Development and application of fuzzy probability to expert witnessing in warning theory

If an expert when questioned in court judges it to be "more probable than not" that a person will read a warning and furthermore judges it "more probable than not" that that person will then heed it, does it therefore necessarily follow that he should conclude that it is "more probable than not" that the person will read and heed a warning? A yes or no answer to such questions commonly figures prominently in multimillion dollar awards to parties seeking compensation for injuries through the courts. Although probability theory would permit such a computation if the individual probabilities were known, no one is likely to be able to defend any particular set of exact values. Or, an expert may opine that the probability of reading such a warning could have been any value between 0 and 1 and the probability of following the warning advice also any probably between 0 and 1 and therefore no conclusion could be drawn one way or another. Does the conclusion "no conclusion could be drawn one way or another" follow from the premises? It is possible to give definitive and reasonable numerical and/or categorical answers to such questions and in the process elucidate the underlying bases for an expert's conclusions.