them out like anything. And we’re now, of course, committed to airplanes in various ways. I wouldn’t be here without one!

Alright. That, I think, is where my pragmatic remarks would focus: on the question of, not the immediate trial and error, which is immediate adaptation, but a long-time, sort of meta-trial-and-error, which would deal with the problem of: is the adaptation one which we can really stand? This would give us some chance of adapting not just to the immediate problem of who dies of what, or the traffic accidents, or what have you, but to the question: if we make this adaptation (in law, in technology, in whatever) to disease, to discomfort, to traffic accidents, to what have you, what will be the implications of that adaptation in the rest of the system, which is all interlinked? And in the end it is the meta-adaptation, the adaptation to the total adaptive system, that is going to kill us or let us live.

May I tell a PhD story, too? Not my own; I don’t have one. When I was working with a B.A. in Palo Alto, there was a young man in the clinical psych department who was doing a dissertation for a PhD in the psych department at Stanford, working at in the clinical psych staff in the VA. His dissertation concerned an educational program to which we were submitting the aides in the hospital. And he started off—the question was: he would test the aides before and after this educational program and find out—it was a rather simple little program designed to teach them to not neck out the patients quite so often, you know, or perhaps alter their attitudes towards patients, whatever that means. He started with 35 aides who hadn’t gone through the program yet. Now, the aides are at the bottom of the totem pole of all the professions of our civilization put together—they’re lower, even, than dishwashers at restaurants—and it’s a floating population. And by the time he finished, he only had 4 aides who had gone right through the program and been tested before and after. So he wrote up his dissertation. And it was turned down by his PhD committee for an inadequate n. The same little thing that John had trouble with. Now, that’s alright, you know? The only question arose: there were four members of that committee. No, that’s alright, too. So we say, alright, there are some questions which can be settled by an n of four, perhaps even an n of one. Other questions which cannot be settled, which require a big n. Which questions? You see, once you acknowledge that there are some that don’t depend on n, that depend on something—what else, I don’t know—you’re in a new universe. And I suspect that’s the universe we’re mostly in in this discussion, in which high n is useful for certain sorts of questions. Mostly, I suspect, the wrong sorts. And there are certain other things which it is possible for a man to have an opinion about without being in agreement with 50 million Frenchmen, who notoriously can be wrong. Okay, that’s all I wanted to say.

1:06:08

Bateson

Between quality and quantity? That’s a very good—I [???] hadn’t mentioned it.

Bateson

Yeah. The thing is quite complicated. There are three things that have to be separated. There is quantity. Quantity is peculiar. You get it by measurement. And it’s always approximate. You cannot, in fact, have five gallons of water, you know? You have five gallons plus or minus an error. No quantity is ever exact in the nature of the case. And if you use quantities as the material for running computers on, for example, you have what are called analogic computers, and these work very differently and have very different sorts of error from what are called digital computers, which are essentially counting computers. They count yes’s and no’s. If you have an analogic computer—say you devise it to do multiplication sums, you will have a faucet, for example, which can be set to a number of gallons per minute. You have a clock which will count minutes. And if you want to multiply three by five, you set it to run one gallon an hour, you set the clock for—what did I multiply by? Three by five. You set it to run three gallons in an hour, and you run the clock for five hours. And you measure the water at the end, and you’ve got fifteen gallons—approximately. If you have a whole sequence of analogic operations, each involving real quantities—the quantities being real, they’re always approximate as I say—you have to be sure that your sequence of approximations will not compound the error of approximation. It’s alright if you design the thing so that the approximations of one step will, on the whole, be reduced by the approximations of the next step. But if they tend veer off and veer of, and maximize and compound, you’ve got trouble.

A digital computer deals with counting. And counting is something rather different. Counting is much closer to quality than to quantity. And for small numbers, you and birds and similar organisms don’t even count in a tick-tick-tick-tick sort of way. When the cards, in nine of spades, is thrown on the table and you’re playing a game, you recognize the pattern, you know four and four and one in the middle—that’s nine. But what the birds do isn’t very clear. The crow, as I understand, can get up to seven. And it looks often as though they count. But let me not go into that for the moment. That’s quite a long story. A rather amusing one, but another day.

Then there’s quality; there’s pattern. Now, you see, what people want to do—and I’m speaking more of popular… the way people talk than of sophisticated physicists in labs—people continually talk as though quantity of energy or money or something could determine a pattern. Now, this seems to be not so. And if you begin to think about this, perhaps the difference between quantity and pattern will come out. Suppose we have a chain and we increase the tension upon that chain. At a certain point it will break, and it will break at what is called the weakest link. That is, by a quantitative change of tension, measured in pounds or whatever, what has happened is that a qualitative difference—a difference, mark you—has been detected. The weakest link was the weakest link before ever the tension was applied in general, and the increase of a quantity has disclosed a qualitative, latent pattern. The development of a photographic image is the same sort of thing. The boiling or crystallization of a fluid is, again, the same sort of thing. That if there is no weakest link, if there is no point in the fluid at which there is a difference between that point and other points—a little bit of dirt, you know; something or other; an already existing crystal, whatever it may be—and FFFT, it’ll go like that when it reaches the proper temperature. Now, if your water is really clean, your beaker is really smooth, and your raising of the temperature is really very gentle, you’ll get almost the sort of thing you get with a frog in hot water. It is very difficult for the water to know when to boil, and it will overshoot the boiling point by several degrees. Then, suddenly, almost explosively, it will get the idea of boiling—presumably from finding some microscopic inequality. But it has to start at a place. And the same is true of freezing. Glass, as you know, is supercooled liquid, in fact—ordinary transparent glass. If you want to get it to crystallize down, you have to raise it and hold it at, what, about 500 degrees [???] of melting point? And cool it very slowly, and you’ll get that iridescence.

And that is—quality is something very different from quantity. Quality… it tends to depend upon discontinuities and it tends to depend upon difference. I embarked a little bit on discussing sense organs and didn’t follow it through. The point about sense organs is that they do not live (therefore you do not live), on the whole, in a world of quantities. Sense organs live in the next logical type up, which is the world of differences. If you want to know something—there is a dot on the blackboard. If I drop my finger on it, I cannot feel it. If I move my finger laterally on it, I feel it at once as a conspicuous difference in level. That is (on the whole, in general) true for all sensations of all sense organs: that what you perceive, the reports which enter your system, are not reports about magnitudes, they’re reports about differences. And if you want to measure length—the position of mercury in a barometer; whatever it is—you put a [???] on, which is a device for making the differences more immediately apparent.

Your eye is continually vibrating. If you so rig it—and this can be done—you build onto the front of the eye a very light structure which will vibrate with the eye, and within that structure you have an image that you’re supposed to look at. You will find that you cannot see that image. You can see bright versus dark, but you cannot see edges or forms unless you have the vibration—which is essentially a saccade, which does the same to the image on your retina that I was doing to that dot on the blackboard: moves the retina relative to the image. And then you see.

In Hawai’i we used to watch the geckos on the window; on the mosquito screens on the window at night. The lizards. The would come out every night for the moths. There’s a gecko here, a moth lands here, gecko turns. Now, nobody moves. The moth makes a little movement, and the gecko gets it. But the gecko can’t make the second move until the moth makes a move because he can’t see the moth. You see, you can see things that don’t move—and you do this by this oscillation of your retina, or by a major movement of your head, and so forth—because you have the phenomenon of reafference. You can discount—in your interpretation of what you see—you can discount the effects of your own movement. If I move my head, you all appear to move, in one sense. But I know that you don’t move, and I can, you see, discount the effects of my movement. The gecko does not have reafference, cannot discount the effects of his own movement, and therefore has to wait for the moth to move before he can catch it and eat it.

Quality. This means that, from a human angle, the essence of quality is the notion of difference. And the essence of quantity, you see, is quantitative. Now, if you want to look at this biologically, one of the simple ways is to take a look at the keys for identification of plants or animals, beetles, I don’t care what—flowering plants. And you will find that a very large number of those key characters to discriminate this group of species in the genus from that group of species are such propositions as ratio between length of leaf greater than width, width of leaf greater than length. Length of toe more or less than one fourth of the length of the total foot. Statements of proportion. Not: how long is the toe? But, what are the proportions of the toe to the rest of the foot, or the nose to the rest of the face, or what have you. And all the time you’ll find the keys are constructed upon that sort of basis, which is looking for difference rather than looking for magnitude. A difference is the first step towards quality, pattern, rather than towards—

1:20:58

Bateson

Yes, indeed! But it is constant and it can be governed in a way that quantity really cannot be governed.

1:21:15

Bateson

I mean, there are big leaves and little leaves, you know? And the qualitative shape persists from the little leaves to the big leaves—with some modification. But on the whole…

1:22:24

Bateson

A quality [???]? This is another use of the word quality.

Bateson

Isn’t it?

Bateson

Hundreds.

Bateson

I’ve now heard two or three different things, and they’re all opposed. By a “quality” sand painting you mean a good sand painting in some sense? Is that the quality, of positive value, of goodness, as attached to a very complex object with thousands [???] how many thousands of differences one is perceiving when one looks at the thing like that. Is it, in the end, reducible to a list of the perceived differences? Is that your question? I would say no [???] to that question.

1:24:26

Bateson

Yeah, yeah. But this is what we’re talking about, yes? Right. Then I would say something like this: take that picture of a frog’s egg on the board, which is, at any rate, simpler than the sand paintings.

To begin with, any organism who’s got half a brain and two eyes will know that that drawing on the board is an organic product. It’s not created by geology, it’s not created by a machine. It’s created by a self-corrective critter. Now, the difference between a line which is ruled and a line which is freehand drawn is just a monstrous, qualitative difference; pattern difference. A ruled line doesn’t look like a line that’s been drawn freehand. You cannot make a line freehand that would look as if it’s been ruled.

Now, this takes us, you see, at once into the strange realm of recognition of one living for another, one product of life to recognize a product of life. The primrose by the river’s brim, a yellow primrose was to him, and it was nothing more, you know? Assumes that he (whose name I’ve now forgotten; who was Wordsworth’s enemy in that poem?) is walking around the world like a scientist whose only organ of perception is a foot-rule, or a colorimeter, you see? And he sees a yellow was to him. He does not see that the lines and symmetries of the primrose—or of that drawing or whatever—are organic. Now, the fact that that sand painting is organic is right in the center of your saying it is [???] a great sand painting or not. It hits you as something alive. It, as a product of something alive, hits you as a living thing. Do I answer your question?

Now, if you want to do an analysis of what are the actual characteristics of a freehand drawn line as against a ruled line, you can list those and you can measure them, and so on and so on, and there isn’t going to be any magic there, you know? Because there’s pattern. A straight line drawn freehand has pattern.

1:28:09

Bateson

If you’re a living creature, then the only thing, really, that you care about much is living creatures. That’s true. [???]

1:28:39

Bateson

Well, I can only say—and must say in self-defense—that to say it depends upon difference is a long step towards what you want from to say it depends upon quantity. Now, you want to go further? I’m willing to go along with you. But I would only claim to have made a first step in that direction. There are then questions of differences of difference, you know? And a lot of stuff out beyond which becomes very rapidly more complex than you can think about. And I think the essence of that sand painting, and such qualities that drawing of a frog’s egg has, gets almost immediately into differences more complex than your conscious analysis can handle—or my conscious analysis. This doesn’t mean it’s in a different physical realm or something like that; that’s another question.

Let me just tease you with a [???]. [???] in the thirties, did a set of experiments on the human perception of difference. He was out to straighten out the phrasings of the [???] law. This is the law which says you can only perceive ratios and you cannot perceive quantities. He has an illuminated disk, and a lamp behind that, a lamp behind that, and he can control the illuminations with rheostat accurately. And he says to the subject, “Are those two different? And if so, which is the brighter?” [???] in empathy for the subject of that experiment. But it’s fairly easy to say which is brighter and whether one is brighter than the other. There will be, of course, very fine edges where you won’t be able to tell. But if you can tell, you can tell. And you can tell which is the brighter. Alright. We now ask you: “You agree there was a difference there, and that was the brighter. You agree there is a difference there, and one of them is the brighter. Which is the greater of those two differences?” And I think you can probably imagine being able to answer that question. We will say that that was the greater of those two differences. And now there is a similar thing here. Now I want to know whether the difference between the differences here is the greater or less than the difference between the differences there. And at this point you begin to find you’ve got a headache! Very soon, the thing gets beyond your ability to feel a meaning to it; the question. Of course, if you want to be a laboratory scientist, all you have to do is to assign values to these, you see. Say, on a ten-point scale, whatever it may be. Get out an envelope and do the arithmetic and you’ll get an immediate answer on the ratios. But that’s different. You’ve now, you see, ignored the fact that it’s really a matter of perception of difference, substituted a perception of quantity, moved into a different universe of discourse altogether, flattened out the logical typing problems of what is the difference between two differences. And you can do that. But when you do that, you’ve moved out of the realm of natural history, out of the realm of the question: what sort of a thing is a human being? Into a different sort of realm and a different sort of problem. And then you’d be fully justified, you see, in saying that I’m talking nonsense about quality.