This was such a fascinating piece to read; I learnt a lot from this! Thank you so much for writing this! Is it true that paradigm shifts are often made by those who are from a completely different background/field of study to the one they make the paradigm shifting discovery in? If true, what could be contributing to this pattern?
Thanks! I think that statement about paradigm shifts is true but I can’t think of good evidence right now. It’s probably something to do with not making the same assumptions as everyone in the field, some of which might be wrong
What do you think about Karl Popper’s falsification criteria for science ?
For me, it draws a useful restriction of science to what can be « predicted », either logically, differentially or statistically, thus excluding most of humanities, dealing with humans that can actually want to game the predictions.
Salut! The falsification criterion is at the core of my thinking here, since I've been reading David Deutsch (who is a Popperian). But I would say that lots of things that are traditionally classified as humanities can be falsified, so it's not actually a compelling distinction. That humans make it messy (and they do!) doesn't mean that truth is theoretically unattainable for social questions.
Man the hair on my neck stood up when you said, "Art and literature wouldn’t be interesting if they were too complex and abstract." And then I felt silly when I realized you were building up that argument just to knock it down.
This is an excellent article. I could mince terms here and there, but there are two places where I take issue.
First: "Some philosophical questions, for instance around morality, are extremely difficult to answer. But there should always exist some true answer if you define the question precisely enough."
I don't know if this is an argument you're making or you're establishing the common argument in this vein of thinking, but I would challenge the idea that there is always a *true* answer if you define the question precisely enough.
Second: "Maybe philosophers, historians and theologians should focus a bit more on making their field progress towards truth rather than write an ever-increasing amount of commentary on the old thinkers."
I think most theologians would be indignant to find someone thinks their work is anything other than pursuing the truth. And I think most philosophers would question (as I immediately did), "what is truth?"
I just started reading your Substack, so I can dig into this a bit later, but it seems to me that this article is written with the presumption that people are reading it are scientists. I think that the above two points reveal the greatest rift between science and humanities. To wit, many scientists think that if you refine a question to an exact enough frame, there is an empirical truth underneath it, and their work is refining their empirical measurements until they can approximate that truth as closely as possible.
I haven't thought about this long enough to offer the alternative view that people in the humanities have--and I can't definitively say many people in the humanities don't share this worldview. I can say that much of the humanities is about approximating truth *while also* questioning the nature of truth. I don't think that most practitioners in the humanities would agree with the statement, "if you refine a question well enough, there is a single empirical truth." I don't think that they'd necessarily disagree either, so much as they would say it's an impossible or poorly formed statement.
I think both sides of the chasm are trying to approximate truth in their own way. Whereas scientist believe that is best done empirically, the humanities believes that other tools are more important. As writer Tim O'Brien said, "That's what fiction is for. It's for getting at the truth when the truth isn't sufficient for the truth." Rewriting this, you might say, "That's what the humanities are for. They're for getting at the truth when empiricism isn't sufficient for the truth."
You're totally right that my essay skirts around the question of what is truth, and sorts of assumes that readers (which I do expect to be primarily scientific-minded people, as am I) agree that true answers exist in many cases.
The "if you define the question precisely enough" part, however, hints at the fact that many questions don't have a truth value simply because they are not well-formed. For instance, "Is the house that I own on the surface of Jupiter blue?" has neither a true nor a false answer; it's just meaningless, since there is no such house. The vast majority of questions are like this. But you can always be more precise and ask, e.g. "If I owned a house on Jupiter, would it be blue?" and then answer the question by making a bunch of assumptions ("to build a house on the gas surface of Jupiter you would first have to build infrastructure which would most likely use such and such materials, which aren't blue, and then you'd have to synthesize some blue paint which would be a poor use of resources, so the answer is probably no").
So sure, not all questions have a truth value. But all can be restated into something that does, with varying levels of interestingness.
I'll end by mentioning that empiricism has mostly nothing to do with any of this. Science can be (and often is) done without direct experience, just logic and ideas. I'm certainly not claiming that truth requires direct experience.
I understand what you're getting at with the truth value. To clarify, when I say empiricism, I may be using it imprecisely (which is kind of funny in a meta way, right?). In a practical sense I think that "is it directly experienced," is not as handy in thinking about empiricism as "could it be measured."
A scientist who makes a claim that couldn't theoretically be measured or quantified in someway would be laughed out of the lab (or journal etc.)
For example, "if I owned a house on Jupiter, would it be blue?" could be measured in a lot of different ways, and so does have a potential truth value and interesting areas to explore through science. But, "Should we build a houses on Jupiter?" isn't necessarily something you would measure, but could be explored endlessly in interesting ways in interesting, generative ways in the humanities.
I don't think the same is for people in the humanities at all. I think that this probably opens up a lot more questions that it answers.
Oh I see. I think empiricism does typically mean that hypotheses must be confirmed or falsified by direct experience. The broader sense that you mean ("could theoretically be measured") is something I agree with. But I'd still say that a "should" question can be theoretically answered, even if it's difficult to do so because of disagreement over the required assumptions.
What it comes down to for me is, could a measurement yield the truth? I think scientists believe that "confirming or falsifying by direct experience," is code for measurement. In a literal sense, you could directly experience La Pieta or the Northern Lights, but the direct experience you relay as impression and feelings wouldn't be scientific (until you talk about increased heart rate, hormones, pupil dilation etc. which for many would be missing the point). I also think that "should" questions could theoretically be answered, but I'm not sure that they can be answered in a measured or quantifiable fashion that scientists believe affirms something as truthful, which is the core of what I was getting at using the word empirical.
I think that you missed one other important difference between science and the humanities. Science is evidence-based. This includes astronomy and cosmology, where experimentation has proven difficult, and is also true of mathematics, where proofs are the evidence required.
It also provides an argument that history should be classified with the sciences. Theology, the antithesis of science, on the other hand does away with evidence completely.
Your point about the classics may show that the humanities value opinions more than evidence. The opinions of people who lived thousands of years ago are still relevant and important, precisely because they have influenced humanity's opinions in general for so long.
I would say that theologians would probably assert that their claims are based on evidence!
You bring up a good point, although not one that is very different from the objectivity-subjectivity discussion — what exactly is evidence? Isn't an (informed) opinion a form of evidence, too? And the fact that it "misclassifies" history suggests that it's also not a convincing criterion for the distinction.
You're also very right that there is a circular quality to the classics! We study Plato because Plato has influence over society because we study Plato... Which doesn't mean it's pointless to do so.
You don't read Thucydides when studying the English Civil War, nor Herodotus to understand the Kennedy Administration. So in that sense would you not say that history is also misclassified by your definition - reading the classics?
No, because history, like other fields, is fractal. You're talking about distinct subfields here: Ancient Greek history, Early Modern English history, and 20th century American history. Each of these subfields has its own "classics." Reading the classics of a different history subfield is no more useful than it would be useful to read a classic paper in astrophysics if you're working on particle physics.
(That is to say, it may be useful in order to get a good understanding of history/physics as a whole, or to cross-pollinate ideas, but it's unlikely to help you directly with what you're working on.)
Then I think it's time to consider whether your definition is useful. If I am studying relativity, it is definitely useful to study Einstein's original papers; these would be the classics for this field.
Are you sure? Do you study relativity? My point is that even the people studying a field like relativity are unlikely to read Einstein's original papers, except possibly out of historical interest; for the purposes of advancing science, they will instead learn about Einstein's ideas from more recent papers and textbooks.
I haven't studied relativity for several years, but I do currently study neural networks. I certainly (as do others in the field) study and refer to the classic works in the field.
Einstein's papers give a excellent summary of what relativity is about and the physical insight behind it. I think you will find them more widely read than you expect.
This is a much less acute distinction than it seems at first.
First, there's a larger epistemological discussion about what constitutes evidence to be had, plus what qualifies as evidence varies within scientific fields. The Wikipedia article for Scientific Evidence is a good jumping off point for this--particularly the middle section, "Philosophical vs. Scientific Views." https://en.wikipedia.org/wiki/Scientific_evidence
Second, there's a large discussion of where the relentless focus on evidence has actually held back science. (I don't have one good link for this so you'll be served well by Googling.) During the height of the pandemic, I heard a lot of smart people talk about the limits of EBM, how it can be really handy for something where you make a single intervention and the results are quickly, widely available. It's tougher for longer-term questions where that's not possible, like epidemiology or weight loss. They suggest that EBM be considered as one of a range of tolls similar to physics, where they soften the focus on evidence and consider models more, so that they can predict outcomes and design better experiments. Given, this still falls under experimentation, but we're considering evidence.
Third, I don't think that, "history should be classified with the sciences," is the right conclusion to draw here. Analogously, it reminds me of the story of Diogenes and Plato, excerpted here from Wikipedia: "According to Diogenes Laërtius, when Plato gave the tongue-in-cheek definition of man as 'featherless bipeds,' Diogenes plucked a chicken and brought it into Plato's Academy, saying, 'Behold! I've brought you a man,' and so the Academy added 'with broad flat nails' to the definition." I feel like reading this article and saying, "well then history is a science," would be like if Plato looked at the chicken and said, "well that's a man." I don't think that the point of the article is that fields were misclassified, but rather that using that classification scheme in the first place creates an unnecessary Chinese wall.
Very fascinating, thank you! It seems to me that the arts have been able to exist in an experimental vein - numerous examples abound of painters, musicians, poets, etc. whose work is considered "experimental", and who devote their careers to the pursuit of one narrowly-defined set of stylistic tendencies (Mark Rothko in painting; Steve Reich and La Monte Young in music; Frank Lloyd Wright in architecture). Perhaps the fight between the "hard" sciences and the humanities occurs because the humanities are acting more like arts than sciences, and the sciences resent / disapprove of that?
I feel like one could write another post in this vein on the differences between the arts and the humanities! It's possible that the arts are more experimental than the humanities and are more alike to the sciences in this way, but not in others (subjectivity, classics, human-centeredness). Also, the main goal of artistic work isn't usually to gain knowledge, which puts them in another class of human activity.
Thanks for that! About reading the classics, though, I feel that it's more of a cliche than the reality that today's philosophers read the classics. I know many professional philosophers who have never read a line of Plato or Aristotle, but have built thriving careers on Gettier problems and such. The language barrier is also a problem. Almost no non-specialist philosopher today can read ancient Greek or Latin (if they don't happen to be Greeks or Italians, for whom it's a bit easier).
I haven't studied philosophy past some introductory courses, so I'll happily defer to your authority on this. But in those introductory courses we did read Plato (though not in the original Greek!), so I'm slightly skeptical that there are professional philosophers out there who have familiarity with Plato only from secondary or tertiary sources. By contrast, nobody reads Newton or Einstein directly, in an introductory science course — or indeed even in an advanced one.
On the other hand, it seems entirely possible that certain branches of philosophy (such as the philosophy of mind, perhaps?) are moving fast towards the science end of the spectrum, with their practitioners being more concerned with the most recent paper than whatever Descartes had to say on the topic. I think this is mostly a good development.
Music theory in the 20th century made a huge lurch toward math. I spent some time as a math/music double major so after reaching grad school in music it was hilarious to watch about 2/3 of my classmates recoil in horror upon first encountering the work of David Lewis and his use of group theory. “It’s maaaaaaath!” And 1/3 would shrug and say “it’s just math…”
(OTOH music/math is a pretty common combination overall. And music studies are not always about “reading the classics” in terms of history/theory, but boy does it involve playing/studying the classics. And there are a whole slough of digressions I could go on here).
This was such a fascinating piece to read; I learnt a lot from this! Thank you so much for writing this! Is it true that paradigm shifts are often made by those who are from a completely different background/field of study to the one they make the paradigm shifting discovery in? If true, what could be contributing to this pattern?
Thanks! I think that statement about paradigm shifts is true but I can’t think of good evidence right now. It’s probably something to do with not making the same assumptions as everyone in the field, some of which might be wrong
Makes sense, thank you!
Hello,
Bonjour,
What do you think about Karl Popper’s falsification criteria for science ?
For me, it draws a useful restriction of science to what can be « predicted », either logically, differentially or statistically, thus excluding most of humanities, dealing with humans that can actually want to game the predictions.
Moving from science to predict,
to conscience of what you want to predict.
Salut! The falsification criterion is at the core of my thinking here, since I've been reading David Deutsch (who is a Popperian). But I would say that lots of things that are traditionally classified as humanities can be falsified, so it's not actually a compelling distinction. That humans make it messy (and they do!) doesn't mean that truth is theoretically unattainable for social questions.
Man the hair on my neck stood up when you said, "Art and literature wouldn’t be interesting if they were too complex and abstract." And then I felt silly when I realized you were building up that argument just to knock it down.
This is an excellent article. I could mince terms here and there, but there are two places where I take issue.
First: "Some philosophical questions, for instance around morality, are extremely difficult to answer. But there should always exist some true answer if you define the question precisely enough."
I don't know if this is an argument you're making or you're establishing the common argument in this vein of thinking, but I would challenge the idea that there is always a *true* answer if you define the question precisely enough.
Second: "Maybe philosophers, historians and theologians should focus a bit more on making their field progress towards truth rather than write an ever-increasing amount of commentary on the old thinkers."
I think most theologians would be indignant to find someone thinks their work is anything other than pursuing the truth. And I think most philosophers would question (as I immediately did), "what is truth?"
I just started reading your Substack, so I can dig into this a bit later, but it seems to me that this article is written with the presumption that people are reading it are scientists. I think that the above two points reveal the greatest rift between science and humanities. To wit, many scientists think that if you refine a question to an exact enough frame, there is an empirical truth underneath it, and their work is refining their empirical measurements until they can approximate that truth as closely as possible.
I haven't thought about this long enough to offer the alternative view that people in the humanities have--and I can't definitively say many people in the humanities don't share this worldview. I can say that much of the humanities is about approximating truth *while also* questioning the nature of truth. I don't think that most practitioners in the humanities would agree with the statement, "if you refine a question well enough, there is a single empirical truth." I don't think that they'd necessarily disagree either, so much as they would say it's an impossible or poorly formed statement.
I think both sides of the chasm are trying to approximate truth in their own way. Whereas scientist believe that is best done empirically, the humanities believes that other tools are more important. As writer Tim O'Brien said, "That's what fiction is for. It's for getting at the truth when the truth isn't sufficient for the truth." Rewriting this, you might say, "That's what the humanities are for. They're for getting at the truth when empiricism isn't sufficient for the truth."
Thank you for this great comment!
You're totally right that my essay skirts around the question of what is truth, and sorts of assumes that readers (which I do expect to be primarily scientific-minded people, as am I) agree that true answers exist in many cases.
The "if you define the question precisely enough" part, however, hints at the fact that many questions don't have a truth value simply because they are not well-formed. For instance, "Is the house that I own on the surface of Jupiter blue?" has neither a true nor a false answer; it's just meaningless, since there is no such house. The vast majority of questions are like this. But you can always be more precise and ask, e.g. "If I owned a house on Jupiter, would it be blue?" and then answer the question by making a bunch of assumptions ("to build a house on the gas surface of Jupiter you would first have to build infrastructure which would most likely use such and such materials, which aren't blue, and then you'd have to synthesize some blue paint which would be a poor use of resources, so the answer is probably no").
So sure, not all questions have a truth value. But all can be restated into something that does, with varying levels of interestingness.
I'll end by mentioning that empiricism has mostly nothing to do with any of this. Science can be (and often is) done without direct experience, just logic and ideas. I'm certainly not claiming that truth requires direct experience.
I understand what you're getting at with the truth value. To clarify, when I say empiricism, I may be using it imprecisely (which is kind of funny in a meta way, right?). In a practical sense I think that "is it directly experienced," is not as handy in thinking about empiricism as "could it be measured."
A scientist who makes a claim that couldn't theoretically be measured or quantified in someway would be laughed out of the lab (or journal etc.)
For example, "if I owned a house on Jupiter, would it be blue?" could be measured in a lot of different ways, and so does have a potential truth value and interesting areas to explore through science. But, "Should we build a houses on Jupiter?" isn't necessarily something you would measure, but could be explored endlessly in interesting ways in interesting, generative ways in the humanities.
I don't think the same is for people in the humanities at all. I think that this probably opens up a lot more questions that it answers.
Oh I see. I think empiricism does typically mean that hypotheses must be confirmed or falsified by direct experience. The broader sense that you mean ("could theoretically be measured") is something I agree with. But I'd still say that a "should" question can be theoretically answered, even if it's difficult to do so because of disagreement over the required assumptions.
What it comes down to for me is, could a measurement yield the truth? I think scientists believe that "confirming or falsifying by direct experience," is code for measurement. In a literal sense, you could directly experience La Pieta or the Northern Lights, but the direct experience you relay as impression and feelings wouldn't be scientific (until you talk about increased heart rate, hormones, pupil dilation etc. which for many would be missing the point). I also think that "should" questions could theoretically be answered, but I'm not sure that they can be answered in a measured or quantifiable fashion that scientists believe affirms something as truthful, which is the core of what I was getting at using the word empirical.
I think that you missed one other important difference between science and the humanities. Science is evidence-based. This includes astronomy and cosmology, where experimentation has proven difficult, and is also true of mathematics, where proofs are the evidence required.
It also provides an argument that history should be classified with the sciences. Theology, the antithesis of science, on the other hand does away with evidence completely.
Your point about the classics may show that the humanities value opinions more than evidence. The opinions of people who lived thousands of years ago are still relevant and important, precisely because they have influenced humanity's opinions in general for so long.
I would say that theologians would probably assert that their claims are based on evidence!
You bring up a good point, although not one that is very different from the objectivity-subjectivity discussion — what exactly is evidence? Isn't an (informed) opinion a form of evidence, too? And the fact that it "misclassifies" history suggests that it's also not a convincing criterion for the distinction.
You're also very right that there is a circular quality to the classics! We study Plato because Plato has influence over society because we study Plato... Which doesn't mean it's pointless to do so.
You don't read Thucydides when studying the English Civil War, nor Herodotus to understand the Kennedy Administration. So in that sense would you not say that history is also misclassified by your definition - reading the classics?
No, because history, like other fields, is fractal. You're talking about distinct subfields here: Ancient Greek history, Early Modern English history, and 20th century American history. Each of these subfields has its own "classics." Reading the classics of a different history subfield is no more useful than it would be useful to read a classic paper in astrophysics if you're working on particle physics.
(That is to say, it may be useful in order to get a good understanding of history/physics as a whole, or to cross-pollinate ideas, but it's unlikely to help you directly with what you're working on.)
Then I think it's time to consider whether your definition is useful. If I am studying relativity, it is definitely useful to study Einstein's original papers; these would be the classics for this field.
Are you sure? Do you study relativity? My point is that even the people studying a field like relativity are unlikely to read Einstein's original papers, except possibly out of historical interest; for the purposes of advancing science, they will instead learn about Einstein's ideas from more recent papers and textbooks.
I haven't studied relativity for several years, but I do currently study neural networks. I certainly (as do others in the field) study and refer to the classic works in the field.
Einstein's papers give a excellent summary of what relativity is about and the physical insight behind it. I think you will find them more widely read than you expect.
This is a much less acute distinction than it seems at first.
First, there's a larger epistemological discussion about what constitutes evidence to be had, plus what qualifies as evidence varies within scientific fields. The Wikipedia article for Scientific Evidence is a good jumping off point for this--particularly the middle section, "Philosophical vs. Scientific Views." https://en.wikipedia.org/wiki/Scientific_evidence
Second, there's a large discussion of where the relentless focus on evidence has actually held back science. (I don't have one good link for this so you'll be served well by Googling.) During the height of the pandemic, I heard a lot of smart people talk about the limits of EBM, how it can be really handy for something where you make a single intervention and the results are quickly, widely available. It's tougher for longer-term questions where that's not possible, like epidemiology or weight loss. They suggest that EBM be considered as one of a range of tolls similar to physics, where they soften the focus on evidence and consider models more, so that they can predict outcomes and design better experiments. Given, this still falls under experimentation, but we're considering evidence.
Third, I don't think that, "history should be classified with the sciences," is the right conclusion to draw here. Analogously, it reminds me of the story of Diogenes and Plato, excerpted here from Wikipedia: "According to Diogenes Laërtius, when Plato gave the tongue-in-cheek definition of man as 'featherless bipeds,' Diogenes plucked a chicken and brought it into Plato's Academy, saying, 'Behold! I've brought you a man,' and so the Academy added 'with broad flat nails' to the definition." I feel like reading this article and saying, "well then history is a science," would be like if Plato looked at the chicken and said, "well that's a man." I don't think that the point of the article is that fields were misclassified, but rather that using that classification scheme in the first place creates an unnecessary Chinese wall.
Very fascinating, thank you! It seems to me that the arts have been able to exist in an experimental vein - numerous examples abound of painters, musicians, poets, etc. whose work is considered "experimental", and who devote their careers to the pursuit of one narrowly-defined set of stylistic tendencies (Mark Rothko in painting; Steve Reich and La Monte Young in music; Frank Lloyd Wright in architecture). Perhaps the fight between the "hard" sciences and the humanities occurs because the humanities are acting more like arts than sciences, and the sciences resent / disapprove of that?
I feel like one could write another post in this vein on the differences between the arts and the humanities! It's possible that the arts are more experimental than the humanities and are more alike to the sciences in this way, but not in others (subjectivity, classics, human-centeredness). Also, the main goal of artistic work isn't usually to gain knowledge, which puts them in another class of human activity.
Thanks for that! About reading the classics, though, I feel that it's more of a cliche than the reality that today's philosophers read the classics. I know many professional philosophers who have never read a line of Plato or Aristotle, but have built thriving careers on Gettier problems and such. The language barrier is also a problem. Almost no non-specialist philosopher today can read ancient Greek or Latin (if they don't happen to be Greeks or Italians, for whom it's a bit easier).
I haven't studied philosophy past some introductory courses, so I'll happily defer to your authority on this. But in those introductory courses we did read Plato (though not in the original Greek!), so I'm slightly skeptical that there are professional philosophers out there who have familiarity with Plato only from secondary or tertiary sources. By contrast, nobody reads Newton or Einstein directly, in an introductory science course — or indeed even in an advanced one.
On the other hand, it seems entirely possible that certain branches of philosophy (such as the philosophy of mind, perhaps?) are moving fast towards the science end of the spectrum, with their practitioners being more concerned with the most recent paper than whatever Descartes had to say on the topic. I think this is mostly a good development.
Music theory in the 20th century made a huge lurch toward math. I spent some time as a math/music double major so after reaching grad school in music it was hilarious to watch about 2/3 of my classmates recoil in horror upon first encountering the work of David Lewis and his use of group theory. “It’s maaaaaaath!” And 1/3 would shrug and say “it’s just math…”
(OTOH music/math is a pretty common combination overall. And music studies are not always about “reading the classics” in terms of history/theory, but boy does it involve playing/studying the classics. And there are a whole slough of digressions I could go on here).