Joined: Oct. 2005
|Quote (GaryGaulin @ May 28 2018,18:00)|
|Quote (N.Wells @ May 28 2018,17:04)|
|The Dinosaur Train definition of hypothesis is passable for three-year-olds, but it has significant shortcomings for older students and for actual scientists, as we have said before.|
You never presented evidence that a hypothesis needs to be more than "An idea you can test."
First, there is the large problem that your stuff fails according to your own standards, as nothing in your pile of rubbish is testable - you still haven't come up with any valid and logically entailed predictions and any ways to test them.
Second, we went over all this back in Nov. 2014. It is true that an hypothesis can be no more than a statement of a fact that is proposed for the purpose of verification or disproof. It is also true that the Dinosaur Train proposal ("you guys are faster because your legs are longer") is indeed an hypothesis. However, those are not the most useful and fruitful forms of hypotheses in science. With some exceptions to be discussed below (primarily in statistics, philosophy, and some areas of theoretical physics) in most of science our hypotheses are our potential explanations for our observations (theories are also proposed explanations, but theories have some degree of support already, while hypotheses are strictly hypotheticals devised for testing and in most cases disproof), and they are accompanied by alternative explanations and proposals for ways to test them. So to be really useful in science, hypotheses generally need to be testable potential explanations, and benefit from being stated in multiple mutually exclusive sets. Sean Carroll describes it very neatly as (paraphrased), ‘Think of every possible way the world could be - those are your hypotheses. Either before or after, look at how the world actually is - that’s your evidence, your data. Where possible, choose the hypothesis that provides the best fit to the data.’ This requires that the hypotheses be cleanly and clearly stated, using explicit and justified definitions, none of which is true for your stuff. Phil Plait says, “Watch the universe, see how it behaves, make guesses about why it’s doing what it’s doing, and then try to think of ways to support or disprove those ideas” - trying to figure out the ways we might be wrong is one of the best ways to improve, and you aren't doing that. The best and most explicit statement about this process is Strong Inference, as formalized by J.R. Platt in 1964. Platt claimed that the best route for fast scientific progress (for exploring the unknown) comprises four steps:
A) Formulate alternative, mutually exclusive, potentially falsifiable hypotheses
B) Plan observations or experiments whose outcomes will exclude one or more of the hypotheses
C) Make the observations or do the experiments as cleanly as possible
D) Repeat the procedure with new hypotheses.
With regards to philosophy and statistics and other versions of hypotheses, let me recopy some of what I said back in 2014:
| There is a longstanding use in logic where an hypothesis is the antecedent of a proposition: it's A in "If A, then B". Obviously, this has analogies to the use of the term in science, where we make predictions from our hypotheses and then test the predictions. |
There is a very different usage in statistics, which we also use in science: Null hypothesis: A is not different from B, at some level of significance; Alternate hypothesis: A is different from B at that level of significance. In this sense, we have stated two mutually exclusive statements of possibility, and we test them against each other. We follow this model this a lot in science when we build mutually exclusive working hypotheses, and, of course, when we do statistics.
Beyond this, in science, hypotheses (other than statistical ones) usually but not always contain significant elements of explanation, rather than just being declaratory statements of alternative realities that are about to be tested. Also, to be really useful an hypothesis has to be testable.
Your Dinosaur Train definition (an hypothesis is a testable idea) is not bad, and it works fine for kids, but scientific practice is a little more complicated.
For all your trumpeting of hypotheses being testable ideas, how come you haven't generated testable ideas and/or tests for the ideas that you have?
You are welcome to ignore our critiques, but you are going to be irrelevant until you resolve the problems that we have been pointing out.
|Yes, the Dinosaur Train simplification is reasonable for little kids, and yes, in science, hypotheses tend to be useful to the extent that they are testable. However, that is not the beginning and end of what an hypothesis is, as they encompass more than just testability. In particular, the philosophy version of hypothesis need not be testable (it can be a hypothetical, a presumption premised for the sake of argument), and quite often in science people are inspired by hypotheses that are not testable or for which they have not yet thought up a test. However, hypotheses that are most useful in science tend to contain (or have implications for) significant explanations. Your stuff by your standards does not even amount to an hypothesis (because by and large it is not testable), whereas by my standards it doesn't amount to a useful hypothesis (plus it is not stated in a form that lends itself to being a decent hypothesis). |