Is what's considered evidence for something ultimately subjective?

Question

Imagine we are trying to conclude whether or not one someone is a prophet. How much evidence will we need to conclude this? Is the answer to this ultimately subjective?

For example, if John makes no prediction, then we are certain that he's not a prophet. If John says there will be a major world catastrophe in 2024, we assign a very low likelihood to him being prophet, and many of us will remain certain. If John says World War 3 will happen in July 3, 2023 and it happens, we are suddenly curious and wonder how he came up with that prediction. If John makes a couple more predictions like that, we might start assigning a high likelihood to become a prophet.

But what is the cutoff? All of those predictions could "technically" have arisen by chance, and it would be difficult to definitely prove that John is a prophet.

The most curious thing about this is that one can assign a higher probability to a prediction being right and hence John being a prophet the more specific a prediction is. But something about that seems off. John is either a prophet or not. There is no "20% chance that he's a prophet" in the real world. It just doesn't exist. So how can we know?

Answer 1

There are two views on what probability means. Under the frequentist view the sentence "There is a 20% chance John is a prophet," would be interpreted as "If we had infinitely many clones of John, then the proportion that were prophets would be 20%." As you pointed out, this interpretation does not really make sense.

However, the Bayesian view offers a significantly more straight-forward interpretation: "The amount of confidence I have that John is a prophet is 0.2, where 0 represents no confidence in the statement and 1 represents absolute confidence." Another advantage to the Bayesian approach is that Bayes's Theorem tells you how a person's confidence should change when new evidence is introduced. This means that every time John makes a prediction, we know how much to increase/decrease our confidence that he is a prophet based on the outcome.

If you look into the math of Bayes's Theorem, you may notice something funny. No matter how much evidence John offers, no matter how accurate and numerous his predictions, the probability that he is a prophet will never become 1. In other words, as long as are belief in John is based on evidence, we can never accept him as a prophet with absolute confidence. David Hume made an argument similar to this, but regarding the existence of miracle rather than prophets. A couple of articles have been published looking at Hume's argument form a Bayesian perspective, but they are all behind paywalls so I can't read them.

It is also worth pointing out that there is a modern fad of applying Bayesian probabilities to places it really shouldn't, leading to a lot of experts getting really annoyed. I could be doing that here, but I don't think so. The main issue these experts have is that there is no known way to perfectly translate a human beings confidence into numbers. Questions like 'How confident were you that John is a prophet before he offered any evidence?' or 'How confident where you that [Event X] would happen before John claimed that it would?' don't have quantitative answers. My point is that if a method exists to extract quantitative answers to these questions, even if I don't know what that method is, I can solve the broader philosophical questions you asked. In other words, although I cannot do the math to calculate how likely it is that John is a prophet, I believe a smarter person could.

EDIT:
I just checked and the SEP has 3 articles that go over Bayesian statistics quite well. From least to most technical I think the ordering would be this, this, and then this.

Answer 2

Having something accepted by a representative or authoritative community implies making science: the truths of experience (empirical truths) are precisely the goal of science.

So, when would it be accepted that a prophet is such thing, or that cats can fly? For many, just one behavioral sample would be enough (e.g. I saw my cat flying! The world should know that cats fly!). But for the authoritative community on empirical truth, that is, the scientific community, a fact becomes scientific (or, a fact is a scientific truth) only if the fact has been tested using the scientific method, giving acceptable results.

Scientific facts are not accepted the first day they are stated. Acceptance of the community follows more or less a typical process: performing tests in strict accordance with the scientific method (basically, demonstrate that successive tests will yield the same results), formally publishing the results (e.g. papers), submitting them to the judgement of the scientific community (which would imply more testing, debating, rebating) and waiting. The scientific community will naturally react in consequence.

There's not a number or percentage of members to convince, there's not a precise number of predictions to hit, or distance a cat must jump to be considered flight. Scientific acceptance is ruled by recognition, respect, meritocracy, logic, philosophy. For example, if Einstein would simply state that cats fly, the scientific community would take it quite seriously. But if we find that Napoleon had stated it, scientists would not give it too much importance.

Answer 3

Is what's considered evidence for something ultimately subjective?

Defines "ultimately"...

For the subject, by definition, all is ultimately subjective, even what is objective.

However, if several individuals are to act somehow collectively, then they have to find a way to agree on what is the case. This is achieved through the notion of objective truth. Individuals have to come to agree between themselves what is objectively true. What the group will decide is objectively true will presumably count as ultimately true for the group. Another group may want to differ.

Thus, if something counts as evidence for each individual in a group, then it can count as evidence for the group. In the process, what is subjectively evidence for each individual is turned into what is deemed objective evidence by the group.

This is arguably somewhat complicated but social life is nothing if not complicated. The saving grace is that every human is a proficient processor, and, overall, it seems to work, even if it is arguably rather laborious.

This explains why any proper judicial process is more protracted than shotgun justice, why democracy is more protracted than autocracy, why science is more protracted than divination.

This also explains why humans seem initially tempted by shotgun justice, autocracy and divination.

Answer 4

There are lots of problems with your thinking here, I will try to spell them out one at a time.

You are confusing Realism, the belief that there is a fundamental reality to our universe, with our knowledge, which can be perpetually uncertain, no matter whether the universe is real or not.

First, YES all our knowledge is subjective. Every single observation we make, even of the most straightforward of things, is a subjective observation. Samke with our conclusions based on those observations. We know that our personal thinking and even observations can be flawed by our preconceptions, and that others can sometimes see where we missed noticing things, or thinking thru consequences, so getting a competent group of others together to help with observations and conclusions, and dialoging to arrive at a consensus, is a very good way to improve on the reliability of both observations and understanding of their meaning. Note that this is denying that we can have objective facts -- the best we are able to achieve for our understanding of the world is intersubjective consensus.

Second, things in the world are also not either/or as you assume. An in dividual may in principle receive valid prophecy, but not realize it is valid. Or receive valid prophecy, AND have ideas of their own that falsely seem to them to be prophecy, and not be able to distinguish them from each other. Or they may only once, ever, get a valid prophecy. Even with assuming Realism to the world, our world can be a complex and messy place in "reality", and it often does not fit our mental categories of either/or. Hence a 20% prophet is entirely possible in our world.

Third, "prophet" is a VERY strong claim. And it relies upon a very disputed set of assumptions, which many people will be very reluctant to accept. The evidence needed to convince skeptical third parties will have to be overwhelming, and will have to include not just evidence for valid predictions, but also evidences for the worldview that makes these predictions plausible as valid prior knowledge.

The two concepts I have encountered that usefully address the "background assumptions" need are "memeplex" and "Research Programme". Memeplex is a concept developed by Richard Dawkins to note how ideas, memes, are really not evaluable in isolation. They tend to come in collectives, memeplexes, which are a mutually reinforcing collection of ideas and assumptions. When one is evaluating two possible interpretations of an observation, the memplex one is working with will radically change one's understanding of what that observation likely means.

Imre Lakatos extended a similar concept to science, noting that single theories or claims are rarely definitive tests of a view, because views are not single theories, but instead an approach to thinking, a Research Programme. All field of science have falsifications of portions their views, but those do not destroy a science claim. Instead, these are accepted as challenges, and those who embrace a Research Programme try to find ways to modify their views to account for the apparent refutation. Lakatos distinguished between "progressive" programs, which have a pattern of resolving apparent problems, and "regressive" ones, which apply unconvincing patches to apparent problems.

An example of a progressive research programme is the Big Bang theory, which was faced with several astronomic timescale problems -- spiral galaxies should have lost their spirals long ago, and the oldest stars in our galaxy appeared to be older than the universe. The old star problem was resolved when we discovered the expansion rate of the universe is growing, and the spiral issue was resolved when it was shown that a galaxy absorbing a nearby micro-galaxy is what produces the spirals, and yes, they DO disappear, over billions of years. A regressive programme would be Ptolemy's cosmology, of an earth centric universe, with transparent spheres circling the universe in complex layers. Add enough custom designed transparent layers, and Ptolemy's model can potentially fit every astronomic observation, but all these patches make it predictively useless, and manifestly absurd now that we can measure distances and velocities of remote galaxies.

Lakatos tried to develop an algorithm to measure progressivity or regressivity, but critics showed how his math did not work. We are left with a subjective judgement call on plausibility/validity of a Research Programme memplex reference frame for evaluating specific issues, such as -- "is this individual making valid prophecies?"

So, your challenge is to outline different reference frames for interpreting a claimed prophecy, and figure out how progressive or regressive they are, and whether an apparent prophecy should just be left on the shelf as a future project, for anti-prophecy programmes, or if it is a strong enough evidence to make supporting programmes much more attractive.

Answer 5

The use of the word evidence in philosophy is strongly grounded in subjectivity in terms of frameworks like:

Russell ... tended to think of evidence as sense data, mental items of one's present consciousness with which one is immediately acquainted. In this, he stood squarely within the tradition of classical empiricism. Quine [said] that evidence consisted of the stimulation of one's sensory receptors. The logical positivists held that whatever evidence there is for a given scientific theory is afforded by observation statements or ‘protocol sentence’, linguistic entities with suitably-restricted contents... According to one recent and influential study, one's evidence consists of the totality of propositions that one knows (Williamson 2000). According to another, one's evidence consists exclusively of one's current mental states (Conee and Feldman 2004). Within contemporary confirmation theory, a prominent version of Bayesianism is naturally understood as identifying one's evidence with those beliefs of which one is psychologically certain. Of course, the suggestion that one might place sense data, sensory receptor stimulations, known propositions, or one's current mental states in a plastic bag (or dig them up from the ground, or send them to a laboratory, or …) is of dubious intelligibility. From the perspective of much ordinary thought and talk about evidence, much philosophical theorizing about evidence would seem to embody a particularly grotesque category mistake.

But that's with respect to claims of what evidence is, not how it can be used. Paradoxically or not, applying the evidence that one has is thought to be more objective in the sense that, "Item A is evidence for claim B," is true only if, objectively, the relevant relation obtains between A and B. As the above-quoted article continues (quoting one Mr. Blanshard):

What is creditable … is not the mere belief in this or that, but the having arrived at it by a process which, had the evidence been different, would have carried one with equal readiness to a contrary belief.

C.f. the description "truth-tracking" in epistemology:

According to “truth-tracking theories”, a subject knows p provided their beliefs “track the truth” of p across modal space. Truth-tracking theorists typically cash this out in terms of subjunctive conditionals, such as “Sensitivity”: *If p had been false, S wouldn’t have believed p (Nozick 1981), or “Safety”: If S were to believe p, p would be true (Sosa 1999; Williamson 2000; Pritchard 2005).

Reliabilism specifically has been confronted with a clairvoyance problem:

One early challenge to reliabilist theories of justification was advanced by Laurence BonJour (1980), concerning a hypothetical clairvoyant named “Norman”. Norman has a perfectly reliable clairvoyance faculty. But he has no evidence or reasons for or against the general possibility of a clairvoyant power or for or against his possessing such a power. One day Norman’s clairvoyance faculty generates in him a belief that the President is currently in New York City, but with no accompanying perception-like experience, just the bare belief itself. Intuitively, says BonJour, Norman isn’t justified in holding this belief. Yet process reliabilism seems to imply otherwise. Since Norman’s clairvoyant power has a high truth ratio, Norman’s belief about the President must be justified. So reliabilism seems to get this wrong. (Similar examples were offered by Keith Lehrer (1990) and Alvin Plantinga (1993).)

So suppose I utter a future-oriented sentence like, "The moon will turn into blue cheese tomorrow." Suppose this happens (the moon visibly becomes bluish tomorrow, and carbon-dating of cheese samples from the moon later reveals that the switch occurred on the uttered date). If I have the ability to foresee when which moons become which cheeses, i.e. if that is my specific prophetic capacity (no other class of claims can be foreseen by my means), it seems as if my capacity well-tracks the truth of the future-oriented sentence. Be that as it may, testing this quite specific form of clairvoyance is rather troublesome (it would be much harder to reach other moons which I may or may not claim to have become cheese of late).

Or suppose instead that I make a list of ten surprising propositions, all of which I claim will come true at some specified time, and on top of that I say that an angel-oracle gave me the list. By the time the 9th item on the list has come true, there's some prima facie reason (or at least motive) for believing that the next item on the list will come true, too, and so if someone obstinately disbelieves me, I am well-motivated (enough) if I later chide them when they prove to be mistaken. (For those who simply suspend such beliefs, well...) However, this is only prima facie, for again, does my acceptance of the list "track the truth" in the overall situation? I would have believed the angel-oracle even had they provided me with some other list; or, since I am not in causal contact now with the resolution of any item on the list in the future, how could I have perceived the future? The future does not exist to be perceived until it becomes present, does it? But now all of this is caught up in the question of how if-then statements are true or false at all:

For instance, I say (*) “If you touch that wire, you will get an electric shock”. You don’t touch it. Was my remark true or false?

Addendum. We also would like to distinguish between:

1. More often than not, when I make a nontrivial future-oriented claim, that claim turns out to be true. I make no claims about where I am getting my information from.
2. The same except I say that my claims are based on analysis (background physics, social trends, etc.).
3. The same except I say that a supernatural being is the original source of my claims; or my claims are informed by the use of a supernatural power that I claim to possess.

(3) covers the case of alleged prophets, then. But there are independent reasons to think that (3) is impossible (it is impossible for a nonspatial, nontemporal, conceptually simple, etc. being to identifiably communicate with spatiotemporal complexes; if we're hearing voices, that's all we know we're hearing, regardless of whether the voices themselves claim to be God), and in turn whatever evidence there is, given a good track record with future-oriented claims on my part, will not be evidence that I am a prophet (although it might be evidence, or at least/rather reason, for me to believe other future-oriented claims that I make).