Standards of Proof
Aug. 30th, 2006 12:07 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
xiphiasI can think of at least four different standards of proof:
1) Mathematical Proof: if you have proven something mathematically, it's true. At least, it's true under the framework of the starting conditions you are working from and the definition of the operators you are working with. But, within the defined framework, it is true, full stop. You gat get all sorts of interesting effects by changing the framework ("Given a line and a point not on the line, only one line can be drawn through the point parallel to the line." What happens if we assume that arbitrarily many lines can be drawn through the point parallel to the line? Hey! We've discovered Reimannian manifolds!), but, within the framework, that which is proven is true.
2) Scientific Proof: this is not as rigid as mathematical proof, because it is always open to reexamination and reinterpretation, but it's damned close. If you have an idea that you can make predictions with, and the predictions come true, then either your idea is true, or something really close to your idea is true, or something genuinely interesting is happening. By the time that a scientific idea gets enough evidence behind it to be called a theory, you can pretty much take it to the bank. Oh, you can certainly make new discoveries, and changes, and find more details and refinements, but, by the time it's called a "theory", it's within spitting distance of reality.
3) American Criminal Standard of Proof: Beyond a Reasonable Doubt. This is a weird one. Because it contains at least one critical undefined -- and, frankly, undefinable -- term: "reasonable." Any definition of "reasonable doubt" that is, um, reasonable, gets uselessly circular immediately.
Still, we can make some assumptions about it. We discount alternative explanations that are vanishingly unlikely, such as "I didn't rob the bank, it was my evil twin," unless other evidence is presented which makes that explanation less unlikely -- such as if the person can present evidence that they actually HAVE an evil twin.
Something is proven to the American Criminal Standard of Proof if we can NOT construct an alternate explanation of the facts which we believe is at least vaguely plausible.
4) American Civil Standard of Proof: Preponderance of Evidence. In the American system, we've got completely different standards of proof for criminal and civil matters. For civil matters, the standard of proof is simply that "the thing we're trying to prove seems more likely than not." A 51% chance of being true is good enough for the civil standard of proof.
What standard of proof should we insist on for various sorts of things?
In general, I think that, if you're going to suggest that a public figure has done something wrong, you don't need to be able to manage a #3 standard of proof, but you ought to be able to manage better than #4.
However -- I think we're at the point where, if your thesis is, "The Bush administration is fucking with us, trying to manipulate the news media, trying to distract folks from their screwups, and give money to their buddies," I think it's fair, at this point, to go with definition #4.
1) Mathematical Proof: if you have proven something mathematically, it's true. At least, it's true under the framework of the starting conditions you are working from and the definition of the operators you are working with. But, within the defined framework, it is true, full stop. You gat get all sorts of interesting effects by changing the framework ("Given a line and a point not on the line, only one line can be drawn through the point parallel to the line." What happens if we assume that arbitrarily many lines can be drawn through the point parallel to the line? Hey! We've discovered Reimannian manifolds!), but, within the framework, that which is proven is true.
2) Scientific Proof: this is not as rigid as mathematical proof, because it is always open to reexamination and reinterpretation, but it's damned close. If you have an idea that you can make predictions with, and the predictions come true, then either your idea is true, or something really close to your idea is true, or something genuinely interesting is happening. By the time that a scientific idea gets enough evidence behind it to be called a theory, you can pretty much take it to the bank. Oh, you can certainly make new discoveries, and changes, and find more details and refinements, but, by the time it's called a "theory", it's within spitting distance of reality.
3) American Criminal Standard of Proof: Beyond a Reasonable Doubt. This is a weird one. Because it contains at least one critical undefined -- and, frankly, undefinable -- term: "reasonable." Any definition of "reasonable doubt" that is, um, reasonable, gets uselessly circular immediately.
Still, we can make some assumptions about it. We discount alternative explanations that are vanishingly unlikely, such as "I didn't rob the bank, it was my evil twin," unless other evidence is presented which makes that explanation less unlikely -- such as if the person can present evidence that they actually HAVE an evil twin.
Something is proven to the American Criminal Standard of Proof if we can NOT construct an alternate explanation of the facts which we believe is at least vaguely plausible.
4) American Civil Standard of Proof: Preponderance of Evidence. In the American system, we've got completely different standards of proof for criminal and civil matters. For civil matters, the standard of proof is simply that "the thing we're trying to prove seems more likely than not." A 51% chance of being true is good enough for the civil standard of proof.
What standard of proof should we insist on for various sorts of things?
In general, I think that, if you're going to suggest that a public figure has done something wrong, you don't need to be able to manage a #3 standard of proof, but you ought to be able to manage better than #4.
However -- I think we're at the point where, if your thesis is, "The Bush administration is fucking with us, trying to manipulate the news media, trying to distract folks from their screwups, and give money to their buddies," I think it's fair, at this point, to go with definition #4.
![[identity profile]](https://www.dreamwidth.org/img/silk/identity/openid.png){kind=small}
(no subject)
Date: 2006-08-30 04:24 am (UTC)(no subject)
Date: 2006-08-30 05:25 am (UTC)Why? Why lower your standards for this particular claim?
II ask this, not as somebody who thinks the statement as wrong, but as somebody who would like to see it easily defended.)
(no subject)
Date: 2006-08-30 10:45 am (UTC)Fool me once, shame on you. Fool me twice, shame on me. Fool me for six years running and you lose the presumption of innocence?
(no subject)
Date: 2006-08-30 03:08 pm (UTC)I guess my take on it was trying to "prove" it to people who don't share my opinion of BushCo's past performance.
(no subject)
Date: 2006-08-30 12:22 pm (UTC)The more likely or in-character an action is, the lower the standard of proof required.
I tend to believe that the majority of people are honorable and take their responsibilities seriously, so require a fairly high standard of proof to believe that someone is shirking their duties for personal gain. However, since the people surrounding Bush have been so consistently willing to do so, that lowers the standard of proof required to get me to believe that they have done so in a particular instance.
(no subject)
Date: 2006-08-30 12:44 pm (UTC)Sometimes stated as "Exceptional claims require exceptional evidence." I can't find the original of that quote -- I suspect it has a different original wording -- but it appears in that form in one of the WP style guides...
One might also add "untrustworthy people must present additional evidence", the expansion of the Boy who cried Wolf.
(no subject)
Date: 2006-08-30 05:33 am (UTC)Yes, I'm being wry.
(no subject)
Date: 2006-08-30 12:42 pm (UTC)(no subject)
Date: 2006-08-30 06:56 am (UTC)6) Historical proof -- the statement that is most consistent with the largest body of surviving primary source documentation is true.
My college epistimology professor would be crabby at me for not remembering more standards of proof, but it's the middle of the night and I'm tired.
(no subject)
Date: 2006-08-30 12:25 pm (UTC)(no subject)
Date: 2006-08-30 01:43 pm (UTC)(no subject)
Date: 2006-08-30 02:10 pm (UTC)(no subject)
Date: 2006-08-30 03:03 pm (UTC)In most cases a consistent system is better (more logical) than an inconsistent system, but sometimes scientific and mathematical proof collide.
The difference between scientific and mathematical proof is experimentation. Science doesn't just have to be logically correct, you have to go out and test it in the real world. And if the real world turns out not to agree with logic, then the real world wins (well sort of... in practice what usually happens is that the axioms are changed, as happened when Einstein recognized the speed of light as a constant, and that measures of distance and time are relative to the velocity of the point of view from which they are measured.)
Anyway... most of classical physics is based on Euclidean Geometry, so in a sense it's "scientifically proven." Non-Euclidean Geometry has not been used this way, so it hasn't been scientifically tested... but, it's mathematically proven to be consistent. Euclidean Geometry, on the other hand, hasn't been proven to be consistent... or at least it hadn't been the last time I was in the loop about these things. That doesn't mean it's inconsistent, only that there isn't any proof of logical consistency.
So, by mathematical standards of proof there is a stronger case for parallel lines crossing (the Non-Euclidean assumption) than not crossing (the Euclidean Axiom). But by scientific standards of proof there is a stronger case that parallel lines don't cross.
I have more to say about the value of consistency in standards of proof, but I sort of need to go now... if I have a chance I will come back and finish this.
Kiralee
The Eighth Standard - Part 1
Date: 2006-08-31 12:48 am (UTC)The truth-testing processes that come with these standards of proof are all different, but all of them, except mathematics, have one thing in common: They depend on competitive doubt. That is 1) the process includes and depends on deliberate formalized debates between opposing viewpoints, called, for example, "trials" or "peer review"; and 2) the "trier of fact" is supposed to start with the assumption that what he or she is being told is false until "proven" otherwise.
I believe that competitive doubt is divisive; it separates people into opposing camps, and escalates conflicts. Does that make competitive doubt a bad thing? Not necessarily. In fact, in the context of scholarship, truth-testing the physical world, economics, politics, legal procedures, public news and information media, and just about anything in the public sphere, I believe that competitive doubt is appropriate. I believe that the methods which use, even depend on, competitive doubt are stronger, more effective, and, in a word, better, than the methods which don't.
However, I believe there are contexts in the private sphere that are different. These contexts all require some method of truth-testing, that is of evaluating differing opinions to determine which is the most accurate or effective; but they also require the participants to maintain a level of connection that the divisive aspect of competitive doubt would destroy. All interpersonal conflict falls into this category.
The Eighth Standard - Part 2
Date: 2006-08-31 12:52 am (UTC)As I've mentioned, all of the standards of proof listed so far, except mathematics, are associated with processes that depend on competitive doubt. So they can't be used (effectively) to solve this kind of problem. But, can we use mathematical standards of proof?
Mathematics depends on the process of logic, not the process of formal debate. Instead of requiring competitive doubt, it requires precise definitions. Now, clear definitions are a good thing to have in any truth-testing process, but...
I was once a computer science and mathematics major, and I learned that computer languages, which are mathematical in nature, are not at all like "real" human languages, despite the superficial appearance of similarity. One difference is that "real" human languages have more reach - a bigger vocabulary, and a greater scope; another is that "real" human languages are fuzzier - many words in "real" human languages are not, and can not be, defined precisely.
I've had nearly twenty years of experience with the English language since then, and I've discovered, by experience, that, with sufficient discipline, one can come very close to mathematically precise definitions; but the closer one gets to precision, the less intelligible one becomes to ordinary people.
The Eighth Standard - Part 3
Date: 2006-08-31 12:59 am (UTC)So, based on this set of beliefs, we can conclude that in order to effectively resolve interpersonal conflicts we need another standard of proof and associated process of truth-testing, that has not yet been defined.
No, I haven't figured it out yet, although I have a few ideas. My scholarly inspiration comes from women's studies, particularly the work of Carol Gilligan and Blythe McVicker Clinchy, and books by Charles Tilly and Deborah Tannen - although in these last two cases I think some of the original research was done by others.
On the other hand, most of my practical inspiration comes from books on writing, and how authors convince readers to suspend their disbelief - that is, prove that their fictional stories are believable. Because clearly there's a standard, and some people are better at it than others.
The closest I've come has to do with the idea of whether a logical system is consistent or inconsistent. Like logical systems, which are based on definitions and axioms, fictional worlds - even the most mundane - are based on a set of core beliefs. When a fact is introduced that's consistent, it maintains and strengthens the suspension of disbelief. When facts are introduced that are inconsistent, the suspension of disbelief collapses... sometimes it only takes a single fact to bring down the entire structure.
The process outside of fiction looks something like this... instead of doubting and debating an opponent you collaborate to discover as many facts as possible about one another's beliefs... "Yes, I'll accept that as an axiom for the moment, can you tell me more?" until one side or the other collapses into disbelief. Although, frankly, the practical results are often much more complex.
The Eighth Standard - Part 4
Date: 2006-08-31 01:03 am (UTC)In the mean time I get very frustrated. Every public school tries to teach it's students about truth-testing. And for the most part, that's a good thing - because if we're going to be responsible citizens it's something everyone needs to know in order to vote effectively, and for a lot of other things as well.
However, one of the most common situations requiring truth-testing that people find themselves in is interpersonal conflict. And when they find themselves in this situation they often do what they were trained to do in school; as a result, they invoke methods of truth-testing that are ineffective, or even destructive to their goals.
And that just seems very sad and wasteful...
...I mean, yes, Bush and his administration have caused a lot of harm. And, at least in some cases, that harm has been the result of violations of the constitution. And we should do something about that, like vote him out of office...
... but in the long run, I think more harm is caused by individual people, doing the best they can to resolve interpersonal conflicts, who don't have the skills or ideas they need to do it effectively; and who substitute processes of truth-testing that are not just ineffective, but, because they depend on the divisive elements of competitive doubt, are actively harmful to their goals.
Kiralee