Well, rather than tell me that I have not put a modicrum of reflection on this as a way to declare yourself right, how about you explain why it's nonsensical?
That's a claim, not an explanation.
I believe I've sketched one reason that the view negative statements get the benefit of the doubt is naive previously, but I will repeat it
below.
First, let's talk about what we mean by "negative" claim. Typically, the proponent of the view that such claims get the benefit of the
doubt is really discussing universally quantified claims, not negative claims. There's no reason to think "Arsenal lost their last game"
gets benefit of the doubt over "Arsenal won their last game." Rather, the proponent means that existential claims have a burden that their negations (universal claims) do not.
Why should we presume (A x)Px is true, and require proof that (E x)~Px? A common argument is that universal claims cannot be proven, and hence should be assumed true. But this is bizarre for two reasons. The first is that universal claims certainly can be proven
in many cases, and even absent proof, evidence can be adduced to show the claim is at least probable. Moreover, why should we think that unprovability entails "likely true"?
There is one sense in which it makes sense to assert a universal claim, barring evidence to the contrary, and that is best described in learning theory. Suppose Px is quantifier-free and suppose furthermore that the following is true:
If (E x)Px then there is some time in the future when we will observe Pk for some k.
That assumption is very strong. It says, in essence, that we are guaranteed to find evidence verifying (E x)Px if (E x)Px is true. It
is rare to know that this is the case. However, if it is the case, then the best strategy is to assume (E x)Px is false and await evidence to the contrary, at which point we will change our beliefs. In this way, we are guaranteed to converge to the truth in the first possible moment.
But the proponents of privileging universal claims pay no mind to either the assumption nor spell out the essential point that Px must
be quantifier-free for this to work[1]. If we allow for nested quantifiers in applying this rule, then we come to ridiculous conclusions such as:
(1) For each a, we assume that (E x)Pax is false.
(2) We also assume (A y)(E x)Pyx is true.
Similarly, if we strictly follow this rule, then we run into issues when we have (E x)Px <-> (A y)Qy, because the left hand side has the
burden of proof, while the right hand side has no such burden.
Example: Suppose we know that the lights were off in the dining room before Joe entered the room. Suppose that if he toggled either of two light switches, the lights are on. Then
(E x)(x is a light switch and Joe toggled x) <-> (A y)(y is a light -> y is on).
Thus, I claim that the view that "negative claims get the benefit of the doubt" (or that positive claims have the burden of proof) (1) has
no good argument in its favor, unless we are quite careful in spelling out the technical details of the learning theoretic claim and (2)
leads to inconsistent conclusions regarding the burden of proof either in the case of nested quantifiers or in the case in which an
existential statement is equivalent to a universal statement.
This is why I regard such principles as naive and nonsensical.
To bring this digression back on-topic, there is no good reason to presume there is no association between Trump and HAS a
priori. Rather, we examine our background knowledge and any evidence at hand to decide whether an association is probable or improbable. If we are unable to come to a conclusion based on what we know or discover, the prudent conclusion is that we simply do not know enough to form an opinion.
(Note that much of what I say here is repeated in the analysis section of the WP article on
Russell's teapot, most notably the bit attributed to Chamberlain. Note as well that Russell certainly did *not* conclude from his example that universal statements get the benefit of the doubt, at least not in the excerpted bits in that article. Rather, it seems to me that he is arguing against a particular kind of appeal to ignorance, whereas the view you espouse is more or less another appeal to ignorance.)
[1] Technically, there can be quantifiers in Px, so long as each of them is existential in the quantifier-normal form of Px.
