HumblePie
Lifer
- Oct 30, 2000
- 14,665
- 440
- 126
Since the establishment of a hypothesis to test is the first step in the scientific method, it is virtually by definition that a hypothesis starts out with no evidence to support it. Once the hypothesis has been established, however, we need to follow through (as you describe) by designing methods to collect data while we test the hypothesis so that on completion we can determine whether or not the hypothesis descibes our real world. Hypothesis that are demonstrated to be true add to our scientific understanding of reality; hypothesis that turn out to be false are discarded.
I'm not sure that the scientific approach to knowledge allows for a "limbo" to hold untested and/or untestable hypothesis. It seems to me that until these hypothesis are tested (e.g. the tooth fairy and string theory?), they need to be banished into the realm of philosophy/religion.
Wrong, that's why it's a method and not a subjective determination. Also you can have some evidence before making a hypothesis easy enough. For example: with my dish washing detergent. In my earlier example I asked you to picutre a person with no knowledge of what that was and given a list of names of brands and asked to pick out which one was best from that list. A person with no knowledge of the subject could still formulate a hypothesis that a given brand, chosen at random, would be best. However a person with prior knowledge may choose the brand that they've used in the past even if they've never used all brands out there. That second person would be creating a hypothesis with a limited set of evidence already obtained, despite obtaining that evidence prior to forming their hypothesis. Why? Much of the time a hypothesis is fomed after evidence for an event has occurred. It makes it easier to create a hypothesis that might lead us to tests that move our understanding int he right direction.
However, one can easily lose sight of the scientific method and only do tests that may only validate and provide the only proof needed for their hypothesis to be correct. Going back to my earlier example of the dish washing detergent. If the second person, who's used dish washing detergent before, is asked to pick the best brand and that person chooses their favorite brand that is fine. What is not fine is if that person ONLY does tests of comparisons between the brand they like and brands they already know they don't like and aren't as good. See what I'm getting at here? Unless all tests are completed the hypothesis will remain unvalidated fully. This actually happens quite a bit too.