It looks to me like (at least here in USA) the court of law does not always require empirical evidence to sentence someone to death. Rather it seems that the person can be sentenced as long as there is a a reasonable assumption that he is the murderer.

Sometimes the evidence seems to be at most of the indirect kind. Yet it seems that this kind of evidence would not be accepted for conclusions in general science.

Is this difference between accepted forms of evidence present on purpose or as a consequence of something else (such as due to laxness of judges, to cut down on backlogs, or some legal philosophical reasons)?

  • (1) Yes, the rules of evidence for criminal trials in the USA are rather distinct from the meaning of evidence in science. (2) BUT In your question, your wording is odd since the standard for capital offenses, which is "beyond a reasonable doubt" a phrase defined in extensive caselaw. (3) Can you make clearer what the question is related to philosophy?
    – virmaior
    Commented Nov 2, 2015 at 1:15
  • Require empirical evidence of what? (Presumably there is empirical evidence that the convicted party committed murder, so if this is not what you want evidence for, what is it you want evidence for?)
    – Ben
    Commented Jan 2, 2023 at 3:35

3 Answers 3


Consider the difference in the situation. In science, where you have a scientific theory and perform experiments to prove or disprove the theory, you have as many attempts as you want to perform these experiments. In a criminal court, there was usually just one crime and just one chance to find evidence, against someone who tried to avoid the existence of such evidence. So we can raise the level a lot higher in science.

"Rather it seems that the person can be sentenced [for murder] as long as there is a a reasonable assumption that he is the murderer." Not at all. There has to be proof beyond reasonable doubt. That's a much higher mark.


In science, one not only has the opportunity to repeat an experiment many times, but is expected to do so, and is expected to do so under controlled circumstances.

In criminal law, an event happened once under un-controlled circumstances, or did not happen at all. Much of the opportunity for "evidence" that scientists have simply doesn't appear in legal situations.

That being said, anecdotally, I recount being told of a paper on particle physics where the final display of data looked like this:

|              2
|        1     *         1
|        *     |         *
|  0     |     |    0    |
   |     |     |    |    |

And with that sample set of 4 points, they concluded the distribution was "Gaussian" and thus the experiment demonstrated their discovery. Obviously someone with the resources to collect more data would be expected to gather more data, but in this case 4 "hits" was enough for evidence in scientific law.

No one even swore in the detector equipment when it recounted the events of the experiment. Sloppy sloppy! =)

  • I suppose the experiment was throwing a dice five times in a row, and forgetting the count for the number six?
    – gnasher729
    Commented Nov 2, 2015 at 23:04
  • My friend (who was the one that actually saw the paper and related it to me) indicated that it was actually a particle physics paper. It was in response to me digging into questions as to how much proving power your experiment needs in order to be published. Obviously major discoveries receive far more rigor than this.
    – Cort Ammon
    Commented Nov 2, 2015 at 23:28

The basis of Common Law extends back well before the advent of reasonable science, and it is entirely based upon belief of the assembled. In U.S. laws, that can be a judge, a panel of experts or judges, or a jury of the peers of the accused. But there is never a case in which material evidence itself is more important than who believes it, and how convinced they claim to be.

We know, from studies of memory and perception, that this is not a reliable standard of evidence from a scientific point of view. We know how memory is rendered harder to retrieve by stress, how our degree of certainty is unrelated to the actual evidence, how memory can be manipulated by the order in which its aspects are later recalled, etc.

But we try to adjust for this with redundancy. We try to make the jury large enough that someone on it will not we swayed by these manipulations, and we generally required unanimity. When there is not a panel or jury, we use the thread of appeal to ensure that a manipulated judge is questioned by more rational authorities above him.

Since statistics matured, science has an objective reference for when one ought to believe, instead of relying directly upon the ability to convince. But that standard arises later than most of the science we were taught in school. So, to some degree, the science you are used to was equally subject to the personalities involved and their political ambitions. But it has survived centuries of repeated attempts to disprove it.

You must log in to answer this question.