But only if they actually read the claim. This is the same as for scholars that cite the source of a claim they found in a paper that is, itself, cited from some other paper. Often you see a game of telephone in these citations. Because it's not looked at well to give a factual claim you found in a review paper, researchers often cite the claim as it was cited in the review paper, but they don't always investigate the claim themselves. This leads to a game of telephone.

A lot of high-profile factoids are like this. The claim that 95% of diets fail, for example, is a specious one that developed after a citation chain like this. The original analysis said that 95% of the sample finished the study above the lowest weight they reached. Through motivated rephrasing and citation laundering, this became 95% of diets fail, often paired with the suggestion that dieters always return to a weight higher than where they started.

Yet, you can find this claim being re-issued again and again in the introductions to papers about all sorts of topics related to dieting.

Another thing I have seen is where the source of the claim gets lost. It starts out as something like "Grainger 2003" and then eventually turns into "Grander 2013", a nonexistent paper with a ton of citations.

So, if you read an article and don't read the cited article, please reference only the article you read.

Except encyclopedias don't have _actual sources_. They are based on source but don't include references as far as I remember.

They very much do have citations as I recall. Maybe not the junior encyclopedias they had your grade school, but any proper encyclopedia had citations.

I was going to tell you that you're wrong, but in looking it up I found far more examples of encyclopedias without sources than with, so maybe that was more common.

I got away with this once. I used wikipedia as the source on a table that everyone in the field knows by heart anyway. When questioned about it (I think briefly?) I said I'd personally edited the article and checked that the table was correct (which I had!) . --~~~~

That is actually common for all encyclopaedias, they do generally cite secondary sources. Especially if the primary sources are not easily verifiable. Similarly they should (I haven't actually checked if they do) cite a translation of an ancient Greek text, not the original Greek text.