Let's say we are acting as a judge in a big farmer contest where the winner will win 5 million USD.
The winner will be the farmer who can convince you that they've produced a green bowlingball-size sallad.
You know that sallads, according to the worlds causal relationships, need water to grow. *If you've guessed that plants will grow without water in the past you've been more wrong than right with your guesses*
There has been a severe drought all this year where no farmer has gotten any water at all on his salladsfields and it is forbidden to water your own fields.
You then get a call from a farmer X who claims to have produced a green bowlingball-size sallad yesterday and eaten it together with 10 other random people he invited from the street and who he claims are independent and not biased towards him. He claims that he and the 10 other people can testify under oath in court(risking perjury) that what X has stated is true and therefor he is entitled to get the prize money since 11 independent witness testimony under oath should be enough justification.
A casual inference, relying on the claim that there are casual structures, makes it very low probability that the farmer has not produced a green bowlingball-size sallad since there has not been any rain at all and other farmers have seen their fields turn into dry deserts. There are not enough ingredients for such a phenomenon to be produced.
You present the low probability for the event to have happened to X but he says that you can't trust that there are reliable casual relationships in the world and therefor you have no JTB and justification to rely on a casual inference to deny him the right to the 5 million USD when he and 10 other people give justification to his statement.
Is X correct that relying on causal relationships and casual inferences is not justified and should not be held more reliable than 11 witnesses?
What are your thoughts?
We have to protect our epistemic vulnerability from mistaken people and charming and convincing fraudsters, who gives us false positives, therefor we can't accept witness testimony as enough justification to give out the price. Accepting witness testimony is risky.
We have a epistemic defeater in that we can't rule out that we are being tricked to believe false positives by X and the other people and we can't know if they are truly independent from him. X can also easily say things that are not true under oath to win 5 million USD and knowing believing that a prosecutor wont be able to prosecute him easily for perjury.
Compare this to the justification we use to refuse a car salesman who claims that we should buy his car which he claims is brand new and perfect but we discover signs that it has severe rust problems in the motor which we casually infer to mean that the car is not going to work properly. The casual inference that there will be a imminent risk of buying a defective car is justification we use to avoid accepting the salesmans testimony?