Do you understand the tides, Colonel Barnes, simply because you know to say ‘gravity’?”
Yes, be cause saying ‘gravity’ in fact means the Newton gravitational law. Aristotle had no idea, that e. g. the product of two masses is involved here.
Does Colonel Barnes? If not, he is just repeating a word he has learned to say. Rather like some people today who have learned to say “entanglement”, or “signalling”, or “evolution”, or...
Except in this case he’s actually saying ‘gravity’ in the right context, and besides, it’s not expected of people in general to know Newton’s laws (or general relativity, etc) to know basically how gravity works.
Although I’d like to know what his answer was to the last question…
I will gladly post the rest of the conversation because it reminds me of question I pondered for a while.
“Do you understand the tides, Colonel Barnes, simply because you know to say ‘gravity’?”
“I’ve never claimed to understand them.”
“Ah, that is very wise practice.”
“All that matters is, he does,” Barnes continued, glancing down, as if he could see through the deck-planks.
“Does he then?”
“That’s what you lot have been telling everyone. <> Sir Isaac’s working on Volume the Third, isn’t he, and that’s going to settle the lunar problem. Wrap it all up.”
“He is working out equations that ought to agree with Mr. Flamsteed’s observations.”
“From which it would follow that Gravity’s a solved problem; and if Gravity predicts what the moon does, why, it should apply as well to the sloshing back and forth of the water in the oceans.”
“But is to describe something to understand it?”
“I should think it were a good first step.”
“Yes. And it is a step that Sir Isaac has taken. The question now becomes, who shall take the second step?”
After that they started to discuss differences between Newton’s and Leibniz theories. Newton is unable to explain why gravity can go through the earth, like light through a pane of glass. Leibniz takes a more fundamental approach (roughly speaking, he claims that everything consist of cellular automata).
Daniel: “<...> Leibniz’s philosophy has the disadvantage that no one knows, yet, how to express it mathematically. And so he cannot predict tides and eclipses, as Sir Isaac can.”
“Then what good is Leibniz’s philosophy?”
“It might be the truth,” Daniel answered.
I find this theme of Baroque Cycle fascinating.
I was somewhat haunted by the similar question: in the strict Bayesian sense, notions of “explain” and “predict” are equivalent, but what about Alfred Wegener, father of plate tectonics? His theory of continental drift (in some sense) explained shapes of continents and archaeological data, but was rejected by the mainstream science because of the lack of mechanism of drift.
In some sense, Wegener was able to predict, but unable to explain.
One can easily imagine some weird data easily described by (and predicted by) very simple mathematical formula, but yet I don’t consider this to be explanation. Something lacks here; my curiosity just doesn’t accept bare formulas as answers.
I suspect that this situation arises because of the very small prior probability of formula being true. But is it really?
Stanislaw Lem wrote a short story about this. (I don’t remember its name.)
In the story, English detectives are trying to solve a series of cases where bodies are stolen from morgues and are later discovered abandoned at some distance. There are no further useful clues.
They bring in a scientist, who determines that there is a simple mathematical relationship that relates the times and locations of these incidents. He can predict the next incident. And he says, therefore, that he has “solved” or “explained” the mystery. When asked what actually happens—how the bodies are moved, and why—he simply doesn’t care: perhaps, he suggests, the dead bodies move by themselves—but the important thing, the original question, has been answered. If someone doesn’t understand that a simple equation that makes predictions is a complete answer to a question, that someone simply doesn’t understand science!
Lem does not, of course, intend to give this as his own opinion. The story never answers the “real” mystery of how or why the bodies move; the equation happens to predict that the sequence will soon end anyway.
One can easily imagine some weird data easily described by (and predicted by) very simple mathematical formula, but yet I don’t consider this to be explanation. Something lacks here; my curiosity just doesn’t accept bare formulas as answers.
I suspect that this situation arises because of the very small prior probability of formula being true. But is it really?
I think the situation happens because of bias. Demonstrating an empirical effect to be real takes work. Finding an explanation of an effect also takes work.
It’s very seldom in science that both happens at exactly the same time.
Their are a lot of drugs that are designed in a way where we think that the drug works by binding to specific receptors. Those explanations aren’t very predictive for telling you whether a prospective drug works.
Once it’s shown that a drug actually works it’s often that we don’t fully understand why it does work.
It’s very seldom in science that both happens at exactly the same time.
Interesting.
I imagined a world where Wegener appeared, out of blue, with all that data about geological strata and fossils (nobody noticed any of that before), and declared that it’s all because of continental drift. That was anticlimactic and unsatisfactory.
I imagined a world with a great unsolved mystery: all that data about geological strata and fossils. For a century, nobody is unable to explain it. Then Wegener appeared, and pointed that the shapes of continents are similar, and perhaps it’s all because of continental drift. That was more satisfactory, and I suspect that most of traces of disappointment are due to hindsight bias.
I think that there are several factors causing that:
1) Story-mode thinking
2) Suspicions concerning the unknown person claiming to solve the problem nobody has ever heard of.
3) (now it’s my working hypothesis) The idea that some phenomena are and ‘hard’ to reduce, and some are ‘easy’:
I know that fall of apple can be explained in terms of atoms, reduced to the fundamental interactions. Most of things can. I know that we are unable to explain fundamental interactions yet, so equations-without-understanding are justified.
So, if I learn about some strange phenomenon, I believe that it can be easily explained in terms of atoms. Now suppose that it turned out to be very hard problem, and nobody managed to reduce it to something more fundamental. Now I feel that I should be satisfied with bare equations because making something more is hard. Maybe a century later.
This isn’t complete explanation, but it feels like a step in the right direction.
“For whatever reason, ” seems like it should be a legitimate hypothesis, as much as ”, therefore ”. The former technically being the disjunction of all variations of the latter with possible reasons substituted in.
But, then again, at the point when we are saying “for whatever reason, ”, we are saying that because we haven’t been able to think of the correct explanation yet—that is, because we haven’t been creative enough, a bounded rationality issue. So we’re perhaps not really in a position to evaluate a disjunction of all possible reasons.
I strikes me his understanding of gravity is on the same level as saying that everything attracts everything else, which is after all not much of a step up on saying that it’s in the nature of water to be attracted to the moon—just a more general phrasing.
You can make more specific predictions if you know that everything attracts everything, and you know more about the laws of planetary motion and so on, and the gravitational constant and the decay rate and so on, but the basic knowledge of gravity by itself doesn’t let you do those things. If your predictions after are the same as your predictions going in can you really claim to understand something better?
Seems to me you need to network ideas and start linking them up to data because you can really start to claim to understand stuff better.
Probably I should’ve added some context to this conversation. One of the themes of Baroque Cycle is that Newton described his gravitational law, but said nothing about why the reality is the way it is. This bugs Daniel, and he rests his hopes upon Leibniz who tries to explain reality on the more fundamental level (monads).
This conversation is “Explain/Worship/Ignore” thing as well as “Teacher’s password” thing.
The reason Newton’s laws are an improvement over Aristotelian “the nature of water is etc.” is that Newton lets you make predictions, while Aristotle does not. You could ask “but WHY does gravity work like so-and-so?”, but that doesn’t change the fact that Newton’s laws let you predict orbits of celestial objects, etc., in advance of seeing them.
That’s certainly the conventional wisdom, but I think the conventional wisdom sells Aristotle and his contemporaries a little short. Sure, speaking in terms of water and air and fire and dirt might look a little silly to us now, but that’s rather superficial: when you get down to the experiments available at the time, Aristotelian physics ran on properties that genuinely were pretty well correlated, and you could in fact use them to make reasonably accurate predictions about behavior you hadn’t seen from the known properties of an object. All kosher from a scientific perspective so far.
There are two big differences I see, though neither implies that Aristotle was telling just-so stories. The first is that Aristotelian physics was mainly a qualitative undertaking, not a quantitative one—the Greeks knew that the properties of objects varied in a mathematically regular way (witness Erastothenes’ clever method of calculating Earth’s circumference), but this wasn’t integrated closely into physical theory. The other has to do with generality: science since Galileo has applied as universally as possible, though some branches reduced faster than others, but the Greeks and their medieval followers were much more willing to ascribe irreducible properties to narrow categories of object. Both end up placing limits on the kinds of inferences you’ll end up making.
Yes, be cause saying ‘gravity’ in fact means the Newton gravitational law. Aristotle had no idea, that e. g. the product of two masses is involved here.
Does Colonel Barnes? If not, he is just repeating a word he has learned to say. Rather like some people today who have learned to say “entanglement”, or “signalling”, or “evolution”, or...
Except in this case he’s actually saying ‘gravity’ in the right context, and besides, it’s not expected of people in general to know Newton’s laws (or general relativity, etc) to know basically how gravity works.
Although I’d like to know what his answer was to the last question…
I will gladly post the rest of the conversation because it reminds me of question I pondered for a while.
After that they started to discuss differences between Newton’s and Leibniz theories. Newton is unable to explain why gravity can go through the earth, like light through a pane of glass. Leibniz takes a more fundamental approach (roughly speaking, he claims that everything consist of cellular automata).
I find this theme of Baroque Cycle fascinating.
I was somewhat haunted by the similar question: in the strict Bayesian sense, notions of “explain” and “predict” are equivalent, but what about Alfred Wegener, father of plate tectonics? His theory of continental drift (in some sense) explained shapes of continents and archaeological data, but was rejected by the mainstream science because of the lack of mechanism of drift.
In some sense, Wegener was able to predict, but unable to explain.
One can easily imagine some weird data easily described by (and predicted by) very simple mathematical formula, but yet I don’t consider this to be explanation. Something lacks here; my curiosity just doesn’t accept bare formulas as answers.
I suspect that this situation arises because of the very small prior probability of formula being true. But is it really?
Stanislaw Lem wrote a short story about this. (I don’t remember its name.)
In the story, English detectives are trying to solve a series of cases where bodies are stolen from morgues and are later discovered abandoned at some distance. There are no further useful clues.
They bring in a scientist, who determines that there is a simple mathematical relationship that relates the times and locations of these incidents. He can predict the next incident. And he says, therefore, that he has “solved” or “explained” the mystery. When asked what actually happens—how the bodies are moved, and why—he simply doesn’t care: perhaps, he suggests, the dead bodies move by themselves—but the important thing, the original question, has been answered. If someone doesn’t understand that a simple equation that makes predictions is a complete answer to a question, that someone simply doesn’t understand science!
Lem does not, of course, intend to give this as his own opinion. The story never answers the “real” mystery of how or why the bodies move; the equation happens to predict that the sequence will soon end anyway.
Amusingly, I read this story, but completely forgot about it. The example here is perfect. Probably I should re-read it.
For those interested: http://en.wikipedia.org/wiki/The_Investigation
I think the situation happens because of bias. Demonstrating an empirical effect to be real takes work. Finding an explanation of an effect also takes work. It’s very seldom in science that both happens at exactly the same time.
Their are a lot of drugs that are designed in a way where we think that the drug works by binding to specific receptors. Those explanations aren’t very predictive for telling you whether a prospective drug works. Once it’s shown that a drug actually works it’s often that we don’t fully understand why it does work.
Interesting.
I imagined a world where Wegener appeared, out of blue, with all that data about geological strata and fossils (nobody noticed any of that before), and declared that it’s all because of continental drift. That was anticlimactic and unsatisfactory.
I imagined a world with a great unsolved mystery: all that data about geological strata and fossils. For a century, nobody is unable to explain it. Then Wegener appeared, and pointed that the shapes of continents are similar, and perhaps it’s all because of continental drift. That was more satisfactory, and I suspect that most of traces of disappointment are due to hindsight bias.
I think that there are several factors causing that:
1) Story-mode thinking
2) Suspicions concerning the unknown person claiming to solve the problem nobody has ever heard of.
3) (now it’s my working hypothesis) The idea that some phenomena are and ‘hard’ to reduce, and some are ‘easy’:
I know that fall of apple can be explained in terms of atoms, reduced to the fundamental interactions. Most of things can. I know that we are unable to explain fundamental interactions yet, so equations-without-understanding are justified.
So, if I learn about some strange phenomenon, I believe that it can be easily explained in terms of atoms. Now suppose that it turned out to be very hard problem, and nobody managed to reduce it to something more fundamental. Now I feel that I should be satisfied with bare equations because making something more is hard. Maybe a century later.
This isn’t complete explanation, but it feels like a step in the right direction.
“For whatever reason, ” seems like it should be a legitimate hypothesis, as much as ”, therefore ”. The former technically being the disjunction of all variations of the latter with possible reasons substituted in.
But, then again, at the point when we are saying “for whatever reason, ”, we are saying that because we haven’t been able to think of the correct explanation yet—that is, because we haven’t been creative enough, a bounded rationality issue. So we’re perhaps not really in a position to evaluate a disjunction of all possible reasons.
“Indeed, Sire, Monsieur Lagrange has, with his usual sagacity, put his finger on the precise difficulty with the hypothesis [of a Creator of the Universe]: it explains everything, but predicts nothing.”
I strikes me his understanding of gravity is on the same level as saying that everything attracts everything else, which is after all not much of a step up on saying that it’s in the nature of water to be attracted to the moon—just a more general phrasing.
You can make more specific predictions if you know that everything attracts everything, and you know more about the laws of planetary motion and so on, and the gravitational constant and the decay rate and so on, but the basic knowledge of gravity by itself doesn’t let you do those things. If your predictions after are the same as your predictions going in can you really claim to understand something better?
Seems to me you need to network ideas and start linking them up to data because you can really start to claim to understand stuff better.
Probably I should’ve added some context to this conversation. One of the themes of Baroque Cycle is that Newton described his gravitational law, but said nothing about why the reality is the way it is. This bugs Daniel, and he rests his hopes upon Leibniz who tries to explain reality on the more fundamental level (monads).
This conversation is “Explain/Worship/Ignore” thing as well as “Teacher’s password” thing.
The reason Newton’s laws are an improvement over Aristotelian “the nature of water is etc.” is that Newton lets you make predictions, while Aristotle does not. You could ask “but WHY does gravity work like so-and-so?”, but that doesn’t change the fact that Newton’s laws let you predict orbits of celestial objects, etc., in advance of seeing them.
That’s certainly the conventional wisdom, but I think the conventional wisdom sells Aristotle and his contemporaries a little short. Sure, speaking in terms of water and air and fire and dirt might look a little silly to us now, but that’s rather superficial: when you get down to the experiments available at the time, Aristotelian physics ran on properties that genuinely were pretty well correlated, and you could in fact use them to make reasonably accurate predictions about behavior you hadn’t seen from the known properties of an object. All kosher from a scientific perspective so far.
There are two big differences I see, though neither implies that Aristotle was telling just-so stories. The first is that Aristotelian physics was mainly a qualitative undertaking, not a quantitative one—the Greeks knew that the properties of objects varied in a mathematically regular way (witness Erastothenes’ clever method of calculating Earth’s circumference), but this wasn’t integrated closely into physical theory. The other has to do with generality: science since Galileo has applied as universally as possible, though some branches reduced faster than others, but the Greeks and their medieval followers were much more willing to ascribe irreducible properties to narrow categories of object. Both end up placing limits on the kinds of inferences you’ll end up making.