Two Different Types of Certainty?

[DavidOdden]

But I have no direct evidence that this life actually exists, other than the fact that I cant' discount it as a possibility.

I would propose to have weak conceptual evidence that life (intelligent or otherwise) off-planet is possible. But that is such weak evidence that the claim for the existence of life elsewhere is on the cusp of being arbitrary. So I would not make the claim that there is life elsewhere, but I also would not deny such a claim. Beyond the weak conceptual evidence that I already know about regarding elementary biochemistry, I simply would not even entertain the claim (but I would entertain new evidence).

The point of most interest, to me, in your post, was that it tries to apply the notion of certainty to something that is not perceptually self-evident. There is a standard Kantian line to the effect that one cannot be certain of anything except that which is formally stipulated and deduced (e.g. "Given P=>Q and P, you can be certain that Q"). We know that certainty can be applied to other things, but the recent discussion here has not, for the most part, squarely tackled the issue of being certain of something that is "physical", and not immediately present, for example whether we can be certain that the freezing point for water at sea level pressures is 32 F. So I encourage trying to move the discussion back in the direction of whether one can be certain about a person's moral character, or something other than axiomatic statements.

[putofftoolong]

>Before I deal with the substantive issues raised by noumenalself, I want to address his opening paragraph.

QUOTE(noumenalself @ Aug 2 2007, 11:51 AM)

Betsy's view that there are two kinds of certainty is completely wrong, and also philosophically dangerous. Her views are almost indistinguishable from mainstream philosophical ideas about certainty, and are inconsistent with the Objectivist view. The reasons she cites for her views have also been refuted countless times in the Objectivist literature, particularly in Dr. Peikoff's lecture courses.

The above charge, even if true (which it isn't), constitutes the fallacy of Poisoning the Well.>

I am trying to understand what is wrong with the above quote, is it the claim of 2 different types of certainty's or the fact that having two could be dangerous. Betsy, is the concept of absolute certainty different from contextual certainty. I though the point of contextual certainty was to allow man to claim knowledge with certainty within a given context, but at the same time allow for the fact that humans can be in error. Thus the genius of objectivist thought is once again shown to be in how its concepts refer to reality and therefore are useful in reality. How would the idea of an absolute certainty be useful in anyway that contextual certainty does not cover. How can man actually acheive absolute certainty in the way you mean it. If absolute certainty were accepted, it would take over the contextual meaning used by objectivists( there is no way they would be allowed to exist together), people would then believe(rightly) that certainty is imposible, and then start writing books about the"pursuit of pure certainty" to be acheived by some allegedly non-imposible means (ie) faith. Absolute certainty leaves an opening(which I think we should shut by refusing to use the word) for someone(religious or other) to take over the word. Is absolute certainty an anti-concept that if used will leave the real certainty without a name?

PS the purpose of me asking is because to me it seems like an important question, I'm not trying to poison any well.

Edited August 4, 2007 by softwareNerd
Fixed Quote-blocks

[noumenalself]

This is a serious negative charge being made, against me personally, in a public forum.

You bet.

It is not my view that "there are two kinds of certainty" so this is an attack on a straw man. What I have said is that other people in the current debate are using the word "certainty" to stand for two different concepts and sometimes both at the same time. For clarity of argument, they should define their terms and use them consistently.

This is nitpicking. The problem is that you think that C1 rather C2 is the proper definition of certainty. That's incorrect, and it's not just a "matter of semantics" which one we pick. It is vitally important that concepts be properly defined, and also that we not multiply our concepts beyond necessity. As I've argued, we need the concept "certainty" to contrast conclusions from those which are merely possible or probable. We already have a concept to cover the senses and the axioms ("self-evident"), so we don't need another concept (your C1 "certainty") to do the work twice over, particularly when "certainty" treated as C2 is already at work picking out a different distinction.

Added: There is also no justification for forming a concept that groups together the sensory given, the axioms, and causal explanations, but not conclusions that are fully supported by the evidence. There is no fundamental difference there. There is also no fundamental similarity between the sensory given, the axioms, and causal explanation that is not already covered by the concept "knowledge." But there is a fundamental difference between the given and axioms on one hand, and causal explanations and other inferences on the other. It is precisely the difference between the self-evident and the inferred.

This is poisoning the well. No argument is presented as to why what is allegedly my view is "dangerous," but it does inject a personal and negative emotional tone into the debate.

Poisoning the well is the fallacy of saying that a person is a bad or unreliable source, therefore her arguments shouldn't be paid attention to. I haven't said anything about your reliability as a source. I have, however, attached a value judgment to my assessment of your position. Judge and prepare to be judged.

Your view is dangerous because, as I suggested, it seriously resembles the analytic-synthetic dichotomy. But to add more: It is also dangerous because it severely delimits the realm of what we can be certain about. As you know, Dr. Peikoff's C2 definition of certainty is essential to refuting skepticism. Skepticism is pretty dangerous, wouldn't you agree? Your view is also dangerous because it suggests that there is a difference between certainty in theory an certainty in practice. C1, theoretical certainty, is all we can really have, but you allow that for practical purposes we can say that things are C2 certain. The theory-practice dichotomy is dangerous.

This is unsupported guilt by association. noumenalself does not say which of my alleged views is indistinguishable from which mainstream philosophical ideas. There is also the implication that my alleged views should be discounted or dismissed merely because they are similar to the views of certain others and not because they are false.

Well, if you think the analytic-synthetic dichotomy is false, and if the theory-practice dichotomy is false, and skepticism is false, then a view that is hard to distinguish from them is going to be false and therefore dangerous.

What evidence is offered that what is allegedly my view is "inconsistent with the Objectivist view?" In this post noumenalself only makes reference to unspecied arguments contained in "Dr. Peikoff's lecture courses" and later to uncited and unquoted material "Dr. Peikoff and Dr. Binswanger have discussed in multiple places, in particular in "The Art of Thinking" (LP) and "The Metaphysics of Consciousness" (HB)." Still later he asserts that my alleged views are wrong because "all of this is covered in various Peikoff lectures" without specifying which lectures and where.

This is more nitpicking. I also mentioned the Objectivist literature, which has also been discussed numerous times in this thread, specifically Dr. Peikoff's section on certainty in OPAR. That is the most relevant source, since it is the best evidence we have about Ayn Rand's view of certainty. I would also think that citing the title of lecture courses is not "uncited." But, incidentally, I believe the material from "The Art of Thinking" is lecture 6. Let's focus on OPAR, though, because that's what most people have access to. And it's all I need to make my point.

Edited August 3, 2007 by noumenalself

[y_feldblum]

What I have said is that other people in the current debate are using the word "certainty" to stand for two different concepts and sometimes both at the same time. For clarity of argument, they should define their terms and use them consistently.

How do you use the term?

[ttn]

The problem is that [betsy] think that C1 rather C2 is the proper definition of certainty.

I agree with the points noumenalself has been making, but I would have put this differently. I don't think the problem involves definitions at all, and it doesn't involve advocating one rather than the other of two genuinely different concepts.

The problem, rather, is thinking that C1 and C2 are "two different concepts." If you read Betsy's descriptions of C1 and C2 which Dan Edge quoted to begin this thread -- read them, that is, from a proper Objectivist perspective -- they say exactly the same thing [*] in different words (as noumenalself already pointed out by noting that, once the standard for proof has been satisfied, denying the conclusion would mean endorsing a contradiction). Only if one holds the premise that there is some kind of dichotomy between the empirical and the logical (or the analytic and synthetic, or some such) could it possibly look like there are two different concepts here which just happen to be named by the same word.

[*] well, leaving aside the (inconsistent) bit in C2 about "high degree of probability". If, as Betsy says later in her formulation of C2, the evidence meets the relevant epistemological standard of proof, then the conclusion is certain (which is to say: absolutely certain, 100% certain) -- not just "highly probable".

[Betsy]

Now to get down to substantive issues.

Betsy's view that there are two kinds of certainty is completely wrong [...]

Betsy says:

The sentence in red was omitted by noumenalself when he quoted me. With the sentence restored, you can see that my view is not "that there are two kinds of certainty." My actual view is that the same word -- "certainty" -- is being used to denote two different concepts.

However, it is simply incorrect to make a distinction between a certainty whose denial would be self-contradictory, and certainty which has merely "full evidentiary support." If a conclusion is supported by all of the evidence demanded by a standard of proof, then it would require self-contradiction—of some of that evidence—to deny it.

This is not necessarily true. There have been cases where all known evidence supported a conclusion as true, but later additional evidence proved it false. The original evidence is not self-contradicted. It is contradicted by later evidence.

It is not the case that a conclusion that has "full evidentiary support" merely has high probability. To say a conclusion has high probability implies that there is still a possibility of error.

In some contexts, there is and the standard of full evidentiary support allows for the possibility of error. That is why a murderer is allowed appeals and judicial review instead of being executed as soon as the jury convicts him.

But as Betsy knows, claims of possibility also require evidence. If none of the relevant evidence supports the possibility that a conclusion is wrong, the conclusion is certain.

Can a person make hasty generalizations based on insufficient evidence and claim certainty? Can he say "I have only seen two lions in my life and they were both females. My observations of lions are relevant evidence and I have no evidence that contradicts that. Therefore, all lions are females."

Of course, sometimes it happens that a claim is supported by all of the relevant evidence, but it still turns out to be false. But as Dr. Peikoff and Dr. Binswanger have discussed in multiple places, in particular in "The Art of Thinking" (LP) and "The Metaphysics of Consciousness" (HB), claims can be certain but still be wrong. "Possibility," "probability" and "certainty" are all epistemological concepts which we need to describe the continuum of evidence.

Quite true. Sometimes the word "certainty" refers to the endpoint of the continuum with "degrees of certainty" used to describe possibility and probability. This is the "C2" usage.

In today's state of political confusion and contradictions, it is difficult to endorse any candidate with any degree of certainty.

With that in mind, it is useful to look at the examples Betsy gives of the kinds of things she would call absolutely certain:

Sorry, Betsy, but you're wrong here on several counts. Once again, all of this is covered in various Peikoff lectures. It is a misapplication of the concept "certain" to call the data of the senses "certain." Sensory perception is the "given," the infallible foundation of our knowlege. But it is a category error to call perception "certain." "Certainty" applies only to the kinds of conclusions that require validation by stages of evidence—it does not apply to the evidence itself. We already have a term for the data of the senses and the axioms: "self-evident." "Certainty" is not the same concept.

That's news to me -- and to Ayn Rand --

The status of automatized knowledge in his mind is experienced by man as if it had the direct, effortless, self-evident quality (and certainty) of perceptual awareness. [Emphasis added]

Also, it's true that deductive conclusions are easier to show to be certain than inductive conclusions, because they involve a mechanical procedure. But there can be mistakes in deductive argument, also, so there is room for a distinction between deductive conclusions that are certain and those which are not. So you're right that deductive conclusions can be certain, but it's not because they uniquely "preserve identity." Inductive arguments do, as well. To contradict the idea that all men are mortal is to contradict an important fact about the identity of men, not an essential fact, but still a real fact. To say that there is a difference in content between "all men are mortal" and "A is A" is a blatant invocation of the analytic-synthetic dichotomy.

I agree. I think some inductive conclusions can be "certain" in the "C1" sense and have made strong arguments elsewhere showing why and when this can be so.

This, at least, is what I understand to be the Objectivist view. If Betsy or anyone else disagrees with the Objectivist view, that's fine, but they should admit it.

Just which of my actual views are you claiming conflict with which of Ayn Rand's views?

[noumenalself]

I agree with the points noumenalself has been making, but I would have put this differently. I don't think the problem involves definitions at all, and it doesn't involve advocating one rather than the other of two genuinely different concepts.

[snip]

[*] well, leaving aside the (inconsistent) bit in C2 about "high degree of probability". If, as Betsy says later in her formulation of C2, the evidence meets the relevant epistemological standard of proof, then the conclusion is certain (which is to say: absolutely certain, 100% certain) -- not just "highly probable".

I agree with ttn that one way to look at this is just to show that C1 and C2 mean the same thing (viz., "non-contradiction" certainty and "full evidential support" come to the same thing). But when you look at the examples Betsy gives of C1, it's pretty clear that she is trying to group different things with it than she is with C2, and that's an independent problem. As I argued in my last post, there's no justification for grouping things in the way she does with C1, so from that perspective it's a bad definition (even if we can charitably interpret her abstract formulation of it to mean the same thing as C2). Her point about probability deepens the trouble.

[Seeker]

I agree with the points noumenalself has been making, but I would have put this differently. I don't think the problem involves definitions at all, and it doesn't involve advocating one rather than the other of two genuinely different concepts.

The problem, rather, is thinking that C1 and C2 are "two different concepts." If you read Betsy's descriptions of C1 and C2 which Dan Edge quoted to begin this thread -- read them, that is, from a proper Objectivist perspective -- they say exactly the same thing [*] in different words (as noumenalself already pointed out by noting that, once the standard for proof has been satisfied, denying the conclusion would mean endorsing a contradiction). Only if one holds the premise that there is some kind of dichotomy between the empirical and the logical (or the analytic and synthetic, or some such) could it possibly look like there are two different concepts here which just happen to be named by the same word.

[*] well, leaving aside the (inconsistent) bit in C2 about "high degree of probability". If, as Betsy says later in her formulation of C2, the evidence meets the relevant epistemological standard of proof, then the conclusion is certain (which is to say: absolutely certain, 100% certain) -- not just "highly probable".

Except for the fact that Betsy's C2 entails doubt, i.e. non-certainty (specifically, as to another person's mental state, because we cannot read minds), while her C1 does not. As far as I know, others' mental states are the only example she provides where conclusive inferences cannot be made). You may be right that C1=C2 once such doubts are dispelled, but then the issue is that, in Betsy's view, such doubts cannot be dispelled because we are not mindreaders. That's the difference.

What got us here was the mistaken allowance that Betsy's C2 is a form of certainty. It isn't. If one cannot resolve her doubts regarding others' mind states, then it is completely improper to allow, as Betsy does, that the criminal law's "beyond a reasonable doubt" standard constitutes a valid form of certainty. I think she should just say what she means: in her view, the criminal law's "beyond a reasonable doubt" standard allows reasonable doubts, i.e. uncertainty, because men cannot read minds. Then this entire mess could have been avoided. To put it another way, debating what "certain" means (as though it is not perfectly obvious) is serving to obscure the real disagreement and put it farther from resolution. I have expressed my view of the core issues here.

Edited August 3, 2007 by Seeker

[dan_edge]

Betsy Writes:

There have been cases where all known evidence supported a conclusion as true, but later additional evidence proved it false. The original evidence is not self-contradicted. It is contradicted by later evidence.

I don't think this is correct. If all known evidence supports a conclusion, and the conclusion fulfills the epistemological requirements for certainty, then the conclusion is absolute within the specified context. When new evidence is found, the context has changed. The original conclusion is still true given the original context. Peikoff's blood type example is an illustration of this.

--Dan Edge

[Betsy]

Second interpretation:

The certainty denoted by C1 means we can corroborate our claims through either noncontradictory identification or sound inductive reasoning, as detailed above.

The certainty denoted by C2 means we have supporting evidence but we lack a sound causal explanation.

Third interpretation:

The certainty denoted by C1 means we have a sound causal explanation that reduces to a tautology.

The certainty denoted by C2 means we have supporting evidence but we lack a sound causal explanation.

Why is this different from the second interpretation?

My statement of C1 is more essential. It states why "we can corroborate our claims through either noncontradictory identification or sound inductive reasoning" in some cases.

[DavidOdden]

My statement of C1 is more essential. It states why "we can corroborate our claims through either noncontradictory identification or sound inductive reasoning" in some cases.

In what way does "a sound causal explanation that reduces to a tautology" state the reason why? In fact, saying that we can "corroborate our claims through either noncontradictory identification or sound inductive reasoning" states why it is a tautology.

[Betsy]

Tautology is an anti-concept: the nihilistic attempt to transform knowledge into non-knowledge. A is A is decidedly not a tautology; and if that isn't, then there isn't anything else in the world that is.

Then you ought to tell Dr. Peikoff he made a big mistake.

It follows that there are no grounds on which to distinguish "analytic" from "synthetic" propositions. Whether one states that "A man is a rational animal," or that "A man has only two eyes"—in both cases, the predicated characteristics are true of man and are, therefore, included in the concept "man." The meaning of the first statement is: "A certain type of entity, including all its characteristics (among which are rationality and animality) is: a rational animal." The meaning of the second is: "A certain type of entity, including all its characteristics (among which is the possession of only two eyes) has: only two eyes." Each of these statements is an instance of the Law of Identity; each is a "tautology"; to deny either is to contradict the meaning of the concept "man," and thus to endorse a self-contradiction.

A similar type of analysis is applicable to every true statement. Every truth about a given existent(s) reduces, in basic pattern, to: "X is: one or more of the things which it is." The predicate in such a case states some characteristic(s) of the subject; but since it is a characteristic of the subject, the concept(s) designating the subject in fact includes the predicate from the outset. If one wishes to use the term "tautology" in this context, then all truths are "tautological." (And, by the same reasoning, all falsehoods are self-contradictions.)

When making a statement about an existent, one has, ultimately, only two alternatives: "X (which means X, the existent, including all its characteristics) is what it is"—or: "X is not what it is." The choice between truth and falsehood is the choice between "tautology" (in the sense explained) and self-contradiction. [Emphasis added]

[noumenalself]

Edited August 3, 2007 by noumenalself

[Betsy]

If unreasoned doubt destroys absolute certainty, then absolute certainty would have no role in man's cognition, which is based only on reason, so I had not expected you to find any grounds for distinguishing "certain beyond a reasonable doubt" and "100% certain".

The grounds are that, with certain entities (those with free will) our knowledge must be inferred from limited evidence and we must form our conclusions with that in mind. With other entities, we can directly perceive the entities and all their relevant causal properties by direct sense perception or introspection.

Hence all that business about mindreading - that supposedly being the necessary (but impossible) step to logically validate and achieve certainty, hence certainty is impossible concerning entities having free will. No matter the evidence, according to Betsy's view, doubts necessarily remain because the step needed to logically validate the conclusion cannot be done.

Let me try to untangle this and clarify what my view actually is.

I am not saying that we can never logically validate our conclusions about other people. We often can make valid conclusive judgments given enough relevant evidence even though we don't have all the relevant evidence necessary for 100% certainty.

Nor am I saying that, since we cannot read minds and directly perceive the motives of others, we should always doubt our judgments of others. That's a false alternative. We should form the best judgments we can, given the limited evidence available to us, and not doubt our judgments until or unless we have a reason to do so.

[Seeker]

You cannot have your doubt and certainty too. It simply makes no sense to say that our conclusions about others can be certain because they can be highly probable but not certain. As to mindreading, here is what you said:

We can be certain of what we see people do but, because we can't read minds, we cannot be equally certain as to the motives that caused them to do it.

We can be certain of what people have done but, because men have free will, we cannot be equally certain about what they will do.

And that my position in essentials.

Again, you say that we cannot be certain. So why bother keeping up the ridiculous claim that we can achieve "certainty", just not 100% certainty? Why not just say what you mean: that such conclusions can be highly probable but not certain?

Edited August 4, 2007 by Seeker

[Betsy]

It is a clearly documented fact, given your posts here, that you have invented or accepted a distinction between 100% certainty and certainty beyond doubt based in reason.

Incorrect. I have neither invented nor accepted any such thing. What I have done is identified that people are using the word "certainty" to mean two different concepts and often conflating the two.

You have persistently failed to identify the essential difference between these degrees of certainty,

That is because it is not a matter of degree but one of two different valid concepts being denoted by the same word.

and have failed to address the point that man has no means of acquiring knowledge other than reason, so to presume that it is possible to go beyond reason in acquiring certainty is incomprehensible.

Why should I address the point that man has no means of acquiring knowledge other than reason when I assume that everybody here -- especially me -- agrees with that?

One can apply reason to the facts and see that you are evading the fundamental flaw in your position. I don't know what your innermost beliefs are, but your words are clear and public.Your view entails that there is a method of cognition for man other than reason: the explanation has been given many times over the past few days.

And what would that non-rational method of cognition be that you claim I am advocating? And where do I advocate it?

I did ask you if you agreed with Peikoff's statement about certainty, and you said you did, in post 82 of this thread. [ed: this post and the one it replies to were moved here from here, so the reference is #82 of that thread. d.o.] You may well now believe that Peikoff's position on certainty is in error and contradicts Ayn Rand's philosophy, as you just implied, but such an accusation regarding Peikoff's explanation of evidentiary concepts is a serious accusation which should be substantiated.

I think Peikoff is correct with regard to the above cited quote. It is also an example of someone validly using the word "certainty" to mean C2.

[Betsy]

Before I deal with the substantive issues raised by noumenalself, I want to address his opening paragraph

Betsy's view that there are two kinds of certainty is completely wrong, and also philosophically dangerous. Her views are almost indistinguishable from mainstream philosophical ideas about certainty, and are inconsistent with the Objectivist view. The reasons she cites for her views have also been refuted countless times in the Objectivist literature, particularly in Dr. Peikoff's lecture courses.

The above charge, even if true (which it isn't), constitutes the fallacy of Poisoning the Well.

I am trying to understand what is wrong with the above quote

It has nothing to do with the truth or the falsity of the argument. The Fallacy of Poisoning the Well consists of beginning an argument by saying negative personal things about one's intellectual opponent in order to cast doubt on their arguments. See the fuller discussion of that fallacy included in my post or Google "poisoning the well" for other explanations and examples.

Betsy, is the concept of absolute certainty different from contextual certainty.

My position -- and I believe it Ayn Rand's as well -- is that all knowledge, certain or not, conclusive or not, is contextual.

PS the purpose of me asking is because to me it seems like an important question, I'm not trying to poison any well.

I know. It was an honest question and I want to give you a helpful answer. If what I wrote wasn't clear, feel free to ask more questions.

[Seeker]

Betsy, please let me ask this. What, in your view, is the difference between the certainty you advocate and the certainty that Dr. Peikoff advocates in OPAR Ch. 5? Because there is clearly a difference. I think that when Dr. Peikoff says certain, he doesn't mean "highly probable" - he means certain - and you do not agree with that. I am trying to get at the linchpin of the disagreement. It seems to boil down to the fact that we cannot read minds, so we cannot know men's motives (what causes people to do things) and that the lack of that causal connection is (for you) a barrier to (contextual) certainty about men's motives. Is that correct?

[DavidOdden]

I have neither invented nor accepted any such thing. What I have done is identified that people are using the word "certainty" to mean two different concepts and often conflating the two.

That is not what you said here. You said "You can't be 100% certain that someone else is moral in his past and present actions either, just certain beyond a reasonable doubt which is as good as it gets." Overtly, you've made a distinction, you did not attribute one version of this belief to someone else, you made this statement yourself, and you refer to both as "certain". The facts speak for themselves. The distinction between C1 and C2 as you defined these expressions here is nonexistent. The meaning of "conclusive evidence, logically validated, that fulfills a standard of proof" is "conclusions derived and inferred from sense perception using rules of logic to preserve identity".

[noumenalself]

That is not what you said here. You said "You can't be 100% certain that someone else is moral in his past and present actions either, just certain beyond a reasonable doubt which is as good as it gets." Overtly, you've made a distinction, you did not attribute one version of this belief to someone else, you made this statement yourself, and you refer to both as "certain". The facts speak for themselves. The distinction between C1 and C2 as you defined these expressions here is nonexistent. The meaning of "conclusive evidence, logically validated, that fulfills a standard of proof" is "conclusions derived and inferred from sense perception using rules of logic to preserve identity".

Nice summary Dave. I think it puts things to rest. Whatever she says elsewhere, Betsy has not only observed the use of two concepts, but used them herself. She endorses both of them as valid concepts, even when I and others have argued in detail that neither of them are valid. And she has not adequately addressed these arguments. I'm certainly not going to pursue this issue any further. Anyone who looks back over the development of this thread can see what Betsy has said, and how she has chosen to parse it, subsequently.

By the way, if calling attention to the dangerous philosophical implications of someone's words is a "negative personal thing," then I think "poisoning the wells" is a fantastic idea. It just isn't a fallacy!!

[ttn]

I agree with ttn that one way to look at this is just to show that C1 and C2 mean the same thing (viz., "non-contradiction" certainty and "full evidential support" come to the same thing). But when you look at the examples Betsy gives of C1, it's pretty clear that she is trying to group different things with it than she is with C2, and that's an independent problem. As I argued in my last post, there's no justification for grouping things in the way she does with C1, so from that perspective it's a bad definition (even if we can charitably interpret her abstract formulation of it to mean the same thing as C2). Her point about probability deepens the trouble.

I had only really looked carefully at the descriptions of C1/C2 that were quoted in post #1 of this thread. (I haven't looked at any of earlier thread where, I gather, she offered this distinction as a route to some kind of skepticism about moral judgments.) In any case, I definitely agree with your analysis in post #28.

[ttn]

This is a bit of an aside, but maybe a more important and interesting topic than what's wrong with Betsy's ideas about certainty.

Dan Edge wrote:

If all known evidence supports a conclusion, and the conclusion fulfills the epistemological requirements for certainty, then the conclusion is absolute within the specified context. When new evidence is found, the context has changed. The original conclusion is still true given the original context. Peikoff's blood type example is an illustration of this.

I think there are some problems here, along the same lines as post #16. First, I sense the view that the phrase "within the specified context" is being treated here as a qualification of "certainty". It isn't, and this view leads immediately back to some kind of Betsy-like distinction between two distinct concepts ("contextual certainty" vs "absolute certainty"). But all knowledge is contextual. There's no such thing as knowledge or certainty which isn't "within [some] specified context." "Contextual certainty" is just a perhaps-helpful name for plain old regular certainty, the only kind there is. (Just like: by "objective reality" we don't mean some other alternative to "regular reality".)

Second, I don't understand why Dan wrote that the conclusion is "absolute within the specified context" rather than "certain within the specified context." Perhaps just a typo.

Third, for a conclusion which was previously genuinely certain to be overturned when new evidence is found, is an extremely unusual situation. Dan's phrasing makes it sound like this is a normal occurrence whenever new evidence is found. If it is normal for some person to be constantly revising claims (earlier held as certainties) when new evidence comes in, that just shows that that person is using the wrong standards of proof.

Fourth, I think it's flat wrong to ever say that a false conclusion "is still true given the original context." Certainty may be contextual, but truth -- correspondence with reality -- is just truth. It is a bizarre perversion of language (and thought) to describe a conclusion which contradicts the facts as (any flavor of) "truth". One of my old philosophy professors (I can't remember which) made up the following parallel to dramatize this point: if a doctor does everything right but the patient still dies, would you describe that patient as "contextually alive"?

And so, fifth, I think Dan is misinterpreting the blood types example from OPAR. I have some questions about this passage, too, and think it could have been made more clear, but I think it's definitely wrong to read it as endorsing any such idea as "contextual truth." Rather, I believe the point is that the original claim (properly understood) is true. That is: the causal factors referred to by "A", "B", etc., do, even in light of the new information, have the compatibility relations that were originally claimed. It's really the same as another example LP mentions elsewhere: that gravity exerts a force which pulls heavy things down toward the earth is, and remains, true, even after one learns about airplanes. In both cases, there's nothing in the original claims (which are about certain causal factors) precluding the existence of other factors entering and yielding surprising outcomes. (By contrast , if the original claim had been "Patients with A blood will never suffer any kind of fatal blood coagulation when injected with A blood from another person" that would be refuted by the "new evidence" -- but that is a very different claim, one which would require different/stronger evidence to support in the first place. In any case, if someone did erroneously accept that claim -- even, implausibly, if he did so without any epistemological culpability -- and then learned about the possibility of rh-based coagulation, he'd have to say that his earlier claim was false -- not "contextually true", but just plain false.)

[Seeker]

After reviewing post #82 and post #98 in the other thread, it is clear that Betsy is invoking different standards of proof that, when reached, constitute certainty (that those standards of proof have been met). C2 is Peikoff certainty from OPAR chapter 5. C1 is a special instance of C2 in which the standard of proof is "absolute" such that it is impossible to be proven wrong later. Because of this, it is highly unlikely that Betsy will acknowledge a disagreement with OPAR ch. 5 (other than her preference that "certainty" refer only to certainty in which the standard of proof is "absolute").

What we have then, according to Betsy's approach, is certainty that a conclusion has been validated "absolutely", or certainty that a conclusion has been validated "beyond a reasonable doubt". So I can see now how she is using the OPAR concept of certainty by plugging in different standards of proof, but then the question simply turns from "degrees of certainty" to standards of proof, and the same objections apply - that there is no distinction between the standards, that all knowledge is contextual, etc.

Edited August 4, 2007 by Seeker

[DavidOdden]

Third, for a conclusion which was previously genuinely certain to be overturned when new evidence is found, is an extremely unusual situation. Dan's phrasing makes it sound like this is a normal occurrence whenever new evidence is found. If it is normal for some person to be constantly revising claims (earlier held as certainties) when new evidence comes in, that just shows that that person is using the wrong standards of proof.

This latter point is the major problem in the certainty business. Although there must be an antidote to the prevaling skeptical view that one can never really know anything, it is equally a mistake to declare that you are certain of everything. Certainty is quite possible, but for most higher-order conclusions it rightly takes major intellectual effort and is difficult to attain. It is easy to be certain that you stubbed your toe, and not so easy to be certain that all crows are black or all swans are white.

Many people do hold the belief that all crows are black, and this can be a belief that is valid in the context of their knowledge. That context is not about crow color itself, but crow species. It is common for people to be aware of only one species of crow (for example, I only knew of C. caurinus and possibly C. brachyrynchos in my youth), and I thought that "crow" meant C. caurinus. In that context, "all crows are black" is true (almost). Your knowledge context can expand, however, and you can learn that there are other species of crow some of which are grey, or black and white, thus your concept "crow" is refined. (The "almost" qualifier above is that there are very rare albino crows which are not black. Most people have no evidence that there is such a thing as an albino crow).

While I agree that overturning certainty ought to be very rare, it is possible. OTOH, Newton's gravitational law while well supported was not certain, because there were well-known counterexamples, and the law was improperly stated as holding of contexts for which there was no evidence that it held).

Truth is often seen as a correspondence between words and fact, which is a mistake. When a person makes a (social) mistake in using the wrong word for a particular concept, his error is not in non-correspondence between idea and fact, but between idea and word.

[MisterSwig]

I think this passage from OPAR, p. 177, is relevant to this discussion:

Now let us note that information about the capacities of a species is not evidence supporting a hypothesis about one of its members. From "Man is capable of murder" one cannot infer "Maybe Mr. X is the killer we are seeking." To validate the latter, one must have grounds to suspect that the human capacity in question was actually exercised by this individual.

We all recognize that man is capable of being dishonest, but from this fact Betsy invalidly infers that maybe Mr. X is being dishonest, when there are no grounds to suspect that Mr. X exercised his capacity to be dishonest.

"It is possible for man ..." does not justify "It is possible that this man ..." The latter claim depends on the individual involved and on the specific circumstances. It must, therefore, be supported by data that are equally specific.

This point has already been argued by a few of us, but I thought it might help to quote Dr. Peikoff's framing of the issue, as well.

Edited August 4, 2007 by MisterSwig