(Bolding added by me for emphasis)
Whew! Okay, then we agree and there's no more reason to argue! My definition works just fine and there's no reason to use yours except:
Which doesn't apply to every atheist as you admit here:
You misrepresent me yet again. What I said is that the appeal of faith does not apply to all New Atheists, not that definition (2) does not apply to all New Atheists.
This one is still off-topic, but what the heck, why not?
That's true, and I said that I have little desire to discuss it in this thread, but you continued to press the issue by asserting against the probabilistic nature of arguments from consensus.
No.Let's take the idea that the Earth is flat. At one point in time, most people thought it was, agreed? At that point in time, most people prior to the people that currently existed also though the Earth was flat. So at that point in time, we take a cross section of history and find that most people throughout history believed the Earth was flat. Therefore, at that time it was more likely that the Earth is flat than it is round. Anyone at that time employing your appeal to popularity would be wrong.
It's true that 3000 years ago this particular argument from consensus would have been false, but a single counterexample does not undermine a probabilistic argument. What would undermine a probabilistic argument is the claim that the argument fails more than 50% of the time, and is therefore generally unreliable.
But this reduces to an absurdity. Suppose that a consensus is incorrect more than 50% of the time. Since a consensus is simply a sum of individual opinions, this means that each individual is incorrect more than 50% of the time. But if each individual is incorrect more than 50% of the time, then all of our perceptions, beliefs, and opinions are generally unreliable and cannot be trusted. It is only if humans are generally reliable (i.e. correct greater than 50% of the time) that we can trust human knowledge. Refined scientific methodologies can increase our reliability, but they cannot create reliability where none previously existed.
A long conversation on this topic--albeit with respect to the burden of proof or rational inclination--took place between myself and Archaeopteryx beginning with
this post. The meat of the discussion is certainly present by
this post.
This also means that the probability for something being true or false changes over time, somehow. It used to be more likely that the Earth is flat, and now it is more likely that the Earth is round. See how silly that sounds?
But consensus does not determine absolute probability, it is only one probabilistic argument. Therefore the probability for something being true or false
under the aspect of an argument from consensus changes over time. This is no different than saying that opinions about what is true change over time, and is no more mysterious. The probabilities for what we believe to be true change as new data is gathered and new arguments considered. There is nothing strange about this.
Now let's expand my example to everything that we've learned over time. For everything we've learned, there's a different belief previously held by, what was at the time, the majority of humanity that was wrong. Since we don't know everything, there are beliefs right now that the majority of humans hold that aren't true. Here's one example you can get on board with, I think: the majority of humans believe that the Bible is not the inerrant word of God. So, according to you, they are probably right and the Bible probably is not the inerrant word of God. But you disagree, I presume, yes?
Again, you're committing the error of assuming that a single counterexample or set of counterexamples undermines a probabilistic argument. That's not true. If the majority of humans believe that the Bible is not inerrant then this constitutes a probabilistic argument in favor of the thesis that the Bible is not inerrant.
That isn't what scientific consensus means. When they say there is a consensus, it isn't because they polled a bunch of scientists and asked them their opinion.
That's exactly what a scientific consensus is.
It means that of all the tested hypotheses about a subject, the majority of those hypotheses turned out in favor of the subject. So you aren't comparing people's opinions, you're comparing actual data.
No, this is manifestly incorrect. It is not a poll of hypotheses or data, it is a poll of
persons, scientists.
If you did poll a bunch of scientists and 90% said global warming is real, it would be an appeal to authority, which is not necessarily a fallacy. If those scientists all studied climatology, then it is appropriate.
It would be a combination of an appeal to authority and an appeal to consensus.
In order for us to accept scientific consensus and deny general consensus, it would have to be the case that scientists are right > 50% of the time (when opining on their field of study) and general people are right < 50% of the time. But this is not possible, for if it were then it would be impossible to transition from being a general person to being a scientist. Someone who is generally unreliable (right < 50% of the time) cannot "pull themselves up by their bootstraps" into general reliability. For example, every attempt they make to arrive at general reliability will be plagued with
de facto general unreliability and untrustworthiness. Every scientific method they develop will more likely be false than true, and every attempt to discern whether something is in fact true will also be plagued by the same unrealiability, thus barring the way to improved reliability.
It only makes sense if the general person is right > 50% of the time and the scientist is right more than the general person. Perhaps the general person is right 58% of the time and scientists are right 82% of the time (in their own field). It is possible to move from 58% to 82%, but it is not possible to move from <50% to >50%.
If those scientists all studied biology, then it would be inappropriate.
Which is why I specified
climate scientists.