
Edge
164— June 27, 2005 
Spiders (2005)
The distinguished Hamburg Bitfilm Festival annually documents how digital technology is revolutionizing filmmaking and photography. For the first time Hubert Burda Media and the HypoVereinsbank, in cooperation with Bitfilm, present this forum for professionals and creative people from all world on 22 July in Munich at "Bundesgartenschau in München 2005" (BUGA 05), the German National Garden Show. In a "Best OF" program bitfilm@BUGA the most creative and innovative digital work of the past year is demonstrated. 3D animations, digital effects and Flash Movies are shown plus films, which develop in the virtual worlds of computer games and small EXE files, which produce in the computer in real time 3DAnimation— an amazing cross section of film, photography, art and technology — inspired! All are invited to enjoy the summer evening in of Munich's most beautiful garden with the Hamburg VJs of Eins23.tv, who presents a Live Remix of the DVD "Brazilectro", underlaid with Digi tieclips from Rio de Janeiro and São Paulo. Followed of scene DJ Nikias Hofmann (P1, secret society) with open END program; plus the digital images of the New York artist Katinka Matson. 
How do the microprocesses of cultural transmission affect the macro structure of culture, its content, its evolution? The microprocesses, the smallscale local processes I am talking about are, on the one hand, psychological processes that happen inside people's brains, and on the other hand, changes that people bring about in their common environment— for instance the noise they make when they talk or the paths they unconsciously maintain when they walk—and through which they interact. Just as the human mind is not a blank slate on which culture would somehow imprint its content, the communication process is not a xerox machine copying contents from one mind to another. This is where I part company not just from your standard semiologists or social scientists who take communication to be a codingdecoding system, a transmission system, biased only by social interests, by power, by intentional or unconscious distortions, but that otherwise could deliver a kind of smooth flow of undistorted information. I also part company from Richard Dawkins who sees cultural transmission as based on a process of replication, and who assume that imitation and communication provide a robust replication system. AN
EPIDEMIOLOGY OF REPRESENTATIONS [7.27.05] 
It's Summer, time to lie on the beach and relax with a wonderful book. Here's a selection of 40 recently published great Summer reads from the Edge community. Read Mandelbrot on "multifractals", Dawkins on "true heredity", Penrose on "Clifford bundles", Marcus on "synaptic strengthening", Searle on "biological naturalism", Leroi on "intersex", Pinker on "biological nature", Garreau on "the telekinetic monkey", Seligman on "avoidant people", Randall on "extra dimensions", Kurzweil on "the singularity", Damasio on "neurotransmitter nuclei", Greene on "quantum weirdness", Dennett on the "Zombic Hunch", Diamond on anthropology to zoology, plus many others. You can't go wrong. Third culture books from: Simon BaronCohen, John Barrow, Paul Bloom, John Brockman, David Buss, Antonio Damasio, Richard Dawkins, Daniel C. Dennett, Keith Devlin, Thomas DeZengotita, Jared Diamond, Niles Eldredge, Kenneth Ford, Joel Garreau, Neil Gershenfeld, Rebecca Goldstein, Brian Greene, Haim Harari, Sam Harris, Gerald Holton, John Horgan, George Johnson, Steven Johnson, Ray Kurzweil, Armand Marie Leroi, Benoit Mandlebrot, Gary Marcus, John Markoff, Earnst Mayr, Ian McEwan, James O'Donnell, Roger Penrose, Cliff Pickover, Steven Pinker, Lisa Randall, John Searle, Martin E.P. Seligman, Michael Shermer, Robert Trivers, Edward O. Wilson. [more] 
GÖDEL
AND THE NATURE OF MATHEMATICAL TRUTH
II
[7.27.05] I doubt that pure philosophical discourse can get us anywhere. Maybe phenomenological narrative backed by psychological and anthropological investigations can shed some light on the nature of Mathematical Truth. As to Beauty in mathematics and the sciences, here speaks Sophocles' eyewitness in Antigone:
[more] 
How do the microprocesses of cultural transmission affect the macro structure of culture, its content, its evolution? The microprocesses, the smallscale local processes I am talking about are, on the one hand, psychological processes that happen inside people's brains, and on the other hand, changes that people bring about in their common environment— for instance the noise they make when they talk or the paths they unconsciously maintain when they walk—and through which they interact. Just as the human mind is not a blank slate on which culture would somehow imprint its content, the communication process is not a xerox machine copying contents from one mind to another. This is where I part company not just from your standard semiologists or social scientists who take communication to be a codingdecoding system, a transmission system, biased only by social interests, by power, by intentional or unconscious distortions, but that otherwise could deliver a kind of smooth flow of undistorted information. I also part company from Richard Dawkins who sees cultural transmission as based on a process of replication, and who assume that imitation and communication provide a robust replication system. AN
EPIDEMIOLOGY OF REPRESENTATIONS [7.27.05] Introduction Dan Sperber is a French anthropologist who has focused on the more cognitive, more naturalist, approaches linked to evolution. "For a long time," he says, "my ideas were not very well received among anthropologists. They’ve been discussed a lot, but I found myself spending too much time with my fellow anthropologists arguing the basics of the field rather than moving forward in research. I got involved in linguistics, experimental psychology, philosophy of science, evolutionary biology, and lots of fascinating topics—and continuing also the conversation with anthropologists. Anthropology is a discipline that has been in crisis all my life." Dan Sperber's parents were both easternEuropean Jews; his father, Manes Sperber, a famous novelist, was born in Galicia, grew up in Vienna, then moved to Germany. He met his mother, who came from Latvia, in France in the 30s . Manes Sperber was a Communist, was very active in the party, but left the party at the time of the Moscow trials. Sperber was born in France. "That's my culture," he says. "I am French. Still, there are French people who are much more French than I am. They have roots as they say, but the image of roots has always made me smile. You know, I'm not a plant."The reason he gives for having become an anthropologist is that he was raised an atheist. There was no god in the family. His father, Manes Sperber, was from a Jewish family, had refused to do his bar mitzvah, and he transmitted zero religion to his son, but at the same time, he had deep respect for religious people. There was no sense that they are somehow inferior. This left the young Sperber with a puzzle: how can people, intelligent decent people, be so badly mistaken? Sperber is known for his work in developing a naturalistic approach to culture under the name of "epidemiology of representations", and, with British linguist Deirdre Wilson, for developing a cognitive approach to communication known as "Relevance Theory". Both the epidemiology of representations and relevance theory has been influential and controversial. He is also known for his early work on the anthropology or religion, in which he tried to understand, in a generalist manner and in a positive way (i.e. without making them into idiots), why people could be religious. He took part in classical anthropological studies but he also argued from the start that you have to look at basic innate mental structures, which, he argued, "played quite an important role in the very possibility of religious beliefs, in the fact that, more generally, beliefs in the supernatural fixate in the way they do in the human mind, are so extraordinarily catching". Sperber's "catchiness", a theory he has been exploring for a generation, connects with Malcolm Gladwell's idea of a "tipping point". "I've never met Gladwell, " he says, "but when his book came out, many people sent me the book, or told me to read it, telling me that here's the same kind of thing you've been arguing for a long time. Yes, you get the kind of epidemiological process of something gradually, almost invivibly spreading in a population and then indeed reaching a “tipping point.” That's the kind of dynamic you may find with epidemiological phenomena. Still, I don't believe that Gladwell or anybody else, myself included, has a satisfactory understanding of the general causes of the dynamics of cultural distribution.” " Now, if I could just write with the slickness of Gladwell, and coin one of his bestselling titles such as Blink! or The Tipping Point. . . but I guess I would also have to give up trying to convey much of the hard substance of my work. Oh well..". Edge is pleased to present An Epidemiology of Representations: A Talk with Dan Sperber. — JB Sperber holds a research professorship at the French Centre National de la Recherche Scientifique (CNRS) in Paris, and has held visiting positions at Cambridge University, the British Academy, the London School of Economics, the Van Leer Institute in Jerusalem, the Institute for Advanced Study in Princeton, Princeton University, the University of Michigan, the University of Bologna, and the University of HongKong.
AN EPIDEMIOLOGY OF REPRESENTATIONS [DAN SPERBER:] What I want to know is how, in an evolutionary perspective, social cultural phenomena relate to psychological mental phenomena. The social and the psychological sciences,when they emerged as properly scholarly disciplines with their own departments in the nineteenth century took quite different approaches, adopted different methodologies, asked different questions. Psychologists lost sight of the fact that what's happening in human minds is always informed by the culture in which individuals grow. Social scientists lost sight of the fact that the transmission, the maintenance, and the transformation of culture takes place not uniquely but in part in these individual psychological processes. This means that if what you're studying is culture, the part played by the psychological moments, or episodes, in the transmission of culture should be seen as crucial. I find it unrealistic to think of culture as something hovering somehow above individuals — culture goes through them, and through their minds and their bodies and that is, in good part, where culture is being made. I've been arguing for a very long time now that one should think of the evolved psychological makeup of human beings both as a source of constraints on the way culture can develop, evolve, and also, of course, as what makes culture possible in the first place. I've been arguing against the now discredited "blank slate" view of the human mind—now splendidly laid to rest by Steve Pinker—but it wasn't discredited when I was a student, in fact the "blank slate" view was what we were taught and what most people went on teaching. Against this, I was arguing that there were specific dispositions, capacities, competencies, in the human mind that gave rise to culture, contributed to shaping it, and also constrained the way it can evolve — so that led me to work both in anthropology—and more generally in the social sciences—,which was my original domain, and,more and more, in what was to become cognitive sciences. In those years, the late 60s, psychology was in the early stags of the “cognitive revolution.” It was a domain that really transformed itself in a radical manner. This was, and still is, a very exciting intellectual period in which to live, with, alas, nothing comparable happening in social sciences, (where little that is truly exciting has happened during this period in my opinion). I wanted the social sciences to take advantage of this revolution in the study of cognition and I've tried to suggest how this could be done. How do the microprocesses of cultural transmission affect the macro structure of culture, its content, its evolution? The microprocesses, the smallscale local processes I am talking about are, on the one hand, psychological processes that happen inside people's brains, and on the other hand, changes that people bring about in their common environment—for instance the noise they make when they talk or the paths they unconsciously maintain when they walk—and through which they interact. Just as the human mind is not a blank slate on which culture would somehow imprint its content, the communication process is not a xerox machine copying contents from one mind to another. This is where I part company not just from your standard semiologists or social scientists who take communication to be a codingdecoding system, a transmission system, biased only by social interests, by power, by intentional or unconscious distortions, but that otherwise could deliver a kind of smooth flow of undistorted information. I also part company from Richard Dawkins who sees cultural transmission as based on a process of replication, and who assume that imitation and communication provide a robust replication system. A good part of my work has been to study, in large part with British linguist Deirdre Wilson, the mechanisms of human communication and show that they're much more complex and interesting than is generally assumed, and much less preservative and replicative and more constructive than one might think: understanding involves a lot of construction, and not just reconstruction, and very little by way of simple replication. When you are told something, the simple view of what happens would be: 'ah! These are words, they have meaning,' and so you decode the meaning of the word and you thereby understand what the speaker meant. A more realistic and, as I said, also a more interesting idea is that the words don't encode the speaker meaning, they just give you evidence of the speaker's meaning. When we speak we want our audience to understand something that's in our mind. And we have no way to fully encode it, and trying at least to encode as much as possible would be absurdly cumbersome. Linguistic utterances, however rich and complex they may be, cannot fully encode our thoughts. But they can give strong richly structured piece of evidence of what our thoughts are. From the point of view of the audience, a speaker is providing rich pieces of evidence, which we interpret in a context of shared background knowledge, drawing on the common cultural, on the local situation, on the ongoing conversation, and so on. You construct a complex representation helped by all these different factors. You to end up with something which will have been strongly guided, sometimes guided in an exquisitely detailed manner, by the communication, by the words used by the speaker, but which end up being a thought of your own, relevant to you, a recognition, to begin with, of what the speaker meant, from which you extract what is relevant to you. We're not that interested when we try to comprehend what others say, in getting in our minds a copy of what they had in mind, we're interested in getting that which is of use and of relevance to us, and we see what others are trying to tell us as a source of insight and information from which we can indeed construct a thought of our own. The same is true of imitation; rarely are you concerned when you imitate other people's behavior in copying them exactly. What you want when you see others doing something that you think is worth doing, for instance, cook a soufflé, it's not to copy the exact gestures and the exact souffle that you saw, with its qualities, and also maybe its defects, your goal is to cook a good soufflé, your good soufflé. The goal of these partly preservative processes of communication and imitation is not to copy per se, but to take advantage of information provided by others in order to build thoughts of our own, knowledge of our own, objects of our own, behaviors of our own, for which we take part of the responsibility. The process is constructive in that sense. Communication is a very broad notion —one should ask whether it makes sense to look for a general theory of communication, given that the notion covers such a variety of processes — processes of communication among machines; biologists talk about communication among cells; by “animal communication” biologists mean also unitentional deception as when the viceroy butterfly has wings mimicking the pattern found on the poisonous monarch butterfly, so as not to be eaten by predator birds, and so on. All these form of communication and many others are communication in a very broad sense where some information—in some broad sense of information too—is provided by one device or organism, and is used by another. There are some commonalities linked to this general definition of communication, and indeed, Shannon and Weaver for instance were interested in such a very basic notion. But if we think of communication in biological terms, it is not clear that we have the subject matter of a useful general theory. Think of locomotion. How much can you get from a general theory of locomotion, even sticking to the biological domain and leaving aside artifacts, airplanes, cars, bicycles. I doubt that there is much to get from a general theory of locomotion that would cover fish swimming, birds flying, snakes crawling, us walking, and so on. If you're studying human locomotion, then you look at the specific organs, the way, for instance, we do it, why we do it, what evolutionary pressure have selected our particular way of doing it. Even more—much more—than human bipedal upright walking, human communication is very special, it's quite unlike the communication you find in other animals. Not just because of language, which indeed has no real equivalent among other species, but also because of another reason which is also quite remarkable but that has not been stressed, and on which Deirdre Wilson and I have been doing a lot of work, namely that if you look at human languages as codes — which in a sense they undoubtedly are — they are very defective codes! When say, vervet monkeys communicate among themselves, one vervet monkey might spot a leopard and emit an alarm cry that indicates to the other monkeys in his group that there's a leopard around. The other vervet monkeys are informed by this alarm cry of the presence of a leopard, but they're not particularly informed of the mental state of the communicator, and they don't give a damn about it. The signal puts them in a cognitive state of knowledge about the presence of a leopard, similar to that of the communicating monkey — here you really have a smooth codingdecoding system. In the case of humans, when we speak we're not interested per se in the meaning of the words, we register what the word means as a way to find out what the speaker means. Speaker’s meaning is what's involved. Speaker’s meaning is a mental state of the speaker, an intention he or she has to share with us some content. Human communication is based on the ability we have to attribute mental state to others, to want to change the mental states of others, and to accept that others change ours. When I communicate with you I am trying to change your mind. I am trying to act on your mental state. I'm not just putting out a kind of signal for you to decode. And I do that by providing you with evidence of a mental state in which I want to put you in and evidence of my intention to do so. The role of what is often known in cognitive science as "theory of mind," that is the uniquely human ability to attribute complex mental states to others, is as much a basis of human communication as is language itself. I am full of admiration for the mathematical theory of information and communication, the work of Shannon, Weaver, and others, and it does give a kind of very general conceptual framework which we might take advantage of. But if you apply it directly to human communication, what you get is a mistaken picture, because the general model of communication you find is a codingdecoding model of communication, as opposed to this more constructive and inferential form of communication which involves infering the mental stateof others, and that's really characteristic of humans. I have been developing my own approach to culture under the general heading of "epidemiology of representations". The first thing to do, of course, is to take away the negative connotation of epidemiology — it's not the epidemiology of diseases — epidemiology is the study of the distribution of certain items or conditions in the population. One can study the distribution of particular pathological conditions, but you can also study the distribution of good habits, or thoughts, or representations, artifacts, or forms of knowledge. I'm not assuming that culture is good — I don't want to have a cultural epidemiology to be on the side of the angels, as opposed to medical anthropology on the side of the demons. What's I like about epidemiology is that it's the one social science that is truly naturalistic in studying what happens in populations, typically in human populations, and it explains the macro phenomena at the level of population such as epidemics, by the aggregation of the micro processes both inside individuals and in their interaction. I believe that the cultural and the social in general should be approached in the same manner. Of course I'm not the only one to do that, a number of people, mostly coming from biology, like Luigi CavalliSforza, Marcus W. Feldman, E.O Wilson and Xharles Lumsden, Richard Dawkins, Bill Durham, Robert Boyd, and Peter Richerson, have developed different conceptions which in this broad sense are epidemiological, or, another way to put it: they are forms of "populationthinking" applied to culture. You take what happens at the population level to results from the microprocesses affecting individuals in the population. Dawkins, who is particularly clear and simple in a good way in his approach, offers a contrast to my approach. For Dawkins, you can take the Darwinian model of selection and apply it almost as is to culture. Why? Because the basic idea is that, just as genes are replicators, bits of culture that Dawkins called “memes” are replicators too. If you take the case of population genetics, the causal mechanisms involved split into two subsets. You have the genes, which are extremely reliable mechanisms of replication. On the other hand, you have a great variety of environmental factors — including organisms which are both expression of genes and part of their environment —, environmental factors that affect the relative reproductive success of the genes. You have then on one side this extremely robust replication mechanism, and on the other side a huge variety of other factors that make these competing replication devices more or less successful. Translate this into the cultural domain, and you'll view memes, bits of culture, as again very strong replication devices, and all the other factors, historical, ecological, and so on, as contributing to the relative success of the memes. What I'm denying, and I've mentioned this before, is that there is a basis for a strong replication mechanism either in cognition or in communication. It's much weaker than that. As I said, preservative processes are always partly constructive processes. When they don’t replicate, this does not mean that they make an error of copying. Their goal is not to copy. There are transformation in the process of transmission all the time, and also in the process of remembering and retrieving past, stored information, and these transformations are part of the efficient working of these mechanisms. In the case of cultural evolution, this yields a kind of paradox. On the one hand, of course, we have macro cultural stability — we do see the same dish being cooked, the same ideologies being adopted, the same words being used, the same song being sung. Without some relatively high degree of cultural stability—which was even exaggerated in classical anthropology—, the very notion of culture wouldn't make sense. How then do we reconcile this relative macro stability at the cultural level, with a lack of fidelity at the micro level? You might think: if it's stable at the macro level, what else could provide you this macro stability apart from the faithful copying at the micro level? It's the only possible explanation that most people think of. But that's not the only one, and it’s not even a plausible one. Dawkins himself has pointed out that each act of of cultural transmission may involve some mistakes in copying, some mutation. But if that is the case, then the Darwinian selection model isunlikely to apply, at least in its basic form. The problem is reconciling this macro stability with the micro lack of sufficient fidelity. The answer, I believe, is linked precisely to the fact that in human, transmission is achieved not just by replication, but also by construction. If it were just replication, copying, and there were lots of errors of copy all the time, then nothing would stabilize and it’s unlikely that the selective pressures would be strong enough to produce a real selection comparable to the one you see in biology. On the other hand, if you have constructive processes, they can compensate the limits of the copying processes. What happens is this. Although indeed when things get transmitted they tend to vary with each episode of transmission, these variations tend to gravitate around what I call "cultural attractors", which are, if you look at the dynamics of cultural transmission, points or regions in the space of possibilities, towards which transformations tend to go. The stability of cultural phenomena is not provided by a robust mechanism of replication. It's given in part, yes, by a mechanism of preservation which is not very robust, not very faithful, (and it's not its goal to be so). And it’s given in part by a strong tendency for the construction — in every mind at every moment— of new ideas, new uses of words, new artifacts, new behaviors, to go not in a random direction, but towards attractors. And, by the way, these cultural attractors themselves have a history. Dawkins, of course, is only one of the people who have proposed new ways of modeling cultural evolution. He's important because he brings it down to the simplest possible version — there's a great merit in simplicity. He sees cultural evolution at the same time as being analogous to biological evolution, and as being an evolution almost independent from biological evolution: it has just been made possible by the biological evolution of homo sapiens, which has given us the mind we have, and which, so the story goes, makes us capable indeed of endlessly copying contents. We are supposed to be imitation machines, “meme machines” to use Susan Blakemore's phrase, and this explains that. Dawkins, in a strange way, presents something very similar to the blank slate view of the mind. The blank slate view, as I was taught it in anthropology, says the human mind is capable of learning anything — whatever content would be provided by culture can be written on the blank slate. Well, the general imitating machine does more or less the same thing. It's capable of imitating just whatever type of content it is presented with, and the relative success of some contents against others, has to do with the selective forces. The idea that the human mind is such a kind of universal imitation machine is hardly better psychology, in my view, than the blank slate story. Others, E.O Wilson and Charles Lumsden, Rob Boyd and Pete Richerson, have asked to what extent the evolved dispositions that both constrain and make possible culture are, in return, affected by cultural evolution itself so as to yield a kind of geneculture coevolution. Instead of having two evolutionary scenarios running in parallel, one biological evolution, the other cultural evolution, you get some degree of interaction, possibly a strong interaction, between gene and culture. The general idea has got to be correct. The details, in my opinion, are still very poorly understood. For a variety of reasons, I believe that memes are not the right story about cultural evolution. This is because in the cultural case, replication is not very successful in explaining cultural stability. I also believe that among the factors we need to take into account to explain cultural attraction of which I was talking before, are evolved aspect of the human psychology. The one type of scholarship and research that has to be brought into the picture, in my view, is evolutionary psychology, as defended in particular in the work of Leda Cosmides, John Tooby, Steve Pinker and taken up in more critical ways by a growing number of developmental psychologists and of philosophers. To understand culture, we have to understand the complexity of the psychological makeup of human beings. We have to go to really deep psychology, understood both in a richly cognitive manner and with a proper evolutionary perspective, to put start explaining cultural evolution. We need a representation of a human mind that's complex in an appropriate manner, true to the empirical data, and rich enough indeed to explain the regularities the, stability, and the variability of culture. This is them a different story, but it’s still a Darwinian story. It's a Darwinian story in the sense that it's an application of population thinking, which tries to explains the macro phenomena in terms of a micro processes and properties, and which doesn’t assume that there are types or essences of macro cultural and social things. Macro regularities are always the outcome of distribution of micro features, evolving all the time. In this Darwinian story however, instead of causal processes in culture as split between robust replication devices and a variety of selection factor, we have a much more promiscuous form of causality. Cultural causality is promiscuous. Constructive processes always interfere with preservation processes. So we need to build models different from standard Darwinian models of selection, in order to arrive at the right way to draw on Darwinian inspiration with regard to culture, that is, we must generalize Darwin to the cultural case, rather than adjust it in a way which twists the data well beyond what is empirically plausible. ~~ The idea of God isn't a supernatural idea. If the idea of God were supernatural, then religion would be true. The idea of God, the idea, the representation of something supernatural is not itself supernatural. If it were, then we would be out of business. Precisely what we're trying to explain is, to quote the title of a book by Pascal Boyer, the “naturalness of religious ideas,” explain, in other terms, how these ideas of the supernatural can occur in the natural beings we are, in human brains and minds and culture, and have the kind of success that they have, in spite of the fact that you can't explain them in the way that you explain so many human ideas, such as ideas that are acquired through experience of the things they are about. We humans have ideas about plants and animals because we experience plants and animals in a special way with the brain we have. We don't experience God, or goblins or witches, because there are no such things. Nevertheless, we have rich complex ideas about them, a richness in many ways comparable to the ideas we have about plants, animals and the natural things around them. How is that possible? The issue is what makes these kind of ideas psychologically, cognitively attractive — "catching", such that they stay with you in your head and you may want to communicate them and to guide your behavior on their basis. And also: which of them, among all the unrealistic unsupported ideas that are possible in infinite variety, are going to be so "catching" as to achieve cultural success, in the manner of the many religious ideas that has been around for centuries? It's not like any blatantly false idea will somehow make it to a cultural success — far from it. Most of them don't stand a chance. What's special about ideas of the supernatural? I argued long ago that it had to do with the fact that they are rooted in our cognitive dispositions, in the way we approach the natural world. Instead of departing from our commonsense ideas so to speak at random, they're like direct provocation — they have always an aspect of going directly against what should be the most intuitively obvious. So for instance it's part of our common sense knowledge of of living forms, that an animal can’t be both a dog and a cat, but the supernatural is full of creatures like dragons that typically belong to several species simultaneously. It's part of our common sense knowledge of the physical world that an entity cannot be in two places simultaneously, but ubiquity is a distinctive trait of supernatural beings. It's kind of again commonsense, in our commonsense psychology which we deploy in everyday interaction with one another, that one's visual perceptions are limited to what's present in front of one's eyes. Supernatural beings typically can see the past, the future, and things on the other side of earth. So supernatural beings are kind of provocations to commonsense. They are really deeply counterintuitive. That's an idea I suggested a long time ago and that Pascal Boyer has developed and enriched in a remarkable fashion, and which I think is one of the cognitive ingredients that helps explain the success of religious ideas. Of course, it's only one little fragment of a kind of complex picture. ~~ I started as an anthropologist. Precisely because they were more cognitive, more naturalist, more linked also to evolution than most, for a long time, my ideas were not very well received among anthropologists. They’ve been discussed a lot, but I found myself spending too much time with my fellow anthropologists arguing the basics of the field rather than moving forward in research. I got involved in linguistics, experimental psychology, philosophy of science, evolutionary biology, and lots of fascinating topics—and continuing also the conversation with anthropologists. Anthropology is a discipline that has been in crisis all my life. When I started the crisis was linked to the end of the colonization. Anthropology had developed during the period of colonization, as a kind of ancillary science for colonial enterprise. At the same time so many anthropologists were actually active in anticolonialist movement, and that was also one of the reasons I came to anthropology. But, the decolonisation, anthropology lost this kind of historical and sociological context. Anthropologists in the 60s, 70s, were asking about their political role, about whether or not we were on the right side. Anthropologists started studying themselves and trying to reflect on their own situation. It was a kind of reflective anthropology, which had a number of interesting aspects. I certainly don't think it was useless although it became a bit obsessive. Parallel to these developments, were the poststructuralist and then postmodernist movements in the humanities and the social sciences, the development of “cultural studies,” and many anthropologists felt at ease in these movements. This produced a new kind of discourse, taking the study of other cultures as much as a pretext as a subject matter to be investigated in a standard scholarly manner. Again, some of the products of this appraoch are of genuine interest, but on the whole more harm has been done than good. While this was happening, others, in part in reaction against this turn toward the literary in anthropology, moved on the contrary toward a more naturalistic anthropology. They became interested in social biology, in biological anthropology. What you find now in anthropology departments is that people can't talk to each other. Some universities have now had two anthropology departments. So anthropology is stilll in crisis, even if it is not the same crisis. You can look at such a crisis from an institutional or from an intellectual point of view. Universities as we know them emerged in the nineteenth century and unerwent major changes, in particular after World War II. It does not make sense to project this short past into an indefinite future. In fact, universities are evolving, transforming themselves beyond recognition. The biggest changes are will be due to new communication technology. There is also now a big and blatant gap between the structure of departments in universities, which have to do with institution of transmission of knowledge, and which seem to define stable domains such as psychology, anthropology, sociology, and the real ongoing research which is structured in new ways — in the form of creative, or dynamic, research programs, that may fall within a traditional discipline, or, more often, across several traditional disciplines. Depending on the productivity of such dynamic programs, they are can go on for ten years, 20 years, 30 years, or more. It is these dynamic research programs that interest me; I've been involved in several, and that's what I find to be intellectually exciting. When we say anthropology is in crisis we're talking about anthropology as defined by academic institutions. And it doesn't matter. It deserves to be in crisis; it deserves to explode, let it do so. 
GÖDEL
AND THE NATURE OF MATHEMATICAL TRUTH
II
[7.27.05] I doubt that pure philosophical discourse can get us anywhere. Maybe phenomenological narrative backed by psychological and anthropological investigations can shed some light on the nature of Mathematical Truth. As to Beauty in mathematics and the sciences, here speaks Sophocles' eyewitness in Antigone:
A true Realist, a true Platonist will not stoop to choose between Beauty and Truth, he will have the tenacity to stick it through until Truth is caught shining in her own Beauty. Sure there are messy proofs, we have to bushwhack trough a wilderness of ad hoc arguments, tours de force, combinatorial jungles, false starts and the temptations of definitions ever so slightly off target. Eventually, maybe not in our own lifetime, a good proof, a clear and beautiful proof will be honed out. That, I think, is the belief of the true Platonist. What Gödel and Einstein were doing when walking together over the Institute's grounds may have been just that; bush whacking, comparing mental notes and encouraging each other not to give up while getting all scratched and discouraged. Yet finding solace in speaking to each other in their mother tongue about their deepest concerns, and the state of the cosmos, the world, the weather and their households too. ~~
Introduction Verena
HuberDyson, a Swiss national born in Naples in 1923, was
educated in Athens before returning to Zurich to study
mathematics (with minors in physics and philosophy), obtaining
her PhD under Andreas Speiser in 1947. She moved to the
United States in 1948, and happened to be in exactly the
right field, the right place, and the right time to witness
her two particular areas of interest — group theory
and formal logic — have an unexpected impact
(via particle physics and digital computing) on the real
world.
She considers herself an Intuitionist, and this prompts the question she is asking herself:
Some
of what follows is a challenge for the layperson...a rewarding
challenge. "People are not prepared to roll up their
sleeves and do some hard thinking and figuring," HuberDyson
says, "and read a book with a a pad of paper and a few
sharp pencils on the side, or their laptop in action. Reminds
me of some of my colleagues, who said 'Oh it's brilliant,
but soooo dense'. But to make up for that I had a
few students who thrived on my dense cooking." VERENA HUBERDYSON is emeritus professor of the Philosophy department of the University of Calgary, Alberta Canada, where she taught graduate courses on the Foundations of Mathematics, the Philosophy and Methodology of the sciences. Before the Vietnam war she was an associate professor in the Mathematics department of the University of Illinois. She taught in the Mathematics department at the University of California in Berkeley. She is the author or a monograph, Gödel's theorems: a workbook on formalization, which is based on her experience of teaching graduate courses and seminars on mathematical logic, formalization and its limitations to mathematics, philosophy and interdisciplinary students at the Universities of Calgary, Zürich and Monash. She lives in Berkeley, California. GÖDEL AND THE NATURE OF MATHEMATICAL TRUTH II Great to find the "Nature of Mathematical Truth" up for discussion on Edge! We do need interdisciplinary dialogue and Edge is certainly offering a fine Forum. My comments will attempt to fill in context and add a touch of reality to the picture of the creative mathematician. Creativity being taken for granted by the profession, I prefer to avoid words that are losing their luster by indiscriminate use. The hard working mathematician is struggling, be it with Platonic reality, the mysterious wiring of his brain or the limited power of human imagination. When at her wits end my mother used to bemoan the fate of man, doomed forever to rattle at the gates of his limitations (an seinen Grenzen zu rütteln). Incompleteness is indeed a limitation phenomenon. Most Edge dwellers will be familiar with what Gödel's proof achieved. But among the general public there is still a maze of deplorable misconceptions. I am incapable of making any comments before laying out the land. Respect for my audience prompting me to avoid cheap short cuts, I will need a few subtle distinctions to place Gödel's achievement into perspective. But, in deference to our IT media rather than to the audience, I will avoid formulas. We need to reach out and talk to each other. That is just what Edge is facilitating and I appreciate the opportunity to make a few clarifying comments to the current surge of Gödel lore. Gödel's notoriety had a slow start. It took more than a decade until the incompleteness theorem gained the attention of the general mathematical community. Onto the popular stage it jumped with a bang in 1979; Douglas Hofstadter's inspired act of popularization. Here is a true story that brings home the need for crossdisciplinary communication: I had heard of the book Gödel, Escher, Bach from some enthusiastic students at Calgary, but not seen it, when I visited the Tatski's in summer 1979 in Berkeley. Noticing the Hofstadter book in Alfred's bookcase I asked him what he thought of it and he exclaimed: "awful, of course it's awful". When I asked whether he had read it he said "of course not". But when I asked him to let me borrow it he got annoyed; the publishers had sent it to him, it belonged to him, and I could not borrow it, of course not. I must add that I had been working with Tarski, that we were old friends and that it is not too far fetched to call Tarski and Gödel the two pillars of modern mathematical logic. They were of the same generation and had come to America during the late thirties from Poland and from Austria respectively. But they were temperamentally totally different personalities. Tarski was an empire builder. Anyway, I went straight to Cody's and bought the Hofstadter book, fervently hoping that I would find it wonderful and could teach Alfred a lesson about his prejudices. Well, Alfred had been right after all. When I returned to Calgary at the end of the summer I found a copy of the Hofstadter book waiting for me with a note from the Canadian Journal of Philosophy asking me to review it. I actually had fun doing that review, I was younger then and instead of getting upset I enjoyed making it into a twin enterprise of a parody and a minimal nononsense account of Gödel's proof. What does get me down nowadays is to find the same kind of rash misconceptions still proliferating in the public arena. The popular writers don't consult the experts, or don't listen to them, and the experts don't encourage the public to ask and listen. Now, let me turn to the Edge 162 of June 8, 2005 headed by a synopsis of the story Rebecca Goldstein is telling:
This may sound interesting, but, disregarding history and context, it is misleading a potentially receptive audience. Ms Goldstein's claim to knowledge of Gödel's personal motivations is presumptuous. Her conjecture is, in fact, incorrect as a glance at Gödel's introduction to his famous paper, quoted below, shows. It is hard to talk about mathematical truth without having battled with it and understanding where we are at now with this conundrum. The distrust of natural language is age old, so is the flight into music, poetry and the abstractions of mathematics, to mention only a few escape hatches. Already Leibniz proposed a characteristica universalis. Russell and Whitehead aimed at an allencompassing formal Foundation in their Principia Mathematica and various versions of symbolic Logic were being investigated at the time of Hilbert's Foundations of Geometry (1899) and his incendiary proclamation, at the international congress of mathematicians in Paris (1900), of a list of tasks for the 20th century; among them a formal proof of the consistency of the axioms of arithmetic under the rules of classical logic. A formal consistency proof is not to appeal to the truth of axioms nor to the truth preserving properties of the deductive mechanism; it is expected to be based solely on the combinatorial structure of the system under investigation. Hilbert was demanding a finitary proof, a proof that could be encoded in arithmetic. Hilbert's dogmatic optimism concerning the powers of formalism is often exaggerated; his own fault, he had a tendency to polemic rhetoric. For balance of perspective consider a quote from his address on infinity of 1925:
Approaching Hilbert with today's demand for rigor and lucidity can be most frustrating. In how many a seminar session did I discuss his infinity paper with my most astute students and it still is not clear to me whether by the solvability of every mathematical problem Hilbert meant the law of the excluded middle or syntactic completeness. I will come back to this distinction in the section on Intuitionism. A brief glance at the three annotated collections of original papers on the foundational issues of that time by Jean van Heijenoort [2], by Martin Davis [3] and by Benaceraff and Putnam [4] will put the current Edge discussion into context for the interested reader. They contain translations of Gödel's first publication in 1931 of his incompleteness theorem: "On formally undecidable propositions of the Principia Mathematica and related systems." [6]: Königsberg, September 7, 1930 The story of the first announcement of the incompleteness theorem is told by Feferman et al. in the commentary to Gödel's completeness paper [5] in the first volume of Gödel's collected works [7]; as fascinating as any current news story of a scientific breakthrough. The occasion was the second conference on Epistemology of the exact sciences organized by the Society for Empirical Philosophy in Königsberg, held on 57 September 1930. On the 6th Gödel presented the results of his PhD thesis, that is, his proof of the equivalence between the formal concept of derivability and the semantic notion of consequence for first order languages. On Sunday the 7th the conference concluded with a discussion of foundational problems aroused by the key addresses of Carnap on Logicism, Heyting on Intuitionistic Foundations and von Neumann on the axiomatic method. Towards the end of that discussion Gödel apparently made his "seemingly casual but dramatically timed announcement" (Dawson) of the discovery of formally undecidable propositions. At first the audience was baffled and Gödel is described as somewhat tentative, but then he really got off the ground and now the editors of Vol 1 themselves are baffled, trying to explain Gödel's surge of boldness psychologically. My hunch is that he had more of a sense of humor and a knack for oneupmanship that he is ever credited for, his legendary aloofness notwithstanding. First Gödel observed that:
After that statement the editors of volume 1 insert three asterisks, to mark a pause of presumably baffled silence broken only by von Neumann's vigorous assent, pointing out that:
To which Gödel responded — after the editors' three asterisks — by announcing that:
adding that:
What a dramatic event that must have been — perfect for a stage presentation by a skilled playwright competent in mathematical logic. A few comments must suffice here, for more see [7] and [9]. First of all the translation of the qualification "inhaltlich" as opposed to formal by "contentual" is literal but awkward. Some translators use the term "material" instead. We all know what is meant, but I am still looking for a more appropriate and gracious word. "Inhalt" stands for content, but is meant here in the sense of cognitive content or meaning rather than stuff, of intension as opposed to extension. Bluntly put, the term refers here to facts of platonic reality. But there is no need to go that far. With his famous sure touch von Neumann narrowed down the term "inhaltlich" to the more specific qualification "intuitionistisch". It is significant that he, not only an exceptionally sharp mind but also an intuitionist in his philosophical views, at least at that time, was the only one to understand immediately what Gödel was talking about. It is most significant that even at that early stage von Neumann opened up the possibility of intuitionist reasoning transcending classical formalism, a point taken up by Gödel in 1933 [8], but quite unheard of at the time when everybody considered intuitionist logic a severe restriction of classical reasoning. Turning my attention to the editors for a moment, let me quote Dawson catching his breath after his three asterisks and commenting on Gödel's feelings (none of the editors are old enough to have been there in 1931):
Is it not most natural that, after having opened the scenario by casting doubt on expectations of completeness, Gödel would spend a few asterisks worth of deliberation on how best to articulate his findings? That settled, it was plain sailing. Von Neumann sure was a welcome second, but a leader? I should think that Gödel was emboldened by what he had to say next. To put the impact of Gödel's announcement into proper perspective I must backtrack a bit. Soundness, consistency and three forms of completeness In the 1930 paper, his PhD thesis, Gödel proved the completeness of first order predicate logic: every logically valid sentence of the pertinent formal language is derivable in this calculus. A symbolic sentence is called logically valid if and only if it becomes true under every interpretation of the language, the concept of an interpretation and truth under an interpretation being precisely defined within the realm of set theory. Whether based on naive or formal set theory, the proof is not constructive; it does not provide a method for deciding whether or not a sentence is valid. In other words, although, if a sentence is valid a proof for it can eventually be found by systematically running through all possible derivations in the calculus, there is no obvious method for finding counterexamples to sentences that are nontheorems. In fact, honing down Gödel's incompleteness proof to a finitely axiomatizable fragment of arithmetic eventually led to the conclusion that there cannot possibly be such an algorithm, at least not for any but the most anemic first order language. Thus, every reasonably expressive predicate calculus is what we call undecidable. The completeness proof for predicate logic ensured that there was nothing missing in the accepted predicate calculus, at least within the frame of classical logic, or so it seemed. It established the equivalence between derivability in the predicate calculus and logical validity, i.e., truth by virtue of logical structure alone. This first kind of completeness marks a link between truth and form. Just like Aristotle's syllogisms, only more so. The next step was to investigate completeness and consistency of theories obtained when extending the formalism by suitable axioms intended to describe fully and correctly certain familiar mathematical structures, in particular the Universe of Sets and the Arithmetic of Natural Numbers, the nonnegative integers. Two types of problems must be distinguished here: the syntactic and the semantic. The syntax treats a formal theory T as an abstract object, a conglomerate of strings of symbols (well formed formulas and sentences, wff's and wfs' for short) that have a certain structure and can be transformed into each other according to certain definite rules. The axioms are the starting wff's, the theorems are the end results of finite chains of wff's put together according to the rules. These chains are the proofs. The syntax of a system is independent of its intended meaning. A formal system is called syntactically consistent if never both a wfs and its negation have proofs, and syntactically complete if, for every wfs, either it or its negation is provable. These are philosophically untroubled notions, but methodologically of key importance; settling these problems can be technically most demanding tasks. There are plenty of theories, such as group theory, that are not meant to be complete in that sense, because they describe a whole tribe of structures. For them decidability is the big syntactic question: does there exist an algorithm for deciding, given a wellformed sentence, whether or not it is a theorem? The semantic notions are the ones that link a formal system to the philosophical problem of the nature of mathematical truth. A formalism is sound or semantically consistent, if only true wfs', i.e., wfs that make true claims, have proofs and semantically complete if all true wfs' are theorems. Unfortunately the unqualified term "complete" is often used for both its versions. In the ideal situation the syntactic and the semantic concepts coincide. There are a few nontrivial such phenomena, e.g., Tarski's completion [10] of Hilbert's formalization [12] of Geometry. But, before these questions could be tackled the notion of a formal system had to be characterized precisely, a task for which Gödel numbering or an equivalent streamlining of syntax is indispensable. It is not a deep idea but an extremely useful tool. The KK (Königlich Kaiserliche) administration of the Austro Hungarian Empire knew as well as the FBI and any efficiently run business that for easy handling it is most convenient to assign ID numbers to the objects of your attention. In a formal system symbols, strings of symbols, words, wff's, wfs' as well as finite chains of these, proofs among them, must all be Gödelnumbered effectively; there must exist a definite program for going back and forth between the language and the Gödel numbers. Talking about a mathematical environment in terms of numbers, just like a government keeping files on its citizenship in a numerical net, the question arises what can be said that way and what is bound to fall between the meshes of the net. Particularly valuable fish to catch are of course the proofs. And that is where Gödel's second insight struck: a reasonable but precise definition of the concept of a calculation is needed, be it to "calculate", given a number, whether or not it is the Gödel number of a proof or, given a nicely axiomatized formal system, to grind out one after the other the Gödel numbers of its theorems. When asked why they were taking Baby Logic or Linear Algebra instead of something more exciting for their electives my freshmen used say "because in Math's you cannot go wrong if you plug in the correct numbers and turn the crank according to instructions." Alas, by the end of the course some felt cheated, others claimed to have seen the light. A foolproof formal system is exactly the pipe dream of those freshmen. But setting it up is a challenge, checking whether or not it lives up to expectations a chore and, once all that is settled one way or the other, the system itself becomes an object of study and starts growing all sorts of new tentacles and barnacles. It is not an end in itself. It is hardly ever used for proving theorems, because that would be too tedious, time consuming and unenlightening. It is the leaps of intuition that shed light on a problem and point the way to a proof that explains a theorem and points beyond it. But the formalization of a discipline may serve as basis for applications and technology. Of course I am thinking now of computers. They can work so fast that they do not need any intuition. But they cannot judge what is a good proof or a significant theorem. In spite of all their garish displays they only know black and white, correct or false; at least so far. Speed, size and quantity of work, diligence, make up for their lack of imagination, common sense and judgment. They cannot leap over into infinity and build castles in the air! What distinguishes a proof from any other tool of persuasion is its decidability. A proof may be messy, dreary, tedious, or look like a joke, but there must be an unequivocal criterion for its validity, even if accessible to but a few specialists. For an arithmetized formal system this means that there must be a procedure for deciding, given the Gödelnumber of a sequence of formal sentences, whether or not it is that of a proof in the system. We want the decision function, which takes on the value 0 when fed the number of a proof and the value 1 when fed a number that is not the Gödelnumber of a proof, to be computable. Thus is raised the problem of characterizing the concept of a computable function. Encountering one it is easy to recognize it as such. When faced with instructions for putting together some gadget, if they are all there, you'll eventually get the job done, but until then you will not know whether you have complete instructions, and all along you'll be wondering: "if something's missing, how will I know for sure it's not just me?" We all know how to follow instructions, but if we want to prove something about calculable functions we need a definition that also tells us what is not a calculable function. We know that there are uncountably many more noncalculable functions than calculable ones, because, no matter how complicated, a computable function must be presented by a parcel of instructions for its calculation in some language or other intelligible to us finite creatures, and of such manuals there can only be countably many. That argument does not give a clue to finding a noncalculable function but is evidence for their abundance. At this point it should be clear that what Gödel needed for the "arithmetization" of a formal language was a precise definition of the intuitive concept of calculability for numerical functions. And that is what he invented recursive functions for, one of his major achievements, a bold stroke for which he would indeed deserve to be declared a hero of our times. The time was ripe for this step; there were alternative definitions by Emil Post, Alan Turing, Alonzo Church and others. These definitions were eventually all proven equivalent; sure confirmation that the definitions were on the right track — and, incidentally, bound for today's computer culture. Within the familiar idiom of computers we now have a precise notion of a Formal System. Couched in a symbolic language it is Gödelnumbered effectively. Shunting back and forth between a language and its Gödel numbers is something that a computer can do with more ease and less tripping than we humans with our limited attention span. A batch of Gödel numbered items is called effective if its Gödel numbers can be enumerated by a recursive function or, in other words, listed by a computer program. The basis for a theory consists of an effective list of axioms, expressed in the symbolic language, as well as an effective list of rules. A sequence of formal sentences consisting of axioms and wff's hanging together according to these rules is called a proof and its last entry is the theorem proved by it. An example of a first order language is what can be said in Elementary Arithmetic about the set of nonnegative integers including the constant 0 under the operations of the successor that transforms n into n+1, addition and multiplication and the natural ordering; a structure so concrete that it practically invites Platonism. We know what those numbers are and how to reach each one of them, in principle if not in practice. It's different with sets, at least to me. While claims like "unless the difference between two numbers is odd, their product equals the difference between two squares," which happens to be true, or "every even integer is the sum of two primes," still a conjecture, can be expressed, the simple fact that every non empty set of natural numbers contains a smallest one transcends first order expressive power. In an elementary (i.e., first order) language you can talk about individuals, about all and about some of them, but not about sets and families of sets of objects, which can be done in higher order. True elementary arithmetic consists of all true wellformed sentences of this restricted language. The axiomatic method builds a formal theory by selecting certain obviously true sentences as axioms and applying to them the deductive apparatus of first predicate calculus to derive ever more intricate theorems; a calculus proved complete in the first sense by Gödel in 1930. Thus the axiomatic method, intended to provide systematic approximations to truth, can only be trusted if the axioms are true and the rules of inference preserve truth. Once the definition of a formal system is articulated, its metatheory is developed within recursive arithmetic, the concept of proof is proved decidable by a recursive function and theoremhood is shown to be effective, all this demanding plenty of nitpicking labor. The rare case of syntactic completeness leads directly to decidability. A syntactically completely and consistently formalizable theory has an effective proof procedure as well as an effective disproof procedure and its decision function is easily pieced together from these two. Alas most of the mathematically meaningful axiomatic systems are doomed to remain approximations. For, Gödel showed how to construct, in any formal system that encompasses a minimal fragment of elementary arithmetic, a sentence, which is true if and only if it cannot be proved, and therefore is true but not provable, unless the system is unsound. His method has been expanded and applied to many systems, but it does not work in the first order theory of the field of the real numbers. This remarkable result by Tarski [10] shows how closely the limitations of a formal system are connected with their expressive power. More can be said and less proved in Arithmetic than in the RealNumberField. The integers are not definable among the reals! Interesting, when you consider how much more concrete, graspable, manageable and real(!) than the reals the integers appear to us. With all this talk about truth, we are not getting any closer to what truth is and how we come to recognize it. Well, if we could say precisely what it is, we would be proving Gödel a liar — and Tarski as well [11]. Consistency is not enough By now the big foundational debates of the twenties have not so much been resolved as swept under by highly technical developments. Hilbert's paper on the infinite [1] vividly illuminates the scene that Gödel found when he entered that quagmire. Hilbert was primarily concerned with consistency. The most unassailable way of reasoning is finitist. Over finite domains of discourse the predicate calculus boils down to the classical propositional calculus, which is indeed a cut and dried affair, alternatives being surveyable. Its extension to infinite domains leads to some dubious practices of the classical predicate calculus such as the acceptance of the absurdity of a universal claim as evidence for the existence of an object. To bring this remark down to earth: if you claim that there are honest people and I ask you why and you reply "they can't all be crooks" I am entitled to ask you to show me an honest person. Hilbert was fully aware of the risks lurking in the makebelieve world created by jumping from our finitary experience to conclusions about the Infinite. He was hoping that this jump would amount to no more than a convenient gimmick. But his mathematical conscience demanded a consistency proof. What kind of a consistency proof? That is where Hilbert tripped up. Wanted was a proof that the system using, what Hilbert called "ideal elements" — reasoning about infinite sets the way we are used to reason about finite ones — was not going to lead beyond the realm of what is justified by finitary reasoning, in other words, that the make believe world of predicate logic was a mere conservative extension of finitism. Alas, he wishfully thought that all he needed was a proof of simple consistency as defined above. That is why the demand for a finitist proof of the consistency of any of the current formal systems was his major concern. Exactly at that point Gödel unhinged Hilbert's dreams, but not his basic plan of proof theory; elaborating his Königsberger revelation Gödel pointed out that
Here we are confronting the subtleties of intuitionist reasoning. There is an important distinction between "all" and "each". It is conceivable that each proof fails to prove a certain sentence G while it is absurd to assume that all proofs fail to prove G. An undesirable situation; conclusive evidence that a mere proof of consistency could not be sufficient evidence for the soundness of a system, even if all its axioms were true. The weak link is in the logic. Here is a loosely technical elaboration:
The nature of that unsoundness is revealing. Assuming G is true:
Since "all
P's fail to prove G" is not provable in S, its negation
can consistently be added to S, resulting in a consistent extension
S' of S in which "not all P's fail to prove G" is a
theorem. According to Gödel's completeness proof [5] every
consistent system has a complete consistent extension. But in
classical logic the negation of a universal claim entails the
existence of a witness for that negation, so here we are led
to the conclusion that
A
ghost such a P would be, a genuine inhabitant of the make believe
world of classical logic
This Gedankenexperiment highlights essential shortcomings of the formalism. There is no way of making sure that "all" means "all natural numbers" and "some" means "some natural number". Just think how easy that would be if we were dealing with a finite set, say all numbers smaller than a million. The trouble comes in via the back door of classical reasoning. No matter how true the mathematical axioms may be, jumping to conclusions from analogy, not heeding Hilbert's own warning that what is OK for finite sets, may not be so for infinite ones, leads the unwary formalist astray, formal consistency notwithstanding. Guided by the wisdom of his intuitionist approach Gödel was on the right track. Remarks
on Intuitionism I have not been keeping track of recent developments, but my general hunch, based on experience, is that once technical interpretations start cropping up the philosophical roots of a way of thinking get overgrown by interesting technical developments and recede into twilight. Small wonder, because truly philosophical ideas are indeed uncomfortably elusive. It may well be true that the job of philosophy is to ask questions, not to answer them. Be that as it may, I'll just give an idea how an intuitionist's understanding of the elementary logical operations transforms the classical rules governing their use: The classical law of the excluded middle practically strangles imagination, freedom and intuition. "Either A or Else!" What do you do when confronted with a menu that says "either fish and chips or else"? The law of noncontradiction, however, makes good sense: never both A and not A. In classical logic the two are connected by boldly identifying the negation of a conjunction A&B with the disjunction: (not A or not B), which gets us to the law (notA or notnotA) , from which it is only a small step to (notA or A). Well that small step is another one of the pitfalls lurking in the shadows of classical logic. If you accept the law of the excluded middle the rule of double negation, which concludes A from notnotA, follows immediately: from notnotA and the make believe law (A or notA) follows A, since by the law of noncontradiction (notA & nonotA) is impossible. In a nutshell: intuitionist reasoning is a refinement of classical thinking. Identifying notnotA with A only when there is good reason to do so is much closer to everyday reasoning than the practice of the classical law of double negation. With a bit of care most of the popular but devious proofs of positive claims by reductio ad absurdum can be replaced by direct arguments. You can teach calculus without recourse to the parlor trick of producing, for every epsilon, a delta with property P out of a proof of the absurdity of the assumption that every delta fails to have that property; just take the trouble to construct, for every epsilon, a delta with property P. That's what doing mathematics constructively means. It is a good clean habit to argue constructively:
The subtleties of intuitionist distinctions make the use of language much more flexible, colorful and expressive than classical reasoning. Allow me my favorite illustration of the power of double negation. Coming from Europe I am still taken aback when my thanks for some gesture or other prompt the response "no problem". Back home I'd get a friendly "gaern gscheh" which would amount to "my pleasure", while I cannot help interpreting "no problem" by what a New Yorker cartoon said on a doormat some time ago: it is apparently a wintry Sunday afternoon, a middle aged couple dressed in furs and mufflers, armed with a bottle, has just rung the doorbell of an apartment when they both catch sight of the doormat which says, in appropriately ornate lettering, "Not Unwelcome", they are staring at it, at each otherwhat are they to do now ? In the classical predicate calculus only conjunction, negation and the universal quantifier are needed. Disjunction, implication and the existential quantifier are definable making free use of double negation. (A or B), (if A then B) and (there exists an object with property P) are classically equivalent to (not(notA¬B)), not(A¬B) and (notallobjects have property notP) respectively. This indeed shows that there is something very artifical about the classical predicate calculus. Indeed some meaning is lost. Double negating the classical predicate calculus you arrive at a weak but valid fragment of intuitionist logic. And so, classical logic is seen to be a fragment of intuitionist logic and obtains a legitimate, if virtual, existence, in the shadow world of double negations, a compelling picture. But more than that: in a gem of a paper [8], concise and lucid, Gödel extends this interpretation to Arithmetic and thus provides classical first order Atihmetic with the seal of consistency, tacitly accepting the fact that intuitionist reasoning is in no need of justification. Ruminations In response to Ms Goldstein's story presented at the beginning of this essay in an excerpt from the Edge interview let me quote from Gödel's introduction to his 1931 paper.
A simple announcement of what the general expectation was and what Gödel had found; a straightforward introduction to a research report in the face of which it is grotesque to picture Gödel rolling up his sleeves and saying to his pal Plato "Ok, now that the PhD is under the hat let's show them something to make them jump out of their smug positivist skins." Several people were actually working on the completeness problem at that time. I do not see how the incompleteness theorem bears on the question of Platonism. Personally I am a selective Platonist, Gödel probably was a pretty radical one, but he was not blind to other approaches to the Foundations as his contributions and comments to Intuitionism show. He may have been guided in his choice of work by his philosophical beliefs, but the results stand on their own. Platonism and Intuitionism are not incompatible, they can be considered as complementary manifestations of mathematical realism, the static and the dynamic, its ontological and its epistemological aspect. The incompleteness theorem uncovered a limitation phenomenon of the tool of formalization. Mechanization or computerization cannot do all the jobs of the working, thinking and visualizing mathematician. Who would expect a Swiss army knife to be capable of tuning a piano? Gödel's incompleteness theorem has not overthrown Hilbert's program but it has caused a shift in it: the call for a Diophantine Decision Procedure turned into the Diophantine Decision Problem, proved unsolvable in 1970, after long hard technical work by Julia Robinson, Martin Davis and Yuri Matijasevic — to mention only the most farreaching result of that nature. As Hilbert himself stressed in the passage cited above in the introduction, there was never any question of setting up a formal system to encompass all of mathematics, at least not by working mathematicians. As we have seen by now, already the choice of language imposes limitations. Once language, axioms and rules are fixed, a formalism is pinned down and there is no tampering with the concept of proof. That seems pretty straightforward. But the question is: what are we talking about in that language, where do the axioms come from, what justifies the rules of inference? We are limited creatures, finite beings with a finite amount of temporal and spatial reach allotted to us. The mind is an unfathomable tool; it projects and imagines. It can conceive of unending processes, it can remember extinct events and it can create the weirdest gadgets. Does it create mathematical objects, rules and axioms out of the blue? Not likely. And surely not by deliberate consensus, just think of the debates, controversies and battles we would be facing if anybody could arbitrarily choose how to see the world — and try to impose this view on her fellow beings. I believe that we are basically wired to perceive and handle the world according to the blueprint of mathematics. With slight variations; careful people won't accept the law of the excluded middle all across the board, common sense will demand constructive existence proofs, others will balk at every use of the axiom of choice, a few deny the Uncountable and some will insist on Finitism. But the squirrels scurrying along my telephone line may well be operating according to a totally different notion of the world although we perceive them as navigating according to the laws of our geometry. I personally do not think that mathematical objects have an existence of their own independent of the human mind. Such a strong Platonic hypothesis does not seem required to explain the inevitability of mathematical concepts and practice. I would not even talk about objects, at least not to start with. They start out as ideas of tools born of a need to cope with the world, just like the idea of a hammer, the wheel, the bed; constructs that take on reality by use and handling. They grow and adjust to our needs. With familiarity and use they pick up reality like moss — and a patina of beauty. They are indeed alive. The finality of a fully formalized theory is that of halfway houses, of temporary stops. Chevalley has likened the Bourbaki volumes to mausolea where beautiful theories are embalmed, while mathematics is going on, developing. No more do we conduct our research inside a formalism than the Egyptians had their orgies in the pyramids. A fully developed formal system becomes an object of study rather than a tool for reasoning. For that it would be much too cumbersome. We cannot do without the agility of welldeveloped mathematical intuition. Consider the example of the complete theory of real closed fields. Soon after that was embalmed the idea of nonstandard models emerged. It very clearly brought home the limitations of the language of first order theories and opened up a new understanding of 17th Century puzzles. Formalisms are not ends in themselves; they serve as archives or become objects of mathematical investigation, or both. Might not Mathematics be a grid developed by the human species to find its way in the struggle for survival? It is cumulative in the sense that, once understood, a theorem or a structure remains true — true to itself. But it is not static, it is growing, its horizons are forever expanding, new concepts are shedding light on old riddles. And it is genuinely a joint venture of humanity, although accessible to but a few in its entire scope, and it sure is hard work. It is remarkable that we can talk to each other, that we do agree about the facts of mathematics and amazing how much use we make of its basics, so much that we could not imagine getting along in daily life without it. And then some are able to reach its remotest areas, which, for all we know in a few more generations will be accessible to all. Others spend a lifetime beating their head against the wall of one piddling problem. There are indeed such cases; casualties of the mathematical endeavor, some really good brains, their stories are not told — not so far. Paul Finsler, one of my teachers, springs to mind in close connection to the topic of our discussion (cf. [9] p 265267). There are particularly vicious problems such as the fourcolor conjecture, mercifully laid to rest by now. Gödel's renowned paranoia may well have been a symptom of a deep depression caused by the frustration of not quite seeing his way through, wherever a hunch was driving him. It may have been an axiom transcending GödelFränkelvon Neumann set theory to settle the continuum hypothesis lurking within grasp only to keep escaping precise articulation, or it may have been some elusive insight that he forever kept expecting just from his next session with Leibniz. Think of all those telltale scribbled records of his Leibniz studies in the Nachlass (the many boxes of notes left in the Gödel Archives at Princeton's Firestone library). Futile and compulsive, one might say. You don't have to be a hypochondriac or prone to paranoia to be unable to feed yourself when in the grips of one of those attacks by what the group theorist Noboru Ito in Chicago used to call a "Great Problem, very annoying, very annoying". Familiarity breeds reality. There is no Königsweg—no regal path for avoiding strife. When a stranger asks a native on the street how to get to the Vienna Philharmonic the answer is "practice, practice, practice". There are no short cuts to those platonic realms of Mathematics, neither to the gut experience of mathematical intuition nor to the vision of its beauty. I doubt that pure philosophical discourse can get us anywhere. Maybe phenomenological narrative backed by psychological and anthropological investigations can shed some light on the nature of Mathematical Truth. (Piaget has made a start with children.) As to Beauty in mathematics and the sciences, here speaks Sophocles' eyewitness in Antigone:
A true Realist, a true Platonist will not stoop to choose between Beauty and Truth, he will have the tenacity to stick it through until Truth is caught shining in her own Beauty. Sure there are messy proofs, we have to bushwhack trough a wilderness of ad hoc arguments, tours de force, combinatorial jungles, false starts and the temptations of definitions ever so slightly off target. Eventually, maybe not in our own lifetime, a good proof, a clear and beautiful proof, will be honed out. That, I think, is the belief of the true Platonist. What Gödel and Einstein were doing when walking together over the Institute's grounds may have been just that; bush whacking, comparing mental notes and encouraging each other not to give up while getting all scratched and discouraged. Yet finding solace in speaking to each other in their mother tongue about their deepest concerns, and the state of the cosmos, the world, the weather and their households too. Gödel probably never found a home in Amerika. Adele gave him shelter as best she, or anyone could. He did have disciples; Georg Kreisel, John Myhill, Gaisi Takeuti, Stanley Tennenbaum, Hao Wang and other visitors to the Institute. But, although one grumbles about the drudgery, there is something wholesome in teaching American undergraduates. It makes for a link with ordinary life and it helps to adjust to a foreign culture. Gödel had a very austere life; partly due to his nature, partly due to his calling and partly due to the way the world was and still is. By far the most illuminating survey of Gödel's work including its methodological and philosophical implications is Kreisel's memoir for the Royal Society [13]. What
kind of a person was Gödel? During the years 195557 of his fellowship at the Institute for Advanced Study Kreisel, originally from Graz, was very close to Gödel. They remained vibrantly in touch for many years, by daily handwritten letters, in German of course, occasional visits and by, most vivid in my memory, interminable long distance phone discussions of ideas, results, pitfalls and technical gossip. Kreisel maintained the same intense style of work and philosophical ruminations with Hao Wang. Although Hao was in close contact with Gödel he never became "family" at the Gödel's like Kreisel did. Their common Austrian background and language must have added significantly to the bond based on philosophical and mathematical kinship. During my postdoc year at the Institute and then while teaching at Goucher College in Baltimore I was working with Kurt Reidemeister, a fellow at the IAS for the two academic years 19481950. The Gödel's and the Reidemeister's had known each other in Europe, maybe in Königsberg, though during the forties Reidemeister was a professor at Marburg. Later he was called to Göttingen, a distinction at his ripe old age. As a topologist and group theorist his bond with Gödel was more cultural than professional. Like Kreisel at the Gödel's I became part of the family at the Reidemeister's and remained so to the end of their days. As Hao Wang and many others observed, somewhat puzzled, Gödel appeared to be more relaxed and in better health during 194850. I am sure this was largely due to the comfortable expansion of his environment by the presence of old time friends, including the wives. Even at that time, mostly from stories that Pinze Reidemeister told, I had a distinct impression that Kurt Gödel never assimilated to American life, much less so than his wife Adele. Pinze was a delightfully spontaneous person. Since Gödel did not like visiting outside his home and the Reidemeister's did not drive, Adele would pick them up for many a social evening at the Gödel's and eventually drive them home to the institute's housing project. Pinze used to tell what fun it was to be driven by Adele after dinner, while Kurtele stayed home. They were all relaxed, telling jokes driving along the dark deserted Princeton back roads. Invariably Adele would heave a deep sigh and declare what a relief it was to be driving such good friends without having to worry about a genius in her car. Pinze enjoyed telling that story at Princeton cocktail parties, preferably within earshot of her Kurt — no one would ever think of calling him Kurtele. By the time Kreisel appeared at the institute at Gödel's invitation I had given up mathematics and was raising a family, not that I considered the two tasks incompatible, but it so happened. It was Kreisel who encouraged me to pick up my mathematics again. As a student in Zürich I had been intrigued by foundational problems, but, unable to find a teacher — Finsler being in bad health and Bernays unfortunately not known to me at the University — I settled on group theory and got quite engrossed in that, still am, though no longer agile and active. Anyway it was natural for me to get into logic under Kreisel's demanding tutelage, and then, to my delight, I found my way to Intuitionism — just what I had been looking for all along! So much so that when I started teaching again I made quite a fuss doing my calculus courses constructively; making a point of presenting my epsilondelta's as explicit functions, calculable ones of course. Probably sometime in spring 1957 Kreisel took me, Esther and George to the Gödel's for tea, a pleasant occasion, a thoughtful gesture on Kreisel's part towards both Mrs. Gödel and me. It was a gracious occasion; old world geniality with George and Esti, aged four and six, in the center of attention, irresistibly cute of course. Eventually Gödel himself came down from his study to join us for half an hour. He was very pleasant and natural, extending his friendly feelings for Kreisel to my children and me. I do not recall what we talked about, I am sure it was mostly small talk, except for a few sharp and witty asides from Adele about Princeton society. At that time I felt a vague kinship with her, both of us balancing, a bit precariously, at the edge of the society we were living in, albeit for totally different reasons. I did appreciate their friendly gesture. Two years later I was on my own starting out on a career teaching and doing mathematics. The whole Gödel lore is beset by psychological myth making. I would like to make a plea not to forget what a conscientious person Gödel was. That goes a long way to explain his reticence, his hesitation to publish or pronounce or do anything that was not honed down to perfect precision and lucidity, and stripped of all superfluous bulk. Kreisel used to have some great quotes of Gödelcomments to news reports, both full of commonsense and — baffling. The familiar story of Gödel pointing out a flaw in the US constitution to the judge on the occasion of his citizenship interview comes to mind. Less known is how diligently, to the point of pedantry, he scrutinized every proposal, every piece of work or mere argument submitted to him. An illustration of Gödel's professional conscientiousness is the case of W. For years, in the 70's I think, a Mr. W. made the circuit of Mathematics meetings claiming to have discovered an inconsistency, first in accepted set theory and finally in all of mathematics. Gödel apparently studied all this gentleman's voluminous and laborious manuscripts, and eventually, to the relief of us all but not to our surprise, found Mr. W.'s claims unfounded. Unfortunately I do not remember the details, but I am sure there are records. Gödel was a very private person. Just looking at him you could see the gentleman of the old world, discretely well groomed, shoes polished as a matter of course, tie in its proper place. Maybe he wore sneakers and jeans while working in his garden, if he ever did work in his garden. But I am sure there exists no public photograph of Gödel in sneakers and jeans drinking a coke out of a bottle. There does, however, exist just such a photo of Tarski in the recent biography [14]. Deeply pondering the question of the nature of mathematical truth and struggling to understand the messages from his own mathematical intuition, haunted by the questions he set himself, Gödel needed total focus to pursue them. I believe that deep concentration on a mathematical problem induces a state of meditation, transcendental meditation in a literal sense. There are times when ill health can serve as a subterfuge to secure the necessary isolation from the bustle of one's human environment, not a deliberate strategy of course. It was said of Tarski that whenever he, such a robust and gregarious character, was complaining of being unwell he was about to reach a breakthrough in his work. Mufflers and over coats, hats and spats are a way of keeping the environment at bay; more symbolic than appropriate in fine weather, but psychologically effective, even if not motivated by hypochondria or paranoia. I had several uncles of the same impeccable appearance and demeanor; very gracious, not at all intimidating; yet it would never occur to me to ask them a personal question, or even to assume they might have secrets that the family would enjoy gossiping about. Just think how lucky we are that Gödel lived long enough to give us so much real stuff to ponder. Time to close this discourse with a pause of respect to Kurt Gödel, a deep thinker and a gracious gentleman. Takeuti in his memoirs [15] recalls:
Let us not cause him too much trouble in his welldeserved peace. __ References, in order of occurrence: [1] David Hilbert, "Ueber das Unendliche," Math Ann. 95 (1926), for translation cf. [2]. [2]
Jean van Heijenoort, From Frege to Gödel, A source
book in mathematical logic, 18791931. Harvard University
Press, 1971. [5] Kurt Gödel, "Die Vollständigkeit der Axiome des logischen Funktionenkalküls," Monatshefte für Mathematik und Physik, 37, 349360, 1930. [6] ............., "Über formal unentscheidbare Sätze der Principia Matematica," ibid, 38, 173198, 1931. [7] ............., Collected works, volume I: publications 19291936. Oxford University Press 1986, edited by S. Feferman, J. Dawson. S. Kleene, C.Moore, R. Soloway and J. van Heijenoort.. [8] ............., "Über intuitionistische Arithmetik und Zahlentheorie," Ergebnisse eines mathematischen Kolloquiums, Heft 4 (1933), Vienna. (translation in [3]) [9] Verena HuberDyson, Gödel's Theorems; a workbook on Formalization. TeubnerTexte zur Mathematik vol 122, Teubner Verlagsgesellschaft, StuttgartLeipzig ,1991. [10] Alfred Tarski, A decision method for elementary algebra and geometry, Rand Corporation, 1948. [11] ............., Der Wahrheitsbegrif in den deduktiven Disciplinen. (The concept of truth_in the deductive sciences) Aqkad. Wiss. Wien, Mathem. Naturwiss. Anzeiger 69, 1932 [12] David Hilbert, Grundlagen der Geometrie, Gauss Weber Festschrift, Teubner Verlag, 1899 [13] Georg Kreisel, "Kurt Gödel, 28 April 190614 January 1978," Biographical memoirs of the Fellows of the Royal Society 26, 1980. [14] Anita and Solomon Feferman, Alfred Tarski; a life in Logic, Cambridge U. P. 2004. [15] Gaisi Takeuti, Memoirs of a Proof Theorist — Gödel and Other Logicians, in Japanese, English translation by Mariko Yasugi (Kyoto Sangyo University, Japan) & Nicholas Passell (University of Wisconsin, Eau Claire, USA) , World Scientific Publishing 2003. 
John Brockman, Editor and Publisher 
Top 