In this [thread=364132]thread[/thread], Liberal expresses the following view:
To me, this view is somewhat disturbing. I thought the hallmark of the scientific revolution was to validate ideas with observations. I do not trust humans to reason perfectly, and trust them even less to come with appropriate axioms and inference rules without observations. I do not even trust humans to understand logic completely; I would say that we did not have a very good understanding of logic until the early 20th century.
Also, although most sciences use statistics and some, such as physics, rely heavily on mathematics, I do not know of one science whose theories are expressed in a formal language. Even mathematicians who do not specialize in logic reason in natural language (augmented with specialized notation). So I do not really understand the high confidence Liberal places in logic when it comes to science.
It is true that data can sometimes be biased or just incorrect because of errors in how it was collected. If it seems unreasonable, you should try to find the error and run more experiments. However, if empirical observations consistently contradict your model, you should re-evaluate your model.
I’m not sure how Liberal is using the word “theory” here. However, I would say that if a physicist encounters data that contradicts his mathematical model, he would not conclude that the mathematics is invalid, but question its applicability to the problem he is studying.
I’m not sure that I understand what Liberal is talking about, but in real science in the real world observation always overrules “logic”. Scientists didn’t reject the photoelectric effect or the orbiting electron models of the atom simply because they contradict “logic”. Those things were firmly established by experimental observation and “logic” was re-written to encompass them. On a more fundamental level there was doubtless a time when standing on the “bottom” of a spherical Earth was considered totally illogical, nonetheless nobody rejected the observations showing that to be the case simply because it is illogical. Instead “logic” was re-written to accommodate it.
It seems that Liberal has some idea that any concept derived logically is somehow a fixed point in science. In fact it is just as falsifiable and transient as any hypothesis. Just because I can prove logically that the Earth must be flat doesn’t mean that I reject any observations that run counter to that. The observation always trumps “logic”.
And the reason it always trumps is that “logic” can so easily be flawed. It may be true that if all our starting axioms are accurate and we have all the relevant data logic must trump all, but in the real world we never have all our axioms correct and we never have all the relevant data. As a result for all practical purposes logic is no more or less than an expression of a hypothesis. In principle it is more than a hypothesis but in pratcie there is no difference.
I have one major question for Liberal, which is how he determines what “logic” is WRT to the observable (and hence objective world? To explain why that question is far from clear, consider that physics tells us that on occasion we can get cause without any effect, multiple mutually exclusive effects from a single cause and cause with no effect. All formal logic AFAIK relies on cause and effect, essentially some variation on modus ponens/modus tollens. So how is “logic” even determined in such a universe? If it is simply determined based on arbitrary axioms then why would such logic not be trumped by observation which falsifies the axioms?
Liberal is correct. No doubt he will be here to explain this better than I will be able to, but here goes.
Empiricism is a fancy word for observation, which is one view of an event. To make it more reliable, we can increase both the number of viewings and number of different perspectives. But we can never be sure to have looked at the event from all perspectives. And we can never be sure that purity and accuracy of any single observation, as we have now introduced human fallibility into the equation.
Deductive logic seeks to eliminate unknowns and questionable “facts” derived from observation techniques and come to conclusions that are true 100% of the time. The example someone gave in that other thread was the relationship of the circumference of a circle to its radius. It will always be a function of 2pi. Given adequately sophisticated equipment one could take every circle and take measurements and—given enough time—come to this conclusion with a very high degree of certainty, but the confidence level will never reach that of a mathematical proof.
We call it an “illusion” precisely observation steers us in the wrong direction. “The pencil in the glass of water is bent.” If we could eschew the use of reason (impossible, of course, because we would need some form of reasoning even to come up with the wrong conclusion) we might say, “The pencil bends when we put it in the glass of water but goes back to normal when we take it out.”
Such a conclusion would superficially match our observations but would not match other things we know or observe about the pencil (e.g., it still feels straight to the touch when it is in the glass of water).
Okay. Fair question, albeit loaded, and there are in my opinion some problems with your analysis that I’ll address. But let me clear up what I think is muddy first, and then let’s see what’s left.
First, I reject the assertion that the observable world is the objective world for a couple of reasons. One, unless you are willing to accept indirect observation, then it cannot be true because we cannot directly observe, say, gamma wavelengths. But for indirect observation (by instruments or particle effects and so forth) you must invoke logic; e.g., that there is some correspondence between the LED you’re reading and what is happening in the laboratory. Therefore, there is no escaping the invocation of reason to supplement observation. Two, your tools of observation (your eyes, your meters, your LEDs, etc.) are themselves a part of the observable world, and therefore their relation to the world is definitively subjective.
Second, you stated twice that there are causes without effects: (1) “we can get cause without any effect,” and (2) “and cause with no effect”. I’m going to go out on what I hope is a very short limb and assume that you meant the second one to be reversed. In fact, we must reverse it because it would be the only one that is true. There are effects without known causes, such as the collapse of an electron’s orbit (and thereby the emission of a particle). However, that stipulation is epistemic in nature and not metaphysical. There is no a priori law stating that there cannot be a cause. Moreover, the collapse is of a waveform, and not of anything real. As Werner Heisenberg has said, “The atoms or the elementary particles are not real; they form a world of potentialities and possibilities rather than one of things or facts.” And as Niels Bohr has said, “There is no quantum world. There is only an abstract quantum mechanical description.”
Finally, I think it is a mistake to characterise logic as a cause and effect system for two reasons. One, it is a rules and formula system. Every logic that I know of (and keep in mind that I’m only a layman in this regard) is built on a set of rules for well formed formulas (called “wiffs”) and how they relate to one another. It is true that, in first order logics especially, one wiff must follow another, but it is not quite accurate to say that one causes another because a logical system can be either complete or consistent but not both. For example, with five axioms, I can prove that 1 + 1 = 2. But those same five axioms condemn me to a system that must of necessity have certain undecidable propositions. If one assertion always caused another, then every proposition should be decidable. And two, logic is not constrained by the unidirectional arrow of entropy. While it is problematic for entropy to reverse in the universe, I can go forwards and backwards from one inference to another whenever implications are biconditional. In other words, if A -> B and B -> A, then A <-> B.
So, logic with respect to the observable world is no different than it is with respect to anything else. The whole purpose of logic is to simplify expressions for examination with systematic symbology. Logic is intended to make an argument, statement, expression, or what-have-you transparent. Some logics (the more familiar ones) are designed to prove things true (or false). These are useful to science, which is NOT designed to prove things true. (It proves things only false.) It seems to me that a lot of scientists today (at least many of those represented here at SDMB) have lost sight of the philosophical underpinnings of their own discipline. This was not always the case. Einstein’s theories, for example, are built on two axioms: (1) the speed of light is constant in a vacuum, and (2) the laws of the universe are everywhere the same. The particulars that followed were deductively derived. Their conclusions HAD TO BE true so long as the axioms were true, and the wiffs were formed correctly. The role that science played by testing Einstein’s theories was to validate the axioms.
I’ll give you that, so long as you give me that by “careful” you mean “through the application of logic”. For example, suppose you see an illusion that makes one circle look bigger than another. So you measure the two circles and determine that both have the same diameter. However, what you left out was the reasoning involved that led you to take measurements. You applied the rules of commutation and symmetry to derive a principle of additive identity: If A = B, then A - B = 0. Without that being true, your measurements prove nothing.
Superficial is exactly right. Lots of things are not as they appear at first glance, but the way we discover this is by glancing a second time, as it were. If I am presented with the Mueller-Lyer Illusion, how do I determine if the lines are in fact the same length? Do I dig out some set of axioms and make deductions from them? No. I take a rule and I measure them. That’s an observation. If I am presented with the Hering Illusion, how do I determine if the lines are in fact straight? Do I use reason? Or do I actually check, by taking a straightedge and making an observation (or by occluding the diagonal lines)?
Deduction cannot tell us anything that we don’t already know, though it might sometimes come as a surprise to us that what we know can be restated in a particular novel way. But if you want to find something out about reality that you don’t already know, you’ll have to actually go out and look. Discovering that the superficial appearance of an optical illusion is misleading is something that will require going out and looking.
Think about what you are saying. Without Euclid’s postulates, what would your straightedge prove? The very reason you reach for a straightedge at all is because there are certain logical principles that are pertinent to what your measurements reveal.
I think there seems to be some confusion of ‘logic’ in this thread.
Yes, and how do you know which is the accurate observation? Your rulers show the same number, but your eyes judge the lengths different. By what process do you decide the authoritative barometer?
This has nothing to do with empiricism vs deduction, though, particularly if this is supposed to be an extension of the discussion of economics in that other thread. All you’re saying here is “If we want to draw conclusions from observations, we must perform logical operations on those observations.” Well, fine. That’s true for reasonable definitions of ‘observation’ and ‘logic’.
But that’s not what you were talking about in the other thread. There you were talking about taking a set of axioms and performing deductions on them, and learning something about the world. Or at least, that’s what I took you to be saying. Perhaps I was mistaken. If my understanding is correct, you are simply wrong. An economic theory deductively derived from a set of axioms could be a very interesting system revealing unanticipated logical relations, much like, say, Euclidean geometry. But it wouldn’t tell us anything about the world. In order to learn how much the world resembles the resulting economic system, we’d have to go out and look to see if things match up, just like we have to go out and look to see if the real world resembles Euclid’s geometry. And, if modern physics is to be believed, the real world resembles Euclid’s geometry quite a lot in most places, but not so much near large concentrations of mass.
Unfortunately for Liberal, I decide based on inductive processes. I have a physical theory about the world, which is built up by observation and induction. I believe that the ruler doesn’t change in length under ordinary conditions, based on my knowledge of its composition and the characteristics of the material its made of. All of which is dependent on observation and induction. And when faced with conflicting observations, I attempt to resolve them by determining which better fits in with the rest of that mass of observation and induction - and so when I measure the lines, and when I observe them with the inward- and outward-facing arrows obscured, they appear to be the same length, then I will conclude that the appearance of differing lengths when the arrows are not obscured is my perceptive apparatus going awry under specific conditions.
There’s lots of reason and logic going on there, but very little deduction from axioms. Lots of induction from observations, though.
Gorsnak, I really think the jig is up on this. Be a good fellow, stick to the topic at hand, and acknowledge that you must apply principles of reason to your observations. Even if your reasoning is nothing more than faith-based induction, induction is still as much a form of logic as deduction, analogics, and all the others.
Give me a break. Look at what you are quoted as saying in the OP.
You are talking about a priori deductive systems, and how they are superior to observation. I responded to your statement about optical illusions as if it was supposed to have some bearing on that issue. If you want now to concede that we know optical illusions are illusions based on observation and inductive reason, then what exactly is your defense of the points you are making in these quotes?
I’m so sorry I took you to mean a priori, deductive reason when you said “If reason did not trump observation, then we could not identify optical illusions.” If I had any inkling you’d mean observation-based inductive reason, then I wouldn’t have contradicted you. And I wouldn’t have had any idea how you thought it was relevant to the OP, either.
The jig that you’re dancing. You are using deductive reason to explain why you’re rejecting deductive reason. Two people have explained to you that your measuring sticks are useless without logic to back them up.
:rolleyes: Oh sweet and merciful Og, take pity on your poor servant.
I have never, ever rejected deductive reason. Deductive reason is a very nice tool which allows us to discover the implications of what we already know. What I deny is your ridiculous claim in the OP, that if somehow it seems based on observation that the claims of Austrian economics fail in the real world, it’s the observations and not the theory that’s wrong.
Just as the observed existence of gravitational lenses demonstrates that Euclidean geometry doesn’t accurately describe the real world, so an observation of someone “breaking” an axiom of Austrian economics would demonstrate that Austrian economics doesn’t accurately describe the real world.
We don’t need to take on roles here. I’m not the teacher, and you’re not the grasshopper. I just don’t have anything more to say to you than I already did if you deny that when you test optical illusions by making measurements, you are applying a priori knowledge that has been deduced from axioms like commutativity, identity, and symmetry. Otherwise, as I said, there’s no reason to take a measurement. I mean, you have a bunch of numbers, but without mathematical logic, you can’t even state an equation or function or inequality. You have nothing.
Then I misunderstood you when your wrote, “I decide based on inductive processes. I have a physical theory about the world, which is built up by observation and induction.”
The claim in the OP is that reason trumps observation. It applies to everything, not just economics. If you believe otherwise, then you shouldn’t be here making so many “if…then” assertions which amount to nothing more than logical implications. You should be trying to convince me by pointing out something for me to observe.
That’s a rather silly example, isn’t it? Euclidean geometry applies only to flat planes. Globes and saddles are what other geometries are for. Luckily, NASA scientists rely on math to send things to Mars and so forth rather than just adjusting their aim over a long series of experimental firings and observations.