I can’t know that I know about many things
To me knowledge is justified true belief. But what is truth? I posit the correspondence theory of truth where something is true if it corresponds to actual states of affairs in reality. Now, we can have evidence-based reasons for our beliefs, and we may say they are then justified. But does lack of evidence, the negation of justification say anything about truth? It does not. I can say that I believe there is no God. I can say there’s lack of evidence. Does the negation of justification justify that belief? Or does justification require positive reasons or justification?
As for truth I can have justified belief and for some things, they are probably true. I may say that I know that the radio is playing. It may be justified by my sense data and by conferring with other people. If others agree that the radio is on, then that is justification. If it corresponds to reality, then it is true. Pretty uncontroversial.
What if I say I know there is no God? Does the lack of evidence provide justification? Maybe. I’m not sure. Negative justification is trickier than positive justification. Furthermore, positive justification deals with the physical world where we can look at the world and draw conclusions. Negative justification is lack of evidence and is much weaker. I can say there is lack of evidence that the multiverse exists, or maybe scanty evidence that is highly controversial. Is it then a justified belief that there is indeed NO multiverse? What if the actual state of affairs in reality is that there is a multiverse, and it turns out there is no way of having a positive justification for it? This example shows that there are many things that could be ultimately unknowable. I think negative justification as the determinant for true belief is weak at best.
As for truth I can have justified belief and for some things, they are probably true. I may say that I know that the radio is playing. It may be justified by my sense data and by conferring with other people. If others agree that the radio is on, then that is justification. If it corresponds to reality, then it is true. Pretty uncontroversial.
What if I say I know there is no God? Does the lack of evidence provide justification? Maybe. I’m not sure. Negative justification is trickier than positive justification. Furthermore, positive justification deals with the physical world where we can look at the world and draw conclusions. Negative justification is lack of evidence and is much weaker. I can say there is lack of evidence that the multiverse exists, or maybe scanty evidence that is highly controversial. Is it then a justified belief that there is indeed NO multiverse? What if the actual state of affairs in reality is that there is a multiverse, and it turns out there is no way of having a positive justification for it? This example shows that there are many things that could be ultimately unknowable. I think negative justification as the determinant for true belief is weak at best.
Comments (44)
I remember when I tripped over a paper on justified true belief (JTB) a few years. I couldn't believe that someone, a professional philosopher, could write so confusingly about something so simple. Then it got worse with all sorts of odd cases - If I consider you a credible witness and you tell me John's car is brown because you saw a brown car parked in his driveway but it turns out that was his sister's car but John's car really is brown - Do I know John's car is brown? And then you get those silly Gettier problems.
So, did Newton know anything about motion and gravity? No, because classical Newtonian physics has been replaced by relativity. So we have to keep going back and retroactively deciding that people really didn't know something at sometime in the past even though they, and we, thought they did. So, nobody can ever really know anything.
Solution. Simple!! Get rid of the "T." It doesn't have to be true, I only have to be adequately justified in believing it is. What does "adequately justified" mean in this context? It means that, given the information I have, the uncertainties in that information, the consequences if I am wrong, and the benefits if I am right, it makes sense for me to make a particular decision. Knowledge doesn't mean anything except in the context of a necessary decision.
It depends on the context. When it comes to the physical world, you need justification just to positively assert an existence or a relation. This is usually expressed as "you can't prove a negative", and in this case we can say that you don't have to justify negatives. The reason for that is simply that our knowledge of the physical world is supposed to explain and thereby predict that world. Everything not part of any specific prediction - everything without justification - is irrelevant for that purpose.
If we go beyond the physical world, the question of justification is less clear. We probably need to know the purpose of metaphysical "knowledge" in order to figure out what a proper justification would be.
That was what I was getting at I think. Science is useful (and destructive). Scientific knowledge can only explain the physical world. A lot of people think everything is physical. And in the case of a multiverse, something physical, it may be that we cannot have positive justification for it. Science will ALWAYS leave MANY things unknown. I hold that the nature of consciousness or what the nature of physical stuff is are things that science can never explain.
I thought I did. In summary again - the words "knowledge" and "know" have no useful meaning if knowledge always has to be true. The only thing that really matters is whether or not I am justified in making a particular decision. If I'm not making a decision, truth is useless, knowledge is useless.
Has anyone ever discovered anything that didn't go on to have any practical consequences? Can something really be knowledge if it has no interaction or impact on the world we live in?
Epistemology as the foundation of good decisions then? Science has made many stupid decisions. One example, in my opinion, the industrial revolution that led to unsustainability. However, I enjoy the many fruits of industrialization. That said, wouldn’t the conscious life of the planet including the many now extinct species been better off without it?
Yes. Exactly. I'm an environmental engineer. My job is to know stuff, know how I know it, know how well I know it, and figure out what to do with that knowledge. I used to joke about putting "Environmental Epistemologist" on my business cards. I have written several times on this forum about the epistemological methods I use in my work.
Quoting Noah Te Stroete
Well, science doesn't make decisions, people do. The jury is still out on whether the industrial revolution was a good thing or a bad thing. The next 30 or 40 years will probably tell. I'm not sad that I won't be around to see. I admit I am worried for my children.
The essence of what we have learned about environmental issues over the last century is that humanity has failed badly to understand the true consequences of its actions. As I indicated previously, a good understanding of the consequences is necessary in order to make any claim to knowledge.
Environmental epistemologist. I like that. People do make decisions. It was ultimately the business class and the desire of the people to make their lives easier that made things so bad because they didn’t think about the consequences. But they weren’t aware of the consequences. No one can KNOW all of the consequences of their decisions. Certainly people continue to fail to see the consequences of new technologies, technology being an application of science. Environmental engineering, though, I must say I don’t know anything about. Do you utilize technology in this field?
That just underlines what I wrote previously - Knowledge only really has meaning in the context of a decision and no one can say knowledge is adequately justified without understanding the consequences of a bad decision.
One only knows a decision is bad after the fact. And bad for whom or what? Some consequences are good for some and bad for others. That’s where values come in. Making decisions has more to do with values than epistemology in my opinion.
All true for a lot of decisions.
One can know what the consequences of of a decision will be before the fact better than we currently do, it just takes more work and attention. It costs money, so there are strong political forces resisting it.
It’s easier to know what the direct intended consequences will be, but there are unintended consequences to every decision. Some good, some bad.
“The road to hell is paved with good intentions.”
That said, I am certain you know as an environmental engineer what is best policy for the environment more so than politicians or corporations.
Environmental engineers apply technology to implement the laws and regulations that reflect environmental policy. I have opinions on what the appropriate policies are, but it's not my job to establish them.
That's not entirely true. Unintended consequences are not unforeseeable consequences, at least not necessarily. We can certainly do better than we do.
I see. It takes a team of experts and a consensus then (always difficult) in order to determine best policy. If only lobbyists didn’t get in the way...
They are political decisions, with all the ups and downs, ins and outs, backs and forths associated with those.
What does “unintended” mean then? If your point is that in the case of governments, institutions, and even individuals, important decisions should be made only after extensive research and deliberation, then I agree.
As an individual, I have a handicap in this regard. I’m impulsive and self-destructive in a lot of ways. I know what things are better for my physical health, but I have vices. I tend to see these vices as necessary for my mental health which is more important to me.
[In retrospect this whole line of thinking is flawed]
Let me ask you if this isn't an alternative formulation of your replacement of JTB with JB.
Quoting T Clark
Assign a factor to the epistemological reliability of your credible witness? God's validated word is 1.0, your credible witness is given a prior of .8. The contention then is that your state of knowledge of this fact is, itself, approx. .8.
Quoting T Clark
In as much as the meaning of knowledge derives from our ability to model and predict the world based on the accuracy of propositions derived from knowledge, knowledge is a measure of the correlation between claims and reality.
John's car is indeed brown and in some other world, his sister's is green. Your knowledge that "John's car is green" remains at .8 until additional information modifies your repository of knowledge or the credibility of the proposition. Truth is correlative of the combined reliability factors involved in a proposition.
Explanatory issues such as those in your first example, while confounding to a discrete model of knowledge, are expressed as true positives in a continuous model while integrating uncertainty in justification. We can actually calculate how many of those true positives are "fortunate" (true, but only by chance). If we know cars only come in green and brown (equally likely), there is a 10% chance (1-.8)*.5 that the match to truth is "fortunate".
Isn't a continuous model of knowledge more intuitive to our understanding of how we know?
When I think I know something (but am not sure) I will hedge my bet or say "Yeah...maybe".
Why is it “(1-.8)*.5” and not “.8*.5”?
If we limit ourselves to a sample of 10 instances, in 8 cases the credible witness's testimony would align with ground truth (in a justifiable way, say, by witnessing the car) - "John's car is brown".
Of the remaining 2, half would mistakenly associate a green car and the other a brown car. In 9 instances the witness would claim "John's car is brown", one of which would be "fortunate".
Totally agree. And I have gotten so much shit over the years for saying the T should be taken out. First, because it implies that there are two criteria for knowledge 'justification and truth. When in fact, there is no second process after evaluating the justification where we then look to see it is true. If we have evidence of a black swan, then the justification that there are lots of white ones is poor justification. When we evaluate the justification we will look for counerexamples and lack of logic, but we cannot determine now if in the long run what we consider true today will be true tomorrow. Adding the true is confused.
Positive and negative justification are computationally different.
The quantifier, "there exists" (?), does not need to traverse the entire domain. It can stop at the first occurrence. The quantifier, "there does not exist" (?), has no other option than to traverse the entire domain.
In abstract, Platonic worlds, computational issues are ignored. So, positive and negative justification are equally strong beliefs:
In classical mathematics, one can prove the existence of a mathematical object without "finding" that object explicitly, by assuming its non-existence and then deriving a contradiction from that assumption.
(constructivism)
In virtual worlds, computational issues cannot be ignored. So, arguing or querying non-existence is allowed but known to be computationally intensive. For example, the SQL "not exists" operator is known to cause serious performance issues:
The problem is when I have many data in my table (like million of rows), the execution of the WHERE NOT EXISTS clause is very slow. I have to do this verification because I can't insert duplicated data.
You can find lots of examples (if you Google for them) of people complaining about the slow performance of the NOT EXISTS operator.
In the real, physical world, support for complete traversal of a domain (unrestricted comprehension) is the exception and not the rule. E.g. the propositions, "All humans own a pair of trousers", but also, "There does not exist a human who does not own a pair of trousers", are not supported. Since the "not exists" quantifier always requiring complete traversal of the domain, you can often assume that you will not be able to expend the energy and other resources required to traverse a relatively large real-world collection.
The quantifier, "there exists", is obviously much better supported in the real, physical world, even for very large domains, because you can just stop at the first occurrence that arises.
So, I agree that in the real, physical world, negative justification is often not even possible.
I can’t know that I know.....wait.....what?????
I know I can’t know many things is a tautology; the negation of a tautology is a contradiction; I can’t know I know many things is a contradiction.
Or....there is some subtlety hidden in the negation that escapes me, and the whole OP actually has some epistemic value.
Belief = something one is inclined to think is the case, with varying degrees of conviction
Truth = a judgment that a proposition has a particular relation (the exact relation depends on the truth theory (or theories) one uses--I use correspondence personally) to other things.
Justification = whatever one considers a good reason to believe that P, and to believe that P is true; most people have a variety of approaches fueling what they consider good reasons to believe various propositions.
A lack of evidence that P is often considered good grounds for a belief that not-P. For example, considering the notion that a normal-sized, non-invisible (etc.) elephant is in my apartment at the moment, the lack of evidence of an elephant in my apartment is a good justification for assigning "true" to "There is no (normal-sized, etc.) elephant in my apartment at the moment."
The same sort of justification works against the notion of a God, though in that case we also have issues like logical coherence re the notion of nonphysical existents.
And if you thought of knowing that way, then, for example, you'd have to say that in terms of the sciences, we categorically can never know anything, because it's a basic tenet of science methodology that any claim is open to revision in the face of new evidence. In other words, it's a basic tenet that any scientific claim could be wrong. It's never impossible that a scientific claim is wrong. Taking a claim so that it's impossible that it can be wrong means that we're no longer doing science.
But obviously, the vast majority of people do not use "know" that way--we know many things in the sciences.
The Gettier problems primarily hinge on the notion that intuitively, people do not want to say that we can know something accidentally.
But the Gettier cases, in my and some other folks' views, typically hinge on misconceived ideas of how the actual contents of belief work with respect to formal logical conventions.
Well.... Two thoughts - First, I don't really like a numerical calculation like this for decision making, although it is not uncommon in my line of work. Example - we put a design out to bid. We establish a set of criteria for the bids, say price, experience of bidder, quality of technical approach to the work, completeness of the proposal, health and safety performance in past projects, company size and resources. Then we weight the criteria by importance. As you might imagine, price gets the highest weight. Then we rank each criteria, multiply by the weighting, and then add up the values. Bidder with the highest ranking gets the project. Problem - it almost never works. We almost never pick the company we think will do the best job.
Second - As I said, you can't really know your "state of knowledge" unless you take in to account the consequences of failure. The question isn't how much uncertainty is there in my decision, it's whether or not the justification is adequate given the risk. The question to ask - what could possibly go wrong.
As I've said, I think you also have to take the consequences of failure into account. If the decision is who has to wash the dishes, who really cares. If it's should we start a war in Iraq, we have to be a lot more careful. We found out what could possibly go wrong with that. Now the people in the White House are itching to do the same thing in Iran.
In the context of my approach to knowledge - justification takes in to account certainty, benefits, negative consequences) - then negative evidence is important. If I have searched carefully for negative evidence and haven't found any, that contributes in a major way to justification.
This brings to mind a relevant quote from Stephen Jay Gould that I've always loved:
“In science, ‘fact’ can only mean ‘confirmed to such a degree that it would be perverse to withhold provisional assent.’” Replace "fact" with "knowledge."
Yeah, we both must be wrong.
Maybe my title was confused. I’m not sure now. I don’t have a lot of confidence in that.
I was thinking that if truth is a requirement of knowledge, then happening to correspond with actual states of affairs in reality cannot be fully known (or certain or proved).
Truth is conditioned by thought, knowledge is conditioned by possibility; both are conditioned by time. I don’t see as one will ever be a requirement for the other. Not all truths are known and not all knowledge is true.
On the other hand, I know I detest Lima beans, so it is absolutely required that it be true Lima beans be something in order for me to know I detest them. When I was 6 it was 18 steps from my bed to the bathroom. When I was 16 it was 14 steps from my bed to the bathroom. The truths and the knowledges of each set of circumstances are exactly the same, but not so are all the states-of-affairs.
One can talk about truth, or one can talk about knowledge, for days. But trying to put them together is a whole ‘nuther can of metaphysical worms.
You’ve given me a lot to think about.
Could you flesh this out for me so I can understand it better?
I think this helps explain your first paragraph, but I would need more of a demonstration or explanation with more examples to fully understand it. That is up to you whether you want to, of course.
Opinions might not be the best way to properly understand much of anything.