It's very common to hear people say that we should only believe claims that are supported by sufficient evidence. Intuitively it seems to make sense, until we realize that this recommendation begs several questions:

  • What is evidence?
  • Can evidence be quantified, and if so, how?
  • What counts as sufficient evidence?
  • Is sufficient evidence a function of the claim X (i.e. SE(X))?

Can these questions be answered objectively, i.e., is there an objective standard of what counts as sufficient evidence?

On the contrary, can these questions only be answered subjectively, meaning that each person has to decide for themselves what is sufficient evidence (for them) and what isn't?

Related: Ought we only form beliefs based on sufficient empirical evidence?

  • 3
    There are many standards of sufficient evidence, and many of them are objective. That does not mean absolute, universal, precise, or anything like that, only that they can be shared and applied publicly. Some are even codified into law: probable cause, preponderance of evidence, beyond reasonable doubt, etc. All the practical ones are more or less vague, and which one is applied depends on purpose at hand. One can quantify it by probabilities (Bayesianism) or degrees of truth (fuzzy logic), but assignment of those numbers, outside of narrow idealized contexts, is itself loose.
    – Conifold
    Sep 7, 2022 at 2:13
  • 3
    @Conifold per your own description, none of those standards are objectively defined and applied -- they are instead intersubjectively applied by a consensus of reasonably minded experts.
    – Dcleve
    Sep 7, 2022 at 3:33
  • Does this answer your question? What is evidence?
    – user14511
    Sep 7, 2022 at 3:41
  • @Mr.White - No. This one includes the nuances of objectivity, subjectivity and epistemic sufficiency, which the other question doesn't.
    – user48437
    Sep 7, 2022 at 4:54
  • 1
    In philosophy of evidence there're widely different views applied from field to field, such as those of externalist Williamson according to whom all and only one's factive knowledge or even some basic belief as emphasized by Plantinga's reformed epistemology constitute evidence, and those of internalists according to whom only one's current intentional mental states such as related beliefs or private experiences/convictions should be considered evidence... Sep 7, 2022 at 5:09

3 Answers 3


Yes, all claims to be "objective" fall to ultimate subjectivity.

  • Every observation, is a subjective observation
  • every threshold of relevance, is a judgment choice
  • every conclusion, is a judgment call about how to evaluate a claim

But "subjective" is mis-treated as a dirty word

All of science, all empiricism, is based on intersubjective consensus of experts.

We know how to create instruments to detect a phenomenon, because the experts in the instruments, and in the phenomena, have reviewed phenomena and the instruments response, and consider the instruments the best way to characterize the phenomena.

We know valid and invalid says to do statistical analysis, because statistics experts have evaluated different methodologies and concluded certain practices give useful methods of discriminating.

We know how to set up experimental protocols (things like double blind with controls, experimental setups like corner box, etc) based on experts having tried a variety of techniques, evaluated their effectiveness, and FOR THAT TYPE OF PROBLEM, recommending a protocol.

For different problems, the experts that one is trying to reach consensus among, vary. For anthropogenic global warming, it is a small group of cross-discipline experts who have been studying this particular problem. For how to fletch an arrow, it would be a small group of craftspeople who are maintaining that skill. For whether a defendant is guilty, it is a small pool of jurors who have considered all the evidence. For who should be our next president, it is all adults of voting age.

So "sufficient evidence", while intersubjective rather than objective, is far from worthless as a criterion for belief.

Note, I too have encountered many who claim to only operate on verified scientific conclusions. In addition to questioning the objectivity of any of this, I also dispute whether any such a claimant actually meets their own standard. I offer the following methodology pursue this question, look at the decisions they made in the first 10 minutes of their day.

If one starts the day upon waking -- what was the sufficient evidence of where they were, of what the feeling in their guts actually meant (full bladder, most likely), of the safety of transferring their weight from bed to floor (floor not a visual illusion, say), etc.? Go thru 10 minutes of their morning, and 99% of what they do will be based upon first person observations and judgments, unverified memories, and prior successful routine -- virtually none of it will be even intersubjectively verified, much less scientifically verified.


Is there an objective standard of sufficient evidence?

No, there is not, in the sense of there being one standard for any decision, regardless of the topic at hand, the gravity of the decision to be made, or the motivation of the decision maker.

A good example is the varying standards in legal proceedings. To prove a civil case, a plaintiff must show only a “preponderance of the evidence”, typically described as 50% plus one.

However, in a criminal matter the burden of proof is “beyond a reasonable doubt”, typically described as the same weight that a person would give to the more serious matters of one’s life. The difference lies in the consequences: a payment of money in a civil case, versus a loss of liberty in a criminal case.

Depending on the type of proceeding involved, there are many other standards of sufficient evidence. A party who wants a preliminary injunction at the outset of a case, for example, must make a showing of “irreparable harm” if the injunction were not granted.

But one overarching standard of sufficiency? I doubt that you will find that anywhere.


If any given standard of "sufficient evidence" is applied objectively to different claims, each with it's own evidence, this standard leads to the conclusions that each of those claims are either true or false.

When considering a standard applied to a set of claims, and whether those claims are objectively true or false, you'll come up with a probability for how likely that standard is to lead to believing true claims.

Now, assuming your goal is to maximise the probability of holding true beliefs, it follows that there is an objective standard of "sufficient evidence".

Of course all human observation and thought is subjective, so we're incapable of knowing what objective reality is, but we can do our best to tend towards that by developing these standards amongst multiple people (to reduce the effect of the subjectivity of any one person), taking into account evidence from many different sources and using this to try to recognise and limit our own subjectivity.

It might also be that your goal isn't to maximise the probability of holding true beliefs - some might favour holding a comforting set of beliefs (whether consciously or not). And you can also potentially define this probability in different ways, given that a belief can either be true or false, and you can hold it or not, which leads to 4 variables - this is much like classification prediction problems in machine learning, for anyone familiar with that. You can also either consider all claims, or some subset. Or you might have a different goal entirely. A court of law, for example, may have a standard of sufficient evidence which is "beyond reasonable doubt", which is biased towards innocence rather than guilt: the goal is not just to have the verdict be correct as often as possible (independent of innocence or guilt), but also to avoid punishing innocent people.

The point is that, given some concrete and (hypothetically objectively) measurable goal, there is an objective standard.

Note: Evidence is roughly just all knowledge related to a particular claim, and you could potentially quantify evidence (on a per-claim basis) based on how likely it is to correspond the truth of some claim. Or you could quantify evidence by saying, in general, this type of evidence is this likely to correspond the truth of applicable claims (e.g. one might say dreams are not likely to correspond to the as-experienced truth of what you're dreaming about). What I said above also applies here: which metric you pick to quantify evidence is subjective, but an objective measurement of that metric exists (if the metric is measurable).

  • A 3rd possible goal for determining beliefs could be pursuing that which is popular - many wish to hold beliefs that maximize social benefits and/or minimize social consequences (e.g. the way many people determine which sports team is "the best"). Sep 8, 2022 at 16:49
  • NotThatGuy and @HoldToTheRod, this question may be interesting as well: philosophy.stackexchange.com/q/93470/48437
    – user48437
    Sep 8, 2022 at 17:02
  • @HoldToTheRod I would probably group social benefits (and comfort, most likely) under irrationality. This could potentially be contrasted with people who admit that something isn't true, but choose to act as if it is anyway (which isn't quite the same as "X is wrong, but I'll do it anyway") ... but maybe this is also just a different form of irrationality. Most people who are motivated by this probably wouldn't admit as much. You may hear hints to this effect, e.g. "if God doesn't exist, then life has no meaning" (I disagree, obviously), from people who otherwise present logical arguments.
    – NotThatGuy
    Sep 9, 2022 at 16:02

You must log in to answer this question.