You guys... I have no hope for
jay, but several of the rest of you don't understand falsifiability either.
The NIST theory is falsifiable because, using the formal definition, it is logically possible for observation or experiment to verify or contradict its conclusions.
The NIST theory is also
testable, which is a stricter version, because one does not need to rely upon observation. One can devise an experiment, and, in theory, conduct the test whenever one wishes.
Let's tackle the muddy case first, that of observation. This does not mean we "observe the Towers aren't standing" or "we observe a meteor didn't do it." That proves nothing beyond the observation itself. Instead, the observation is as follows: If, in the future, an identical structure should happen to be struck by an aircraft, in a materially similar manner, will we observe consistent behavior or not? This question can be asked, makes logical sense, and provides a definite answer. Therefore, the theory
is falsifiable.
Now, since we have no expectation that anyone will build a near copy of a WTC Tower, nor that it will be hit by a passenger airliner, we may have a very long time to wait. This is why the stricter definition
testable is more useful.
So, how do we test the NIST theory?
Newtons Bit has already given one correct answer -- we can build our own finite element model and do experiments on that. And, guess what, many people have. That list includes Weidlinger Associates, Purdue University, University of Edinburgh, Arup, and University of California at Berkeley. The results, incidentally, are unanimous: Apart from fine details, all confirm NIST's result.
The Truthers like to trot out this ridiculous canard that they cannot test NIST's theory unless NIST gives them the model, and until NIST does so, they are justified in maintaining their idiotic beliefs. This excuse (a
call to perfection logical fallacy) fails for multiple reasons. For starters, the NIST models
are available, you just have to pay for them. Some are in the public domain. But beyond this, if you wanted to test the theory, you wouldn't use their models in the first place. If you get their models and re-run them, you will get the same answer, unless NIST actually never ran the model in the first place and made up results out of thin air. (They did not.) Instead, you are better off making your own models, because if you use the same models you expose yourself to systematic error. All of the researchers I listed above built their own models, each with its own tradeoffs and limitations. Personally I favor the Purdue model above all others, including NIST's, for reasons I expounded upon in my
whitepaper -- but any one will do to prove the basic hypothesis of fire + impact + time = collapse.
Other ways to test NIST's theory include first principles analysis and scale modeling. I described this process in my first
appearances on
Hardfire. While the calculations I ran through in these three shows are very simple, they nonetheless are valid tests one can apply to NIST's theory. But one can do much more detailed tests by constructing models based on these calculations, and some, notably Dr. Quintiere at Northwestern, have done so. Results? Again, small discrepancies from NIST's results, but utterly consistent with the overall theory.
You can also see that NIST's theory is testable because it has been tested, specifically to compare it against other competing theories. The best known is the "pancake hypothesis." In brief, this hypothesis is similar to NIST's in terms of impact and fire, but differs with respect to structural response. In the "pancake hypothesis," voiced by FEMA early in the investigation, the collapse mode initiates when floor connections fail, leaving columns with longer unsupported lengths, leading to elastic buckling. In NIST's theory, the floor connections must remain intact, as this delivers an eccentric load after the floor trusses buckle and sag. NIST provides numerous tests and compares these to both hypotheses, in the end concluding that the "pancake hypothesis" is wrong -- it cannot be reconciled with video evidence, and it is a failure mode that the Towers were relatively resilient against. NIST's theory, in contrast, survives these tests. But, were this not the case, NIST would have discarded this theory. Hence, the theory is testable. It's been tested!
In general, to decide whether or not a theory is testable, just ask yourself: How would I check this to see if it's true? What would a different result look like?
Here's a practice deal. One genuninely unfalsifiable theory is the Creationist theory of evolution. How would you test that some unseen, ill-defined supernatural force guides evolution? Can't do that? All right, let's give up on testability and go for falsifiability -- is it possible to observe evolution that definitely does not fit this definition? The answer is no.
Quite obviously, this has nothing in common with NIST's theory. Virtually every bit of work that they did, from digging through old documentation to running computer models under different scenarios to destructively evaluating components, is a test designed to confirm or disprove one of their core assumptions. Read it with this mindset and it will become much clearer to you
why they did what they did.
Let's not talk about this anymore. It's all in the reports. All you have to do is read them. Pay particular attention to NCSTAR1-2B chapter 9.12 and NCSTAR1-6 chapter 9.4.4.
