First, thanks for responding. I value your perspective.
So you start off by saying that we don’t have a great definition of “democracy,” meaning that it’s difficult to come up with a definition that includes what we want and also excludes what we want.
That’s fine. I think Plato said the same thing about chairs. We all know what a chair is. But it’s difficult to come up with a definition that includes everything we would call a chair and excludes everything we would not call a chair.
The same goes for science, and that’s fine.
This all started off with your question of what is a “useful theory.”
In your blog post on that topic, I believe you said that a “useful theory” makes you smarter. I think that my response to your Tweet is along the same lines:
“Useful theories let you do things. They don’t have to be true.”
This is just describing how science actually works on the actual ground.
What do I mean?
Well, I assume that some physicists study “electrons.” I assume that physicists spend very little time studying whether “electrons” actually exist. Electrons fit in our model of how reality works, but models change and are refined and are sometimes upended completely.
The “problem” is that we all walk around saying that electrons are real.
We don’t know that. So saying that electrons are real is pretty close to saying that Zeus created the moon.
Electrons make sense based on our current model of the physical world, similar to how phlogiston made sense to chemists of the past.
So is this a problem? In many ways, it’s not.
We can do a lot of cool stuff with our current model of the physical world.
First, however, we need to acknowledge that our model of the physical world lets us do useful things. It is not the physical world.
It is a model.
We don’t have direct access to whatever the physical world is because the only way that we can perceive the physical world is through our senses.
Yada yada yada.
So we have a model.
Science that is carried out within the confines of this model is “Normal Science.”
Let me ask you this: How many scientists will get funded who propose research outside of our current model of the physical world?
Let’s use the well-worn example of helio-centric versus geo-centric models of the solar system.
Copernicus wasn’t the first guy to say that the Earth revolves around the Sun. He was just the first guy anyone really listened to, though, because the “scientific community” was more willing to consider the idea.
So I think your view that Kuhn is advocating a “Heroic” model of science is flawed. Nobody listens to the “heroes” until they are ready to hear them.
On that note, few people want to be a hero because heroes are often crushed by the establishment.
So maybe scientists are trying things out, they find some strange results, and they just shrug and move on.
Now, let’s get to revolutionary science.
“Revolutionary Science” happens when someone is pushing our model of the physical world closer to “reality” by creating a model that is fundamentally different than the one it is trying to replace.
So to use your Newton and Einstein example, it’s not that all of the scientists are in the dark until someone saves them. It’s that all of the scientists are busy taking Model X^100 to X^101, and then someone says, “Hey guys, it’s actually Y.”
In regard to your claim that “Kuhn was wrong matters because… because if he were right then you could turn around science funding and simply identify the “revolutionary science†and support that instead of wasting time with “normal scienceâ€. But, of course, if there was some way to tell apart “normal science†from “revolutionary scienceâ€, we would know.”
The problem is that when things from outside the prevailing paradigm often sound crazy. Crazy things often don’t get funded.
So saying that electrons are real is pretty close to saying that Zeus created the moon.
We have “electron guns” (X-rays) that are difficult to comprehend if we do not have the concept of an electron. Meanwhile, Zeus does nothing to explain the Moon.
Let me ask you this: How many scientists will get funded who propose research outside of our current model of the physical world?
In Canada, the humanities are about as well funded as science. There is little difference in how humanity professors secure funding versus physics professors. And, of course, the humanities have little to do with science.
How science gets funding in the present system (which was created in the 1970s) tells us little about the nature of science, especially when this same system is applied to fund the arts, for example.
“Revolutionary Science†happens when someone is pushing our model of the physical world closer to “reality†by creating a model that is fundamentally different than the one it is trying to replace.
Sociology is not a science. Let us substitute “sociology” for “science”:
“Revolutionary Sociology†happens when someone is pushing our model of reality closer to “reality†by creating a model that is fundamentally different than the one it is trying to replace.
Sounds good?
There is nothing particularly insightful to the term “revolutionary science”, at least nothing that informs us about the nature of science. You can talk about revolutions in finance, engineering, politics, business… and, yes, in science too. Revolutions are not a fundamental characteristic of science anymore than it is a fundamental characteristic of any other human field.
Reproducibility pre-dates science. If I have a recipe to create a given sword, and I want to outsource the production to you, then I need to make it reproducible… I have to make sure you can generate the same result. That, by itself, does not buy us science. In fact, lots of stagnant pre-scientific civilizations had excellent skills at reproducibility.
It is uncommon to attempt to reproduce results in science. I invite you to open any science journal and look for the phrase “we reproduced results from Y showing X”. You will find that, overwhelmingly, authors present “novel” work, with experiments “never done before”.
It is actually not necessary to have good reproducibility in science. The important characteristic is falsifiability.
Many scientists do not have a clear idea what “science” means, because the concept does not exist in their native languages. For example, the closest Finnish equivalent is “tiede”, which is more closely related to the German concept “Wissenschaft”. It encompasses all fields of research that could exist in a reputable university.
I have a general idea which fields are usually considered science, but I am not aware of any features that can reliably differentiate them from other fields of research. Natural sciences are definitely science. Parts of many social sciences are often considered science, as they use similar methods to many natural sciences. On the other hand, some natural sciences such as astronomy do not use such “scientific” methods that much. Mathematics and statistics are probably not science, while aesthetics and theology are definitely not science. Even computer scientists cannot agree what computer science is.
1) String theory/M-theory is, as I dimly understand the situation, not falsifiable. Therefore, it is not science. Or?
2) Given any mathematical statement I can come up with a set of axioms within which it can be proven true, a set of axioms within which it can be proven false, and a third set within which the statement is undecidable. Is mathematics a science? (Actually I don’t think mathematics is a science any more than French is a science. Mathematics is a language and I don’t know what falsifying a language would mean.)
Anyone who uses the language of “undecidability” is clearly outside of science, in my mind.
Greg Lindensays:
Is it fair to say that you’re arguing that most science is iterative, small steps, by everyone? That what seems revolutionary is more a lot of drudgery, the pile of knowledge accumulating, until one of the many standing on that growing pile has the right timing and inclination to take a leap?
I don’t quite see the point you made about “a clearly stated hypothesis is often the end result, not the starting point”, but I think what you’re saying is that iteration on many speculative hypotheses eventually leads to a much more clever, clearly stated, and valuable hypothesis? In the absence of information, we can’t have a scientific breakthrough, so it’s very much a process of the accumulation of knowledge to get to the point where we understand enough to have that breakthrough?
Is it fair to say that you’re arguing that most science is iterative, small steps, by everyone? That what seems revolutionary is more a lot of drudgery, the pile of knowledge accumulating, until one of the many standing on that growing pile has the right timing and inclination to take a leap?
That’s well stated but I’d go even further. It is only in retrospect that we can tell this story… that we can pinpoint the “leaps”. These leaps are not objectively identifiable as they happen, and this makes the whole narrative suspect.
We have to be careful because human beings are great at making up stories about what happened.
I don’t quite see the point you made about “a clearly stated hypothesis is often the end result, not the starting pointâ€, but I think what you’re saying is that iteration on many speculative hypotheses eventually leads to a much more clever, clearly stated, and valuable hypothesis? In the absence of information, we can’t have a scientific breakthrough, so it’s very much a process of the accumulation of knowledge to get to the point where we understand enough to have that breakthrough?
Let us take anything we do not understand. Let us take the accelerated aging of astronauts, including low-orbit astronauts. We can quibble about whether it is “aging” but for my purpose it will do to call it aging. So what causes it? It seems that people first thought it was radiation, but then if you do the math, many airline pilots are exposed to just as much if not more radiations. And we now work a lot harder to protect astronauts from radiations. Microgravity seems to be a possible culprit. But how would you go from microgravity to early cataracts and diabetes? What is going on there? Let us try to get our astronauts to do a lot of intensive workouts, maybe it will help. Ah. No, it does not seem to stop the accelerated aging. This is a subproblem of a large problem: aging. Your body ages. Ok. So how does this cell at the end of your finger knows how old it is? Logically, there must be a clock, but what is it? If you workout a lot, will it slow or accelerate this clock? Do all human beings have the same clock? And it is not just human beings. Many plants die on schedule, though you can often perenialize them. If there is a schedule, there must be a clock… where is the clock? Is it the same clock as human beings? Or are they many different clocks. There are vaguely stated theories… but nothing clear-cut that you can test.
What is consciousness? Is it a thing or an illusion?
Let us take deep learning… the latest craze. Even the very best experts will tell you that they are surprised by how well it can work in some instances. They have vague theories as to why it happens… but we are still lacking the precise theorems… and they may never come.
Let us take another one. We know without a doubt that cigarettes cause cancer. Fine. Why? There are vaguely stated theories, but it is unclear.
So maybe you are interested in astronaut aging. Maybe you want to go to Mars one day, so you care about that. You don’t want to launch with the body of a 35-year-old and get there with the body of a 65-year-old. What are you going to do? You may start by taking small mammals in near space and measure all sorts of things… then you may try again, this time changing a few parameters… and so forth. Each time you hope to learn something. Knowledge piles up. In time you will build up a nice hypothesis that people can check easily enough because it is concrete and testable… but you usually do not start there.
First, thanks for responding. I value your perspective.
So you start off by saying that we don’t have a great definition of “democracy,” meaning that it’s difficult to come up with a definition that includes what we want and also excludes what we want.
That’s fine. I think Plato said the same thing about chairs. We all know what a chair is. But it’s difficult to come up with a definition that includes everything we would call a chair and excludes everything we would not call a chair.
The same goes for science, and that’s fine.
This all started off with your question of what is a “useful theory.”
https://twitter.com/lemire/status/811277657467944960
In your blog post on that topic, I believe you said that a “useful theory” makes you smarter. I think that my response to your Tweet is along the same lines:
https://twitter.com/carlroberts_us/status/811368423301251072
“Useful theories let you do things. They don’t have to be true.”
This is just describing how science actually works on the actual ground.
What do I mean?
Well, I assume that some physicists study “electrons.” I assume that physicists spend very little time studying whether “electrons” actually exist. Electrons fit in our model of how reality works, but models change and are refined and are sometimes upended completely.
http://nautil.us/issue/40/learning/-why-science-should-stay-clear-of-metaphysics
The “problem” is that we all walk around saying that electrons are real.
We don’t know that. So saying that electrons are real is pretty close to saying that Zeus created the moon.
Electrons make sense based on our current model of the physical world, similar to how phlogiston made sense to chemists of the past.
So is this a problem? In many ways, it’s not.
We can do a lot of cool stuff with our current model of the physical world.
First, however, we need to acknowledge that our model of the physical world lets us do useful things. It is not the physical world.
It is a model.
We don’t have direct access to whatever the physical world is because the only way that we can perceive the physical world is through our senses.
Yada yada yada.
So we have a model.
Science that is carried out within the confines of this model is “Normal Science.”
Let me ask you this: How many scientists will get funded who propose research outside of our current model of the physical world?
Let’s use the well-worn example of helio-centric versus geo-centric models of the solar system.
Copernicus wasn’t the first guy to say that the Earth revolves around the Sun. He was just the first guy anyone really listened to, though, because the “scientific community” was more willing to consider the idea.
So I think your view that Kuhn is advocating a “Heroic” model of science is flawed. Nobody listens to the “heroes” until they are ready to hear them.
On that note, few people want to be a hero because heroes are often crushed by the establishment.
Check this out: http://mobile.nytimes.com/blogs/well/2016/04/13/a-decades-old-study-rediscovered-challenges-advice-on-saturated-fat/
So maybe scientists are trying things out, they find some strange results, and they just shrug and move on.
Now, let’s get to revolutionary science.
“Revolutionary Science” happens when someone is pushing our model of the physical world closer to “reality” by creating a model that is fundamentally different than the one it is trying to replace.
So to use your Newton and Einstein example, it’s not that all of the scientists are in the dark until someone saves them. It’s that all of the scientists are busy taking Model X^100 to X^101, and then someone says, “Hey guys, it’s actually Y.”
In regard to your claim that “Kuhn was wrong matters because… because if he were right then you could turn around science funding and simply identify the “revolutionary science†and support that instead of wasting time with “normal scienceâ€. But, of course, if there was some way to tell apart “normal science†from “revolutionary scienceâ€, we would know.”
The problem is that when things from outside the prevailing paradigm often sound crazy. Crazy things often don’t get funded.
So saying that electrons are real is pretty close to saying that Zeus created the moon.
We have “electron guns” (X-rays) that are difficult to comprehend if we do not have the concept of an electron. Meanwhile, Zeus does nothing to explain the Moon.
Let me ask you this: How many scientists will get funded who propose research outside of our current model of the physical world?
In Canada, the humanities are about as well funded as science. There is little difference in how humanity professors secure funding versus physics professors. And, of course, the humanities have little to do with science.
How science gets funding in the present system (which was created in the 1970s) tells us little about the nature of science, especially when this same system is applied to fund the arts, for example.
“Revolutionary Science†happens when someone is pushing our model of the physical world closer to “reality†by creating a model that is fundamentally different than the one it is trying to replace.
Sociology is not a science. Let us substitute “sociology” for “science”:
“Revolutionary Sociology†happens when someone is pushing our model of reality closer to “reality†by creating a model that is fundamentally different than the one it is trying to replace.
Sounds good?
There is nothing particularly insightful to the term “revolutionary science”, at least nothing that informs us about the nature of science. You can talk about revolutions in finance, engineering, politics, business… and, yes, in science too. Revolutions are not a fundamental characteristic of science anymore than it is a fundamental characteristic of any other human field.
How about science is a discovery of statements supported by procedures that everyone can repeat and get the same factual results?
Reproducibility pre-dates science. If I have a recipe to create a given sword, and I want to outsource the production to you, then I need to make it reproducible… I have to make sure you can generate the same result. That, by itself, does not buy us science. In fact, lots of stagnant pre-scientific civilizations had excellent skills at reproducibility.
It is uncommon to attempt to reproduce results in science. I invite you to open any science journal and look for the phrase “we reproduced results from Y showing X”. You will find that, overwhelmingly, authors present “novel” work, with experiments “never done before”.
It is actually not necessary to have good reproducibility in science. The important characteristic is falsifiability.
Many scientists do not have a clear idea what “science” means, because the concept does not exist in their native languages. For example, the closest Finnish equivalent is “tiede”, which is more closely related to the German concept “Wissenschaft”. It encompasses all fields of research that could exist in a reputable university.
I have a general idea which fields are usually considered science, but I am not aware of any features that can reliably differentiate them from other fields of research. Natural sciences are definitely science. Parts of many social sciences are often considered science, as they use similar methods to many natural sciences. On the other hand, some natural sciences such as astronomy do not use such “scientific” methods that much. Mathematics and statistics are probably not science, while aesthetics and theology are definitely not science. Even computer scientists cannot agree what computer science is.
1) String theory/M-theory is, as I dimly understand the situation, not falsifiable. Therefore, it is not science. Or?
2) Given any mathematical statement I can come up with a set of axioms within which it can be proven true, a set of axioms within which it can be proven false, and a third set within which the statement is undecidable. Is mathematics a science? (Actually I don’t think mathematics is a science any more than French is a science. Mathematics is a language and I don’t know what falsifying a language would mean.)
Anyone who uses the language of “undecidability” is clearly outside of science, in my mind.
Is it fair to say that you’re arguing that most science is iterative, small steps, by everyone? That what seems revolutionary is more a lot of drudgery, the pile of knowledge accumulating, until one of the many standing on that growing pile has the right timing and inclination to take a leap?
I don’t quite see the point you made about “a clearly stated hypothesis is often the end result, not the starting point”, but I think what you’re saying is that iteration on many speculative hypotheses eventually leads to a much more clever, clearly stated, and valuable hypothesis? In the absence of information, we can’t have a scientific breakthrough, so it’s very much a process of the accumulation of knowledge to get to the point where we understand enough to have that breakthrough?
Is it fair to say that you’re arguing that most science is iterative, small steps, by everyone? That what seems revolutionary is more a lot of drudgery, the pile of knowledge accumulating, until one of the many standing on that growing pile has the right timing and inclination to take a leap?
That’s well stated but I’d go even further. It is only in retrospect that we can tell this story… that we can pinpoint the “leaps”. These leaps are not objectively identifiable as they happen, and this makes the whole narrative suspect.
We have to be careful because human beings are great at making up stories about what happened.
I don’t quite see the point you made about “a clearly stated hypothesis is often the end result, not the starting pointâ€, but I think what you’re saying is that iteration on many speculative hypotheses eventually leads to a much more clever, clearly stated, and valuable hypothesis? In the absence of information, we can’t have a scientific breakthrough, so it’s very much a process of the accumulation of knowledge to get to the point where we understand enough to have that breakthrough?
Let us take anything we do not understand. Let us take the accelerated aging of astronauts, including low-orbit astronauts. We can quibble about whether it is “aging” but for my purpose it will do to call it aging. So what causes it? It seems that people first thought it was radiation, but then if you do the math, many airline pilots are exposed to just as much if not more radiations. And we now work a lot harder to protect astronauts from radiations. Microgravity seems to be a possible culprit. But how would you go from microgravity to early cataracts and diabetes? What is going on there? Let us try to get our astronauts to do a lot of intensive workouts, maybe it will help. Ah. No, it does not seem to stop the accelerated aging. This is a subproblem of a large problem: aging. Your body ages. Ok. So how does this cell at the end of your finger knows how old it is? Logically, there must be a clock, but what is it? If you workout a lot, will it slow or accelerate this clock? Do all human beings have the same clock? And it is not just human beings. Many plants die on schedule, though you can often perenialize them. If there is a schedule, there must be a clock… where is the clock? Is it the same clock as human beings? Or are they many different clocks. There are vaguely stated theories… but nothing clear-cut that you can test.
What is consciousness? Is it a thing or an illusion?
Let us take deep learning… the latest craze. Even the very best experts will tell you that they are surprised by how well it can work in some instances. They have vague theories as to why it happens… but we are still lacking the precise theorems… and they may never come.
Let us take another one. We know without a doubt that cigarettes cause cancer. Fine. Why? There are vaguely stated theories, but it is unclear.
So maybe you are interested in astronaut aging. Maybe you want to go to Mars one day, so you care about that. You don’t want to launch with the body of a 35-year-old and get there with the body of a 65-year-old. What are you going to do? You may start by taking small mammals in near space and measure all sorts of things… then you may try again, this time changing a few parameters… and so forth. Each time you hope to learn something. Knowledge piles up. In time you will build up a nice hypothesis that people can check easily enough because it is concrete and testable… but you usually do not start there.