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-misc...
[2] http://www.nature.com/news/2009/091026/full/4611181a.html
That is what the Constitution says. The policy of the government is very, very different. [0]
Remember, don't talk to government agents without a lawyer.
However, certain statements are also acts. While the content of the statement is protected, the act itself is punished. For example, perjury, public endangerment, etc., are acts which can take the form of speech.
This isn't true. False statements of fact are a major exception to the First Amendment [0].
Besides perjury, you can prosecute people for defamation or libel. Usually it's a civil matter, but there's also criminal defamation. Finally, as WildUtah mentions below, making false statements is a crime.
In my opinion, it's a bad idea to predicate crimes on people lying. 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. A good example of this is the securities laws. Fundamentally, they are predicated on the simple idea that it is illegal to lie in connection with the sale of securities. In practice, it invites endless litigation over whether, say, a CEO "lied" when he said in an analyst call that a company would meet certain benchmarks when in fact the company failed to do so.
These sorts of crimes are not just conceptually complicated, but suffer from difficult proof problems. The difference between a "lie" and a "mistaken statement" is simply a person's state of mind at the time a statement is made. This sort of thing is devilishly difficult to prove.