SEPTEMBER 23, 2016
IN TOM STOPPARD’S PLAY ARCADIA, a precocious student in early 19th-century England reflects on a novel scientific worldview:
If you could stop every atom in its position and direction, and if your mind could comprehend all the actions thus suspended, then if you were really, really good at algebra you could write the formula for all the future; and although nobody can be so clever as to do it, the formula must exist just as if one could.
Many philosophers, scientists, and artists have pondered the existence of such a formula. Nobel Prize–winning theoretical physicist Steven Weinberg not only believes one does exist, but has argued — optimistically — that its discovery is in sight, though he has lately backed off somewhat from this latter claim. At the same time, however, in his book Dreams of a Final Theory, he laments that “the more the universe seems comprehensible, the more it also seems pointless.”
If you “could write the formula for all the future,” then nothing exists fundamentally except atoms, moving according to deterministic, eternal laws — and everything else either reduces to atoms or is to some degree illusory. So much for our experience of free will, our sense of agency and intentionality, the sensations of colors and sounds, and the notion that we have responsibility for the choices we make.
And so we are left with the existential question of how human beings are to find meaning in a universe in which they are nothing but machines.
Sean Carroll, a physicist at the California Institute of Technology, attempts to tackle this question in his elegantly written new book, The Big Picture. A highly literate and sophisticated natural philosopher, Carroll writes in an engaging voice; he has an established record as a popular nonfiction writer (including of a blog and two previous books). Sometimes fair to a fault, he scrupulously acknowledges competing views on key questions, and almost always draws the line between established scientific fact and speculation. I have no hesitation recommending this book to laypeople who want to get a taste of one view of the big picture. That said, I have to dissent from the conclusions Carroll draws from his “big picture.”
His book is built around a simple message. Science, he argues, has now reached a point when we comprehend — or can anticipate comprehending in the near future — enough of the universe to answer the fundamental questions that have long troubled human beings: What is the meaning or purpose of the universe? What is our place in that world? As living, intelligent agents, what are we in a vast universe? What is the basis of ethics and morality?
Carroll’s tour of science is worth the price of admission, but the book disappoints when he offers his answers, which are by no means new. They have been discussed by philosophers since the Newtonian revolution in the late 17th century, if not before. To be sure, scientific knowledge has progressed substantially — but, in my opinion, his confidence is striking, even foolhardy, given the deep, unsolved problems science still faces across the board — from cosmology to physics to biology.
Carroll advocates for what he calls “poetic” or “cheerful naturalism.” “Naturalism” stands for his belief that only the elementary particles and forces described by the laws of physics exist, and “poetic” implies his attempt at humanizing this reality. The solidity of a table or the smoothness of water are arguably illusions that hide a truer description in terms of atoms. Although Carroll asserts that sensations (or what philosophers call qualia), choice, values, and almost everything we associate with being human are indeed illusions that have no place in a fundamental description of the world, he concedes that it is nonetheless useful to talk about them as if they do exist — hence the “poetic” side. In fact, he encourages us to talk as we like so long as we recognize that fundamentally we are just robots playing out the laws of physics. All of which is another way of saying: He embraces the old existentialist advice that it is up to human beings to create values in a universe that cares neither for our fate nor for its own. Similar ground has been covered by the likes of Steven Weinberg, quoted above, and the modern atheists, including Richard Dawkins and Daniel Dennett.
An example Carroll discusses, which helps clarify his perspective, is the strong artificial intelligence hypothesis. Carroll embraces it when he states that, “[u]nder naturalism, there isn’t that much difference between a human being and a robot.”
It is remarkable that this assertion almost passes for common knowledge nowadays. We know exactly how to make a robot, but have no idea how to design or construct a human being (except the old fashioned way). How can we confidently claim that they are the same? The argument seems to be that the universe is a machine, so humans and robots — as subsystems of that universe — must be machines too. This begs the question, because different machines operate by different principles, and the principles needed to make our minds comprehensible remain unknown.
Optimism is generally a good thing — unless it leads to discounting the seriousness of obstacles. Carroll’s Big Picture reaches that tipping point by seeming not to recognize that science is still far from solving certain big mysteries that pose consequential challenges to the completeness of his “big picture.”
Like Dawkins, Carroll concertedly attacks religion and recommends science as a replacement for faith. He goes so far as to offer “ten considerations” as substitutes for the 10 Commandments. But I wonder how convincing these arguments are to people of faith.
Indeed, science offers no replacement for the clear sense of purpose and meaning that follow from the hypothesis that our universe was created by a conscious, loving god in order both to nurture and test us. Nor does it offer anything to replace the consolation religion offers against the ubiquity of pain, sorrow, evil — and death. This is not an argument for religion; the point is rather that, once these comforting beliefs are rejected, science cannot make up for them.
What is new in Carroll’s version is the framing of his argument against religion within Bayesian probability theory, in which probabilities are not objective facts but merely measures of our beliefs. You start by making a subjective guess and then, as new evidence surfaces, adjust the probabilities according to certain rules. The rules embody a sense of rationality — hence, this is a kind of belief-therapy.
The problem with this approach, however, is that, even if you accept its premises, what people take as evidence for a proposition such as “God exists,” and how we weigh that evidence, varies widely — even among those who value rationality. Some people of faith take “as evidence” revelations they have experienced or events they believe to be the result of prayer. If they follow Carroll’s recommended rules of rationality, they will in fact reinforce their belief in God. Other people accept that ordinary deduction from public evidence cannot establish belief in the existence of God, so they embrace faith as an act taken of their own free will, which proves, in itself, that there are truths beyond science. Neither kind of believer will be moved by Carroll.
I share the hope that we are on the road to a better world in which fewer of us surrender to the easy consolations of fundamentalist religion, but I fear that, to get there, we scientists are going to have to acknowledge the limitations of science in addressing the fears and dilemmas inherent in human life.
A brief list of problems that scientists have tried for decades to solve, without consensus or success, includes: The issue of measurement in quantum mechanics, how the initial state of the universe at the big bang was chosen, what chose the laws of physics, and how life began.
There are several plausible answers to each of these, but they remain conjectures because they are incomplete, and we lack convincing evidence to choose among them. A rational observer therefore has to acknowledge the possibility of major gaps in our understanding of nature; filling these may require future scientific revolutions we cannot presently imagine. This is not a reason to sneak in an appeal to a god, but it is a proper source of wonder.
One such “gap” example, to which Carroll devotes ample space, is the question of how the brain gives rise to a mind. More specifically, how can sensations such as colors and sounds be aspects of natural phenomena. David Chalmers calls this the “hard problem of consciousness” because, unlike the ones just mentioned, it is not even possible to imagine what an answer within science might look like.
I am not making this an argument against naturalism, although many do. Chalmers’s hard problem is trivial, if you are willing to go beyond naturalism and embrace the concept of the natural world as only part of what exists. It is “hard” only for a naturalist.
In addressing this problem, Carroll considers several alternative views, and ends up embracing one according to which sensations and consciousness are not fundamental to reality, but are nonetheless real, because they are useful to talk about. As he observes, “There is only one, unified physical world, but many useful ways of talking about it, each of which captures an element of reality.” He further explains that “a poetic naturalist has no trouble saying that conscious experiences exist. They are not part of the fundamental architecture of reality, but they serve as essential pieces of an emergent effective theory.”
The key words here are “emergent” and “effective.” Fundamentally, argues Carroll, when we think we are perceiving the sensation of the color red, our brains are simply exhibiting a propensity to behave in a certain way — namely, a propensity to feel and assert that we have seen red. Because he is “poetic,” Carroll allows us to report our sensation of color as if it were real, so long as we know that it is not a fundamental property of the underlying physics. In other words, our sensation of color is real in the same sense that temperature and pressure are real. They are helpful concepts, and useful within a restricted domain, but eliminated in the fundamental description of reality: elementary particle physics. Not only does Carroll believe that the natural world alone exists, but also that things in that world have no fundamental properties except as described by the laws of physics. The former defines naturalism, but not all naturalists believe the latter view. There is a middle ground in naturalism, occupied by those who question whether properties like the sensation of color can be explained by reduction to the purely physical properties of neurons — another way of asking whether matter may have fundamental, irreducible properties beyond those described by the laws of physics.
Philosophers in this middle ground include David Chalmers, Galen Strawson, and, before them, Alfred North Whitehead. They would claim that the “big mistake” (in Strawson’s words) is the claim that we know enough about the properties of matter to deny that sensations cannot be among them. We can call this group “open-naturalists.” One can get a sense of the disagreement between a poetic-naturalist and an open-naturalist by considering philosopher Frank Jackson’s so-called knowledge argument:
Mary is a great neuroscientist who knows everything science can learn about the physics and biology of the brain, but she has been confined since birth to a black and white environment, completely without color. One day she opens the door, steps outside, and looks at a cloudless sky. “Now I know what the sensation of blue is like!” she exclaims. “I knew all about the electro-chemistry of neurons as they respond to light, but today I have learned something completely different: what blue looks like.”
If you are an open-naturalist, then you can believe that Mary’s new knowledge concerns a particular property, which the matter in her brain enjoys and which can’t be captured by the properties mentioned in the laws of physics. What blue is like is not going to figure in the laws that govern the motions of the atoms making up a neuron — but it might nonetheless be an aspect of what matter essentially is. If, on the other hand, you believe matter has no properties except those that play a role in the laws of physics (a view philosophers call physicalism), you have to believe Mary is fooling herself.
Carroll essentially believes that Mary is fooling herself. His argument is based on the assertion that “[a] person has knowledge about something if they can (more or less) answer questions about it correctly, or carry out actions associated with it effectively.” This behaviorist definition makes it impossible even to contemplate that we can have knowledge of what the qualia or direct experience of blue is like that goes beyond an automatic reporting — such as “I see red,” which a robot might be programmed to make based on the wavelength of light. Carroll defines away the very possibility that Mary can know something from her direct experience that cannot be deduced from a scientific description of the activity of her neurons.
This is one of those cases where even smart rational people, trained in critical thinking — such as professional philosophers and scientists — fall into distinct camps, separated by a fundamental lack of understanding. Each side expresses wonder at how the other side “just doesn’t get it.”
Another prominent example Carroll discusses is the so-called “Chinese Room Argument” proposed by the philosopher John Searle. Assume that you know no Chinese (otherwise fill in a language you’ve never heard spoken). You work in a room that contains a large number of books, each with lists of pairs of sentences written in Chinese. Through a slit in the door, you receive slips of paper with sentences in Chinese. Your job is to look up the sentences in the books, and write out the corresponding phrase on a slip of paper, which you then pass through the slit. Because you don’t know Chinese, you have no understanding of what the phrases mean. The books were prepared by a computer analyzing a large data set of conversations among Chinese speakers, and the books merely recorded the most common responses to common questions. (This is a bit like Google Translate works.)
Someone outside the room may have the impression of having a conversation with someone in Chinese. Searle argues the person would be mistaken, because the person in the room has no understanding of any of the sentences. Nor is he conversing with the computer that prepared the books — the computer did nothing but record correlations. Nor did the programmers or operators of that computer understand Chinese.
Searle uses this example to argue that fluent speakers “understand” sentences in a way that “Chinese room” operators do not. If you agree, then you believe human beings understand in a way that machines, as they are currently understood, cannot duplicate. This example is also often used to argue against the strong artificial intelligence hypothesis discussed earlier. Hence we are not like those machines. This does not deny that we are natural and so made out of atoms. It is merely to assert that the principles that explain how we work have yet to be discovered.
Carroll disagrees. He argues that, “if the world is purely physical, then what we mean by ‘understanding’ is a way of talking about a particular kind of correlation between information located in one system […] and conditions in the external world.” Again, he defines the issue away.
While these two thought experiments highlight different issues — because understanding is not a sensation — people tend to line up predictably. If, like Carroll, you believe Mary is fooling herself, then you probably also believe the Chinese room understands Chinese in essentially the same sense people do. If, on the other hand, you believe Mary has learned something not deducible from her knowledge of physics, you likely believe we understand language but Google Translate does not.
The gulf between poetic-naturalists like Carroll and open-naturalists extends to other puzzles of contemporary science. I’ve often been struck that radical physicalists, who side with Carroll on these two thought experiments, often believe in the following concepts: “the many worlds interpretation” of quantum mechanics; the strong artificial intelligence idea that computers, such as the ones we use, may become conscious and intelligent in the same sense humans are; the multiverse hypothesis that our universe may be one of a vast number of unconnected universes; and the view that time and its apparent directionality is emergent and not fundamental. Carroll argues in this book for all these views.
On the other hand, I’ve noticed that if a scientist or philosopher rejects one of them, then she likely rejects all of them.
For example, here is Carroll’s defense of the multiverse: “What should be our credence that there is such a multiverse? It’s difficult to say with our current level of understanding of fundamental physics and cosmology. Some physicists would put the chances at nearly certain, others at practically zero. Perhaps it’s fifty-fifty.” That’s it … that is almost the whole argument! In other words, he has no idea. Indeed, how could he have one, since the proposition that a vast set of universes causally unconnected to our own exists is, almost by construction, untestable?
To make the idea of the multiverse plausible, Carroll tells us that “it is a prediction made by other theories — theories that were invented for completely different purposes.” But this is not the case. The theories he refers to are string theory and inflation. But string theory doesn’t predict the multiverse — it comes in a vast number of different versions, and so doesn’t make any predictions at all. There is no reason to think that the different possible universes described by the different string theories actually exist. That you can write an equation describing a theory on a blackboard in no way implies the existence of a world described by that equation. Inflation also comes in a large number of versions, and some of those predict more than one region of spacetime bubbles emerging out of the inflating universe. But other versions of inflation describe just a single universe. You can imagine combining string theory and certain versions of inflation to yield a hypothetical picture in which inflation leads to a large number of bubbles that are populated by the different string theories. But this is in no way a prediction made by the two theories; were this not to be the case, one could continue to believe in them. It is rather a move or gambit that attempts to wrest an explanatory success from a combination of two explanatory failures.
All of the positions Carroll embraces imply that we know enough to settle any ostensible mystery about the world. As he comments in an interview with Wired magazine, “I want readers to know that the world is understandable. We are not there yet, but it’s not a fundamental mystery.” From the origin of the universe up to the nature of thought and perception, there are, suggests Carroll, no longer any essential mysteries. The open-naturalist, by contrast, admits the likelihood that big gaps remain in our knowledge, to be filled in by future revolutions that may introduce novel concepts as yet unavailable to us. The mind, we assert, is an aspect of a brain, and part of nature. But neuroscience, despite decades of detailed work, is in a state akin to biology before Darwin: it lacks the central principles necessary for understanding. Similarly, quantum mechanics requires completion based on novel concepts. The principles governing the choices of laws and conditions at the origin of the universe have also yet to be articulated. Presently unknown principles from either physics or neuroscience may one day afford us new opportunities for resolving the puzzle of how a living brain gives rise to experience.
This is a very good book, and I found myself rooting for the author even as I disagreed. Nonetheless, I fear that some readers may come away with a misguided impression of how science works. Carroll may attempt to frame scientific knowledge as the result of a rational process based on probabilistic rules of inference, but plenty of mysteries and surprises remain. I fear it is a mistake to present science as an alternative to religion, because science can never provide the kind of certainty religion offers. Science is not a fixed set of facts — it is a collection of methods for finding errors in our thinking and hence is structurally self-correcting. Scientific knowledge is always provisional and always subject to radical revision based on revolutionary new insights. We understand a lot, but in the future that understanding will be couched in terms of radically different concepts and principles that illuminate questions that only confuse us now.
The faith that current knowledge will suffice to explain how a brain gives rise to a mind, or how the laws of nature in the universe were chosen, will likely appear, in hindsight, to have been as shortsighted as the idea that Newtonian mechanics sufficed to explain atomic physics. At the end of the 19th century, Lord Kelvin spoke for many over-confident scientists of his era when he declared that science was essentially complete, with only two worrying clouds on the horizon. Those two clouds gave rise to the quantum and relativity revolutions. Despite tremendous progress in the century since, our horizon remains definitively cloudy. By neglecting this cloudiness, Carroll communicates a sunny view of science that is sadly and oddly conservative.
The answer to the unwarranted confidence of religious dogma is not an unwarranted confidence in a current scientific paradigm. Rather, we must embrace the radical uncertainty of scientific knowledge, which is always subject to revision and conceptual revolution. In short, to be a true naturalist requires that we admit the limited nature of the progress we have made, while also celebrating and anticipating with wonder and gratitude how much more remains to be known.
Lee Smolin is a theoretical physicist who was educated at Hampshire College and earned his PhD at Harvard University. He is a founding and senior faculty member at Perimeter Institute for Theoretical Physics.