> It can be devilishly difficult to distinguish between someone who is lying and someone who is misinformed, and that task becomes exponentially more complicate

> It can be devilishly difficult to distinguish between someone who is lying and someone who is misinformed, and that task becomes exponentially more complicated when the underlying "fact" is itself uncertain. Another good example (IMHO) is the case of academic fraud. Often, the researchers' assumptions bias the entire study, leading to a result that an unbiased researcher wouldn't get. For example, the biased researcher pools groups that shouldn't be pooled, or the biased researcher doesn't log-normalize data that should be log-normalized, leading to "wrong" results. Edit: But most of the time, the researcher doesn't even realize that (s)he is biased. On the other hand, we have researchers who clearly have monetary interests behind their publications: Séralini comes to mind who publishes biased books on GMOs but also publishes extremely biased research, or all the "bad pharma" as described in Ben Goldacre's book of the same name. There have been cases where "obvious" scientific fraud has been investigated by the police, for example in Italy [1] or South Korea [2], but the slope is truly slippery. Where do you draw the line between "honest mistake" and "lie"? Furthermore, most countries already have a problem with more and more scientists dropping out of science, which will only be exacerbated if you can go to prison for a huge mistake. [1] http://retractionwatch.com/2013/12/06/should-scientific-misconduct-be-handled-by-the-police-its-fraud-week-at-nature-and-nature-medicine/ http://retractionwatch.com/2013/12/06/should-scientific-misc... [2] http://www.nature.com/news/2009/091026/full/4611181a.html http://www.nature.com/news/2009/091026/full/4611181a.html