Ranting About the End of Science
In 1996, John Horgan wrote a silly book called The End of Science. In October 2006, he published an equally silly article in Discover on the same topic. Horgan believes that, somehow, science has already made all the important discoveries that exist and that all that's left to do is fill in the details.
Most of Horgan's arguments resemble a fallacy I call, "The argument from personal incredulity:" "I can't imagine how this could happen, so it can't/didn't/won't happen." This fallacy says a lot more about the limitations of its author than it does about the topic said author is discussing. In Horgan's case, he can't seem to imagine how science might make important, new, fundamental discoveries, so he assumes that they must not exist.
In doing so, he's missing the most important aspect of novel discoveries, their unpredictability. Scientific revolutions are unpredictable by their nature, because if anyone could know the content of a revolution before it happened, they would have, in a sense, already brought it about. Einstein's relativity is a good simple example: the content of the theory was the abandonment of absolute space and time. If at some earlier time, someone had realized that a scientific revolution would come from abandoning absolute space and time, they could have invented the theory themselves. Particularly in the case of special relativity, the equations aren't that hard to derive, it's the concept that's difficult to invent. Similar examples pervade the history of science: for instance, if someone had thought of natural selection before Darwin, they could have revolutionized biology instead of him. I don't mean here to denigrate the detail work required to validate and explain an essential discovery: without the follow through, none of these ideas would rate as anything more than another crazy idea. However, knowing that an idea will be revolutionary is tantamount to having done the work already, so the point stands: knowing about a scientific revolution is the same as making such a revolution.
Extending this fallacy, Horgan makes some rather flimsy, and sometimes disingenuous, arguments, several of which go only to show his lack of understanding of the topic. He accuses his opponents of "faulty inductive reasoning" based on the past expansion of science, but his own analogies that science are ending have deep weaknesses. The fact that special relativity and quantum mechanics limit certain kinds of knowledge has nothing to do with whether science can continue on forever: those theories concern only narrow kinds of physical information, not knowledge about how the world works. His analogy to the exploration of the Earth is a subtle form of begging the question. Because the Earth is a finite object, we know that it's possible to finish exploring it, to some approximation. Horgan is contending that science is likewise finite, so comparing it to the Earth is assuming his entire argument.
In more incorrect reaching based on current science, he observes that natural selection designed humans for reproduction, not "discovering deep truths." However, this vague statement could as easily mean that humans are incapable of science at all, which is obviously untrue, as they are incapable of making further discoveries. Moreover, even if it were true, that doesn't mean that humans can't develop ways to surpass their own limitations, whether through changing themselves or building other things that transcend those.
Horgan addresses the important unsolved problems in science by suggesting some of them might unsolvable and ignoring the rest. He calls research on the former, "ironic science." While I agree that there are in-principle unanswereable questions and might concede one of those he mentions, he goes too far. He argues that string theory is ironic science, but it doesn't even come close to meeting his definition. I share his skepticism about the chances of string theory's ultimate success: the hope behind string theory is that there's a single possible theory of everything based on what we know now about our universe, and that we only need to find the right mathematics for it. Given its notable lack of success, I'm beginning to lean towards the possibility that string theory may not work, for one of three reasons: string theory's mathematical structures are wrong or inappropriate for the task; or the program itself is hopeless, because we need more experimental data to determine what a theory of everything for our universe looks like; or there is no theory of everything, only increasingly good approximations. Only the third possibility makes string theory "ironic science," but even if the pursuit of a theory of everything is vain, that doesn't predict the end of physics because it's still possible to discover those better and better approximations.
Horgan brings up another issue in physics, the increasing expense of, and increasing unwillingness of governments to pay for, high-energy experiments. Again, I agree with his pessimistic summary of the difficulties in the field without agreeing with his conclusions. With the cancellation of the SSC, the era of the classical accelerator had reached its end. The LHC is more or less a cooperative effort between all the high-energy physicists in the world, not to mention all the research grants they've been able to get from all the governments. The LHC's replacement is probably decades off, at best. There's just not much future for large accelerators. That doesn't necessarily mean high-energy physics is dead, though, since someone could invent a much more efficient accelerator, creating even higher energies a fraction of the cost of current designs, or a clever way around needing to use accelerators to study high-energy physics at all. I don't know if either is possible; but more to the point, Horgan doesn't know that both are impossible.
When Horgan tackles a silly argument against his proposition-that scientific knowledge is all provisional-he makes an even sillier conflation of Popper and Kuhn with postmodernism. Popper was the first to point out in a formal way an important logical feature of science: it's possible to falsify, but not verify, a predictive or explanatory theory. An event that doesn't agree with a theory disproves it, but no matter how many events agree with a theory, it's always possible the next won't. Kuhn invented paradigms to explain how revolutionary theories overturn fundamental concepts. However, neither of these ideas have a damn thing to do with postmodernism, nor do they say anything about the limits or lack thereof of science. For an example, if physicists were to discover the true and final theory of everything, Popper would say that it would be impossible to know for certain that it is the true and final theory, and Kuhn would explain how its new concepts replaced old concepts.
Horgan's sections about applied science are particularly lacking in his key deficiency, imagination. He states that nanotechnology will never amount to anything, then cites the dismal failure of fusion power, without accounting for the vast differences between the two situations or showing the faintest recognition that good practical applications, like good theories, are surprises. The fundamental problem with fusion, in my opinion, is that it's a technology searching for a need. People already have plenty of power from burning dinosaurs, so the only possible justification for fusion is cheaper power, but fusion has never even come close to fulfilling that. On the other hand, nanotechnology promises new capabilities with immense value, so there are actual incentives for making it work. One way to see the contrast is that few private companies have ever invested in fusion power, but there are many investing in aspects of nanotechnology. However, even if nanotechnology joins the great graveyard of overhyped technologies, that doesn't mean there won't be others to take its place. In his section on biological science, he opens with an attack on a strawman-crackpot, Aubrey de Grey, and then commits the same error in logic he earlier his opponents accused of: faulty inductive reasoning. He talks about the failures of medicine with cancer and gene therapy, but only implies that these failures will continue, without providing any actual evidence.
In the fields where his case is so weak he can't invoke false analogies, he resorts to special pleading. He argues that neuroscience has never disproved Freud-but the number of academic psychologists who take Freud seriously is vanishingly small, and you will find few articles in reputable journals based on stereotypical Freudian ideas. As evidence of Freud's endurance, he offers a Newsweek headline, but that's comparable to saying that evolution never defeated creationism because of the latter's devoted following among certain Christians, or that Einstein's ideas justify belief in Eastern religions because of the vacuous popularizations of that idea. He also implies the rather naive argument that if there doesn't exist a neural code similar to the genetic code (and there doesn't), that there aren't still deep results possible in the field. For mathematics, since Godel's theorem refutes his position prima facie, he falls back to contending that the theorems mathematicians are working on now are too difficult for humans to comprehend or find interesting.
I find Horgan's case thin. However, rather than presenting a wholly negative counterargument, I'll note some important outstanding problems that I have no reason to believe aren't soluble.
We know that the kinds of matter we're familiar with only make up about 4% of the matter in the universe. Another 22% is probably dark matter, and 74% is dark energy. There's lots of speculation, but little hard data, on what the dark matter is. The most common speculation for the dark energy is that empty space carries some energy, but no known theoretical model predicts anything close to the observed value. It's also possible that our interpretations of the data as showing dark matter and dark energy are wrong, and something else altogether is going on. The only theories about the origin of life are all speculation. We have a variety of ideas about how it might have happened, but only the most circumstantial evidence for any of them. Discovering life from a separate abiotic-to-living transition would shed light on this problem, as would any steps to making living systems or partial living systems from abiotic components in the lab in a way that simulates something that might have happened without intelligent intervention. Meanwhile, this also touches on several important related questions. How many different types of living systems can exist in our universe? Only based on carbon? Only using DNA/RNA or minor variations as genetic codes? Etc. How common is life? How does the genetic code work? We have a rough idea of the biochemical details and the outlines of the regulatory system, but next to no understanding of the coordinations that transform a single cell into an adult multicellular organism. Concomitant with this, we only have no real understanding of how genes affect biological systems except for corner cases. How does the brain create the mind? We have some idea of how the individual components, neurons, work, what large portions of the brain do, and how the mind works, but we lack detailed understanding in all cases, and there's no consensus at all. This also ties in with the question of hard AI: how do you give a computer intellectual abilities that match a human's in areas like natural language comprehension? Sociology and related fields are still well in their stamp collecting phase. We have lots of data about human behavior in small groups, but nothing resembling a cohesive theory at all.
I could go on, but I think that sample shows well enough that science is not anywhere near an end; and that's not accounting for all the strange and weird things we haven't stumbled over yet.
Most of Horgan's arguments resemble a fallacy I call, "The argument from personal incredulity:" "I can't imagine how this could happen, so it can't/didn't/won't happen." This fallacy says a lot more about the limitations of its author than it does about the topic said author is discussing. In Horgan's case, he can't seem to imagine how science might make important, new, fundamental discoveries, so he assumes that they must not exist.
In doing so, he's missing the most important aspect of novel discoveries, their unpredictability. Scientific revolutions are unpredictable by their nature, because if anyone could know the content of a revolution before it happened, they would have, in a sense, already brought it about. Einstein's relativity is a good simple example: the content of the theory was the abandonment of absolute space and time. If at some earlier time, someone had realized that a scientific revolution would come from abandoning absolute space and time, they could have invented the theory themselves. Particularly in the case of special relativity, the equations aren't that hard to derive, it's the concept that's difficult to invent. Similar examples pervade the history of science: for instance, if someone had thought of natural selection before Darwin, they could have revolutionized biology instead of him. I don't mean here to denigrate the detail work required to validate and explain an essential discovery: without the follow through, none of these ideas would rate as anything more than another crazy idea. However, knowing that an idea will be revolutionary is tantamount to having done the work already, so the point stands: knowing about a scientific revolution is the same as making such a revolution.
Extending this fallacy, Horgan makes some rather flimsy, and sometimes disingenuous, arguments, several of which go only to show his lack of understanding of the topic. He accuses his opponents of "faulty inductive reasoning" based on the past expansion of science, but his own analogies that science are ending have deep weaknesses. The fact that special relativity and quantum mechanics limit certain kinds of knowledge has nothing to do with whether science can continue on forever: those theories concern only narrow kinds of physical information, not knowledge about how the world works. His analogy to the exploration of the Earth is a subtle form of begging the question. Because the Earth is a finite object, we know that it's possible to finish exploring it, to some approximation. Horgan is contending that science is likewise finite, so comparing it to the Earth is assuming his entire argument.
In more incorrect reaching based on current science, he observes that natural selection designed humans for reproduction, not "discovering deep truths." However, this vague statement could as easily mean that humans are incapable of science at all, which is obviously untrue, as they are incapable of making further discoveries. Moreover, even if it were true, that doesn't mean that humans can't develop ways to surpass their own limitations, whether through changing themselves or building other things that transcend those.
Horgan addresses the important unsolved problems in science by suggesting some of them might unsolvable and ignoring the rest. He calls research on the former, "ironic science." While I agree that there are in-principle unanswereable questions and might concede one of those he mentions, he goes too far. He argues that string theory is ironic science, but it doesn't even come close to meeting his definition. I share his skepticism about the chances of string theory's ultimate success: the hope behind string theory is that there's a single possible theory of everything based on what we know now about our universe, and that we only need to find the right mathematics for it. Given its notable lack of success, I'm beginning to lean towards the possibility that string theory may not work, for one of three reasons: string theory's mathematical structures are wrong or inappropriate for the task; or the program itself is hopeless, because we need more experimental data to determine what a theory of everything for our universe looks like; or there is no theory of everything, only increasingly good approximations. Only the third possibility makes string theory "ironic science," but even if the pursuit of a theory of everything is vain, that doesn't predict the end of physics because it's still possible to discover those better and better approximations.
Horgan brings up another issue in physics, the increasing expense of, and increasing unwillingness of governments to pay for, high-energy experiments. Again, I agree with his pessimistic summary of the difficulties in the field without agreeing with his conclusions. With the cancellation of the SSC, the era of the classical accelerator had reached its end. The LHC is more or less a cooperative effort between all the high-energy physicists in the world, not to mention all the research grants they've been able to get from all the governments. The LHC's replacement is probably decades off, at best. There's just not much future for large accelerators. That doesn't necessarily mean high-energy physics is dead, though, since someone could invent a much more efficient accelerator, creating even higher energies a fraction of the cost of current designs, or a clever way around needing to use accelerators to study high-energy physics at all. I don't know if either is possible; but more to the point, Horgan doesn't know that both are impossible.
When Horgan tackles a silly argument against his proposition-that scientific knowledge is all provisional-he makes an even sillier conflation of Popper and Kuhn with postmodernism. Popper was the first to point out in a formal way an important logical feature of science: it's possible to falsify, but not verify, a predictive or explanatory theory. An event that doesn't agree with a theory disproves it, but no matter how many events agree with a theory, it's always possible the next won't. Kuhn invented paradigms to explain how revolutionary theories overturn fundamental concepts. However, neither of these ideas have a damn thing to do with postmodernism, nor do they say anything about the limits or lack thereof of science. For an example, if physicists were to discover the true and final theory of everything, Popper would say that it would be impossible to know for certain that it is the true and final theory, and Kuhn would explain how its new concepts replaced old concepts.
Horgan's sections about applied science are particularly lacking in his key deficiency, imagination. He states that nanotechnology will never amount to anything, then cites the dismal failure of fusion power, without accounting for the vast differences between the two situations or showing the faintest recognition that good practical applications, like good theories, are surprises. The fundamental problem with fusion, in my opinion, is that it's a technology searching for a need. People already have plenty of power from burning dinosaurs, so the only possible justification for fusion is cheaper power, but fusion has never even come close to fulfilling that. On the other hand, nanotechnology promises new capabilities with immense value, so there are actual incentives for making it work. One way to see the contrast is that few private companies have ever invested in fusion power, but there are many investing in aspects of nanotechnology. However, even if nanotechnology joins the great graveyard of overhyped technologies, that doesn't mean there won't be others to take its place. In his section on biological science, he opens with an attack on a strawman-crackpot, Aubrey de Grey, and then commits the same error in logic he earlier his opponents accused of: faulty inductive reasoning. He talks about the failures of medicine with cancer and gene therapy, but only implies that these failures will continue, without providing any actual evidence.
In the fields where his case is so weak he can't invoke false analogies, he resorts to special pleading. He argues that neuroscience has never disproved Freud-but the number of academic psychologists who take Freud seriously is vanishingly small, and you will find few articles in reputable journals based on stereotypical Freudian ideas. As evidence of Freud's endurance, he offers a Newsweek headline, but that's comparable to saying that evolution never defeated creationism because of the latter's devoted following among certain Christians, or that Einstein's ideas justify belief in Eastern religions because of the vacuous popularizations of that idea. He also implies the rather naive argument that if there doesn't exist a neural code similar to the genetic code (and there doesn't), that there aren't still deep results possible in the field. For mathematics, since Godel's theorem refutes his position prima facie, he falls back to contending that the theorems mathematicians are working on now are too difficult for humans to comprehend or find interesting.
I find Horgan's case thin. However, rather than presenting a wholly negative counterargument, I'll note some important outstanding problems that I have no reason to believe aren't soluble.
I could go on, but I think that sample shows well enough that science is not anywhere near an end; and that's not accounting for all the strange and weird things we haven't stumbled over yet.