what are our (science-types) ideals and expectations on how we want people to participate in this debate and, more generally, in democracy and life as a community?
I expect people to be honest and listen to other people. You don't have to agree at all, but if you don't, then I also expect you to explain why you don't agree. Otherwise, there's not all that much to debate.
I expect people to understand the difference between facts (things you can only disagree with if you can show them to be incorrect) and opinions (things people can disagree with for hardly any reason at all). Almost the same is the difference between objective and subjective, but not always...
I expect people to act with respect. When something is factually wrong there are different ways to go about indicating the correction. Take a moment to consider going about such a thing in a tactful manner.
When someone has a different opinion, it is possible to make fun of it or explain your own opinion. Different circumstances call for different strategies.
First step: understanding better what exactly science-types think of as "knowledge"--what is it and where does it come from?
Knowledge: a human understanding of the external reality we find ourselves existing within. -goal is not to be exact in some over-arching, absolute sense (this is impossible... we do not have a "reality" measuring stick in order to test if we're "absolutely correct" about anything) -always based upon available information (data... evidence... observations...) from that external reality, not opinions or desires from our imaginations -goal is, therefore, to expand our knowledge as much as possible in order to be accurate as much as possible -as the available information changes or is added to, our knowledge similarly 'is changed or is added to' as a reflection of the new pool of information
When we request each other to provide evidence in an argument, should we be referring to source papers that contain original data, or is referring to authorities good enough?
Depends on the subject and the audience. In general, though: referring to authorities can be just fine but source papers that contain original data will always trump referring to authorities with no source papers and no data. And again... referring to actual data that can be repeated (regardless of any source paper or authority) will always trump everything.
The important factor is getting the information from the external reality. If you can show that to be valid... then you're the trump card. Always. Source papers with original data is a very good way to do this, but certainly not the only way... and not necessarily the best way.
If it's good enough, when and why is that the case?
When all audience members are satisfied that the data is an accurate description of the external reality as much as possible.
Wanting-more-information-but-currently-being-unable-to-obtain-any is not a valid excuse to prevent the current best possible explanation from the information that actually is currently available from being the most rational option.
Not being satisfied because you don't like the explanation is not a valid excuse either.
The only valid objection is if you can show the best-possible-explanation-from-the-information-that-actually-is-currently-available to be false using some of the information that actually is currently available... otherwise, you're only talking about hopeful possibilities... and there's billions of those as long as you have a good enough imagination.
Is scientific knowledge the set of all source data in the literature, or is it the set of inferences and conclusions that have been generally agreed upon by the scientific community, based on those data?
A bit of both, as it should be. Just not all of the picture:
Scientific knowledge is the set of all source data in the scientific literature, yes. We then use this data to derive additional sets of inferences and conclusions, yes.
But, you cannot forget the next, very important, step:
These additional sets of inferences and conclusion are then tested against the external reality in order to gain more information.
This "more information," regardless of whether it confirms or denies the inferences and conclusions, is then assimilated into the literature as part of "scientific knowledge" and the process of deriving new sets of inferences and conclusions from the current level of knowledge is begun again. And it's then tested again. And new guesses are then derived again. And they're tested again. And new guesses are then derived again. And they're tested again. And new guesses are then derived again. And they're tested again. ... ...until we have all the knowledge (that's going to take a while).
if I'm not a nerdy book-worm whose primary interest is learning about the natural world, how do I participate in your democracy?
Take a look at the external reality around you. Does it make sense with what the nerdy book-worm's say? If so... then explain why you think the current explanation is satisfactory. If not... then show what you think is incorrect about the current information.
If you can't understand/learn/educate-yourself-about what the nerdy book-worm's say... then you have a few options: -ask some questions in order to help with your learning -don't participate
It's the same thing as any other high-end endeavor. If you can't understand/learn/educate-yourself-about what the professional basketball players are doing on the court... then you have a few options: -ask some questions in order to help with your learning -don't watch basketball on TV
Or more generally, how do you expect a general public who is not necessarily compelled by knowledge to interact with you?
If you aren't interested in "the truth about how things in this universe work" then I don't expect you to interact with me on the truth about how things in this universe work.
The same way that if you're not interested in basketball, I don't expect you to interact with me about basketball.
Does it have to be on your terms? If so, why?
No, it certainly doesn't have to be on my terms. You can want to gain an accurate description of our external reality on whatever terms you'd like. The important concept is that you want to gain an accurate description of the external reality we find ourselves existing within. If you do want to... then that's all that's necessary. The reason why you want to is up to you and is something you need to figure out for yourself. If you don't want to do that, however... then there's no reason for you to discuss such things, why would you?