Is there a "correct" set of criteria for classifying "opinions" (e.g. based on accuracy)?


The concept of opinion is more diverse than subjective:

  • "opinions" can evolve to as far as scientifically backed facts.
  • "opinions" can exist as strong group beliefs (such as anti-slavery)
  • an expert opinion should be distinguished from a non-expert opinion

Thus merely "an opinion" is a very limited view on information. But there's a reason to ask for more criteria to be added for opinions, because otherwise rising beyond subjective or "the worst opinion" would make little sense.

Some practical concern:

For example, if media is still "an opinion", then not much of "general sorts" (more objective) could be said. Thus people who write articles, may be wasting their time. Because e.g. the choice about "what to write about" is still subjective. The way it's written contains subjectivity (word usage for example).

But does/can there exist a "correct" set of criteria for opinions? That is, a categorization/classification system for opinions, which can distinguish between their accuracies?


Initially one'd think that maybe not, since the criteria may deviate between individuals.

But there are examples, when they coincide. Thus there exists some criteria, but none of them is necessarily the ultimate one.

But then when one starts to compare criteria, they may form hierarchical relationships. Scientific criteria ought to be more valid than religious criteria. And then one could say that "scientific criteria is (more) correct".

But this may again fall to at most group subjective. So then would a reasonable view on opinions be:

subjective or group subjective beliefs that have varied measure for truth/accuracy

But what can one reference that truth/accuracy against? To compare to something better, doesn't it mean that one may have to "lock" the "correct criteria" as reference? How is the "legitimacy" of this locking motivated? On subjective basis? But what about views that concern more than the subject?

What could be the "best" epistemologies of our current times? That can serve as references?

Practical example:

If one "valuates" people differently based on e.g. education, physical features etc. Then how are these to be "accurized"? Some would claim them as "mere opinion", whereas basing on cultural research, things such as attractiveness and fitness as well as "achievement" do have general cultural patterns. Thus, for accuracy, it feels an underestimation to treat them as "mere opinions". Rather, they could be called "informed opinions", with some caution with not being too subjective. While being a questionable topic (since people possibly hate being judged), I think it can serve as an every day example of the difference between "mere opinion" and "some more".

  • 2
    What's a "sphere", in the context of your question? – MichaelK Dec 4 '18 at 14:43
  • Have you ever had a look into the private/public distinction in e.g. Dewey, Arendt, Habermas? They explicitly speak about the public sphere as constituent of factual and normative truth with intake from particular opinion. – Philip Klöcking Dec 4 '18 at 18:24
  • AFAIK, it is more common to speak of scientific, political, and religious discourse or framework instead of sphere. – Philip Klöcking Dec 4 '18 at 18:26
  • This question is extremely vague. "Sphere" is still not very clearly defined, it is only mentioned as a concept but there are not practical examples of what a sphere could be in practice. A more troubling problem though is that you have not set up any criteria for what constitutes "correct";you are even putting that in quotes as if you do not know that yourself. You need to clarify more what a sphere is and what you mean by an opinion being correct. – MichaelK Dec 4 '18 at 19:51
  • both my and @elliot svensson's answers have been down voted, presumably because they fail to address the salience of your question. Yet together the answers appear to address the question's most likely referents. So, with MichaelK, I must conclude that the question, as posed, which suggests that you seek a standard with which to determine the accuracy or validity of an opinion ("What could be the "best" epistemologies of our current times? That can serve as references?"), is likely too vague to be understood, answerable. – gonzo Dec 4 '18 at 20:26

The places where your opinions matter the most surround the individual person. What is the best way to spend Friday night? What is the best city to live in? What would be the best profession? Who would be the best spouse? Which religious organization is the best?

Your opinion is completely non-transferable for these kinds of decisions, and your opinion is also completely essential. You can't improve anything by superimposing your opinion on somebody else's decisions, and it's not an improvement for an opinion other than yours to be placed on your decisions.

| improve this answer | |
  • I don't think "you can't improve anything by superimposing" can be always true. It's true in some cases due to "separation of concerns". But there are situations, where a more correct view ought to out rule the more worse for "common good" for example. If one couldn't superimpose, then laws would not make sense. – mavavilj Dec 4 '18 at 23:31
  • Also, I believe that "improvement" or "advancement" could be essentially about superimposing better views over lesser ones. – mavavilj Dec 11 '18 at 2:36

I seem to be playing in a narrow range of keys recently, but...

I find Wittgenstein's notion of the language game, and the psychoanalytical concept of the unconscious group mind together make to be a good model for how opinions, beliefs and facts interact. There is not a distinction to be made, other than how central those things are to ongoing games.

All facts rely upon opinions to establish their meaningfulness, which makes them opinions to a small degree. Do you trust the person doing the measuring? Is the peer network that verifies those facts large enough, and does it include a diverse enough perspective? Those remain opinions and whether something really is a fact, even a scientific fact, depends very directly upon the answers. This collapses what seems to be a set of positions into a continuum of belief without actual endpoints. No belief is entirely opinion, and none is entirely a fact.

Individuals and groups take part in a continuous renegotiation of all the aspects of life that are socially constructed. But those social constructions occur at multiple levels. Often those levels have their own 'bylaws' or local rules for negotiation. But ultimately these meet and overlap with more global negotiations. Ultimately, truly global negotiations do, in fact occur -- we have a global consensus that we should not kill one another without a good enough reason. We tend to agree that manipulating the truth too aggressively is bad, even when it is productive. The roots of ethics get established and live in this domain. Things get put into it and taken out: we are now pretty certain that self-determination is a positive thing, internationally, and have stopped trying to convert everyone to something. There is still dissent, but it seems clear what there is to dissent from, and that the dissenters have really lost for the moment.

The local rules are most relevant when they reflect a position that chooses the membership of a group according to whether they are influenced by a given group mind. A religious sect or ideological resonance has decided to think together, their beliefs don't face revision, because they live in games with a given local rule-set that protects them. And the ability of that local mind to mature in given ways has been an overall productive force in history, even though the groups of people it establishes get co-opted to play different roles in different games. (e.g. until Bismarck ready for universal male suffrage, Lutheranism is used to decide who is really German enough.)

Abuses of this system occur when information that is only consensus in one game intrudes into other games. Wittgenstein's Philosophical Investigations is full of places where physics intrudes into metaphysics, or language intrudes into religion, etc. in ways that can be untangled and seen as attempts to cheat according to local rules by imposing rules from games centered on a different local set of meanings. (The use of Catholicism to decide how to divide up the German-speaking part of the world was such an abuse. It succeeded, sort of: Austria exists, but is now independent of any association with Catholicism. But in retrospect, we are not sure that isn't something that encouraged Nazism.)

So the answer is 'yes', but that decision needs something closer to a map than a set of rules. Opinions are valid where they remain relevant to the domain they concern. And that relevance depends broadly upon the actual opinion.

| improve this answer | |
  • The answer to what question is "Yes." Is it: "Is there a correct sphere for opinions?" If so, did u take the term sphere to = something like domain, and opinion to be on the other side of a continuum whose binary would be fact? If so, what did u make of "What could be the "best" epistemologies of our current times? That can serve as references [stadardsds?]?" I liked you answer, though, regardless of what the question was. Particularly the fact that you explicitly started with W's language games, and implicitly answered in the context of rule following within forms of life. – gonzo Dec 4 '18 at 23:20

If the question is are there better and worse ways of thinking about and formulating "better" opinions that (i) arise in one as opposed to another domain of inquiry, and/or (ii) to address one or another type of problem/issue, explanatory strategies, so to speak, then consider that AC Crombie identifies six different styles of reasoning in his Styles of Scientific Thinking in the European Tradition:

  1. The simple method of postulation exemplified in the mathematical sciences;
  2. The experimental exploration and measurement of complex observable relations;.
  3. Hypothetical construction of analogical models;
  4. The ordering of variety by comparison and taxonomy;
  5. The statistical analysis of regularities of populations, and the calculus of probabilities;
  6. The historical derivation of genetic development.

So the question becomes, is there be a meta-reason, a meta explanatory strategy of justification/warrant for choice of a particular style of thinking/reasoning [formulating opinions] in this but not that domain, for this but not that problem/issue? For determining whether Might that be success? Consensus within the relevant community? (You may want to peruse Ian Hacking’s musings about styles of reasoning>)

Because the focus of your query appears to be identifying the proper foundations for strictly speaking “opinions,” rather than truth or knowledge (sometimes characterized as justified true belief), I’m not sure that any of this is relevant or helpful. But it may provide a useful framework for considering the question.

| improve this answer | |
  • The problem with the numbered things that you give is that they're still a bit tied to subjective epistemologies. A scientist would agree on those, a non-scientist could find them irrelevant and theoretical. Thus the spheres don't coincide, or do they? The scientist and the non-scientist are somehow "different"? – mavavilj Dec 4 '18 at 23:32
  • Which is why an answer requires you to be much more specific in defining the term "sphere." For instance, consider that you tell an atheist medical doctor and an empathetic layman Christian scientist that you have cancer. The MD might recommend radiation or chemo, whereas the CS might render a very different opinion (eg, tell your children you love them and make peace with your enemies). No? Is your question how (what rule to follow) to decide between them? – gonzo Dec 4 '18 at 23:49
  • No, I'm asking about the spheres. I don't have the answer. I'm asking e.g. how can the scientist and non-scientist find common ground, or the CS and the MD. Or should it be possible to also "win over" another sphere and when? E.g. science is neglected by religious people, thus "validity" does not hold, unless it's in subjective epistemologies. What kind of stance can one take, if there's dissonance between epistemologies? Ignorance? This would form a kind of sphere, where some things are meant to be ignored. – mavavilj Dec 4 '18 at 23:59
  • But what about e.g. subject-to-subject agreements? E.g. it makes sense for informed people to listen to the expert opinion of the MD and not the CS. But how do they "know" that this is the way? Science is tested, yes. But the MD is still a person, not a 100% accurate machine. Also there are more questions than what science can measure. Political, every day, human rights etc. They also need opinions, but how can one choose over them? Just setting a "context" already contains subjectivity. I believe it's impossible to set non-subjective contexts, even if naively one may claim so. – mavavilj Dec 5 '18 at 0:02
  • The political context is particularly challenging. The very nature of politics is "overarching". Yet it may rely on very subjective, non-hard beliefs with naive measures (such as popularity). E.g. the "real power" is on the physically strongest. Yet there's a claim about limiting it, even though it's reliant on "considering something wrong with the rule of the strongest", which is a subjective belief. The measure by popularity vote is again subjectively chosen measure, there are other measures. – mavavilj Dec 5 '18 at 0:13

Not the answer you're looking for? Browse other questions tagged or ask your own question.