MemberlistThe Library Index  FAQFAQ   SearchSearch   RegisterRegister   ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 
The Truth is Always Simple (APPLIED EPISTEMOLOGY)
Reply to topic Goto page Previous  1, 2, 3, 4, 5, 6, 7  Next
View previous topic :: View next topic  
Ishmael


In: Toronto
View user's profile
Reply with quote

This is a section I have just composed for the new book I am co-authoring with Mick on Deserts and Climate. This section is the first part of a discussion of the "Truth is always simple" axiom.

Comments welcome.

So it is that, after attributing a few deserts to trade winds--those that are consistent with the universal climate model based on latitude--the half dozen or so deserts that remain can be accounted for only by inferring several additional, desert-making mechanisms: Continentality (Turkestan), Coastal cold currents (Atacama), Monsoons (Thar), Rain-shadows (Patagonia), and, in some cases, a confluence of these factors is required (Australian, North American, and Taklimakan/Gobi).

Quite a patchwork of solutions have been applied to solve this problem. Who would imagine that several different inputs would all produce precisely the same output. In many cases, two adjacent deserts are imagined to be the product of completely different forces.

This is all terribly complex and would certainly violate Occam's Razor, if any less complicated explanation was available for the desert phenomenon. But no such alternative explanation exists. This is the best one we've got.

But why is that so? Why is there no simpler explanation?

One possibility is that the explanations we've got are correct. As problematic as that conclusion might be, in this particular case, perhaps the universe really is as complicated as climate scientists suggest.

Another possibility is that these explanations persist only because no better ideas are on offer.

If so, Occam's Razor may be partly to blame.

Occam's Razor is the perfect instrument for evaluating alternatives. Occam instructs: Given multiple solutions of equivalent explanatory power, choose the simplest. Unfortunately, his axiom can't comment upon the merits of a singular idea. If there's only one proposal on the table, so long as it's consistent with the facts, Occam gives it his blessing, regardless of complexity.

But what if that only solution is wrong? If that proposal were rational and appeared consistent with the facts we knew, how could we tell it was in error? How might we know the idea was a bad one?

Science says that bad ideas are eliminated by falsification. A bad idea will show itself eventually, when illuminated by a fact with which it is inconsistent. Unfortunately, 'eventually' sometimes fails to arrive for a very long time and, no matter how quickly 'eventually' comes, it always arrives too late when a better idea might have been had in the meantime.

Sometimes, the arrival of 'eventually' is delayed by the limits of observation; we simply lack the data by which to refute a bad idea. So it sticks around until we do. As soon as the data shows up, the bad idea is discarded.

But that's not typically what happens.

Scientific theories are not discarded the moment some observation is made that doesn't quite fit. Indeed, there's a convincing argument that ideas shouldn't be so readily tossed aside. The nature of an observation may be misunderstood or the relationship it has to the model may not be as contrary as believed. Typically, a large number of inconsistencies must accumulate before an otherwise useful theory might be said to be falsified.

But even this is not what typically occurs.

In practice, a prevailing theory can be abandoned only after the development of an alternate theory of equivalent explanatory power, and only after that alternate idea has won widespread social acceptance within institutional academia. The status quo must persist until such an alternate is available.

And thus the scientific 'Catch-22' stands revealed.

Occam's Razor will tell us only which of two or more theories is to prevail in a contest. After the winner is declared, Occam will find no fault with it. No fault in the prevailing theory; no need to develop an alternative; so no alternate theory is available to challenge the prevailing wisdom when inconsistencies arise. The anomalies merely accumulate: Ignored where possible; rationalized where necessary. The ruling model is immune from falsification and virtually immortal.

With age, accepted theories only gain resilience. They grow ever-more capable of accounting for anomalies. So many 'ifs', 'ands', 'buts' and other exceptions accumulate that there's generally a slot available somewhere in which to shoehorn any observation that fails to fit the original mold. Still worse: The more necessary it is that this be done; for a dozen fields of study are now dependent on the model for its utility. Oh yes; a Nobel prize could be awarded for preserving the model.

Hence all the famous anecdotes of scientific discoveries following from laboratory accidents. Someone tips over a jar and, 'Oh my god! Penicillin!' Accidents are almost the only way to break free of the closed loop.

Science, it is said, is conservative. What works is reluctantly cast aside. But isn't it about time we asked whether it ought to be so conservative? Whose interests does that conservatism serve?

Shouldn't science be more.... progressive?

Occam's Razor empowers us with the means by which to resolve a conflict in the marketplace of ideas. We need other tools to generate that conflict and challenge the status quo.

It is Applied Epistemology that gives us those tools.

Applied Epistemology is a method by which to evaluate the merits of ideas without obvious fault.
Send private message
Tilo Rebar


In: Sussex
View user's profile
Reply with quote

Theories are only useful if they have predictive power.

Is there somewhere in the world that doesn't have a desert, when using the criteria from the current model says it should?

I expect that if it exists it will have been hidden away to avoid awkward questions being asked.
Send private message
Ishmael


In: Toronto
View user's profile
Reply with quote

More excerpts from the book...
Applied Epistemology

Applied Epistemology, as presented in this book, is not to be confused with the academic practice of epistemology that sometimes operates under the same name but without the bother of actual demonstrated applicability. We are not here concerned with a meta-analysis of the validity of the scientific method. We are concerned with contributing to that method; honing it; applying it more faithfully and more vigorously; and proving the worth of our methods with results that are open to critique.

Applied-epistemologists are primarily concerned with what they know about the universe in the absence of data. Typically, academics reach conclusions by use of inductive reasoning, arriving at general rules based upon specific observations. Applied Epistemology rejects this approach.

Applied-epistemologists do not argue from evidence. They don't approach a problem with an open mind and follow the data where it leads. Not at all. Applied-epistemologists approach problems with very narrow minds: Minds narrowed to a single, rational probability. This rational probability is the one possibility reason demands presumed until falsified by observation. This method of falsification of the otherwise probable is the only secure method by which to determine truth.

Only when observations render impossible the otherwise rationally probable are other possibilities considered. The most likely of these is then presumed.

Applied-epistemologists do not entertain multiple probabilities. In a contest of possibilities, only one probability operates. The one reality most likely is the probable reality. All other possibilities are relatively improbable by implication.

Rational probability differs from statistical probability. We are not assigning percentile chance to outcomes. Applied Epistemology operates where data is absent. The rules of Applied Epistemology are used to evaluate potential realities and to determine which is more probable absent the influence of evidence. The rules of Applied Epistemology tell us where to place the burden of proof and what we are obliged to presume true except where demonstrably impossible.

Evidence cannot tell us what is true. Evidence can only falsify what is to be otherwise presumed.
Send private message
Ishmael


In: Toronto
View user's profile
Reply with quote

Epistemological Rule 1: The Truth is Always Simple

Applied Epistemology would replace Occam's Razor with a blunter instrument: A declarative statement. We posit that the truth is always simple, excepting where it is necessarily complicated.

How can such an assertion be justified? How can we be so certain of the inherent simplicity of the universe?

We needn't be; the principle can stand on the basis of demonstrable usefulness. The presumption is tested by continued application. If the universe is a complicated place, we would never know it to be by presuming it so. However, by presuming the universe simple, we are able to establish just how complicated it must be; for, to the extent that we fail to falsify the presumption of simplicity, to that extent we establish the degree of complexity necessarily present in the system.

For in a very real sense, we know that the universe is simple. We know this by the same rules of reason that underlie Occam's Razor. In a contest of models, says Occam, we are obliged to choose the simplest. This obligation exists because the simplest model is probable.

This is only true because simplicity itself is probable.

Where anything is possible, simplicity is probable; complexity, improbable. Therefore; we must assume the universe simple except to the degree that simplicity proves impossible.
Send private message
Ishmael


In: Toronto
View user's profile
Reply with quote

Epistemological Rule 1, Corollary 1: There is Only Ever One Cause.

Complexity knows no bounds. There is no upper-limit to how complicated something may be. But simplicity has its theoretical limit at the monist principle: All is one.

In a cause/effect chain, that principle is expressed; 'If X then Y.' There can be no logical expression of lesser complexity. Nothing more simple. Nothing more likely where nothing is known to contradict it.

'If X then Y' is the probable reality where causes are opaque or where a proposed cause and effect model fails. When applied-epistemologists assert that the truth is always simple, they mean that, absent convincing evidence to the contrary, X causes Y; any given observation is assumed to be the product of one and only one factor. If a given observation is the product of multiple causes, we could never learn this by presuming it so. Only by presuming one cause to be at work may we determine how few are truly necessary.

This means that effects may not be attributed to confluences of factors except where absolutely necessary. It also means that, where an effect is attributed to such a combination, we have reason to suspect that this attribution is in error. The fewer inputs required, the simpler the theory and, as the truth is always simple, we presume one input sufficient cause for any effect we seek to explain. If we do not yet see that factor, it is likely that we have not looked hard enough.

Any weakness in the arguments favouring an established, complex model and probability again favours one cause for the effect. This is true even where we haven't yet any notion of what that one cause might be; for it is here, where Occam's Razor cannot operate, that Applied Epistemology is most effective.
Send private message
Grant



View user's profile
Reply with quote

Where anything is possible, simplicity is probable; complexity, improbable. Therefore; we must assume the universe simple except to the degree that simplicity proves impossible.


I wonder if this concept fits in with the currently popular multiverse theories of cosmology. If there is an infinity of parallel universes each with a different set of physical laws, then in a tiny percentage there will conscious creatures trying to understand their world. Most of the valid explanations uncovered by these creatures will be of the simple if X then Y type, because in that set of the multiverse where intelligent creatures exist there will be more simple explanations than complicated ones.
Send private message
Ishmael


In: Toronto
View user's profile
Reply with quote

This was my first draft of these sections. They have already undergone one significant revision and will be revised several more times.

Trying to formalize the A.E. methodology has been a dream of mine from the beginning but this is the first time I have succeeded to any large degree in expressing how the method works and why it is rational.
Send private message
Tilo Rebar


In: Sussex
View user's profile
Reply with quote

Ishmael wrote:
...Trying to formalize the A.E. methodology has been a dream of mine from the beginning...

Very good Ishmael, this needs to be done.

My old Grandma used to tell me, "You'll need to spend many, many long hours boiling off and discarding the flesh from the body of knowledge you've been taught, before you will behold the skeleton of connected facts which will allow you to create wisdom."
Send private message
Wile E. Coyote


In: Arizona
View user's profile
Reply with quote

Ishmael wrote:
Evidence cannot tell us what is true. Evidence can only falsify what is to be otherwise presumed.


What type of evidence would you accept to falsify a theory?

Let's take as an example the chronology proposed/presumed by Fomenko.

What type of evidence would you accept, from me or anyone else, that would prove to you that Fomenko was wrong..........
Send private message
Ishmael


In: Toronto
View user's profile
Reply with quote

Wile E. Coyote wrote:
What type of evidence would you accept to falsify a theory?


This is an excellent question; precisely the sort I was hoping to get, as it is not a question that had occurred to me to answer before receiving it from you. Now that I see it, I recognize the absolute necessity of answering it.

The reason being that, in reality, very few theories are ever falsified, whether academic in origin or whether "Epistemologically Correct."

Instead, specific incarnations of theories are falsified, forcing alterations in the model. However, each alteration generally increases the complexity of the model, making it (if ever so slightly) less likely.

In the traditional terms by which science proceeds, eventually a theory is rendered more complex than a competing notion, by virtue of an accumulation of complications. For Applied-Epistemologists, even sooner a point is reached at which the tension between the ideal solution and the actual model becomes so overwhelming that the original proposal fails on epistemological grounds, though no alternative theory is yet available.
Send private message
Ishmael


In: Toronto
View user's profile
Reply with quote

Wile E. Coyote wrote:
Let's take as an example the chronology proposed/presumed by Fomenko. What type of evidence would you accept, from me or anyone else, that would prove to you that Fomenko was wrong..........


No evidence would falsify so broad a contention as that made by Fomenko, on the whole: i.e., "History is much shorter than it appears." But accumulations of data can so render specific versions of this notion so complicated that they become unsustainable, especially so if the orthodox model of history remains comparably simple.

And let's not forget that Fomenko begins with a situation that makes his idea very complicated from the start: He has somehow to account for all the reams of "false data!" There's simply so much of it as to be nearly overwhelming. Accounting for it is his most serious problem (and mine too, where I would similarly revise History).

But Fomenko has one tiny advantage. He has an observable phenomena -- a data set -- that orthodoxy positively refuses to even try to account for. This is his statistical analysis of historical documentation. This is the foundation of Fomenko's work.

If the phenomenon is legitimate, it must be accounted for and Fomenko's work can be seen simply as a theory necessary to account for these observed patterns in historical documentation.

Fomenko need claim only one thing: Given that there are statistical patterns in the historical record that may be observed to repeat faithfully over two or three millennia regardless of the historical sources consulted, historical records must be composed of two-to-three duplications of original sources.

Because Fomenko already has so much complexity to account for, his thesis becomes immediately untenable if his opponents are able to do just one thing: Find any other way to account for the patterns.

That's it.
Send private message
Wile E. Coyote


In: Arizona
View user's profile
Reply with quote

"Any intelligent fool can make things bigger and more complex... It takes a touch of genius --- and a lot of courage to move in the opposite direction." - Albert Einstein
Send private message
Mick Harper
Site Admin

In: London
View user's profile
Reply with quote

Einstein Schweinstein, what's it matter as long he loves his mother.
Send private message
Ishmael


In: Toronto
View user's profile
Reply with quote

The new book I am writing (with Mick) is proceeding very well. I have now completed the second and fourth Chapters (skipping most of the third).

Chapter Two concerns the principles of Applied Epistemology. It is extremely difficult to hone my language and discipline my thinking sufficiently to express these ideas and to argue successfully in their favour.

I present below (in installments) the complete Chapter. Some of this was presented above already but all of it has undergone massive revision.

I seek criticism. I will accept praise.
Send private message
Ishmael


In: Toronto
View user's profile
Reply with quote

Chapter Two: Applied Epistemology

The Limits of Occam's Razor


But why is that so? Why has a better theory of desert formation never been found?

One possibility is that deserts do exist where they do for precisely the reasons identified by climatologists. As problematic as that conclusion might be, it remains possible that Earth's climate really is as complicated as climate scientists suggest and that deserts can appear when any of several different atmospheric conditions arise.

On the other hand; a better solution is rarely found when nobody is looking for it--and nobody is.

Partly to blame, is Occam's Razor.

Occam's Razor is a scientific tool, perfectly shaped for evaluating and choosing between alternative models. Occam instructs as follows; given multiple solutions of equivalent explanatory power, choose the simplest. Unfortunately, he can make no comment upon the merits of a solitary idea. If there's only one proposal on the table, so long as it is rational and consistent with the facts, Occam gives it his blessing.

But what if that sole solution is wrong? If rational and factually consistent, how can we tell whether it is in error? How might we recognize the idea as a bad one?

Science, in principle, relies upon falsification to weed out bad ideas. Eventually, a bad idea will show itself, when illuminated by a fact with which it is inconsistent. Unfortunately, 'eventually' sometimes fails to arrive for a very long time and, no matter how quickly it comes, a better idea could always have come sooner.

Sometimes, the delay is down to the limits of observation; we simply lack the data by which to refute a bad idea. Once the data does show up, the bad idea is discarded.

That's what they teach us in school, but that's not what typically happens.

Scientific theories are not thrown aside the moment some observation is made that doesn't appear to match predictions. Indeed, there's a convincing argument that useful ideas shouldn't be so readily discarded; the nature of an observation may be misunderstood or the relationship the new data has to the model may not be as contrary as first believed.

Sometimes, the troublesome data is just ignored. It sits there like an open secret; and sometimes a few young and foolish scholars, who don't mind the ruin of their burgeoning careers, obsessively employ it in quixotic attacks upon the prevailing model. More often, some way is found to account for the inconsistencies by altering the original theory. Some new wrinkle is added; some new exception is made, and everything proceeds again much as it did before; no one much paying attention to the ever increasing complexity of the standard model so long as it is preserved.

Either way, inconsistencies and complications must accumulate in enormous numbers before an otherwise useful theory might be said to be falsified.

But even this is not typically what occurs.
Send private message
Display posts from previous:   
Reply to topic Goto page Previous  1, 2, 3, 4, 5, 6, 7  Next

Jump to:  
Page 2 of 7

MemberlistThe Library Index  FAQFAQ   SearchSearch   RegisterRegister   ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in 


Powered by phpBB © 2001, 2005 phpBB Group