Tuesday, September 28, 2004

Definitions, Take 2

Well, Nicolas, I don't think we are sharing the same definitions quite yet.

I'll update my terms in the hope that we'll agree to some shared definitions.

Level of Trust - a level of expectation that a model or proposition is true (level on a scale from zero to one). Every meaningful proposition can be assigned a degree of trust.

Strong Belief - a proposition for which one has a high level of expectation that the proposition is true (level on a scale from zero to one, but typically greater than 0.5). Only a subset of meaningful propositions are categorized as strong beliefs.

Relative Belief - a proposition, that while one may have a low level of expectation that the proposition is true, is estimated to be most likely true (level on a scale from zero to one, often less than 0.5). Example: My sports team will win the championship.

Knowledge - that subset of strong beliefs justified by empirical facts and by verified models.

Faiths - that subset of beliefs unjustified by empirical facts or by verified models.

Heuristics - that subset of faiths unjustified by, yet consistent with, empirical facts and verified models.

Superstitions - that subset of faiths inconsistent with empirical facts and verified models.

(Man, I need to make a Venn Diagram.)

Every person has a different quantity of experience, scientific expertise, and capacity for reason. This explains why one person's knowledge is another person's faith, or one person's heuristic is another person's superstition.

Here are some examples of my terminology:

All men are mortal - Trust does not apply because the proposition is not meaningful.

The Chicago Cubs will win the World Series this season. - I have a 4% level of confidence that this proposition is true. Therefore, it is not a strong belief. However, I may think they have the best chance of any team of winning, so I may call this a relative belief.

The energy of a photon is proportional to its frequency. - Knowledge based on empirical facts and confirming experiments.

The world is subject to physical laws. - This is a faith, specifically, this is a heuristic.

God made humans in different races and intended that the races not interbreed. - This is a faith, specifically, this is a superstition.


Now, suppose I want to plan a project. I will estimate risk and reward.

I may decide to base my plan on a relative belief, e.g., I may invest in Microsoft stock on the estimation that it is most likely to be a good investment.

I may decide to pray that I recover from a treatable disease instead of seeking medical attention. This is a decision based on superstition.

When I started my business, I believed it would be successful. I figured I had better than a 50% chance of success (though not much better). I also thought that I had a very small chance of great success. Overall, the expectation value of my return on investment was high enough to pursure the business. Furthermore, if the business failed, I knew I could return to the work I was doing before (hey, it was the nineties!). I would say that I believed that my business would be a success, but I did not have faith it would be a success (by my definitions).

There are times when we want to convince ourselves to believe in something that we do not trust so that we can make proposition more likely. If I do not believe that I can compose music, I will decrease my ability to compose music by reducing my motivation. This is a heuristic. If I doubt myself, my performance will probably be worse. If I believe in myself, my performance will be better. Whether this heuristic is a good idea or not depends on the consequences of failure. If I can't compose music, I've wasted some of my free time. No big deal. If I falsely believe that I can win the Ansari X Prize and invest my life savings in a rocket, there is a high cost of failure.

One other scenario. I may not plan my actions at all. Action without planning accurs more frequently than we would like to admit. I can throw a punch or say something cruel as an emotional outlet, without planning for the consequences. When we act on impulse like this, the only thing that we are trusting is that the resulting emotional release will be pleasurable or cathartic.

Perhaps you want to re-label some of my definitions? Or add new definitions for new terms?

doctor(logic)

Monday, September 27, 2004

Symbols and Language

Nicolas,

You reject my claim that, in language, we are mapping symbols to empirical measurements.

Yet, I maintain this is effectively what we are doing, even if it may not have a structure as formal as mathematics or science.

In mathematical models, we explicitly map symbols to empirical measurements. Our brains evolved to do the same thing without using raw symbol manipulation. We can estimate distances, speeds and rates of closure. We can perceive ranges of temperature. We can detect colors and hear sounds. Let's take your example: I am seeing a blue object. Though we can distinguish two similar colors, we cannot quantify color in terms of spectra without the aid of instrumentation. However, the principle is not weakened. We are able to recognize the predominant frequencies in the color as blue. We have similar limitations with other senses. For example, I cannot estimate the a car's length to better than about six inches of precision, even when I' m looking at it.

Our brains may not naturally use symbolic algebra, but it is creating an approximate mathematical model nonetheless. It must do so in order to perceive the world. The average human brain is a microcosm of scientific research. Certainly, the average human is no scientist. Working primarily from instinct and emotion, humans are not naturally suited to rigorous science.

Here is a plausible scientific model about how our brains actually work. For each concept the brain associates with a measurement (e.g., speed, color, auditory volume) or thing (bee, car, planet), there is a region of our brain that is activated (or 'spikes') when we observe or imagine the corresponding thing. Language is an approximate association between these activations and vocalizations. Our language lacks the precision or clarity of mathematics, but it is optimized for social interaction and brevity. Voila! A scientific model of language. Okay, so I skipped a few details, but I don't doubt that we will one day create a very effective model of both thought and human language.

When this happens, there will be only science, and abstract theories about language (e.g., Wittgenstein's Tractatus) will be cast aside.

doctor(logic)

Sunday, September 26, 2004

Knowledge, Belief and Faith

Nicolas,

This is difficult, but let's try to narrow down our definitions again. These are my working definitions. I'm not averse to altering them in reasonable ways.

Belief is a model or collection of models that are trusted by an individual as a consequence of reason and prior experience. Here, trust is not a binary value (yes or no). Trust is a continuous value between zero and one. I believe 97% that I will to walk to lunch later this morning. This is belief because I do not have verification.

Knowledge is trust in the verification of a belief. Verification is rarely complete, but it is statistically possible. I will know whether I walked to lunch or not this afternoon with 99.999% confidence (maybe I was hallucinating). I may believe that the universe is supersymmetric (symmetry between matter and energy), but it is not knowledge until I have empirically verified my belief.

Faith is different. Faith is a model or collection of models that are trusted by an individual a priori, independent of any prior experience, reason or process of verification. Again, trust here is a value between zero and one.

Let's subdivide faith into that which is the result of reason (reasoned faith), and that which is not (blind faith). Faith, in the traditional sense, does not require reason.

Does reasoned faith include heuristics?

If you include heuristics in faith, then the principle of verifiability is a form of faith. The principle is a priori. Note that, by including heuristics, we have expanded faith beyond its traditional sphere.

If you drop your keys at night, do you have reasoned faith that you dropped them under the street light? Or do you only search under the street light because that's the only place you'll be able to find them? Is this faith at all? I would say this is not faith. It is a choice based on knowledge and belief, so not all heuristics should be considered faith.

Is faith in God heuristic? Perhaps, if a man sees eternal life as the ultimate goal, then subscription to a particular religion is the only way to have even a small chance at eternal life (again, discounting transhumanism). There are few people people of faith who would claim this line of reasoning. Besides, this line of reasoning is flawed because there are plenty of alternative religions to choose from. People do not have religious faith because of reason. They don't need reason to have blind faith.

Your definitions of knowledge, belief and faith may be different. That's okay, but we need to get the definitions in sync.

Nicolas, you have mentioned faith several times. Is it reasoned faith or blind faith? If it is reasoned, what are the reasons?

doctor(logic)

Saturday, September 25, 2004

The Role of Science

Nicolas,

I am not saying that science can choose for us what kind of society we want to live in. Science cannot choose anything by itself. Science is simply a way to know the rules of the universe. It does not tell us what technology we should deploy. Science makes technology possible, and tells us what the effects will be when deploy a given technology.

We can (and often do) choose to deploy a technology without analysing the consequences or by ignoring the predictions of science. This happens all the time.

Should we colonize the galaxy? There are advantages and disadvantages. Science can tell us what they are. It cannot choose for us. However, science can tell us what strategies or actions are consistent with our goals.

Logical positivism doesn't command us to create a communist or fascist society. Not at all. It is a false argument to place blame on science and logical positivism for the Soviet Union or Nazi Germany. The fact that much of the scientific rationale claimed by the Nazis wasn't science is beside the point. Science is completely neutral when it comes to intent. Science provides technology to facilitate intent.

So how do we decide to make "value" decisions? For any given objective, we will probably have many possible courses of action. Each course of action has a cost and a reward. We must incorporate human emotion in our model when we evaluate each possible course.

To me, this is plainly scientific. There is a lot of uncertainty involved when assigning a "weight" to human emotion. We cannot completely ignore human emotion - science predicts failure in these cases. We also cannot assign infinite weight because many goals (including survival) can never be met when assigning infinite weight to our emotions.

Transhumanism aside, you are correct that science cannot "control" emotions in a personal or social sense. However, science tells us the price we pay for placing emotions above other factors.

In summary, science tells us what the facts are, how the world works, and what will be the consequences of our actions. What we do with that information, science does not say (though it may still provide an explanation for our choice).

doctor(logic)

The early church's choice

Axiologist. As it happens, I am also waiting for a book I should receive from Amazon in a few days. It is a big tome on the Vienna Circle and is supposed to contain some material on M. Schlick's murder. And, so I hope, on Nelböck's paper and the reasons it was rejected by Schlick. Also, I have found a site that contains the full text of the Austriacus article. But it is in German. So if you know somone who would like to translate it for us...

Regarding the choices made by the early church, I think they need to be put in context. Church fathers were learned men and they had probably more than 100 times more philosophical material than what has survived to reach our time. Virtually all possible combination of faith and reason had been tried by the countless schools that freely mixed greek traditions with semitic, egyptian, or persian thought. So all they had to do was to pick the right combination. But they had to do it quick and make the right choice. The barbarians were at the gates and the empire badly shaken. Around 250 CE, Goth made raids in the west and in Asia Minor and Greece while the imperial troops were thoroughly routed by the Persians in the east (an emperor was even captured and taken captive to the Parthian capital, something absolutely unheard of). For several decades, insecurity mounted and terrible epidemics ravaged the whole eastern half of the empire. Estimates based on fiscal documents seem to show that Alexandria, a city not directly impacted by military events, lost up to a third of its population in 25 years. To top it all, this was also the time where the persecutions were at their highest and bloodiest.

So the minds of the church Fathers were probably quite concentrated and my impression is that they did a pretty good job given the circumstances. The image of God they chose was not a Nazi thug. It was just the mirror image of the people they were facing. They thought they were not going to convince visigothic chieftains, or bloodthirsty centurions about to boil them alive, with kind words. So they built on the jewish God (quite strong already) to create a mighty creature capable to impress even these hardened men. And they succeeded. While Rome had been unable to tame the barbarians on its borders by force, the church managed to convert most of them in a few centuries.

And what is amazing is that, while doing this extremely tough job in the most difficult of times (the war in Irak is a piece of cake in comparison), they managed not to lose track of the more subtle philosophical issues. In turning down the Gnostic, and later Arian or Nestorian, vision of God, I believe they preserved christianity's potential as a healthy civilization-building ideology for after the difficult period would have been seen through. Indeed, the history of the middle ages seems to support this. The west was rebuilt in five centuries, to the level it reached in the Renaissance, largely by the church. I do not say that the church Fathers forsaw the disasters that befell us because of our modern rationalist and scientist excesses. But I believe they were witnesses to what happened in the various Gnostic and other sects that were common in their time and did not like what they saw. An interesting job could be done in that direction, I believe : decrypt such great Gnostic-bashing Fathers such as Irenaeus of Lyons or Augustine and see what we could apply to LP or even the Unabomber.

So God is not a Nazi thug, it is just something strong enough to impress a IIIrd century thug. Not that I think it is any better. And this is indeed the biggest problem the church has faced for the last few centuries. The church Fathers made practical choices and the church, and everyone else, had only to thank them for that during the first 1500 years of its existence. Now, in the very mild and refined atmosphere of our present time, the result of these choices has become an increasingly crushing embarrassment for christianity. So be it. Either christians manange to purge their faith of what has become unacceptable to us (I doubt they can), or we have to get to work ourselves, just as te church Fathers did. But, in that case, we have to be conscious that being less good at it than they were would be impardonable. We have far more material at our disposal and, above all, we have their example to learn from. In that perspective, I doubt that just replacing the Apocalypse by a milder text (and a Gnostic one, of all things) will be nearly enough. Building a TOE for the next 1500 years requires more than that.

Regarding what books to burn, I basically agree with what you mean. Except that buring a book is probably a bad move because it is only too easy to "rediscover" the contents of a banned text. It is probably more effective to keep it in full but tightly laced in a straightjacket of counter arguments and eye-witness reports on its consequences. If the church had preserved the contents of the Nag-Hamadi library together with its own refutations of it and added reports on what was going on within Gnostic sects, it might have saved us from some of the tedium of having to reexperiment the effects of rationalism for ourselves. I am trying not to repeat this mistake with LP here. Count on me to leave for future generations a big fat CD-ROM of material on LP : their complete works, what their oponents have said, what we say here and as much information on the Unabomber, Nelböck, Gödel and other such cases.

To finish with a note on the Unabomber, and to make a connection to what you say on emotivism, I found the following excerpt on page 40 of Harvard and the Unabomber :

Somewhat paradoxically, he thought of himself as a scientist, embracing what philosophers call Logical Positivism [...] Further, he believed in positivism's parallel theory of ethics, sometimes called "emotivism", which holds that moral and spiritual judgements, being scientifically untestable, are mere "cognitively meaningless" expressions of emotion. To him, ethical and religious scrupules are [...] what he called "brainwashing". [...] Like the Nazi doctors who performed sadistic tests on concentration camp victims, Kaczynski called each of his bombings of human targets an "experiment".
Let us see what Wilks can do to salvage emotivism.

Friday, September 24, 2004

Statements about the future

Let us take two statements from your last two posts :

  1. 'God will show up in X years'
  2. 'Ted Williams will be up and running in 20 years'

They are exactly of the same form, right ? So which one belongs to science and which one does not ? Now conduct a poll asking this question. More than 99% of respondents will answer (correctly, in my opinion) : (2) belongs to science and (1) does not. So we are all agreed on this. But if we are, it cannot be on the basis of form, since the form of the two sentences is the same. So on what basis do we all make this judgement on which we are so well agreed ?

I think this is not a simple problem and we should approach it with great care. Indeed, this is probably one of the most difficult problems of modern philosophy. One that people like Carnap, Ayer, Quine and Wittgenstein have been struggling with for the best part of their lives. In that debate, I have chosen my camp. It is that of the later Wittgenstein (after 1930, not to be confused with the Wittgenstein of the Tractatus, which was written just after WWI).

Its answer to this particular problem would probably be to say that you cannot use form to distinguish between the two statements because form has no direct relationship with reality. It is just a set of rules, a 'grammar', that only apply within language itself, to govern how we use it, but cannot say anything to us about the objects to which we apply it. In fact, the later W is in full agreement with you, Doc, when you say that language is nothing more than a protocol. But this has a consquence I am not sure you see as clearly. If you look at it, Logic, in its modern, mathematical sense, is nothing more than form. It is made of syntactical rules. So it can only be used as a set of rules, a programming language to transform statements into other statements. You cannot use it to say things about the things the statements talk about; or even about the statements themselves. Logic is not a language in the sense that it does not say anything about anything. It is just a mechanical device to transform bits of language into other bits of language. This is the idea behind the equation of Logic with Grammar (one of W's famous stunts). No one would be tempted to call grammar a language. Would you ?

So what is language ? W claims that we cannot answer this question precisely because the word applies to a variety of human activities that only share a 'family ressemblance' between them and no hard and fast property. Instead, what there is is a variety of 'language games', with overlapping domains. None of the domains constitutes the whole of language. Everyday language within a given family is one such game. Everyday language within a given company is another one; very similiar but different. This is why we have company-specific jargon. Science is yet another language game or, rather, a cluster of related language games.

Confused ? Well yes, it is difficult. Because it is not the way we are used to think about science and language. Another way to put it (my own, this time) is to say that language is something that happens between people and not between people and objects. We definitely are not 'mapping symbols to empirical measurements'. We are making noises or putting together strings of graphical marks that others may use as clues to imagine what kind of empirical experience we might have had. If I say 'I am seeig a blue object' you are almost certainly going to imagine a different color than the one I am currently seeing. And the point here is not whether I am being precise or not. It is the process you use to decode my message : you imagine a color at random and picture it in your mind because it is not incompatible with the word 'blue'. So there is no direct 'mapping' between the sign and the empirical experience. What there is is a set of reflexes, acquired through training, which make us accept or reject a certain group of empirical experiences or imagined empirical experiences, in the presence of a given sign.

The consequence is that science cannot be defined on the basis of the statements it uses. 'Ted Williams will be up and running in 20 years' is a scientific statement if it is uttered by someone who has the relevant credentials to be considered a scientist. If it is uttered by a snake-oil peddler, it is not a scientific statement even though it is exacly the same statement in both cases. Science will always elude you if you try to define it as a 'body of knowledge' or a set of statements of a certain kind. You can only catch it if you accept to consider it as human activity (a language game) practiced by a recognizable group of human beings with definite boundaries in the social space. In a word, it is a human institution, a tribe. The only way to get a grasp on what science is is sociology (Kuhn), not Logic.

Once again, we cannot have knowledge because we have no access to things in themselves (Kant). But, hopefully, we have institutions whose activity tend to accumulate justifications for certain beliefs. Science, or rather, the scientific establishment, has used a certain method over the past few centuries that has proved extremely efficient at justifying a certain cluster of beliefs. This method might have been also efficent at justifying other beliefs, like those we use as a basis for ethics. The history of the XXth century is mainly the result of a number of experiments at doing just that. I think everyone agrees that these experiments (Nazism, Communism, Logical Positivism, the Murray and CIA mind-control attempts) are failures. The result is that scientists are no longer credible outside science itself, that is outside the domain of expertise within the bounds of which our culture considers them to be reliable.


More than Science?

...science is science, it has its own methods and is amazingly sucessful in its own domain, but it does not provide means to control our emotions.

Nicolas,

Are you saying that there is something outside science that is more successful at describing parts of the world? Specifically, that you believe that there is knowledge outside of mathematics and science?

You would not be alone in this belief. This belief dominates theology, philosophy and is probably accepted by more than 95% of the world's population. It is often said that science has its domain, and the supernatural has another.

Of course, the number of people subscribing to an idea is not a measure of its correctness.

What must one sacrifice to say that there is something more than science and mathematics? One must sacrifice either logic or the assumption that the universe is governed by physical laws.

Some people are happy to make these sacrifices. I am not one of those people. I have given a lot of thought to this. Suppose there are two domains, one that follows physical laws and one that does not. I can't see a way to bridge the two domains with human experience without destroying causality in the scientific domain. Thus, I have not been able to escape the conclusion that either the world is totally causal and accessible to science, or it is not at all causal.

As axiologist points out, emotion is perfectly within the realm of science. We each have statistical models of behavior that we have learned from our relationships with ourselves and with others. We can predict with better-than-even accuracy what the effects of various events will be on the average person. The more we know about someone, the better we can predict their behavior. At a physiological level we can look at blood pressure, blood chemistry, facial expressions, EEG's, brain scans and probably many other observables.

So, propositions about emotion are meaningful to logical positivists. Similarly, propositions about many other things are equally meaningful to the degree that they fall under the domain of science:

  • synthetic ethics

  • game theory and trust

  • language

  • human delusion

Logical positivism is not the bogeyman!

You also say that "science is honest because it is coherent with its own ethos." What you mean to say is that science is okay as long as we don't use it to actually make any decisions about personal or social matters.

But now we need science more than ever. The world is wracked by war and ignorance. Political powers use the irrationality and emotion of the people to gain support for their foolish plans. The people are suckered into supporting policies that are clearly going to hurt them in the future. The people will always be cheated this way as long as they are convinced that there is some supernatural power that will ensure everlasting life and final justice.

Perhaps philosophy was never going to have much impact on the average person. However, if it was to have any effect, philosophy would need to have come to some reasonable consensus. The philosophers' war on positivism has robbed philosophy of any beneficial effect it might have rendered to popular culture. Instead, philosophy's message to the common man is that it's okay to have faith instead of reason.

doctor(logic)

Language is still not supernatural

Nicolas,

In your critique of my logical positivism you argue that we can never truly know what any of our propositions say with infinite precision because we can always claim that a word is not defined well enough (infinite precision) to perform the corresponding falsification test.

I will assume from your critique that you believe my process is equally flawed when applied internally within the mind of a single individual. That is, the flaw exists independent of there being two intellects who might have different interpretations of the same language.

I also assume from your posting that you believe that, while we cannot have infinite precision in meaning, we can have some precision. For if we could not, then language itself would be utterly useless and all science would be impossible.

These assumptions are not an unreasonable starting point. The question is whether language can be developed to arbitrary precision to address the falsification of a specific scientific proposition.

So now we must look at how we gauge the precision of language.


Let's take a little detour and consider a sample proposition:

Heat from a flame is reduced with distance from the flame.

This proposition contains several components: "heat", "flame", "reduced", "with distance".

Informally, we see that we are correlating heat, distance and flame. These three things are patterns of sensation. A flame is a recognized pattern of sensation that incorporates heat, light, sound and possibly smoke. Heat and distance are directly perceptible with our senses. The word "reduced" relates a set of empirical facts together - specifically relating the degree of sensation. The reduction in heat with distance is a mathematical model. In the model, the two degrees of sensation are related to one another in a given context. So we see that this proposition expresses a scientific theory, and it appears to succeed because we are simply mapping symbols to empirical measurements.

Informally, language terms in this proposition are symbols either for empirical measurements and categories or for mathematical relationships.


Can this be made formal? Yes it can. It's called science.

We create a formal language by associating mathematical symbols with the output from sensors, and create mathematical models that relate the different symbols. Imprecision in such a language is exactly the same imprecision we have in scientific theories. The crudity of our sensors is the crudity of our language. As we learn to improve our sensors, our definitions become more refined. As our mathematical models become more complex, our vocabulary of verbs becomes larger.

My claim is that the limits of precision of language in describing the real world are exactly the limits of science. Hence, the ultimate meaning of a proposition is the description of the falsification test (or tests) for the proposition.

Our brains are engines of science at the most basic level. Without this scientific ability, we would have no understanding of our world at all. Our personal scientific ability is what allows us to create language and to understand human subtleties like love, deception, and ethics.

doctor(logic)

P.S. Let's return to your favorite case: 'God exists, he will show up in X years'. Is this scientific? Can this proposition be falsified? If it cannot, it is outside of science and mathematics. Devise a scientific experiment that will falsify your proposition. If you can devise such an experiment, then the proposition is significant (at least in a purely logical sense).

The pleasure of fighting

Dear Axiologist. You correctly noticed there is a fight going on in this blog and suggest we might want to declare a truce. I do not think I want a truce now because I realize I enjoy fighting, provided it is for sports and not an actual war (which it is'nt).

What are we doing here ? we fight while exercising our reason. Is'nt this an instance of the connection between emotions and reason that you point out (rightly, in my view) ? It is quite clear that a large part of our emotional world is related to fighting and we only need to consider our ancestor's way of life to understand why. So let us fight, but only in socially accepted ways that minimize harm and produce as much indirect benefits as possible, like aggressive blogging, sports or business.

Indeed, I am not sure we have a choice. A well known experiment goes like this. You put a rat in a cage where he gets non lethal electric shocks at random intervals. It is properly fed and the environment is in all aspects ok, except that the rat has nothing to to but wait in idelness for the next jolt. As a result, most test subjects become depressed. They developp imunological weaknesses, lose weight, etc. Now put two rats in the same cage and subject them to the same treatment. Very soon, they will start fighting between them. But they will not become depressed. None of the symptoms previously mentioned will appear in most cases. Declare a "truce" (in the form of a wall in the cage between the two rats, for example), and they wil become depressed again. So do not try to prevent me from fighting with my cage mate, since it is probably good for both of us.

I believe there is also a connection with the unabomber. My impression is that he did not have enough opportunities in his life to fight in order to cure himself of what he had suffered. I just recieved 'Harvard and the Unabomber' from Amazon this morning, so I will be able to check if I am right shortly. But, from what I know, Kaczynski went on to pursue an accademic career after he graduated from Harvard. As I know from direct experience, there is little opportunity for a good healthy fight in accademia. There is hostility indeed, but it mostly translates into low intensity warfare and the kind of dirty tricks that are probably not the best thing for mental health. This is in marked contrast with business of sports. Perhaps Kaczynski would not have become the Unabomber if he had done more of these.

How to control our emotions

Axiologist, I agree with you that there is more important business to do than to bicker about completeness, falsifiability and so forth. The problem is that there is no point in moving 'beyond Logical Positivism' until we are not reasonably sure that these questions are not going to come back in their unreconstructed LP form, and throw everything into confusion again.

Many people are nowadays ready to admit that emotions and reason are not separate, and I am among them. But realizing this provides only a very short relief as the question of how we control our emotions comes back with a vengence. As we have just shown, the urge to fight and and other potentially nasty things feature prominently in our emotional world. So we need some way of restraining them. Cartesian dualism belongs to a long tradition in our culture that tries to point to such a way through a posit that there is something called 'reason' that can be stronger than our 'passions'. I agree that this position fails and it is important to say that it does because there nothing more dangerous than the sense of false security that an undetected failed protection device creates. It is partly for this reason that I attack LP. Like cartesian rationalism, LP masquerades as a method to avoid harmful emotional statements. It is therefore important to show its claims to be empty.

Now what about the claims on science ? Everyone wants to have science on its side. Since science has become so sucessful, there are many who are ready to 'courir au secours de la victoire' (run to the help of victory), as we say in french. Quine, like LP, is among those brave people. Indeed, the fight between them really looks like a sordid tussle between the inheritors of a big fortune. Apparently, Quine came out on top and ran away with most of the loot. I find Wittgenstein's position more courageous : science is science, it has its own methods and is amazingly sucessful in its own domain, but it does not provide means to control our emotions. It is just not its busines. Moreover, I believe that one can show (and that is what I tried to do in this blog) that trying to apply science methods (what I call the mistrust game) to that end makes matters worse. Let us just leave science alone and move on. We have to accept to rely on ourselves to find the means to tame the threatening maëlstrom of our emotions.

So the problem remains to find those means or, rather, to chose from the myriad of possibilities that are being peddled to us. The genealogy of ideas you outline (speculative philosophy then natural philosophy then social mathematics then back to speculative philosophy) could be interpreted as the nihilistic account of absurdity going around eternally in vacuous circles, leaving us nothing but despair. Hopefully, I am convinced we can break free of this circle. Tipler has some interesting things to say, I think, but 'religion is part of science' is not among them. 'Science is part of religion' might be a better start. Indeed, if you interpret 'religion' as 'the sum total of what we believe in' then this is exactly how I picture things. Wittgenstein's statement that 'philosophy and science are separate' fits in this view as well : you just need to consider philosophy as another component of this grand total.

But we have not really started yet. If we want to, I think looking more closely at the first term of your genealogy might provide a starting point. You mention 'speculative philosophy' as being composed of religion and metaphysics. But these could, just as well, be seen as irreconcilable opposites. The first few centuries of chistianity were shaped by an internal struggle between rival factions one can more or less identify with them. On the one hand, gnostics (and, later, manicheans) held that it is knowledge that saves (hence their names). On the other, Paulininans, and other currents which eventually coalesced into the orthodox church, contended that only faith can save. Gnostics were direct descendants of the greek philosophical tradition which had been transmitted to them mostly through neo-platonism. From it, they had inherited a devotion for sophia (wisdom, knowledge, from which philo-sophia). Their opponents held for pistis, which is the greek word for faith, but their traditional roots were more jewish than greek. Eventually, pistis won and went on to create christianity in the form that persists to this day. Sophia went underground for several centuries and started to reappear, timidly at first, as scholastic christian theology then more openly and eventually morphed into the countless avatars of modern philosophy, including Logical Positivism. So, when you mention religion and metaphysics, I see the ghosts of the old pistis and sophia.

Pure sophia leads to failure, as is demonstrated by the flaws of cartesian Dualism, Logical Positivism and all the orther isms whose bones get whiter in the sun every day. The conclusion is not very difficult to see : we have to rediscover pistis, faith. Yes, I dare to utter the F-word, and I do so because I have discovered that 'faith' does not necessarily imply 'faith in God' in the sense given to the word 'God' by the three main monotheistic religions. Actually 'God' is a placeholder that has been used to embody faith because it was too difficult to visualize something so abstract without imagining it as some sort of person. In short, God is an hypostasis of faith, but an entirely disposable one, in my view. This is where I find interesting things in Tipler : he proposes another hypostasis for faith (which, by the way, he also confuses with God). To be more precise, I believe it is helpful to equate faith with trust, and this is why I spent so much energy defending it here. You mentioned someone saying that 'anthropology is the secret of religion'. Here we are : trust is what human societies produce. It is more than a necessary condition of their survival, it is their very essence. Let us put faith/trust back to the center of our preocupations because it is where it belongs. Trust is what we do collectively because we are indvidual embodiments of longevity. Our emotions, tamed by a trust producing culture, contribute to our longevity both as individuals and as a collective. Science does not help us in that, or at least not in its present form. So let us keep at bay those who see nails everywhere because they hold a hammer, and use whatever tools we have (philosophy, poetry).

Pistis crushed sophia by excomunication and church-enforced discipline. These means are repellent to us now. But I venture to say that we should probably be grateful to the church to have saved us from gnosticism. Our modern societies probably owe their existence to protestantism more than to anything else, including philosophy (read Weber). Protestantism, though it is anti-church, is unmistakbly christian in that it is squarely pro-faith. And a strong pro-faith current in our culture is what we owe to the choices of the early christian church. We need now to move on beyond God because it is no longer credible (as Doc says, we do not see the motivation any more). But it would be a grave mistake to throw faith overboard in the process. If we were to become gnostics again we would indeed be running in circles.

Thursday, September 23, 2004

Is Logical Positivism a Church or a Method ?

Doc, as I assumed, you knew that I knew. But did you knew that my trap had a double action mechanism ?

Let us come back to your definition of a 'complete' statement.

(D) A proposition is complete if it contains all the necessary assumptions and definitions that will make it a falsifiable model.
What is the purpose of (D) ? Ist it a method for determinig what statements are commplete from those that are not or is it just a metaphor giving a broad idea of what a complete statement is. What I tried in my previous post was to use (D) as a method. Since I noticed (D) contained the phrase 'all [...] definitions' I tried to see if my test statement contained all the definitions required to test it in practice. It obviously does not contain them (it refers to them). So I tried to see if I could make use of your definition of 'completable' to turn it into a complete statement that would meet criterion (D) by replacing the names of notions by their definitions. If you reject this way of using (D), it means you do not consider (D) as a method. And indeed it would be ridiculous to call it such, as my absurd use of it shows.

But if (D) is not a method, in what sense is it still Logical Positivism ? What differentiates LP from other philosophical currents is precisely that it wanted to provide a method to discriminate between valid and invalid (meaningless) statements. Says Ayer :
A complete philosophical elucidation of any language would consist in enumerating the types of sentence significant in that language, and then displaying the relations of equivalence that held between sentences of various types [...] [T]he deduction of relations of equivalence from the rules of language is a purely logical activity; and it is in this logical activity that philosophical analysis consists.
If 'logical activity' is not synonymous with 'method' I really do not see what it is supposed to mean. So Logical Positivism says it is an effort to provide us with a method to tell the 'significant' sentences from those that are not. And this method will take the form of rules of equivalence between statements (what I tried to do with (D)). Carnap genuinely tried to deliver on that promise (Aufbau). And, so far, we have assumed that your theory of 'completeness' is another such attempt.

Now consider what happens if LP cannot, or will not, come forward with such a method but still lays a claim to existence, to be something ? What kind of thing will it be ? When I read your last post and some of your earlier comments, I have the impression that whether a given statement will count as complete or not is always ultimately decided by you. I propose various statements and high priest Doc of the LP church decides whether they will be accepted or rejected. Well you can do that. But this is probably not the way you see things.

Actually, it was not the way LP saw them either. They wanted to reject certain propostions (the title of the first chapter of Ayer's book Language, Truth and Logic reads : 'THE ELIMINATION OF METAPHYSICS') but did not want to use the methods of a church to reach that goal. So they had to say that they would do it by providing a method. Indeed, when it became clear that doing so would be so much more difficult than anticipated that it looked an extremely remote prospect, at best, all honest Logical Positivists retreated. Hence Ayer's admission that it had been 'all wrong'.

Now the question I lay before you, Doc, is this : are you going to be an honest Logical Positivist or not ? If you are, you must provide us with a method fast; or accept that you have no reason to reject statements like 'God exists, he will show up in 101.000.000.000 years' other than because you do not like them or do not believe in them (that is fine with me, I do not either). Failing which, you will have to put up with my saying that you belong to a church, and a dishonest one at that, because it pretends not to be one. There is nothing new here. LP attracted a lot of flak along these lines in the 1950s and 60s and it is what ultimately undid it : no one (especially not a british gentleman like A.J. Ayer) wants to be seen to belong to a dishonest cult.

A last remark : contrast this with science. Science is honest because it is coherent with its own ethos. It never says that a statement is to be believed without providing you with a method to test that statement. 'The mass of the proton at rest is 0.938 +/- 0.001 GeV' is indeed a statement of science in that sense because I can learn what it means from books and courses and test it on my own. But when science cannot provide you with a method, it just stops; it does not say anything. But scientism does. It says many (negative) things about metaphysical statements, religion, traditions, etc. Does it have a method one can use to test these claims on his own ? No.

Language is not supernatural

Nicolas! So you knew that I knew that you were setting a trap. But did you know that I knew that you knew?! ;-)

I did not define "complete" to mean that the proposition must be written in a way that is comprehensible to someone in a completely different language. If I say that "the acceleration of an object is inversely proportional to its mass", a man who knows only Japanese will not understand any of my words or verbs. This does not render my proposition meaningless or untestable. Similarly, a claim that "neutrino mixing is responsible for the deficit of solar neutrinos" is no less complete just because you have not learned how to interpret it. The fact is that you can learn to interpret it and perform the experiments that will falsify the proposition.

From my perspective, your counterclaim is trivial. It basically says that in order to understand a proposition it must be expressed in a language you have learned.

I have a problem with the linguistic school (Wittgenstein et al). They mistake language for something more fundamental than a protocol.

First, I would answer by saying that perception of the world does not require language. Other species (and a few individual humans) have demonstrated the ability to perceive the world and even use tools without using language at all.

Second, scientific progress in neuroscience, language and information theory have made the linguistic school obsolete. If language was fundamentally linked to perception and understanding of the world, it would be invulnerable to scientific explanation. However, it is straightforward to see how language works from a scientific perspective. With science, we can gauge the limitations of our language.

There's a good analogy here. Science accepts that a universe that is subject to physical laws is perfectly compatible with the fact that experimental measurements have finite precision. Similarly, we can say that conceptual representation of the world (meaning) is perfectly compatible with language that has finite precision. In both cases, we can make propositions about the world with an arbitrarily high degree of precision, and falsify those propositions to a correspondingly high degree of precision.

Language is simply a technology for communication. It is not beyond science, and it is not metaphysical.

doctor(logic)

Wednesday, September 22, 2004

Complete ? Really ?

Forgive me Doc, but, as you no doubt have guessed, my question was a rethorical trick. And it seems you fell into the trap.

Let us review your example of a complete statement :

(S) The mass of the proton at rest is 0.938 +/- 0.001 GeV
in light of your definition of 'complete' :
(D) A proposition is complete if it contains all the necessary assumptions and definitions that will make it a falsifiable model.
Let us suppose (S) indeed contains all the necessary assumptions. It certainly cannot be said to contain all the relevant definitions. What is a proton ? What is a GeV ? What does 'at rest' mean (for a proton) ?. These are more than rethorical questions, this time. As far as I am concerned, I roughly know what a proton is supposed to be and that a GeV is a Giga electron Volt. But I must confess that I do not remember very well what a Giga electron Volt is, and especially how to measure it. The term 'at rest' is probably even trickier since, if I am not mistaken, protons are never at rest in the kind of experiments we can do. So the 'mass at rest' is not something you measure directly but deduce (how ?) from measurements under different circumstances (which ones exactly ?). So (S) is definitely not complete according to (D).

Now is it 'completable' ? You can concievably transform (S) into another statement (S1) where 'proton', 'GeV' and 'at rest' have been replaced by short definitions (several lines long) of what these notions mean. But these definitions will inevitably contain other words like 'spin' or 'quark' or 'particle accelerator' or 'magnetic field' that will, in turn, require definition. Repeating the expansion process will yield a third statement (S2) where these words will have been similarly replaced by definitions. But why should (S2) be more complete than (S1) ? There is, in fact, every reason to believe (if you don't, just try) that (S2), being longer, will contain even more terms requiring definition. So from (S2) we will derive (S3), then (S4), etc. Now, is this series of statements going to stop at some point ? Logical Positivists believed it could and this is why R. Carnap (Der Logische Aufbau der Welt) or Otto Neurath tried to build very detailed systems to decompose any statement into a set of sense experience data readouts + predicitive structure. This project is generally considered to have failed and, indeed, has been abandoned. The coup de grâce seems to have come from W.V. Quine who said :

our statements about the external world face the tribunal of sense experience not individually but only as a corporate body.
and

The unit of empirical significance is the whole of science.

which means that Quine (and everyone else nowadays) believes that the sequence (S), (S1), ... (Sn),... cannot be meaningfully stopped until it contains 'the whole of science'. Now, 'the whole of science' is not a statement because it is not even precisely defined (what counts as science ? must we include definitions of ordinary words as well ?) So, none of the statements (S), (S1), ... (Sn), ... is a complete statement and what could be considered complete ('the whole of science') is not a statement. Ergo, there are no complete statements, since the above argument can be repeated for any starting statement (S).

Honestly, Doc, do you still believe your defintions of 'complete' and 'completable' are taking us anywhere ?

Completion

Here are some examples of complete propositions.

Scientific propositions are the easiest:

The mass of the proton at rest is 0.938 +/- 0.001 GeV.

Two objects with different masses will accelerate at the same rate in a uniform gravitational field in which external forces are negligible. Link


In these scientific examples, an experimentalist knows enough to devise an experiment that will falsify the proposition.

Let's move on to less scientific propositions.

I paid my taxes for fiscal year 2003.

Volkswagen makes the cheapest car in each category of automobile, with categories defined as groups of vehicles with comparable performance and luxury appointments.

I think that tartar sauce smells repellent.


In each case, we can test the proposition and potentially prove it false. The last proposition is the most difficult to argue as complete, but I think that there are very few people who could lie on this subject and still evade all of the potential falsification tests. To the extent that it is an empirical proposition about personal feelings, it may not need to be falsifiable. I'll discuss this more below.

The following natural language statements are incomplete, but are completable (--> indicates completion):

Unemployment is bad. --> Unemployment in contemporary American society is bad for economic growth, tends to increase crime and depresses morale.

Tartar sauce smells repellent. --> I think that tartar sauce smells repellent.

There is no bread at the supermarket. --> There were no salable packages of bread in edible condition on the bread shelves at the supermarket when I was there 15 minutes ago.


Again, an incomplete statement is completable if we can agree that it is equivalent to a complete statement.

Let's now look at statements that are generally intended as non-completable:

God exists.

God is good.

God created the universe.

Killing is absolutely wrong.


Each of these statements suffers from the problem that the person who speaks these things generally does not intend them to be equivalent to falsifiable propositions.

For example, most people would not be willing to translate "God is good" as "I think of my wife is a goddess, and she makes me feel good," or translate "God exists" as "There is an extraterrestrial life form that possesses technology vastly superior to our own."

Most statements that are "incomplete by intention" are the result of confusion. For example, they may be attempts to explain the unknown ("the universe") in terms of the more unknown ("god").

Empirical facts are not themselves falsifiable. If my photodetector records two events between 1pm and 2pm, this is an empirical fact: "Photodetector #12 detected two events between 00:13:00 GMT and 00:14:00 GMT with the time measured by clock #27". Note that this does not necessarily mean that there were no other photons passing through detector #12 during that same time period. Other detectors may see the events differently. Sometimes, you only have one detector, and it may give you an inaccurate result.

The scientific view of the universe as a machine that follows physical laws is perfectly consistent with imperfect measurement.

We humans are also detectors. If you see something that no one else could have seen, that empirical fact may not be falsifiable. However, this does not mean that your measurements correctly reflect physical events. You may have imagined the events. Furthermore, humans are more than just detectors. Humans are theorizers as well as detectors. To make a measurement directly, a human must fit the observation into his or her neural model of reality. This means we may express a measurement in terms that are not reliable or consistent. We are also gamers, so we can lie about what we observe (consciously or subconsciously).

Empirical propositions are readouts from scientific equipment, or expressions of sense observations by humans:

I saw a large bird of prey soaring over my home town this afternoon.

The electrical potential across the terminals of this battery reads 5.25 volts.


Due to known flaws in human perception, we must rely on methods that minimize these flaws. Double-blind experiments are one such method.

doctor(logic)

The Metaphysics of Scientism

Doc, I do not say scientists pursue metaphysical goals. I say scientism does. And I am not alone.

Look at your definition of knowlege. Its point No. 2 reads "Empirical facts.". Presumably, these empirical facts must be expressed as completable statements, which, according to your definition of 'completable', can then be decompressed into a complete equivalent statement. Now, let us put that to the test : can you show us a complete statement ?

Tuesday, September 21, 2004

More Definitions

Nicolas, I think we need to focus on definitions again.

First, let's return to the ethics. The scientific view is that ethics is a set of rules that people abide by in order to live together harmoniously.

In theory, we might create a new society by devising specific goals for that society, and then dictating the corresponding ethical rules to all of its constituent persons.

If humans were programmable computers, I might be in favor of this sort of purely synthetic architecture. In the real world, people are not programmable, and we have to share a very small planet. It's virtually impossible to create a new society that is independent (non-interacting) with others. In effect, the world is one big society, and the constituents can't agree on what the goals of the society should be. In fact, most people wouldn't say there were any goals at all.

At this point, we want a system of ethics that enables us to thoughtfully alter our society's goals. When people are fighting for survival, they cannot be thoughtful. This means ethics that respect human rights, improve our standards of living, promote education, and allow society to survive long enough to avoid extinction.

Knowledge

According to LP (in my formulation), the understandable world follows laws of physics. Science is the way to determine these laws. Once a scientific model is found to work in a given domain, it always works. Newton's laws are still used to build cars and bridges, despite being less applicable than Einstein's theories of relativity.

What is knowledge? Knowledge consists of three things:


  1. Analytic facts that are derived from axioms using logic.

  2. Empirical facts.

  3. Scientific theories that are shown to predict accurate experimental outcomes over a certain domain and with a certain level of precision.



Belief

What is belief? Belief is a personal decision to accept knowledge as adequate for decision-making. I might believe that I will get my car back after I lend it to my brother. We believe flying is safe because a lot of people do it and we trust the FAA to police the air transportation industry. If we have participated in experiments ourselves, we can believe the theories we verify. This definition of belief meets your requirement that verification be possible.

Trust

As you say, trust in another person is a measure of the expected behavior of that person. Trust is never total. For example, I may say that my trust in John Kerry is 98% when it comes to nuclear non-proliferation. That means I estimate a 2% chance that he won't help and may hurt. Depending on what is at stake, I might want better odds. That's when I need verification. When the stakes are high we demand verification. That's why we have independent organizations as watchdogs.

Can I afford trust my friend to pay for his share of our restaurant tab? Sure. I can afford to lose money once or twice before I no longer trust him. At an interpersonal level, we frequently neither want nor require verification.

One more comment about this. There is a cost to verification. We may choose to trust something or someone because we deem the cost of verification too high AND the cost of bypassing the trusted party too high. I may not trust the quality of the food at the supermarket. However, I don't have the resources to test the food for toxins (verify), and I don't have time to grow my own organic food (bypass). Instead, I opt to trust the supermarket food.

You argue that verification is detrimental to trust. Do you mean that verified action requires no trust? For example, do you mean that if I never use a credit card, I never build a credit rating?

Is this bad because, if the verification mechanism breaks down, there is no trust to back it up?

If so, does this hold for any aspect of our society that depends on modern conveniences? Should we all have vegetable gardens in the event that the normal food supply fails?


Completeness

Completeness is the decompression of a natural language statement into a scientific one. It's really that simple. If a statement is not completable, then we cannot agree on an equivalent scientific statement. If a person cannot agree that their proposition is equivalent to some scientific proposition, then their proposition is poetry. This is not metaphysical. It is an acknowledgement that the proposition has nothing to do with experience (empiricism).

Scientists are not "pursuing metaphysical goals". They assume that it is possible to learn about the universe, then follow the only course of action consistent with the fact that the universe can be understood. There is simply no way to understand the world otherwise.

Miracles

This posting is brought to you by the miracle of the transistor!!!

If miracles are natural (i.e., not supernatural), then yes, it is part of physics. If aliens land on Earth and use some technology to bring a dead person back to life, hey, that's a miracle.

Unfortunately, miracle has another meaning. People use it to mean that the laws of physics are suspended by a supernatural force. According to LP, propositions about the supernatural are meaningless.

Did the Jesus of the New Testament walk on water? If he did, then he had some serious technology. To the extent that we can talk about this, we are talking about feats of technology, not the supernatural.

Now, religionists will insist that their stories are not about technology. After all, who wants to bow down before a technological wizard, even if he did create this particular universe?

[I'm off on a tangent here, but why would a being so powerful want to force us to acknowledge that he is superior when it's so blatantly obvious? That would be like me forcing a two-year old to acknowledge that he can't beat me at tennis. I just don't understand any of the supposed motivations here.]

The Unabomber

I'm sorry, but I don't think that LP has anything to do with this. As you say, it is social maladjustment that is the cause of the problem. A person such as the Unabomber will use any justification that is at hand. If a man watches a surgical procedure on TV and is inspired to perform illegal surgeries on unwilling victims, is surgery to blame? I don't see it.

doctor(logic)

Metaphysical poison or cure.

Axiologist, I see you are puzzled by my previous post. Well, I was just trying to make things clearer and I still hope we may achieve this goal. I know It is difficult to let go of 'knowledge' in the sense of 'knowledge about the world' that has become so familiar and reassuring (you talk of permanence...). But I think it is worth doing so since it is the cause of so much confusion. I would be interested to know what you find about that in the works of Quine. Many people contrast him with Wittgenstein so it is certainly well worth trying to confront them once more.

Your allusion to christian miracle myths (Christ walking on water) is an additional occasion to attempt to make myself clear. In my view, a belief in miracles is just like the beliefs we derive from physical theories, such as 'the big and the small sphere will reach the ground simultaneously'. Both are of the same nature. They are beliefs about the world and as such may turn out to correctly predict events we may experience. The big difference is the degree of justification they have in our eyes. In the case of miracles, it is very thin. For Galileo's theory of free fall, it is very strong (though not perfect, as we have seen previously). What seem clear to me is that none of them belong to the realm of metaphysics. In a sense, we may say that beliefs in miracles are part of physics. They are just a branch of physics whose credibility rating has fallen quite low of late. This is another instance of the advantage of doing away with 'knowledge' in the traditional sense. If you accept all this, you no longer need to trouble your mind with the question whether 'metaphysics [can] become physics', which is really an unnecesary and artificial problem. As Wittgenstein would put it, this is another case of 'bewitchment of our mind by means of language'.

The other cases you mention can be dealt with in the same manner. Of course, the smallpox vaccines is BETTER than the smallpox gods. And I like the BETTER very much. Indeed, the beliefs behind the vaccines have a much BETTER justification than the beliefs in the smallpox gods. Jenner's belief in his vaccine became better justified when he had tested it and shown it to work. Our beliefs in Jupiter's anger at the sight of a lightning becomes less justified (to the point of fainting completely) when we start to adopt beliefs in electromagnetism and justify them with a lot of empirical evidence, etc.

Metaphysics, in my view, is something else. It is also a category of beliefs, or ontological commitments if you wish, but in abstract entities. The Christ of the miracle myths is not abstract, since he can walk on water. Indeed, the Christ is supposed to be God embodied as a human being; a very concrete creature. He is therefore a piece of physics; religious physics but physics nonetheless. By contrast, platonist ideas, Leibnitzian monads, Whitehead's events or, in my view, knowledge in the scientistic sense, are unmistakably metaphysical. Beliefs in abstract entities are a very powerful medicine because they act largely unseen (abstract = invisible) but can impact all of our decisions (is it good or bad ?, should I check ?). As you know, a powerful medicine can be a blessing or a curse.

Now the question of LP and the Unabomber. Obviously, there is no necessary relationship between the two. The way I see it is more like the relationship between tobacco and Lung cancer. Something that, coupled with a predisposed terrain, and some additional stress, will greatly increase the likelihood of the dreaded outcome. However, I do not think that the importance of stressful events, like the psychological experiments, should be overestimated. Everyone's life includes some kind of stress, sometimes far worse than that confronted by the Unabomber (think of concentration camp inmates or torture victims). A sound cultural backround is what normally helps you overcome stressful events without loosing your balance. Very few concentration camp victims became terrorist madmen. What I think LP did to the Unabomber was that it dissolved whatever background he had (probably not very strong) and replaced it with a perverse 'mistrust everyone' nightmare. He was then naked in front of stress, probably suffered from it inordinately, and therefore overreacted. Those who escaped that fate either had a stronger background to begin with, or were helped along the way by others or, like Wittgenstein, managed to reinvent everything by themselves. In most cases, it is probably a combination of the three, in varying proportions.

Ethics, beliefs and knowledge II

Doc, reading your last comment, I feel the need to ask you for some clarifications about your position on ethics. In a comment, some time ago, you said :

(A) Ethics is also a technology for utility. [...] Given a set of goals, we can synthetically construct a system of ethics from our scientific knowledge of human behavior.
Then, more recently :

(B) We can study what laws will make society stable, prosperous, and accomodating to individual pleasure and self-actualization. We can call this "ethics" if we want...
But, in your latest comment, your position seems slightly different :

(C) You talk about two alternatives: [(1)] working from the current status quo, and [(2)]using a "scientific" approach to reinvent ethics from scratch. [...] I don't see approach (2) as being the "scientific" one. It is unrealistic and idealistic...
Also, in the comment from which (A) came, you said :

(D) Ethics are not absolute, and cannot be derived purely from theory.
I think these remarks are very interesting because they faithfully represent a fairly widespread view of ethics, with recognizable roots in LP and other forms of scientism. (A) and (B) represent the original, full fledged, scientistic view of ethics. In the mid XIXth century, thinkers like Marx, already had similar views. I think we can summarize it as the pre-WWII scientistic take on ethics. Indeed, Nazism, Communism and other recent nasty discoveries seem to have forced people to be more sober and prudent. Hence (C) and (D), which are, nowadays, the only formulation mainstream thinkers allow themselves in public.

Now, correct me if I am mistaken, but it seems to me that (A) and (C) are contradictory, since to 'synthetically construct a system of ethics' looks pretty close to 'reinvent ethics from scratch'. My interpretation of this contradiction (if this is indeed one) is that, in your case, (A) and (B) remain what you think at heart, even when you say (C) or (D). An additional indication of this can probably be found in the fact that, just before (C), you say : "Even if we decided to radically overhaul ethics...". So being 'radical' is not something you completely rule out, or do you ?

In any case, even if full blown tabula-rasa is indeed ruled out, as it is, most of the time, nowadays, I think using a scientistic approach to ethics is harmful nonetheless, even if you take the status quo as a starting point and take a gradualist stand. This takes us back to the title of this post. The scientist strand in our culture rests on making a strong distinction between knowledge and beliefs. I am going to try to expose first why this is detrimental to ethics and seconds why I think it is just not helpful to make such a distinction on more general grounds.

Within what I choose to call the 'scientistic tradition' (others would call it otherwise), knowledge is presented as something you can rely on because it has been empirically verified. Scientific knowledge is thought of as the prototype and, sometimes, as the only form of knowledge. By contrast, beliefs are what is not knowledge and hence, the word takes on an automatic negative connotation : mere beliefs, or worse. What consequences does this has in practical situations ? When presented with an assertorical statement, a person under the influence of this view will ask himself : 'Is this a belief or a piece of knowledge ?', and then 'how can I know ? I must check'. In a very natural way, the next step will be that he will start to verify, or, at least attempt to do so. He will behave like a scientist faced with a new theory. Now we may ask a question : is what is obviously good in the lab also good in other situations ? I believe there is no reasons to believe so and that it is the root of the scientistic mistake : to make the implicit assumption that what has proved sucessful in science can be equally profitable in all areas of human life.

Why is this detrimental to ethics ? Ethics is a set of behaviour rules. In practice, they often take the form of 'tit for tat' statements like 'if you follow the dress code, people will be nice to you' or 'if you report a gang member to the police, your whole family will be murdered'. All these rules rely on an anticipation of the future behaviour of others. In the absence of trust, this anticipation cannot work. Indeed, one of the main use of the word 'trust', in our language, is to describe just such anticipations. But as we have already amply shown, verification is detrimental to trust. The ability to verify is good for trust, provided it is kept unused most of the time. Frequent actual verifications have the opposite effect. Ergo, excessifve verification is detrimental to Ethics.

So, where are we now ? We have shown sucessively that :
  1. To oppose knowledge and beliefs prompts people to verify the assertorical statements they are presented with.
  2. Frequent verification of statements kills trust.
  3. Trust is necessary for behaviour rules (Ethics) to function.

Hence, I believe we have shown that opposing beliefs and knowledge is bad for ethics. Doing it a lot (statements (A) and (B) above) is very bad. Doing it a little (gradualist, post WWII, approach) is less bad, but bad nontheless. It cannot overthrow civilization as a whole but it can still produce the likes of the Unabomber.

Now, let us turn to the second argument announced above : what can we think, in general, about the idea of separating knowledge from beliefs ? First, as you pointed out, Doc, there is a powerful urge in us to go for knowledge because we crave for security (survival, avoidance of pain, etc.). Knowledge has an air of certainity that seems to fill that need. Secondly, When we enquire about how to define knowledge, the answer that comes back most of the time is 'justified true belief'. There seem to be a more or less complete consensus among analytic philosophers about this definition. Very well, but what dose true means in the above statement ? One can try to appeal to logic but (you know my opinion about this) this is open to criticism. Another way, that you have tried yourself, is to appeal to the notion of 'complete' statement. Your own theory of 'completable' statements is basically that : it accepts as valid those statements that can be transform into a complete one in a finte number of steps. As I discovered a few days ago, other have already tried that route :

the final problem is to conceive a complete fact...

This is a citation of A.N. Whitehead I found in an article of Charles Hartshorne, one of the leading thinkers of process theology and a vocal advocate of metaphysics. The aim of this citation is to describe the goal Whitehead set himself, in his later (post Russelian) philosophical period, and in the pursuit of which he designed his system of events and process. This system is undeniably metaphysical both in aim and in methods and, indeed, Hartshorn commands Whitehead as 'the greatest metaphysican since Leibnitz'. Now, I do not mean to say it is bad to do metaphysics, though I must say I am none too convinced by Whitehead's system. What this just shows, in my view, is that trying to define 'complete' facts, or statements, is a metaphysical pursuit.

So we were led from knowledge to true statements, then to 'complete' statements (through your definition of 'completable'), which happen to be metaphysical objects.

Scientism starts by separating knowledge from belief and treat metaphysics as mere beliefs that should be abandoned for knowledge (Axiologist's Marx quotation). Then we discover that the scientists core pursuit is itself metaphysical in nature ! If you go back to the first point above, you even realize that it is a metaphysical pursuit we want to believe in because it helps us feel secure. In other words, it is the opium of the people.

Faced with this, we can choose to describe all of it as a 'difficulty', to be explained away, or we may choose to realize that the very notion we have of knowledge is defective : it is a source of problems and of no real benefits. The solution I propose is to drop the word 'true' from 'justified true beliefs'. My view is that the only things we have are beliefs, some of which are more justified than others. Scientific theories are extermely well justified beliefs and have great value as such. But ethics supporting beliefs are also valuable as we just cannot live together without them. The scientific method is inappropriate for justifying such beliefs for the reasons we have said. Metaphysical beliefs are unavoidable because imagining metaphysical entities is the only way we have to formulate the justification of ethical rules. The tricky bit is to choose the right kind of metaphysics. As I said, my view is that the scientistic picture of knowlege is bad metaphysics (dishonest metaphysics, for that matter) and, thus, should be rejected.

This does not mean that I find no use of the word 'knowledge' legitimate. For example, I have no objection to 'I know this play by heart' or 'I know the proof of the theorem X'. In these cases, what is known are pieces of purely formal content, not statements about the world. While you stay within the bounds of the a-priori, it is perfectly reasonable to say that one knows or does not know something. However, there are also, in everyday usage, statements like 'I know so-and-so' or 'I know he will come tomorrow'. I believe the trouble started from those statements, which express intimate conviction more than certainity, when people, impressed by the achievements of science, tried to find a way to obtain statements about the world that would be as certain as those we can have about a-priori content. Philosophers like Kant or Wittgenstein have tried to cure us of this mistake ever since. With mixed success.


Monday, September 20, 2004

Ethics, beliefs and knowledge I

Doc, I think you are doing just what is needed when you attempt to define ethics. Coming to an agreement often requires no more than clarifying what we mean when we use certain loaded but vague words. Especially so in philosophy.

When I read you, I realize we give very different meanings to the word 'ethics'. You say it is about 'good' versus 'bad'. I would take a different starting point and say it is about all the rules of behaviour that we learn from others, as opposed to those that we form ourselves through experience or those that are dictated to us by our instincts. Of course, there is a link with 'good' versus 'bad' because, when one wants to express a behaviour rule, he will often say 'behaviour X is good (or bad)'. Also, when a rule needs support, for example because it is being criticized, people will often try to justify it by showing it promotes behaviour which produce 'good' outcomes. However, as we will shortly see, there are cases when behaviour rules and 'goodness' are not so obviously linked.

I think it is helpful to define ethics as learnt behaviour rules because it clarifies certain problems. First, there is the case of moral relativism. For a criminal gang member, it is a serious offense to report a crime committed by another gang member, even if he is from an enemy gang. Is it good or bad ? Well first we can say that the gang member's ethical system includes a rule to that effect, while others (like ours) do not. If 'ethical' is defined as above, everyone will agree. Then we may go on debating whether that rule is good or bad, but as a separate, second issue. Another reason for liking such a definition of 'ethics' is the realization that our definition of 'good' and 'bad' is very specific to our culture, while having behaviour rules is not since all cultures have some. For example, Indic cultures have a notion of 'Dharma' that may vary according to caste or age. The result is that these cultures prescribe many behaviour rules to their members while it is debatable whether they have anything that corresponds to our notion of 'good'.

As far as I am concerned, I agree with you that trying to define 'good' once and for all is not going to take us very far. However, I am pretty sure that we need behaviour rules. First, it is quite obvious that when established rules start to fail, the short term outcome is invariably something we do not like : violence, abuse of the weak by the strong, poverty and so on. I agree that this is like smell. What I say is that no one likes the smell of chaos when he is unfortunate enough to actually experience it (as opposed to closet revolutionaries like the unabomber). A second reason is that I believe learned behaviour rules do work. Look, for example, at the behaviour of the criminal described above. Empathy, risk evaluation and interest do play a part in people's decisions. But so does the anticipation of other people's judgement. And this judgement largely rests, in practice, on whether one has followed the accepted norms of behaviour or not. Is this good or bad ? Is this conformism ? This is not what I am trying to determine. I just observe that it is the way it works.

To make a long story short, I believe that 'ethics', in the sense of learned rules, do influence our behaviour because it is chiefly how we get praise or blame, and that it helps protect us from chaos. So we now have a case for 'ethics' which is compelling enough, while making no appeal to any notion of 'good', in the traditional way. However, once we have agreed on this, the hardest work is still ahead of us; that of choosing among the rules that are presented to us (and, in our multi-cultural world, there are many) those we want to follow and those we want to reject. I think that we are agreed that there are roughly two methods to go about this :

  • We may choose to rely on tradition; see what rules our forebears did follow and try to adapt them incrementally and marginally according to our intuition, experience and judgement.
  • Or we may try a more radical approach. We may attempt to establish scientific laws of human behaviour and then frame a set of rules based on this theory.

Our difference lies in the fact that you espouse the second view while I definitely choose the first. I agree that the second option looks attractive. As I said earlier, I would like, as anyone would, to have nice clean ethical building blocks that I could rearrange according to my needs, and those of my comunity, without the fear of nasty side effects. Compared with this rosy view, tradition looks arbitrary and baroque indeed. However, as I also said earlier, we should resist the attraction such an optimistic view of ethics has on us because experience has tought us to be careful. My reasons for saying so are exposed at length in my "Trust vs. verification" post. I will restate them shortly here. The scientific attitude crucially depends on the scientists playing the mistrust game because it is the only way science can be empirically grounded. But this makes this approach unfit for dealing with ethics because :

  1. of the importance of trust to ethics.
  2. of the fact that, in the case of human behaviour, it is impossible to separate the scientist from its experiment.

2. entails that what scientists do and say has an impact on their experiment, which is the whole of human life if they deal with human behaviour (you cannot hold humans in a lab to prevent the experiments you have done on them to affect the rest of humanity). Since scientists play the mistrust game between them, they are bound to propagate mistrust if they study human behaviour. Because of 1., this is bound to be detrimental to ethics, as examples like the unabomber seem to show.

Therefore, since the scientific approach to ethics seems to fail, we are left with the tradition based approach, which, when you look at it in this new light is not that bad.

When I started that post, I wanted to deal with beliefs and knowlege since I think their opposition sheds some light on ethics. In particular, it is another way to show why trust is so central to ethics. Since this is already quite long, I will leave that for a later post.

Sunday, September 19, 2004

Why do we need ethics?

I can see Nicolas X shiver as he reads the title of this posting!

So far, this conversation has been very productive for me.

I'm quite pleased with my formulation of logical positivism, and I'm writing up a more formal description of my ideas that I will share with you. I feel that rejecting LP will require one to reject some very fundamental things (e.g., objective reality, logic, etc.).

Similarly, I was hoping that I might come up with some ethical system that relies on a principle so fundamental, that denying the ethical system would be absurd.

Instead, I have come to have a better understanding of ethical questions and why they may not matter (in the classical sense) at all. This is not to say that one's decisions do not matter. I just feel that ideas about "good" and "bad" are biological illusions.

The Idealists, et al

First, to your philosophers. In particular, Bradley and Royce are trying to solve a problem that doesn't exist.

Nicolas, I know that we disagree on this, but I don't see any contradictions in the scientific world at all. Not one. So when these philosophers say that thought (or ultimate intellect) unifies the world so that the contradictions disappear, I'm not even convinced of their motivation, let alone their argument.

Here is an excerpt of a summary about what Royce had to say about thought:

According to the correspondence theory of knowledge an idea (or judgment) is true if it correctly represents its object; error obtains when an idea does not correctly represent its object. It is indisputable that finite minds do sometimes entertain erroneous ideas. Royce pointed out that in such a case the mind must contain an (erroneous) idea and its (false) object, while simultaneously intending, or "pointing toward," the idea's true object. If the mind is able to intend the true object then that object is somehow available to the mind.


Of course, this is very silly. The first error here is that it is not obvious that the mind understands what it is intending, so the argument fails before science even reaches it. Today, we can discount this entire class of argument purely on scientific grounds. The brain is a large neural network learning about its surroundings. We are certain of the physical basis for our intelligence (we are biological machines). The brain doesn't have foreknowledge physical reality through "intention". It learns about physical reality through our five senses.

Beyond these mistakes, the idealists did what we have been doing up until now. They tried to reverse engineer "good" to find out what first principles would imply the "good". They were handicapped by their lack of scientific insight.

However, maybe we too have been wrong about this. Perhaps it is a mistake to assume that defining "good" is a philosophical question at all.

Ethics

What exactly are ethics? Ethical behavior is behavior that is "good". Suppose for now, I accept that this is meaningful. What is the point?

This blog approaches the question of ethics as if its answer has something to do with how people will behave, e.g., "if only the Unabomber had not been exposed to LP...". I don't think this is reasonable.

First of all, about 90+% of the planet's population doesn't study philosophy. To the extent that they encounter philosophical ideas, those ideas are dictated to them by religious leaders. These leaders are not going to be influenced by the likes of us. They are either deeply delusional or would find our approach to philosophy unprofitable (financially and as a source of political power).

Secondly, even if we were to convince people to learn and understand our philosophy, they would still not behave completely ethically. Again, we come back to game theory and psychology. We can say what kinds of behavior are good, but if an individual can gain from deviating from socially good behavior, they will. The vast majority of people never make a decision based on ethics. They make decisions based on habit, empathy, risk and personal gain.

I do not accept that the traditional "good" has a clear definition stemming from reason alone. "Good" is basically a region of the brain that is activated by thoughts and experiences that produce pleasure and avoid pain. For example, we might find it "good" to forego a pleasure to avoid some future pain (e.g., deciding to study before partying, or deciding not to sleep with the boss's wife). The concept of the "good" depends on our life experiences, our environment, and our DNA. It is no more fundamental than any other trained organ of the body. One might say that our sense of goodness is similar to our sense of smell. Ethics, as a fundamental study of what is good, is no more sensible that a fundamental study of what smells bad.

So, in my view, we can see that any system that rigidly defines "good", is comparable to a system that rigidly defines what smells nice. I may find smoked kippers delicious, but others find them repellent. To say that I am a social deviant because I like Kipper Snacks would be absurd. I am a social deviant for completely different reasons! :)

So what is good? That which will lead to optimum survivability of Earthly intelligence? A chemical means to keep us happy all the time? Who is to say which is more "good"?

We can study what laws will make society stable, prosperous, and accomodating to individual pleasure and self-actualization. We can call this "ethics" if we want, but it is based on a goal framed in a scientific way, not on an absolute idea of the "good".

doctor(logic)

Saturday, September 18, 2004

What about these ?

Searching for possible sources of synthetic-aprioritical inspiration, I found two areas of philosophy which are boradly contemporary with LP and have strong links (of opposition) to it.

The first one is the neo-Hegelianism of F.H. Bradley and the other is the so called process philosophy of A.N. Witehead. The first one was apparently the dominant school in english-speaking philosophy before Russell and the "linguistic turn" of the early Wittgenstein and the Vienna Circle, which were largely a rebellious reaction against it. The second has had little impact in philosophy itself but has fueled an important current of modern theology (so called process theology). F. Tipler's Omega Point Theory makes numerous references to it.

I find none of them satisfactory, though I feel they both contain interesting ideas (in particular the second one). What do you think ?

Thursday, September 16, 2004

Questions of Survival

I'm not yet ready to accept an a priori ethical system, but I am intrigued by axiologist's comments about game theory, survival and immortality.

According to Plato's Symposium, all lusts stem from the will to eternity and immortality through creation of things, and even the begetting of children, as this is the only victory over death.

Is this anything more than a statement about human nature? Perhaps it is just the survival instinct manifesting itself on our higher personality.

Nature is game theory on a grand scale. Evolution tosses aside entire species just because new species are better competitors. In nature, survival is the only virtue.

On the other hand, human brains are very complex. For most of us, survival is not our only consideration. We take risks every day, and most of us can conceive of a "fate worse than death".

Can we say that it is always better to survive than to become extinct? Is this true only for our entire species, or for individuals as well? What will happen when we invent technology that makes an individual immortal?

Does any of this have any bearing on ethics, or have we just reviewed game theory?

Game theory tells us how to win, but not whether we should try.

doctor(logic)

Wednesday, September 15, 2004

The end of this Blog ?

Axiologist, Doc, I want to thank you for your participation to this Blog. Your input contributed to shape a debate on a subject which was, from my own perspective, quite uncertain at the start and is now much clearer. Of course, nothing is settled. But, at least, we now know where we stand and each of the individual ingredients of the debate has gained in precision by a large amount.

Doc, I do not lump you with nazis. At least not with the nazis you commonly see in movies, cold, sadistic, black uniform, and all. What I had in mind are all the vast numbers of very nice people who were nazi party members in the early 1930s because they thought Hitler was genuinely doing the right thing to take Germany out of the Great Depression and away from civil war (in a way, he did), or the equally large numbers ov very nice palestinians living in the occupied territories now and who are members of Hamas because it happens to be the only organization able to run any kind of public services there. One may be genuinely certain to be doing the right thing in that sort of circumstances, and this is a frightening thought.

Behind Logical Positivism I sensed we could touch on something deeper and here we are. We have a glimpse of the familar tragedy of human life. We need ethics and to have ethics we need beliefs. But we are blind as to where our beliefs lead us. If only you were right, Doc and it was possible to manufacture clean, emission-free, beliefs.

I think I have done my job and clarified the question I wanted to clarify. Now we could go on and try to create new beliefs. If possible, uplifting beliefs. But this is not what I feel like doing now. If one of you (Axiologist ?) wants to do it, I would be happy to contribute.

How to Choose a Creed ?

Doc, thanks for your last comment (Tuesday 8:54 pm). You ask the right questions and I think you put them the way they should be.

Now how to answer them ? I may indeed come up some day with an "synthetic a-priori" or "ethical system" or, as I would prefer to call it, a belief system. But this is not essential here as many other people can do it and there are many other such systems around already (See for example the God-u-like site, which, by the way, includes an entry on Logical Positivism).

What matters is how you choose one, and, in a way even more vital to one's own survival and suffering avoidance, how do you notice and reject those like militant Islam, Nazism or (in my view) Logical Positivism, which are really toxic. Making the right choice is important because, as I said earlier, it is impossible not to have a belief system (more on this later if you like). It is also fiendishly difficult because, unless you have a massive event like WWII, the link between a faith and its impact may be hard to pinpoint.

Because a belief system controls you in large part, and not the other way around, it cannot be deliberately manufactured nor fully analysed. Otherwise, it would mean you are in control, and the thing under review would certainly not be your true belief system. So no one can construct a belief system. Beliefs emerge in a continuous and largely random process and eventually combine into systems of varying degrees of tightness. People who "come up" with fully explicit systems rarely invent anything. They are assemblers and rarely manage to encompass really complete systems. If you want to see them you have to take a larger view through history, litterature, etc.

So I am not going to construct anything, but yes, if I do come up with something, it will be according to my own feelings and preferences because what else do I have ? Your own feelings and preferences are what drive you to like Logical Positivism.

So, what I say is that belief systems are given, they are not built (no engineering here please). But I do not say that, once a given system is there, in front of you, you have to swallow it in one gulp or run away. You may discuss it and, that is what philosophy is about. Philosophers are like the accountants of beliefs (perhaps this is why they may be so pusillanimous and irritating at times). They count points in the battle between systems. Because it is a battle. Since beliefs control people, defending or attacking them engages the whole of yourself, ego, feelings, education and all. Therefore, the process of "discussing" beliefs is a necessarily messy affair involving all kinds of human postures like cool, rational, discourse but also appeals to sensitivity, to authority, to poetry, art, you name it. In some cases, of course, it may have to involve threats and even violence. WWII, as we said, was largely a "discussion" about beliefs and values.

All this is bad news. But pretending that it can be avoided is either foolish or deceitful. There is no nice way to synthesize cleanly modular ethical building blocks that people could assemble in various ways according to their needs, while being fully aware and in control of the process. It is a little bit like the Gödel Theorem : you cannot build a belief systems whith those very hands it is its purpose to control.

The last question is that of human nature. Well, I do not know what it is. If certain belief systems seem to work well and produce desirable results (like 'Be free so trust but ensure you can verify'), we may surmise that they have a good fit with human nature. But there is little else we can say. So if "human nature" changes (and I see what you mean here), belief systems will have to as well. But this need not worry us since there is a constant supply of new systems. So we are always guaranteed that at some point in the future one will emerge that will have a better fit with the changed circumstances than the one we had previously.

So, and to summarise, there is no point in talking about 'a priori ethics' since all ethical/belief systems are a-priori. Yours no less than any other.