Barbour, Weinberg and the Anthropic Principle
Although Barbour’s The End of Time started this triangular correspondence between friends, I should like to begin with Weinberg’s article in the New York Review of 21.10.99, which was sent to me asking for my views. These, I must emphasise, are my own, influenced, admittedly, by my friends’ ideas but in no way constituting a Quakerly compromise ‘feeling of the meeting’.
Weinberg begins by saying that in considering whether the universe has been designed one must have a notion of what a designer might be. For example, a chaotic universe "could be designed by an idiot". I deny this, on grounds which I have given in a book I have yet to publish, called Born Under the Pleiades, in which I declare that to exist at all the universe must have come out of something good. These grounds, however, are not arguments at the end of which one can put a "QED"; but neither are they irrational ramblings, or "mystical mumbo jumbo", to use Weinberg’s phrase. They are, like so many considerations that matter to us, appeals to common sense or human intuition. And here, in these days when "metaphysics" has returned to being a term that is treated with respect, I must fly my flag and express my view that metaphysics is nothing but an attempt to formulate appeals to common sense and human intuition in order to tack onto the end of them a "QED" that I regard as entirely spurious.
Many such appeals can be made, and of an enormous variety of kinds. Among these, the kind that might tell us that no idiot could produce any sort of universe, with or without laws and regularities, needs a special term, and I wish I could think of a better one than (dropping Weinberg’s "mumbo jumbo") the word "mystical". I should like to make this term respectable by pointing out, first, that no arguments follow from a mystical intuition and that no-one can be obliged to pay attention to any account of a mystical experience, let alone draw any conclusion from it, and second, that many respectable people have had such experiences or appealed to such intuitions or come to such conclusions. Freddy Ayer, for example, describes in his autobiography experiences that could well be called adolescent mysticism, and near the end of his life he had another, more impressive one. He described it in something that has been published, but I have never read this account, and quite definitely I bet my bottom dollar that he was too good a logician to end it with any kind of "QED". Much more recently Ludovic Kennedy, sharing a platform with three other atheists (including Nigella Lawson, Freddy Ayer’s step daughter), lost his temper with the arguments put up by the audience and said "you people do not understand the true nature of spirituality", and described an experience from his war-time naval years, when, on deck at night, he had felt that he, the sea and the stars were one --- and here I must bet my bottom dollar again and confidently doubt whether, in remembering this moving experience, it would occur to him to draw the conclusion that he and the universe were identical. I propose to call the kinds of mysticism that Ayer and Kennedy could feel sympathy with "sane mysticism", even if they part company with me when I elaborate my own kind.
This brings me back to Weinberg’s opening remarks. My own personal view is that it is useless to attempt even a "vague idea of what a designer [of the universe] would be like". Now religious people will probably take me to mean by this that the designing God is so ineffable that we cannot begin to apprehend what he "would be like". No: for I do apprehend something definite --- namely, that the term "design" is completely inappropriate to any consideration of how the universe originated. It is an anthropomorphism which can no more be applied to the universe’s origin than a white beard can be. The one conviction common to the people I regard as sane mystics is that anthropomorphisms of all kinds have to be abandoned. As Meister Eckhart put it: God is not good, I am good.
As a spiritual exercise, this abandonment of anthropomorphism is called the via negativa, and it usually takes the form of a gradual shedding of the religious and theological superstitions one has been brought up in --- a weaning process, in other words --- but it does not usually prompt its followers to break with their past and buttonhole their ex-co-religionists and say "look --- you’ve got it all wrong". In most cases, they go to Midnight Mass, or whatever their equivalent might be, wish their fellows well and then go out and look at the stars and feel at one with them.
What intrigues me is what form the via negativa might take if it were practised by atheists. I do believe there are a few around. What their substitute for an occasional friendly trip to Midnight Mass might be, we shall just have to wait and see. Here I am only concerned with their terminology. Personally, in contradiction to Meister Eckhart, I would continue to use the word "good" of the universe’s origin. I do not see any alternative to it. Naturally, we must still eschew anthropomorphic interpretations of this word and not, for example, get trapped in tangles of the form "if the universe’s origin was good, how can this and this and this happen in it?" In the negative tradition one might say "the universe isn’t at all bad", or even go as far as "actually, it’s quite decent". Personally, I go further, and have an overwhelming conviction that the universe is so bloody good that it never needed to be designed, or rather, to put the horse properly in front of the cart, that the universe’s origin is so good that it never needed to design it.
When the baby has gone out with the bath water like this, there is no need to avoid the word "evil" in discussing the old quandary. "How can all these evil things happen in a universe that is in origin good?" They have happened. That is part of the way the universe is. The task is to stop any more of them happening. Short of world-wide suicide, we can’t wash our hands of the universe because we detest certain aspects of it. We need to clean it up. We have to work with it.
To readers who have struggled so far in spite of their repugnance for the way I have expressed myself, I can perhaps offer my conclusion in a simple nutshell that they may accept. All talk (hypothetical just as much as theological) of the universe having been designed is a mere anthropomorphism. The question cannot arise: and so, a fortiori, Weinberg’s second question, whether its designer has a form consistent with any of the world’s major religions, cannot arise either, and neither, equally, can whether "some cosmic spirit of order and harmony" was behind its non-design. That goes out of the window with the bathwater too, and in so far as this phrase can be justified as being metaphor or poetry I think that better metaphor, and better poetry, will soon be found.
Now I must move to the ‘anthropic principle’, and say that my first thought on reading Weinberg on it was that everything that needed to be said was contained in Barbour’s two quotations from Boltzmann, in his notes on pages 341-2. These were written some decades before the term "anthropic" was introduced, but in embryo they contain the whole argument. In a universe so extended (and, mainly, blandly probable) as Boltzmann took for granted ours must be, there is nothing inherently improbable in supposing that some locations in it should be, as it were, islands of improbability in which such excessively improbable beings as ourselves might evolve; in which case, finding ourselves alive and conscious in such an island, we have no grounds for claiming that our improbability represents an overall improbability that requires some special explanation.
This beautiful exposition of Boltzmann’s needs to be viewed in the light of recent cosmological theories, which I do believe modify it while also enhancing its force; but then, I believe, an argument that has come to my mind from Cicero’s De Divinatione needs to be considered as a comparison. I will leave this, however, until I have considered modern cosmology and biology, lest readers think it too improbable that an argument from Cicero could be relevant.
I still remember Hoyle, on the Third Programme in the early days of his steady state theory, saying that somewhere, out there, there must be an eleven who could beat the Australians. Since then, a bewildering succession of discoveries, both astronomical and biological, has made the apparent likelihood of there being anyone like us "out there" fluctuate extraordinarily, between the overwhelmingly and the minisculely probable. I hope to show that a reasonable modification of Boltzmann’s idea can take these fluctuations in its stride.
Biologically, it seems accepted now that life of some kind will evolve wherever it can, but it does not follow that life like ours will. In the fifty years since steady state, while the probability of life evolving in any particular place has seemed (on the whole and with fluctuations) to have increased, the chances that have led to our own form of life have come to appear more and more improbable. The notion that if the dinosaurs had never been extinguished we might by now, or soon, have evolved as intelligent saurians is strictly for science fiction fans. My views on this point, I must admit, have been much influenced by my schoolboy reading of Huxley’s Evolution, where in his last chapter (now the penultimate chapter, since another has been tacked on) he argues the need for generalised forms to be available to take each next evolutionary step, because once a species has become over-specialised it can only evolve in the direction of that specialisation. I don’t know what is thought of this argument now. I’d be very grateful if any biologists who read this would enlighten me.
The reason why I can afford to be nonchalant about where this pendulum of the probability of intelligent life will settle down is that recent cosmological theories have come to lean towards a multiple universe paradigm. This has nothing to do with the notion floated by quantum theorists of ‘alternative’ universes that arise (one hopes only theoretically) each time alternative quantum-events are paired as ‘waiting to see what actually happens’ --- let alone the science fiction alternative universes in which alternative characters in a television series are recognisable caricatures of each other.
An early form of multiple-universe-theory was related to me by Dieter Peetz in the early seventies, as told him by astronomical colleagues at Nottingham: the expansion after the big bang is eventually followed by a big crunch, whereafter a new big bang occurs with a new random assignment of physical constants, and so ad infinitum, allowing the constants we enjoy to be considered as coming about by chance. Our universe will thus be just one of a potentially infinite series, its constants simply the ones that allow evolution to work. No universe so resulting can communicate with any other universe, and so the inhabitants of the lucky one, not realising that they live in just one universe in a cosmos of many, will inevitably be tempted to over-interpret the anthropic principle.
From the point of view of people wishing to put the anthropic principle into perspective, the obvious disadvantage of this simple paradigm is that after the next big crunch there is nothing to stop the next assignment of constants from bringing about a universe that eternally expands, putting an end to the series. This fault, however, has been overcome by the latest paradigm, which assumes a quasi-simultaneous explosion of big bangs, each with its own assignment of constants, and no resulting universe capable of communicating with or influencing any other. One has to say "quasi" because this very hypothesis precludes any comparison of relative time between universes. And since there is no series to be broken by a universe that goes on expanding for ever, that possibility is no drawback to the paradigm.
A variant of this paradigm is mentioned by me in my Iris and Other Minds (smuggled by me into my wittgenstein.co.uk Internet site, as this present essay may be when Wittgenstein honour is satisfied), namely the budding theory: a black hole buds a new universe which, once free, has no temporal correlation with its parent universe, and so the result is a cosmos which is as quasi-simultaneous as the one sketched above.
So convenient are these paradigms for resolving the apparent paradox of the anthropic principle that one cannot help wondering whether they were invented for that purpose. While I am not so sceptical myself, I do find them convenient for my own purpose of, as I have put it, putting the anthropic principle into perspective. For this, my first remark is that, granted that empirical evidence is in question, it must be of a very interesting kind. My reading is that cosmologists, in trying to explain the workings of the universe, encounter problems, and in resolving them construct theories requiring multiple mutually incommunicable universes. These theories are therefore incapable of either verification or falsification by empirical research: their only possible corroboration is the elegance with which they resolve the problems which gave rise to them. Nevertheless, the propounders of these theories are certainly assuming that our companion universes are real, and in case this assumption sticks in the throat of positivists I offer a contrasting idea where the word "real" would quite definitely not be in place. This idea is so useful to my analysis that I should like to pause here to expound it.
Suppose a stastically minded cosmologist who wishes to consider the ‘inherent probability’ of a posited cosmos of universes. To do this he invents a hyper-cosmos of such cosmoses, with a stastical distribution of characteristics. This enables him to compare one posited type of cosmos with another type, and thus, in examining the cosmos which best satisfies the problems which he hopes to resolve, he will be able to assign to it a probability of being the one that actually exists. Of course, the truth could be that a slightly less probable cosmos (understood: less probable on his assumptions) is the one that actually exists --- he can never know, but the whole bent of his enquiry is the requirement that one such cosmos, containing our own universe as a member, does exist, while the others are theoretical constructs, and the hyper-cosmos that embraces all of them is even more theoretical.
This brings us to where Cicero is relevant. He wanted to argue that extremely improbable events could very occasionally occur by chance. He imagines an excavator finding what looks like a sculpted head or bust among the rubble of a quarry. The excavator naturally concludes that it has indeed been carved by some sculptor: but Cicero argues that it could, by extraordinary chance, have been formed by random movements of the stones around it. This argument reads with exhilarating modernity, but it leaves a question open. If this lump of rock bore a rough resemblance to a human head, Cicero’s point is acceptable. If, however, he is thinking of a head that might, with slight subsequent damage, have come from the studio of Praxiteles, the probability is too minute for us to accept it as a serious possibility. The same is true of monkeys producing the complete plays of Shakespeare (I notice that no-one was ever quite so ambitious as to think they might have produced the sonnets as well). Similarly when one is told by physicists that it is only extremely improbable, not impossible, that a kettle of water put on a fire to boil will freeze instead. I don’t know the order of improbability here, but I understand it is something like being only likely to happen once or twice in the course of the whole universe. If it should be likely to happen a dozen times in the course of a universe I should still say the same, and for typescripts and sculptures that could expect a dozen universes to turn them up by chance I would argue even more strongly: this is asking the browbeaten layman to misuse the words of our language.
Yet my argument is not tending towards saying that the meanings of our language require us to say that the universe is somehow quasi-designed by some kind of quasi-deity. Quite the contrary, and that is why I have brought Cicero into it. We know what it is like to find intriguing objects and wonder whether they are artefacts or improbable consequences of natural causes, and we also know what it is like to find something so remarkable that we refuse to be browbeaten by statistical arguments. We do not know what it is like to find intriguingly improbable universes --- let alone compare them with universes that are bona-fide designer-labelled.
Weinberg mentions the apparent improbability of the universe’s having a ‘cosmic constant’ close enough to zero to avoid uncreative expansion or uncreative contraction. Our value is conveniently close to zero, and that is why Weinberg invokes the multiple universe paradigm to avoid the anthropic principle getting out of hand. Shortly after his article was published, however, hot news came from the science front that there was now evidence that this constant was not just conveniently close to but damn near exactly zero, which means that the expansion of the universe will eventually cease but never be followed by contraction: it will settle in equilibrium and leave us enjoying it for as many billions of years as we could want.
This news does not seem to me as stupendous as it has been made out. Who wants more than a billion years of human future? Nevertheless, since the anthropic principle is my topic I have to provide some response to the idea. So I will take the most extreme case of improbability I can imagine and assume that our universe has characteristics that make it next door to unrepeatable. Moreover, granting that these characteristics do give a reasonable likelihood of life occurring, the requirements for intelligent human-like life to evolve are still so remote that we might expect it to occur in at most, say, half a dozen of the universe’s galaxies, none of which we can ever hope to reach. As to our own galaxy, let us allow human-like life to occur somewhere else in it once, just to keep people cheerful, and let us allow ourselves to meet it at some time in the next billion years. On top of that, the improbability of the universe having a precisely zero cosmic constant and thereby lasting for ever is mere gilt on the gingerbread.
Together, those suppositions constitute the maximum temptation to over-interpret the anthropic principle, but I insist that a temptation is all it is. My anti-Cicero argument holds, and it holds in both directions. We are neither entitled to say that this excessively improbable universe must have been designed, nor that there must be some paradigm of random events that enables it to have happened by chance. All we know is that it is there and that we are alive in it. Without any possibility of comparing our universe with others, the relevance of Cicero’s thought experiment simply does not arise.
Now consider an opposite case, that the way the cookie of random assignment of constants crumbles, universes conducive to life occur frequently in this supposed cosmos of universes, and frequently enough for intelligent life to evolve in many of its universes, and consequently in many galaxies in each universe and thus, with luck, in many solar systems in each galaxy. Now here, I insist, we have to keep our heads. Since this is only a thought experiment we are free to imagine whatever conditions we like, and posit a universe positively crawling with intelligent life, and elevens capable of beating the Australians all over the place. That is not at all what seems to be the case in this particular universe. I think my ‘opposite case’ ought to be restricted to having some resemblance to what we can expect given what evidence we have. Intelligent life "in many solar systems in each galaxy" will still leave difficulties of exploration so extreme that any one civilisation will be extremely lucky to encounter one or two others before it decays.
I don’t want to get bogged down in discussing the practical problems of intra galactic exploration (let alone inter galactic problems), but I do wish science fiction writers would take them more seriously. For my present purposes it is enough to state my belief that they make the practical upshot between a cosmos whose universes tend towards intelligent life and one whose universes tend against it much of a muchness. In other words I do not believe the question is going to be settled by counting the intelligent species we meet, because I don’t think we are going to met any, but (if "settled" is the right word for this) by cosmologists finding convincing solutions to the problems that led them in the first place to invent multi-universe paradigms. So I have to ask: supposing that the most convincing and elegant solution is universally accepted to be one which makes our particular universe excessively improbable among possible universes, what implications will that have for the anthropic principle? My answer can only be: none. It does not matter whether we are considering a merely statistical distribution of possible universes or a distribution of real universes in a real cosmos. In neither case is there any possibility of empirical comparison. Without that, Cicero’s thought experiment fails to be an analogy. The universe is what it is.
Indeed, I can make the point even more strongly by elaborating my scenario of cosmological paradigms. Suppose that a solution, much as sketched above, is accepted as so convincing that for a century or more our improbability is taken for granted. Nevertheless, sanity prevails. Our universe in its cosmos is regarded as corresponding to an island of improbability in the mainly bland universe supposed by Boltzmann, so that his argument of overall probability is preserved. Then a revolutionary cosmologist propounds a new paradigm. Grasping the old problems (like the metaphorical bull by the horns) he --- or I think it will be more likely to be she --- shows even more convincingly that they can be solved without the assumption of multiple universes. The universe is no longer a Boltzmann island. It is all there is. Unfortunately, old paradigms die hard. Humanity has lived so long with the old one that everyone is now convinced that an excessively improbable universe is all there is. Our revolutionary struggles to disabuse her fellows, but not until her granddaughter grows up and takes over does the message get across: this ‘all there is’, this universe, is simply what there happens to be.
In it, we are alive, and capable of enjoying our lives, and we shall enjoy them even more if we set to and start to clear our particular corner of it up a little. And as to that need, perhaps I could quote another friend who made the rather important point that we should do best to tidy it up ourselves and not wait for our governments to do so for us. Admirable things in themselves, of course, and very useful once they are given the right idea, but they do need us to set them an example.
By way of postscript I should like to add that, although the above has nothing to do with Wittgenstein, I hope my completion of it, on Barbour’s End of Time itself, will count as applied Wittgenstein.
In adding this postscript, however, I should like to say something more about the unlikelihood of finding intelligent life elsewhere in the galaxy, or our being found by other intelligent life. Recently there has been another spate of murmurings about the rich possibilities of life elsewhere, even of ‘non-DNA’ life, and I should like to sketch, without getting bogged down in the subject, why, even if all that is true, it won’t increase significantly our chances of ‘close encounters’.
First, when science fiction writers once realise that, because of the properties of travel at near the speed of light, it need not take a thousand years to travel a thousand light years, they lose all sense of moderation. They fail to take account of the corollary --- namely that relativity requires their explorers to come home, albeit after only a short period in their own time-frame, forbidding periods of time ahead in the time of their home planet. Gamow calculated many sample schedules, based on his assumption that a feasible neutron-drive could sustain acceleration of exactly g (thirty two feet per second per second) for unlimited periods. These schedules were quite encouraging, provided one saw no psychological problem in coming home many generations after one’s own children had died.
There is, however, another problem that arises from the Gamow idea, and to the best of my memory he did not discuss it himself. He put no upper limit to the closeness of approach to the speed of light which his drive could achieve. This meant that, theoretically, one could even travel to other galaxies if one did not mind coming home when not only one’s descendants but the solar system itself had ceased to exist. Yet in fact, this aim is unattainable even theoretically. Approach to the speed of light entails approach to infinite mass. No problem, Gamow’s ghost replies. This is so only in the frame of reference of the background galaxies. In the astronauts’ frame of reference mass will be standard (and weight will be what goes with an acceleration of g, namely the weight we are used to). One’s blood will not acquire mass that makes the heart unable to pump it. One will be able to use one’s navigation rockets to change one’s orientation with respect to the background --- by 180 degrees, for example. All this will mean is that one will be decelerating by g in one’s own frame of reference. For example, one will experience normal gravity as one stands with one’s feet towards Gamow’s drive. Given the mass one has acquired in relation to the background one will be nowhere near slowing down by thirty two feet per second per second in that relation, and given the fact that this mass is also in relation to any objects one is passing, a collision with even a particle of dust would be catastrophic.
I do not know what limitations this problem would impose on exploration within the galaxy, and I leave details to astrophysicists who are interested in science fiction. Common sense, however, suggests to me that the limitations would be extreme, and so I shall stick my neck out and predict that even brave astronauts willing to return home to be greeted by their ageing grandchildren will never be able to go beyond the Pleiades.
So I repeat: the most optimistic assumptions about the prevalence of intelligent life elsewhere will still not mean that any one civilisation has any reasonable likelihood of meeting even one other in the course of its development, which we might extend, for the purposes of fiction, to allowing us to meet just one. I did once sketch a science fiction film trilogy based on this just-one-encounter hypothesis. Because of recent discoveries in paleoanthropology, and because I unwisely tied my plot to a disproved hypothesis of my own invention, it will need a lot of revision; but until I reach retirement and have time for that I am happy to offer the idea to the science fiction industry. I’d put up a prize if I could afford to. Since I can’t, I offer it simply as an alternative science fiction genre. The just-one-close-encounter-and-don’t-hope-for-another-one model.