What distinguishes things people say from the truth is the fact that true things are actually true - that is, there are aspects of objective reality that show the claims to be correct.

When a person systematically believes things that are true, it involves a causal chain from that truth to that person's cognitive state. While it is true that the truth is a cause of the true belief, one should not dismiss the intermediate links in the chain. In this case, an important link is citation of the sources that you, presumably, have already gathered and read in order to be able to write your article. This allows readers to distinguish in their own minds, in a manner correlated to how reality distinguishes truth from falsity, well supported arguments ultimately based on empirical evidence, from nonsense theories with no grounding in reality.

No one can free you from the responsibility of checking claims

That misses the point, which is to tell people where they can look to verify the claim. It is not like the mere existence of a citation makes your claims correct, the cited source still needs to be evaluated.

How would you know I wasn't just fabricating the sources, or citing sources that were themselves based on fabrications?

I would expect to get to a first hand account of the actual observations within a few steps of following citations. (If I don't, I would worry that sources themselves are not grounded in reality.) I would judge an undisputed account of first hand observations in a refereed journal of the appropriate topic to be unlikely to be a fabrication. Though, I would expect the original article I am evaluating to at least show me the first step of this process, since it is so easy for anyone who has actually based their discussion on relevant sources and is therefore likely to be correct.