Our discussions here may have become what Thomas Kuhn (in The Structure of Scientific Revolutions) described as "incommensurable" - that is, we are not discussing the issue using common terminology.

It seems to me - from my view - that you are conflating "hypothesis" with a "final hypothesis" or "conclusion". Your definition of a hypothesis seems to be a statement based on "evidence" that describes what probably happened during an incident. I don't understand how data becomes "evidence" other than as an evaluation against a hypothesis - for example I observe beading on an electrical conductor. That is data that, in and of itself, may have no meaning - it's just a bead on a conductor. It becomes "meaningful" when we evaluate it against the hypothesis "Beading on electrical conductors occurs when the conductor is energized." Now, the observation of "THE BEAD"(the data) becomes "evidence" in support of the hypothesis that the conductor was energized. The beading on an energized conductor may gain more significance if we hypothesize that the beading occurred as a result of fire attacking an energized conductor.
It seems to me that, if I have understood you correctly, you contend that I could not have formed the hypothesis that "Beading occurs on energized conductors" without first having observed the "evidence" - the bead. The hypothesis (Beading occurs on energized conductors") is formed after the "evidence" (the bead) is observed. I struggle with that methodology. I can go into a fire scene without evidence but with hypotheses. (In fact, I argue that the more hypotheses you form, the better your conclusions will be.) For example, I can go to a fire scene with the knowledge that "beading occurs on energized electrical conductors" without knowing whether there is even electricity or conductors at the scene. I can form the hypothesis "What if this fire resulted from an electrical event?" - along with hundreds of other hypotheses. The formation of the hypothesis does not make it "invalid". If I formed conclusions based solely on that hypothesis, without making observations - collecting data- and analyzing that data against the hypothesis- then I have problems. However, when I observe the bead, then I have some support for the hypothesis that the conductors were energized - the data now becomes "evidence" that "beads only form on energized conductors" and that energized conductors would be necessary for electricity to be part of the ignition sequence. (There are other issues that develop if I ONLY look for confirmational data but I am not suggesting that you ONLY look for data confirming or refuting one hypothesis at a time.)




Consider the "knowledge" that beads form on energized conductors. How do we know that and where does that knowledge come from? The research that supports that knowledge came from a hypothesis. Granted that beads were observed on fire scenes - data - but the assignment of meaning to that data came from research based on the hypothesis "How do beads form on conductors?", "what if beads only form on energized conductors?" and its counter "what if beads can form on unenergized conductors?" From those questions, test plans can be developed. If beads form on energized circuits, that hypothesis is supported. If they form on the "dead" conductors, then that hypothesis is supported and the counter hypothesis refuted (at least under those test parameters.) What if the researcher who formed the hypothesis "what if beads only form on energized conductors" had been told that his hypothesis was invalid and not based on "evidence" - how could they have attempted to acquire the needed evidence if they couldn't even ask the question without the evidence? Ask all the questions you want (a hypothesis is just a question - a "what if" statement.)

"Negative Corpus" is a long discussion, perhaps for another thread (again.) But, I believe it is an overused term and pulled out as a blanket attack to opposing opinions. Negative corpus used to be applied only to the concept that "I have no idea how this fire started so it must be an incendiary fire" with no more support other than an "ipse dixit" assertion. It has morphed since then.

Steve