|
Author
|
Topic: The One Paper All Scientists Should Read
|
John Bracht
Member
Member # 5
|
posted 03. September 2002 20:12
Hi everyone,
I don't do this very often, but I wanted to recommend a PHENOMENAL paper I just read as a way to strengthen my own scientific reasoning. It's titled "Strong Inference" and published by Science magazine in 1964; it's a must-read for any scientist (in my opinion). Here's the citation:
Platt J. Strong Inference. Science 146(3642);16 Oct 1964:1-6
The author outlines why some scientists do great science and others are only mediocre. Particularly, he cautions against doing "method-based science" in favor of "problem-based science" and the need to design the crucial experiment to rule out competing hypotheses.
It is this issue of formulating multiple, competing hypotheses that really seemed relevant to much discussion over intelligent design and evolution (though that's not what the paper was about):
quote:
Chamberlin says our trouble is that when we make a single hypothesis, we become attached to it.
"The moment one has offered an original explanation for a phenomenon which seems satisfactory, that moment affection for his intellectual child springs into existence, and as the explanation grows into a definite theory his parental affections cluster about his offspring and it grows more and more dear to him...There springs up also unwittingly a pressing of the theory to make it fit the facts and a pressing of the facts to make them fit the theory..."
quote:
Some cynics tell a story, which may be apocryphal, about the theoretical chemist who explained to his class.
"And thus we see that the C-Cl bond is longer in the first compound than in the second because the percent of ionic character is smaller."
A voice from the back of the room said, "But Professor X, according to the Table, the C-Cl bond is shorter in the first compound."
"Oh, is it?" said the professor, "Well, that's still easy to understand, because the double-bond character is higher in that compmound."
To the extent that this kind of story is accurate, a "theory" of this sort is not a theory at all, because it does not exclude anything. It predicts everything, and therefore does not predict anything. It becomes simply a verbal formula which the graduate student repeats and believes becauset his professor has said it so often. This not science, but faith; not theory, but theology. Whether it is hand-waving or number waving, or equation-waving, a theory is not a theory unless it can be disproved. That is, unless it can be falsified by some possible experimental outcome.
I think this paper is a great reminder (if not a wake-up call) to many in science (myself included) to really be thinking rigorously about how to falsify theories, and how to design crucial experiments such that we can exclude alternative hypotheses. I've definitely experienced the tendency of people (and myself) to grow "affectionate" toward their (and my) own theories, and it's good mental exercise to think of alternative hypotheses and crucial experiments to distinguish between them. I highly encourage all scientists out there: get this paper and read it! It will be worth your time, and could revolutionize how you do science (and the kinds of results you get)!
Finally, I just wanted this to be a gentle reminder to everyone in the debate over the origins of complex systems (see, I had to get this post on-topic somehow!) to not be too attached to your favorite hypothesis, no matter what your position, to think seriously about alternative hypotheses (and how to exclude them). By this I mean going beyond the stereotypical "design" hypothesis and "evolution" hypothesis, but rather the many different design and evolution hypotheses that are out there (there's plenty of variety of each, and they certainly aren't mutually exclusive either). This whole exercise requires a good does of humility on both sides of the aisle, as well as a lot of hard thinking about detailed, testable scenarios, but I think it could revolutionize our science of complex systems of all types.
John Bracht
[ 03 September 2002, 20:18: Message edited by: John Bracht ]
IP: Logged
|
|
Mark Szlazak
Member
Member # 391
|
posted 03. September 2002 22:56
Thanks for the reminder but this isn't anything new! In fact it has a name, "eliminative induction", and has been know for over two centuries. My scientific mentor who prided himself as a good experimentalist always operated this way. In fact, he learned this from his mentor along with the saying that "data has at least two interpretations, but often its many more", a rough mark of good experiment(s) in his mind was the number of control groups it/they had to account for these.
A book that discusses this approach to general relativity testing is Earman's "Bayes or Bust?", chapter 7: A Plea for Eliminative Induction.
By the way, this makes "falsifications" relative to the competing hypotheses, this is not the falsificationism Popper wanted. Thank God for that! Poppers falsification is known to be a failure, simultaneously being too permissive and too restrictive. It countenanced as scientific many claims that palpably were not. Equally, it denied scientific status to many claims that were patently scientific. Thus, Popperians had to deny the scientificity of singular existential clams--for instance, that there are black holes.
[ 04 September 2002, 00:48: Message edited by: Mark Szlazak ]
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 04. September 2002 09:45
While it is useful to be reminded of the benefits of "problem-based science" and the value of considering a range of testable/falsifiable hypothesis, it also important to recognize that the there are basic, fundamental, technical issues associated with evolution/ID question which are not addressed by any known theory formulation/theory validation process.
The processes responsible for ‘changes in biological design’, (whether the changes involved are called evolution, learning, or some other form of adaptive change) are complex processes. There are no generally accepted procedures and standards for formulating ‘scientific models and theories’ of such complex processes. And if someone does formulate a complex model or theory of such a process, there are no generally accepted procedures for testing/validating/falsifying.
The idea of ‘going back into the literature’ and finding useful techniques which will help resolve the evolution/ID issue must be very comforting to many individuals. It is, however, useful to note that no one in the past ever found reliable ‘hard science’ solutions to any of the ‘processes responsible for changes in biological design’. If humility is beneficial, then it should be truly humbling to recognize how little progress has really been made in constructing scientific theories and models of the complex processes responsible for changes in biological design.
While I agree wholeheartedly with the benefits of looking back at things that worked in the past, I believe it is also critical that we recognize that the set of techniques developed in the past have so far failed, and failed rather miserably, at producing hard science models of complex processes such as design. If we ever hope to resolve the evolution/ID issue (or even just the evolution issue) we not only need to look to the past, but we also need to look beyond the box created by our past successes.
The point of my remarks is not to open up an unproductive debate about the usefulness/effectiveness of existing scientific techniques. I simply want to point out that there are a set of techniques currently being developed under the heading ‘design science’ which appear to address some of the fundamental/technical problems associated with constructing and validating theories of complex processes or paradigms.
The design science approach involves formulating and testing/validating theories in terms of 'solution spaces and restrictions on solution spaces’ rather than in terms of ‘mathematical algorithms’. The ‘new’ approach appears to have two ‘interesting’ features. First, it appears to provide almost immediate practical results with respect to formulating and testing new theories and models of changes in design science. Second, although the terminology and concepts used in the new approach are very different, the ‘new’ approach appears to be simply a ‘reinterpretation of’ rather than a ‘change in’ the scientific procedures that worked in the past.
I would strongly suggest that anyone interested in solutions to ‘problem based science’ take a quick gander at the developing design science techniques.
IP: Logged
|
|
nobody
Member
Member # 145
|
posted 04. September 2002 12:38
Good morning Warren,
That sounds very good. Do you have any links that I could start with?
Thanks in advance.
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 05. September 2002 11:36
Nobody,
Since you express an interest, I would be happy to send you additional information privately.
IP: Logged
|
|
|
Printer-friendly view of this topic