jonathanjong
Headphoneus Supremus
- Joined
- Aug 4, 2008
- Posts
- 4,041
- Likes
- 12
Here's something I'm working on on the D v. ID issue. It's illustrative of my views on defining science. I don't know if you can be bothered, but I'd like some constructive criticism on where I'm going wrong:
What are the aims of science? This answer to this question depends, of course, on whether one is a scientific realist or a scientific anti-realist (or instrumentalist, etc.). I’m a scientific realist, and so are most scientists. I submit that scientific realists believe the aim of science to be the discovery and explanation of facts about the world. We observe facts, collect phenomena, and propose explanations (or theories) that (if they are good theories) both explain previous known facts and generate true novel predictions. If we are also methodological naturalists, then we assume that all scientific explanations must be naturalistic explanations. So, what should the scientist do when no naturalistic explanation is available for a given phenomenon? When asked, “What explains X (where X is a phenomenon for which there is no good naturalistic explanation)?”, she should reply, “We don’t know…yet.” But what if someone proposes an explanation (or theory), which explains X, but which involves supernatural events and entities? Then, the methodological naturalist should say, “That theory is nice, but I reject it on methodological grounds, and would rather wait for a purely naturalistic explanation.” But this is dogmatic.
Perhaps qua scientist, the methodological scientist might say, “No thanks, I’ll wait for a naturalistic explanation”, but once she steps out of the lab, she whispers, “Actually, strictly off the scientific record, I do think that’s a reasonable explanation.” This allows her to keep her methodological naturalism as a scientist, without being dogmatic. But this is odd and schizophrenic. It’s requires the scientist to compartmentalize her life, so that what is reasonable to believe as a person (in general) is not reasonable to believe as a scientist.
So, I submit that science should be in the business of discovering and explaining facts (and phenomena), without being committed to only naturalistic explanations. Scientists should look for the best explanations, whether or not they involve gods, goblins, and gnomes. But then, do we still have a principle of demarcation between science and non-science? I’m not sure we do, actually. And this is troubling, but I think it’s acceptable. But hold on, what about Popper’s widely accepted principle of demarcation: falsifiability?
To see if falsifiability does the trick, let’s take a test case: The neo-Darwinian theory of evolution by natural selection (NS) v. Intelligent Design (ID). In the U.S., many recently sought a principle of demarcation between “scientific” and “non-scientific” theories to adjudicate on the issue of whether ID should be taught alongside NS in schools. For the most part, falsification was thought to fit the bill. The claim is that NS is falsifiable and therefore scientific and therefore appropriate to be taught in high school science classes, and that ID is unfalsifiable and therefore unscientific and therefore inappropriate to be taught in high school classes. I think this is false. NS is no more falsifiable than ID is. Why not?
This is a difficult question. Both Kuhn and Lakatos have pointed out that most scientific theories are made up of a “hard core” (Lakatos’s term), as well as a “protective belt” (Lakatos again), such that the “protective belt” is always blamed from successful falsifications of hypotheses generated by the theory (so that the “hard core” survives unscathed). This seems to be true. Newtonian physics was not abandoned when it was found that Mercury’s behaviour did not conform to it. Instead, astronomers assumed that there was an unseen planet or asteroid belt, etc. Neither was NS abandoned in light of eusocial insects with sterile castes (which seemed to contradict NS). Instead, Darwin proposed group selection. Indeed, it is difficult to see what would constitute a successful falsifier of NS. I doubt Dawkins would give NS up even if a mammalian skull was found in a stratum of rock that dates 200 million years back. He might claim it’s a fake, or something. Indeed, even if potential reproductive success was a poor predictor of actual reproductive success (remember: this was the solution against the charge that NS was tautological), NS would not be falsified. Or so I claim. On the flip side, there are falsifiable versions of Intelligent Design. Or at least versions of ID that are as falsifiable as NS. A corollary of ID is that there is “specific complexity” in organisms. That is, there are some features of some biota that are “irreducibly complex”, defined in some mathematical terms I don’t pretend to understand. Now, this is an empirical claim that is either true or false, and we can check whether it is true or false. That is, we can falsify it. So, ID generates falsifiable hypotheses. Whence the difference? Not in falsifiability.
Now, the U.S. got rid of Creation Science on religious grounds. It’s unconstitutional because it favours a particular religion. ID is much more slippery in this regard. ID need not be committed to any religious doctrine. It only states that there are phenomena (i.e., specific complexity) that are inexplicable in NS terms, and that an intelligent designer (whatever that might be) is necessary (or the best explanation) for such phenomena. Methodological naturalism notwithstanding, this is a fair enough theory.
So, should we teach ID in high school science classes? Well, it depends. It depends, not whether ID is science or not (as I don’t think we have good principles of demarcation for doing so), but whether or not it’s good science. Is “specified complexity” is sound concept? Are there “irreducibly complex” features of biota inexplicable in NS terms? If not, then it shouldn’t be taught because it’s bad science. We no longer teach Ptolemaic astronomy and Phlogiston theory, because we currently have better physical and chemical theories than these. If NS is (currently) a better theory than ID, then why take up time to teach ID? Sure, ID might be true (and NS might be false), but at the moment it doesn’t seem like we have good reasons for believing so. We can’t teach everything, and NS is difficult enough to teach as it is. So, no ID. Not even my rejection of methodological naturalism and falsification as principles of demarcation can save ID.
What are the aims of science? This answer to this question depends, of course, on whether one is a scientific realist or a scientific anti-realist (or instrumentalist, etc.). I’m a scientific realist, and so are most scientists. I submit that scientific realists believe the aim of science to be the discovery and explanation of facts about the world. We observe facts, collect phenomena, and propose explanations (or theories) that (if they are good theories) both explain previous known facts and generate true novel predictions. If we are also methodological naturalists, then we assume that all scientific explanations must be naturalistic explanations. So, what should the scientist do when no naturalistic explanation is available for a given phenomenon? When asked, “What explains X (where X is a phenomenon for which there is no good naturalistic explanation)?”, she should reply, “We don’t know…yet.” But what if someone proposes an explanation (or theory), which explains X, but which involves supernatural events and entities? Then, the methodological naturalist should say, “That theory is nice, but I reject it on methodological grounds, and would rather wait for a purely naturalistic explanation.” But this is dogmatic.
Perhaps qua scientist, the methodological scientist might say, “No thanks, I’ll wait for a naturalistic explanation”, but once she steps out of the lab, she whispers, “Actually, strictly off the scientific record, I do think that’s a reasonable explanation.” This allows her to keep her methodological naturalism as a scientist, without being dogmatic. But this is odd and schizophrenic. It’s requires the scientist to compartmentalize her life, so that what is reasonable to believe as a person (in general) is not reasonable to believe as a scientist.
So, I submit that science should be in the business of discovering and explaining facts (and phenomena), without being committed to only naturalistic explanations. Scientists should look for the best explanations, whether or not they involve gods, goblins, and gnomes. But then, do we still have a principle of demarcation between science and non-science? I’m not sure we do, actually. And this is troubling, but I think it’s acceptable. But hold on, what about Popper’s widely accepted principle of demarcation: falsifiability?
To see if falsifiability does the trick, let’s take a test case: The neo-Darwinian theory of evolution by natural selection (NS) v. Intelligent Design (ID). In the U.S., many recently sought a principle of demarcation between “scientific” and “non-scientific” theories to adjudicate on the issue of whether ID should be taught alongside NS in schools. For the most part, falsification was thought to fit the bill. The claim is that NS is falsifiable and therefore scientific and therefore appropriate to be taught in high school science classes, and that ID is unfalsifiable and therefore unscientific and therefore inappropriate to be taught in high school classes. I think this is false. NS is no more falsifiable than ID is. Why not?
This is a difficult question. Both Kuhn and Lakatos have pointed out that most scientific theories are made up of a “hard core” (Lakatos’s term), as well as a “protective belt” (Lakatos again), such that the “protective belt” is always blamed from successful falsifications of hypotheses generated by the theory (so that the “hard core” survives unscathed). This seems to be true. Newtonian physics was not abandoned when it was found that Mercury’s behaviour did not conform to it. Instead, astronomers assumed that there was an unseen planet or asteroid belt, etc. Neither was NS abandoned in light of eusocial insects with sterile castes (which seemed to contradict NS). Instead, Darwin proposed group selection. Indeed, it is difficult to see what would constitute a successful falsifier of NS. I doubt Dawkins would give NS up even if a mammalian skull was found in a stratum of rock that dates 200 million years back. He might claim it’s a fake, or something. Indeed, even if potential reproductive success was a poor predictor of actual reproductive success (remember: this was the solution against the charge that NS was tautological), NS would not be falsified. Or so I claim. On the flip side, there are falsifiable versions of Intelligent Design. Or at least versions of ID that are as falsifiable as NS. A corollary of ID is that there is “specific complexity” in organisms. That is, there are some features of some biota that are “irreducibly complex”, defined in some mathematical terms I don’t pretend to understand. Now, this is an empirical claim that is either true or false, and we can check whether it is true or false. That is, we can falsify it. So, ID generates falsifiable hypotheses. Whence the difference? Not in falsifiability.
Now, the U.S. got rid of Creation Science on religious grounds. It’s unconstitutional because it favours a particular religion. ID is much more slippery in this regard. ID need not be committed to any religious doctrine. It only states that there are phenomena (i.e., specific complexity) that are inexplicable in NS terms, and that an intelligent designer (whatever that might be) is necessary (or the best explanation) for such phenomena. Methodological naturalism notwithstanding, this is a fair enough theory.
So, should we teach ID in high school science classes? Well, it depends. It depends, not whether ID is science or not (as I don’t think we have good principles of demarcation for doing so), but whether or not it’s good science. Is “specified complexity” is sound concept? Are there “irreducibly complex” features of biota inexplicable in NS terms? If not, then it shouldn’t be taught because it’s bad science. We no longer teach Ptolemaic astronomy and Phlogiston theory, because we currently have better physical and chemical theories than these. If NS is (currently) a better theory than ID, then why take up time to teach ID? Sure, ID might be true (and NS might be false), but at the moment it doesn’t seem like we have good reasons for believing so. We can’t teach everything, and NS is difficult enough to teach as it is. So, no ID. Not even my rejection of methodological naturalism and falsification as principles of demarcation can save ID.