I think I suffer from scientism. Philip Goff has accused me on X of being captured by it (as well as having an anachronistic attachment to Positivism). I think this must be true, because I always have to re-look up the definition of scientism, and I don't really see why there can't be a provisional version of positivism. I think the whole "it's self defeating" can be a way to leverage a categorical error into a pseudo debunk. Something akin to saying "oh yeah, well what is the fundamental thing made of?" or "What happened before the Big Bang?"
They covered the topic of scientism on Nullius In Verba, but I think they are a little under-read on the topic, so I recommend the series more than the particular episode.
I have followed your exchanges with Goff with great interest. I think Goff suffers from what I call epistemic avarice, an unfounded expectation that there be satisfying explanations for things. But I also think he's an honest broker overall. His friendship with Keith Frankish, and Disagreeableme and his encouragement of his own graduate students to publish against panspychism makes me think he's not a bad egg. I think he's worth engaging. Though his recent turn to heretical theism kind of gives up the game he's always been playing.
Snake, I agree, I think Goff genuinely believes what he says and is generally a nice guy. But I still think he's seriously misguided about the nature of evidence and the weight of non-empirical metaphysical claims. Interesting about his turn to heretical theism, I did not know about it, and yes, it kind of gives the game away...
I will have to challenge you on anecdotal evidence. In short, the type of empiricial evidence you are talking about is to support or refute propositions that are universally quanitifed.
Example: all stars greater than 30 stellar masses collapse to a black hole.
Anecdotal evidence helps to validate a universally quantified proposition but it is limited.
But when it comes to an existentially quantified propostion, anecdotal evidence is all you've got.
Example: there exists a Savior
(Luke 1:1-3)
1 Many have undertaken to arrange a record of the events that have been fulfilled among us
2 just as those who from the beginning were eyewitnesses and servants of the word passed them down to us
3 it also seemed good to me, having carefully followed everything from the start, to write an orderly account for you, most noble Theophilus.
This type of anecdotal evidence can be validated, ironically, through universally quantified propositions. I will, in that case refer you to the Federal Rules for Evidence:
Antony, if I understand you correctly, you just validated my point. I don't believe in any savior, precisely because his existence is predicated entirely on the basis of witnesses / anectodal evidence, which is far too weak to support the extraordinary claim being made.
As for the federal rules, I don't recognize the government as an authority in epistemology. For instance: "Rule 610. Religious Beliefs or Opinions." Nope.
I didn't really validate your point. I just used a negative example. The reason I did that was to call in the legal rules of evidence because they are a starting point for validating (or in this case invalidating) anecdotal evidence. This determination is different from that of evidence to validate a univerally quantified proposition.
For a positive case, I asked my friend and advisor claude.ai this question (since I am no history expert):
"What evidence do we have to prove that Julius Caesar was assassinated?"
Claude comes up with 5 cases of "Anecdotal Evidence": Cicero, Nicolaus of Damascus, Suetonius, Plutarh and Appian. Cluade then gives the criterion by which you can judge this anecdotal evidence to show that this event occurred.
What makes this particularly convincing is the convergence of independent sources on core facts:
The date (Ides of March)
The location (Theater of Pompey, during a Senate meeting)
The key conspirators (Brutus, Cassius, and dozens of others)
The manner (23 stab wounds)
The immediate aftermath (conspirators fleeing to the Capitol)
The sources sometimes disagree on minor details (exact words spoken, specific sequences), which actually suggests they're drawing on different authentic traditions rather than copying each other.
---
Anecdotal evidence usually implies a human to tell the anecdote. In the sciences, we tend to strive for universally quantified propositions. It is interesting to ask if there is an equivalent in science to anecdotal evidence for an existentially quantified proposition. I can come up with two examples:
1. The 5 degree Kelvin Cosmic Background Radiation from the Big Bang
2. The iridium boundary that indicated that a meteor struck the earth around the time the dinosaurs disappeared.
I would suggest that these to events are attested to by multiple readings, but they may be impersonal examples of "anecdotal evidence". They tell a tale of a particular event. They are not like the observational tables of Tycho Brahe which are essentially not limited to a particular event.
As to the Federal Rules for evidence, I would respectfully submit that they are a good example of applied epistemology in regards to anecdotal evidence. If you want a more "scientific" approach (I am not using scare quotes here, I mean to emphasize that this is legitimate science versus ad hoc legal rules) I asked Claude this question:
"What psychologists have studied the reliability of eyewitness testimony in court?"
And it proceeded to tell me all of the psychological studies where eyewitness testimony is NOT reliable. I gave a follow-up question:
"These are studies of unreliability. I want studies that show when testimony is reliable."
And it then proceeded to list studies that indicated when anecdotal evidence is reliable:
"optimality hypothesis"—the idea that under optimal encoding conditions (good lighting, adequate viewing time, low stress), confidence and accuracy show significant positive correlations
witnesses can accurately recall stressful events even months later Simply Psychology, and that misleading questions don't necessarily corrupt memory formed under real-life conditions the way lab studies suggested.
retrieval effort cues as positive indicators—statements delivered fluently without hedges or delays are significantly more likely to be accurate
And then Claude notes:
"The philosophical tension here is striking: the field constructed its identity around exposing fallibility, perhaps because establishing positive reliability requires demonstrating ground truth—which is harder than showing error. The focus reveals an implicit epistemological bias toward skepticism over establishing conditions of justified confidence."
In short, sometimes anecdotal evidence is all you've got, especially for existentially quantified propositions. Instead of discounting all anecdotal evidence out of hand, it pays to determine when anecdotal evidence is likely to be reliable.
Antony, as a matter of courtesy I ask all my readers not to write comments that are of essay length, and not to quote extensively AI. I try my best to answer every comment, but there is a limit to the time I can spend doing so.
Historical documents and witnesses are not in the same category as anecdotal evidence. For one thing, you mention five, not one, source, and those sources are congruent. The rest of Claude's analysis strikes me as broadly correct, but not at all at odds with what I write in the essay we are discussing.
I would warn against telling Claude, or any other AI, to look for evidence to confirm your point of view. They are known to cherry pick or make shit up. My understanding of the relevant literature (and I looked, since I teach a course in criticla thinking) is that eyewitness testimony is considered of very low quality. Which doesn't mean it's useless. Same goes for anecdotes: low reliability, but not useless.
Contra to your claim, I don't discount all anecdotal evidence out of hand. I simply insist that it is very low on the rang of kinds of evidence, and that whenever possible we need to do better. It also depends on what exactly one wishes to prove by way of anecdotal evidence: I have anecdotally observed that I like red wine during the winter time. No problem. But if I told you that I have anecdotal evidence of miracles you should be properly skeptical.
Thanks. Helpful to remember facts count and how to try to sort.
Furst, AI has taken over media. Ascertaining by image ir speech pattern may not be enough. I fell into a cavern. Google no longer help. Pundits may simply be puny.
Second. A way to keep my brain functioning I read in the New England journal of medicine, as well as ground truth in Substack. A systemic analysis post discussed the possible difficulties with double blinded studies of medications lots of studies in GLP-1 meds if the study is to help in tension to treat, research needs to identify those who drop out or take drug after trial ended. Outliers are many. Perhaps few ever follow instructions. It requires researchers be more diligent Second, as a doubter, a study comparing one drug with another and without. From man places around the world . A editorial group asked why they had not evaluated with heart failure all men, Italian nephrologists of they did not in failing kidney the heart goes first. So important later, a physician sent in to editor and said the manufacturer of the second drug…. Drug vs Placebo—supported rhe research. My question, did they get data if bad outcomes and did not bother to post them. If the endpoint blood pressure, physicians should be aware of data as patients. . The NEJM editorial said include data in study, or do another with data. Thank the good David
Thank you Dr. Pigliucci for the clarification and explanation on evidence. As a beginner in critical thinking and philosophy it is important to understand what evidence is given and should be provided in day to day arguments. It is a learned skill to evaluate the evidence. Especially scientific and logical. Identifying logical fallacies, insufficient premises etc. takes time to learn. As does going to primary source scientific studies and making sure one is not Cherry picking, misrepresenting and misinterpreting the research. To me learning these critical thinking skills are crucial to develop "clear sighted" wisdom especially in our modern world.
I am leaving a question that I wanted to make in our upcoming seminar in Rome:
I’ve noticed that Stoics and even Academic Skeptics often rely on anecdotal evidence to anchor their arguments. Cicero, for instance, frequently cites virtuous or non-virtuous examples through legendary figures like Hercules, Romulus, Hippolytus, or Theseus. None of these figures actually existed; even if they had, they represent isolated cases rather than a statistically significant sample.
Seneca goes even further in De Clementia, using the behavior of bees to justify the role of a king when advising Nero: 'Nature herself has conceived the idea of a king, as you may learn from various animals, and especially from bees...'
While these cases obviously predate Hume and modern scientific inquiry, it makes me think that if we were to dismiss this type of anecdotal argumentation, their works would lose much of their power. There is also the element of pathos that these philosophers employ to truly persuade their readers.
Since this blog post aligns well with your own philosophical style. What is your approach to studying these authors, especially for your upcoming book on Cicero, where you will likely encounter many of these anecdotal arguments?
Demian, I think you have already at least implicitly answered the question: examples from mythology clearly cannot count as evidence, not even of the anecdotal type. They are rhetorical devices, and both Seneca and especially Cicero were excellent at rhetoric. They understood Aristotle's advice to accompany logos with ethos and pathos if one wishes to be persuasive.
As for actual anecdotal evidence, as I say in the essay, it does count, but at the lowest rang of the scale. It's a starting point, not the end of the matter. One of the reasons I practice Stoicism is because there is a lot of *modern* evidence, from systematic research in cognitive and behavioral science, that its general approach works.
If one were to justify the claim that 'virtue is the only good' based on the premise that the human telos is 'living according to nature,' one could take an empirical path. By relying on biological research into what distinguishes humans from other species, one might conclude that acting courageously, justly, temperately, and prudently is the best way to align with our prosocial and rational nature.
Alternatively, one could take a purely rational path, arguing that anything truly 'good' must be universal—there is no scenario where acting with such virtue could be considered 'bad.' Here you don't need to rely on science or empirical data (although you might have to, if you start to gather different scenarios to start to create a "sample").
This is an oversimplification, of course, but it illustrates the point. While both appear to be valid paths, would you prioritize the empirical approach over the purely rational one? I am also wondering if this style has to do with your own background (that combines science and philosophy), and if you see other academics still favoring the purely rational/logical one for philosophical inquiry.
Demian, I simply don't believe that reason by itself leads to discoveries about the nature of the world, and since I consider ethics to be an empirical discipline, no, I'm not satisfied with rational arguments by themselves. They need to be coupled with empirical evidence. Another way to put it is that ethics is about solving the problem of how human beings can live together and flourish. Surely this is an empirical issue, depending on human nature and the nature of human societies.
Thanks Massimo! This is also a good starting point for a subsequent discussion in Rome, so I am leaving more questions for the seminar directly. Thanks for taking the time to reply and see you very soon to continue discussing this and other topics.
Although I certainly appreciate that the article explains that there are other kinds of evidence than empirical, there is also an implied reductionist argument that at the very least I find unhelpful. There is no doubt to any reasonable person that empirical evidence has expanded human knowledge exponentially. But there is also a lack of recognition of what we have lost in abandoning a more holistic approach that split disciplines into neat categories.
There are many other forms of evidence. Cumulative evidence arising from aggregation or, although maybe weak on its own, argumentum ex silentio. There is also coherence evidence, which may be evidentially relevant and contextual evidence where background knowledge constrains plausibility independently of new data. Comparative or pattern evidence is absent. Historical evidence is reduced to authority when it is far more than that. Even modal considerations can add evidential force to explanations. Abductive reasoning is more than Bayesian updating and has its own evidential mode.
Science has replaced philosophy as the dominant authority on truth. Yet science is not oriented towards wisdom. The universal approach sought understanding that informed how to live. Science only shows how explain or manipulate. We seem to have gained precision and vast amounts of knowledge but to do what with? We seem to have lost integration, any sense of necessity and more importantly any shared account of why any of it matters. This is probably why we see the world on such a destructive path. We blind our understanding by granting authority to science without bounds, instead of recognising it as the remarkable tool that it is, rather than as a source of human meaning.
Great post and I am also enjoying the comments section. As I read this, I thought about the discussions that I have with friend every few months where we debate politics and life. In our discussions, we often reach an impasse. We actually wade through these different types of evidence. Empirical and logic evidence are often dismissed in favor of anecdotal and authoritarian claims. The impasse occurs when I challenge his authorities. For example, JFK Jr is not an authority.
Just for the record, I know that I am right and he is wrong. :-)
Sakshat, ah, that touches on the very interesting and complex issues of which authorities one trusts and why. I would say, again, that a reasonable answer should depend on the available evidence. And the evidence is pretty clear: JFK Jr. is a dangerous lunatic.
Edgar, you seriously think the problem with the world is too much science?
All the other kinds of evidence you mention are sub-categories of the ones I list, so I honestly don't think I missed anything.
As for reductionism, setting aside that it is simply a method that breaks down big problems into digestible chunks, and is therefore extrenely useful, even "holistic" approaches ought to be based on evidence of some kind.
Wisdom too is no exception: unless based on evidence it is utter and dangerous nonsense, as seen in plenty of self-professed gurus and prophets throughout the history of the world.
Not too much science, but too much authority granted to science, especially when it is asked to answer questions about why, what matters and how we ought to act or think.
I am not arguing against the use of evidence, but against an overly narrow account of what counts as evidence.
For me wisdom is not technical know how. It is what we do with what we know, believe, experience and think. Without that orienting function knowledge alone tells us nothing about how to live.
Edgar, could you give me some examples of this alleged excessive authority given to science? Because much of what I see is the opposite: people making decisions, at all levels, with the utmost disregard for scientific evidence.
Again, my account of evidence is not narrow. All your counter-examples can be treated as sub-categories within the ones I have outlined.
I agree that wisdom is not technical know-how. But then what is it? If it is the ability to act appropriately under whatever circumstances, don't you think that ability is fed by evidence?
Science or empirical evidence is routinely mistaken for, or presented as, the truth itself. But science is provisional. It always has been. Worse, much of what is presented as “science” in public discourse is correlation drawn from small or limited studies, then promoted as decisive.
As truth, it has been given the authority to invoke and justify actions and decisions that are, in reality, ethical, political, or economic choices. It is used to privilege one claim over another for particular ends. It is used by governments to justify policies, by corporations to justify actions, by courts to justify decisions putting innocent people away. There are thousands of examples of each today but seen even clearer historicaly, which I leave nameless as I do not want to get into the politics of these.
If you are asking for named cases, that becomes a political argument rather than a philosophical one, which is exactly the authority problem I am pointing to.
What I am saying is that science and empirical data are evidence. There are other forms of evidence. Science is probably the best kind for technical know how, but it is not the most useful for all enquiry. Science and data are provisional and if we think of them as the truth they can justify anything. This is because science is partial, incomplete and provisional and always will be. Does that mean we ignore science or empirical data? Of course it does not. But neither should we give it undue authority.
Although I understand clearly the concept of using what we do know and have to understand the world, and science gives us so much of that, it is not sufficient on its own to understand the world or our place in it. And furthermore, it is dangerous to do so.
If you are interested, I wrote an article a while back on how I see this:
Edgar, you keep making generic statements without examples, so all I can do is to say that I disagree, I don't see too much deference to science in modern society, especially in the US. Quite the opposite, I see quacks and pseudoscientists and bullshitters at the highest levels of government.
As for scientific conclusions being provisional etc., I doubt any scientist or philosopher of science would disagree. So what? Do you have access to some other kind of knowledge that is not provisional? If so, please share it with the rest of us.
Should political, ethical and economic decisions be made on the basis of the best science? You bet! On what else would you base them? We both known that science is in the business of answering questions of facts, not of value, but the two are deeply interrelated, unless one plucks one's values out of nowhere and for no particular reason. Which I wouldn't recommend.
So are you defining the science you disagree with as that of “quacks and pseudoscientists and bullshitters”? If that is the case then we are not really talking about science at all. We are talking about about failures of judgement and authority. Who gets to decide what counts as “best”.
Yes, all knowledge is provisional. Only pointing out that science is too and therefore must be understood and used within its limits. Provisionally was not a criticism only a description. Science establishes facts and causal relations. Explaining what is the case and what happens if we intervene.
The existence of scientific evidence that something works does not explain what it should be done or what is “best”. When you ask on what else such decisions should be based my answer is not about certainty only practical rational judgement about human flourishing. Judgements grounded in forms of evidence that science does not ask. We can look to diachronic evidence. What sustains rational agency? Intersubjective evidence so we can decide what is presupposed for moral justification or social cooperation. Coherence is a nice one if something can be applied consistently without undermining. And not limited to these either.
They include long-term historical experience of what corrodes or sustains human communities. The cross cultural stability of human “goods” and the internal coherence of principles when applied consistently over time. Or whether a principle can be justified to those subject to it as rational beings rather than imposed on them as systems to be managed.
Science informs these judgements and constrains what is possible. It tells us what works and at what cost. But it does not tell us which costs are acceptable or which ends are worth pursuing or how a human life should be oriented.
Science one criterion of evidence for how we judge, act, and think and an important one. But it is not sufficient for questions of value or meaning or how to live well.
As I argued in my Answers for Aristotle, there is no need for primacy: the best approach is a combination of science and philosophy. As I note in the essay, logical arguments (the bread and butter of philosophy) are a type of evidence, just like empirical data (the bread and butter of science).
The consideration of evdence, no matter how convincing is foolish without considering alternative hypotheses. Thus JS Mill urges understanding you opponents position in order to understand your own. This of course is a matter of logic (probability theory) not merely normative. The conversation beween Jordan Peterson and Cathy Newman is a wonderful example of seeing no need for alternative hypotheses due to ideological blindness.
Terry, well, yes, but I fail to consider a given hypothesis it is up to my collaborators or even opponents to put forth alternatives for me to consider. So long as they are backed up by evidence.
As for Peterson, sorry, I have absolutely no stomach to watch him.
this essay is paywalled, however, I've discovered other essays by Donald on JP. I can't evaluate his criticism of the books - I have 12 rules and its sequel but have been unable to get through them. Meanwhile his conversations favourably reveal his moral compass for example with Abigail Shrier, Nigel Biggar, Lee Jussim, Andrew Doyle, Gad Saad, Katherine Birbalsingh, Jonathan Haidt, Greg Lukianoff ... I pass by the religious ones. I have never thought of JP as a Stoic - some people do apparently.
Thank you that was an excellent summary
Glad you enjoyed it!
I think I suffer from scientism. Philip Goff has accused me on X of being captured by it (as well as having an anachronistic attachment to Positivism). I think this must be true, because I always have to re-look up the definition of scientism, and I don't really see why there can't be a provisional version of positivism. I think the whole "it's self defeating" can be a way to leverage a categorical error into a pseudo debunk. Something akin to saying "oh yeah, well what is the fundamental thing made of?" or "What happened before the Big Bang?"
They covered the topic of scientism on Nullius In Verba, but I think they are a little under-read on the topic, so I recommend the series more than the particular episode.
https://nulliusinverba.podbean.com/
Snake, I would consider being so insulted by Goff to be a badge of honor…
I have followed your exchanges with Goff with great interest. I think Goff suffers from what I call epistemic avarice, an unfounded expectation that there be satisfying explanations for things. But I also think he's an honest broker overall. His friendship with Keith Frankish, and Disagreeableme and his encouragement of his own graduate students to publish against panspychism makes me think he's not a bad egg. I think he's worth engaging. Though his recent turn to heretical theism kind of gives up the game he's always been playing.
Snake, I agree, I think Goff genuinely believes what he says and is generally a nice guy. But I still think he's seriously misguided about the nature of evidence and the weight of non-empirical metaphysical claims. Interesting about his turn to heretical theism, I did not know about it, and yes, it kind of gives the game away...
Excellent article. A lost opportunity of sorts for the french, they should have called him " David Humaine" .
👍
I will have to challenge you on anecdotal evidence. In short, the type of empiricial evidence you are talking about is to support or refute propositions that are universally quanitifed.
Example: all stars greater than 30 stellar masses collapse to a black hole.
Anecdotal evidence helps to validate a universally quantified proposition but it is limited.
But when it comes to an existentially quantified propostion, anecdotal evidence is all you've got.
Example: there exists a Savior
(Luke 1:1-3)
1 Many have undertaken to arrange a record of the events that have been fulfilled among us
2 just as those who from the beginning were eyewitnesses and servants of the word passed them down to us
3 it also seemed good to me, having carefully followed everything from the start, to write an orderly account for you, most noble Theophilus.
This type of anecdotal evidence can be validated, ironically, through universally quantified propositions. I will, in that case refer you to the Federal Rules for Evidence:
https://www.law.cornell.edu/rules/fre
Antony, if I understand you correctly, you just validated my point. I don't believe in any savior, precisely because his existence is predicated entirely on the basis of witnesses / anectodal evidence, which is far too weak to support the extraordinary claim being made.
As for the federal rules, I don't recognize the government as an authority in epistemology. For instance: "Rule 610. Religious Beliefs or Opinions." Nope.
I didn't really validate your point. I just used a negative example. The reason I did that was to call in the legal rules of evidence because they are a starting point for validating (or in this case invalidating) anecdotal evidence. This determination is different from that of evidence to validate a univerally quantified proposition.
For a positive case, I asked my friend and advisor claude.ai this question (since I am no history expert):
https://claude.ai/share/c818669a-e92f-4e8c-bcd4-1b685741804f
"What evidence do we have to prove that Julius Caesar was assassinated?"
Claude comes up with 5 cases of "Anecdotal Evidence": Cicero, Nicolaus of Damascus, Suetonius, Plutarh and Appian. Cluade then gives the criterion by which you can judge this anecdotal evidence to show that this event occurred.
What makes this particularly convincing is the convergence of independent sources on core facts:
The date (Ides of March)
The location (Theater of Pompey, during a Senate meeting)
The key conspirators (Brutus, Cassius, and dozens of others)
The manner (23 stab wounds)
The immediate aftermath (conspirators fleeing to the Capitol)
The sources sometimes disagree on minor details (exact words spoken, specific sequences), which actually suggests they're drawing on different authentic traditions rather than copying each other.
---
Anecdotal evidence usually implies a human to tell the anecdote. In the sciences, we tend to strive for universally quantified propositions. It is interesting to ask if there is an equivalent in science to anecdotal evidence for an existentially quantified proposition. I can come up with two examples:
1. The 5 degree Kelvin Cosmic Background Radiation from the Big Bang
2. The iridium boundary that indicated that a meteor struck the earth around the time the dinosaurs disappeared.
I would suggest that these to events are attested to by multiple readings, but they may be impersonal examples of "anecdotal evidence". They tell a tale of a particular event. They are not like the observational tables of Tycho Brahe which are essentially not limited to a particular event.
As to the Federal Rules for evidence, I would respectfully submit that they are a good example of applied epistemology in regards to anecdotal evidence. If you want a more "scientific" approach (I am not using scare quotes here, I mean to emphasize that this is legitimate science versus ad hoc legal rules) I asked Claude this question:
https://claude.ai/share/22d80334-403e-4d8e-b0c3-3947adc06187
"What psychologists have studied the reliability of eyewitness testimony in court?"
And it proceeded to tell me all of the psychological studies where eyewitness testimony is NOT reliable. I gave a follow-up question:
"These are studies of unreliability. I want studies that show when testimony is reliable."
And it then proceeded to list studies that indicated when anecdotal evidence is reliable:
"optimality hypothesis"—the idea that under optimal encoding conditions (good lighting, adequate viewing time, low stress), confidence and accuracy show significant positive correlations
witnesses can accurately recall stressful events even months later Simply Psychology, and that misleading questions don't necessarily corrupt memory formed under real-life conditions the way lab studies suggested.
retrieval effort cues as positive indicators—statements delivered fluently without hedges or delays are significantly more likely to be accurate
And then Claude notes:
"The philosophical tension here is striking: the field constructed its identity around exposing fallibility, perhaps because establishing positive reliability requires demonstrating ground truth—which is harder than showing error. The focus reveals an implicit epistemological bias toward skepticism over establishing conditions of justified confidence."
In short, sometimes anecdotal evidence is all you've got, especially for existentially quantified propositions. Instead of discounting all anecdotal evidence out of hand, it pays to determine when anecdotal evidence is likely to be reliable.
Antony, as a matter of courtesy I ask all my readers not to write comments that are of essay length, and not to quote extensively AI. I try my best to answer every comment, but there is a limit to the time I can spend doing so.
Historical documents and witnesses are not in the same category as anecdotal evidence. For one thing, you mention five, not one, source, and those sources are congruent. The rest of Claude's analysis strikes me as broadly correct, but not at all at odds with what I write in the essay we are discussing.
I would warn against telling Claude, or any other AI, to look for evidence to confirm your point of view. They are known to cherry pick or make shit up. My understanding of the relevant literature (and I looked, since I teach a course in criticla thinking) is that eyewitness testimony is considered of very low quality. Which doesn't mean it's useless. Same goes for anecdotes: low reliability, but not useless.
Contra to your claim, I don't discount all anecdotal evidence out of hand. I simply insist that it is very low on the rang of kinds of evidence, and that whenever possible we need to do better. It also depends on what exactly one wishes to prove by way of anecdotal evidence: I have anecdotally observed that I like red wine during the winter time. No problem. But if I told you that I have anecdotal evidence of miracles you should be properly skeptical.
What about the internet? Isn’t that evidence? 😂
Great essay! ✍️👍
🙏
Thanks. Helpful to remember facts count and how to try to sort.
Furst, AI has taken over media. Ascertaining by image ir speech pattern may not be enough. I fell into a cavern. Google no longer help. Pundits may simply be puny.
Second. A way to keep my brain functioning I read in the New England journal of medicine, as well as ground truth in Substack. A systemic analysis post discussed the possible difficulties with double blinded studies of medications lots of studies in GLP-1 meds if the study is to help in tension to treat, research needs to identify those who drop out or take drug after trial ended. Outliers are many. Perhaps few ever follow instructions. It requires researchers be more diligent Second, as a doubter, a study comparing one drug with another and without. From man places around the world . A editorial group asked why they had not evaluated with heart failure all men, Italian nephrologists of they did not in failing kidney the heart goes first. So important later, a physician sent in to editor and said the manufacturer of the second drug…. Drug vs Placebo—supported rhe research. My question, did they get data if bad outcomes and did not bother to post them. If the endpoint blood pressure, physicians should be aware of data as patients. . The NEJM editorial said include data in study, or do another with data. Thank the good David
Thank you Dr. Pigliucci for the clarification and explanation on evidence. As a beginner in critical thinking and philosophy it is important to understand what evidence is given and should be provided in day to day arguments. It is a learned skill to evaluate the evidence. Especially scientific and logical. Identifying logical fallacies, insufficient premises etc. takes time to learn. As does going to primary source scientific studies and making sure one is not Cherry picking, misrepresenting and misinterpreting the research. To me learning these critical thinking skills are crucial to develop "clear sighted" wisdom especially in our modern world.
Exactly!
Thanks Massimo,
I am leaving a question that I wanted to make in our upcoming seminar in Rome:
I’ve noticed that Stoics and even Academic Skeptics often rely on anecdotal evidence to anchor their arguments. Cicero, for instance, frequently cites virtuous or non-virtuous examples through legendary figures like Hercules, Romulus, Hippolytus, or Theseus. None of these figures actually existed; even if they had, they represent isolated cases rather than a statistically significant sample.
Seneca goes even further in De Clementia, using the behavior of bees to justify the role of a king when advising Nero: 'Nature herself has conceived the idea of a king, as you may learn from various animals, and especially from bees...'
While these cases obviously predate Hume and modern scientific inquiry, it makes me think that if we were to dismiss this type of anecdotal argumentation, their works would lose much of their power. There is also the element of pathos that these philosophers employ to truly persuade their readers.
Since this blog post aligns well with your own philosophical style. What is your approach to studying these authors, especially for your upcoming book on Cicero, where you will likely encounter many of these anecdotal arguments?
Demian, I think you have already at least implicitly answered the question: examples from mythology clearly cannot count as evidence, not even of the anecdotal type. They are rhetorical devices, and both Seneca and especially Cicero were excellent at rhetoric. They understood Aristotle's advice to accompany logos with ethos and pathos if one wishes to be persuasive.
As for actual anecdotal evidence, as I say in the essay, it does count, but at the lowest rang of the scale. It's a starting point, not the end of the matter. One of the reasons I practice Stoicism is because there is a lot of *modern* evidence, from systematic research in cognitive and behavioral science, that its general approach works.
Thanks Massimo! This makes sense.
As a follow-up question. To use an example:
If one were to justify the claim that 'virtue is the only good' based on the premise that the human telos is 'living according to nature,' one could take an empirical path. By relying on biological research into what distinguishes humans from other species, one might conclude that acting courageously, justly, temperately, and prudently is the best way to align with our prosocial and rational nature.
Alternatively, one could take a purely rational path, arguing that anything truly 'good' must be universal—there is no scenario where acting with such virtue could be considered 'bad.' Here you don't need to rely on science or empirical data (although you might have to, if you start to gather different scenarios to start to create a "sample").
This is an oversimplification, of course, but it illustrates the point. While both appear to be valid paths, would you prioritize the empirical approach over the purely rational one? I am also wondering if this style has to do with your own background (that combines science and philosophy), and if you see other academics still favoring the purely rational/logical one for philosophical inquiry.
Thank you.
Demian, I simply don't believe that reason by itself leads to discoveries about the nature of the world, and since I consider ethics to be an empirical discipline, no, I'm not satisfied with rational arguments by themselves. They need to be coupled with empirical evidence. Another way to put it is that ethics is about solving the problem of how human beings can live together and flourish. Surely this is an empirical issue, depending on human nature and the nature of human societies.
Thanks Massimo! This is also a good starting point for a subsequent discussion in Rome, so I am leaving more questions for the seminar directly. Thanks for taking the time to reply and see you very soon to continue discussing this and other topics.
Happy to be helpful! See you soon!
Although I certainly appreciate that the article explains that there are other kinds of evidence than empirical, there is also an implied reductionist argument that at the very least I find unhelpful. There is no doubt to any reasonable person that empirical evidence has expanded human knowledge exponentially. But there is also a lack of recognition of what we have lost in abandoning a more holistic approach that split disciplines into neat categories.
There are many other forms of evidence. Cumulative evidence arising from aggregation or, although maybe weak on its own, argumentum ex silentio. There is also coherence evidence, which may be evidentially relevant and contextual evidence where background knowledge constrains plausibility independently of new data. Comparative or pattern evidence is absent. Historical evidence is reduced to authority when it is far more than that. Even modal considerations can add evidential force to explanations. Abductive reasoning is more than Bayesian updating and has its own evidential mode.
Science has replaced philosophy as the dominant authority on truth. Yet science is not oriented towards wisdom. The universal approach sought understanding that informed how to live. Science only shows how explain or manipulate. We seem to have gained precision and vast amounts of knowledge but to do what with? We seem to have lost integration, any sense of necessity and more importantly any shared account of why any of it matters. This is probably why we see the world on such a destructive path. We blind our understanding by granting authority to science without bounds, instead of recognising it as the remarkable tool that it is, rather than as a source of human meaning.
Great post and I am also enjoying the comments section. As I read this, I thought about the discussions that I have with friend every few months where we debate politics and life. In our discussions, we often reach an impasse. We actually wade through these different types of evidence. Empirical and logic evidence are often dismissed in favor of anecdotal and authoritarian claims. The impasse occurs when I challenge his authorities. For example, JFK Jr is not an authority.
Just for the record, I know that I am right and he is wrong. :-)
Sakshat, ah, that touches on the very interesting and complex issues of which authorities one trusts and why. I would say, again, that a reasonable answer should depend on the available evidence. And the evidence is pretty clear: JFK Jr. is a dangerous lunatic.
Edgar, you seriously think the problem with the world is too much science?
All the other kinds of evidence you mention are sub-categories of the ones I list, so I honestly don't think I missed anything.
As for reductionism, setting aside that it is simply a method that breaks down big problems into digestible chunks, and is therefore extrenely useful, even "holistic" approaches ought to be based on evidence of some kind.
Wisdom too is no exception: unless based on evidence it is utter and dangerous nonsense, as seen in plenty of self-professed gurus and prophets throughout the history of the world.
So give us an expanded table, please! The one offered is insufficient.
Do I have to do all the work?? 🤣
Not too much science, but too much authority granted to science, especially when it is asked to answer questions about why, what matters and how we ought to act or think.
I am not arguing against the use of evidence, but against an overly narrow account of what counts as evidence.
For me wisdom is not technical know how. It is what we do with what we know, believe, experience and think. Without that orienting function knowledge alone tells us nothing about how to live.
Edgar, could you give me some examples of this alleged excessive authority given to science? Because much of what I see is the opposite: people making decisions, at all levels, with the utmost disregard for scientific evidence.
Again, my account of evidence is not narrow. All your counter-examples can be treated as sub-categories within the ones I have outlined.
I agree that wisdom is not technical know-how. But then what is it? If it is the ability to act appropriately under whatever circumstances, don't you think that ability is fed by evidence?
Science or empirical evidence is routinely mistaken for, or presented as, the truth itself. But science is provisional. It always has been. Worse, much of what is presented as “science” in public discourse is correlation drawn from small or limited studies, then promoted as decisive.
As truth, it has been given the authority to invoke and justify actions and decisions that are, in reality, ethical, political, or economic choices. It is used to privilege one claim over another for particular ends. It is used by governments to justify policies, by corporations to justify actions, by courts to justify decisions putting innocent people away. There are thousands of examples of each today but seen even clearer historicaly, which I leave nameless as I do not want to get into the politics of these.
If you are asking for named cases, that becomes a political argument rather than a philosophical one, which is exactly the authority problem I am pointing to.
What I am saying is that science and empirical data are evidence. There are other forms of evidence. Science is probably the best kind for technical know how, but it is not the most useful for all enquiry. Science and data are provisional and if we think of them as the truth they can justify anything. This is because science is partial, incomplete and provisional and always will be. Does that mean we ignore science or empirical data? Of course it does not. But neither should we give it undue authority.
Although I understand clearly the concept of using what we do know and have to understand the world, and science gives us so much of that, it is not sufficient on its own to understand the world or our place in it. And furthermore, it is dangerous to do so.
If you are interested, I wrote an article a while back on how I see this:
https://therationalcosmos.substack.com/p/the-truth-of-science-and-the-science
That takes a look at a lot of this as I am limited for space here.
Edgar, you keep making generic statements without examples, so all I can do is to say that I disagree, I don't see too much deference to science in modern society, especially in the US. Quite the opposite, I see quacks and pseudoscientists and bullshitters at the highest levels of government.
As for scientific conclusions being provisional etc., I doubt any scientist or philosopher of science would disagree. So what? Do you have access to some other kind of knowledge that is not provisional? If so, please share it with the rest of us.
Should political, ethical and economic decisions be made on the basis of the best science? You bet! On what else would you base them? We both known that science is in the business of answering questions of facts, not of value, but the two are deeply interrelated, unless one plucks one's values out of nowhere and for no particular reason. Which I wouldn't recommend.
So are you defining the science you disagree with as that of “quacks and pseudoscientists and bullshitters”? If that is the case then we are not really talking about science at all. We are talking about about failures of judgement and authority. Who gets to decide what counts as “best”.
Yes, all knowledge is provisional. Only pointing out that science is too and therefore must be understood and used within its limits. Provisionally was not a criticism only a description. Science establishes facts and causal relations. Explaining what is the case and what happens if we intervene.
The existence of scientific evidence that something works does not explain what it should be done or what is “best”. When you ask on what else such decisions should be based my answer is not about certainty only practical rational judgement about human flourishing. Judgements grounded in forms of evidence that science does not ask. We can look to diachronic evidence. What sustains rational agency? Intersubjective evidence so we can decide what is presupposed for moral justification or social cooperation. Coherence is a nice one if something can be applied consistently without undermining. And not limited to these either.
They include long-term historical experience of what corrodes or sustains human communities. The cross cultural stability of human “goods” and the internal coherence of principles when applied consistently over time. Or whether a principle can be justified to those subject to it as rational beings rather than imposed on them as systems to be managed.
Science informs these judgements and constrains what is possible. It tells us what works and at what cost. But it does not tell us which costs are acceptable or which ends are worth pursuing or how a human life should be oriented.
Science one criterion of evidence for how we judge, act, and think and an important one. But it is not sufficient for questions of value or meaning or how to live well.
It seems to me to be a question of what has primacy - science or philosophy. Where ultimately I see it as the sum is greater than the parts.
As I argued in my Answers for Aristotle, there is no need for primacy: the best approach is a combination of science and philosophy. As I note in the essay, logical arguments (the bread and butter of philosophy) are a type of evidence, just like empirical data (the bread and butter of science).
Hume is a deserved hero in Scotland. A great gastronomic giant too!
Le bon Davide!
The consideration of evdence, no matter how convincing is foolish without considering alternative hypotheses. Thus JS Mill urges understanding you opponents position in order to understand your own. This of course is a matter of logic (probability theory) not merely normative. The conversation beween Jordan Peterson and Cathy Newman is a wonderful example of seeing no need for alternative hypotheses due to ideological blindness.
Terry, well, yes, but I fail to consider a given hypothesis it is up to my collaborators or even opponents to put forth alternatives for me to consider. So long as they are backed up by evidence.
As for Peterson, sorry, I have absolutely no stomach to watch him.
Ah you miss something special - Cathy is very aggresive and carries the arrogance of ideology. She is tripped up by simple logic.
As opposed to Peterson??
yes in this instance. But do you have particulars in mind so I can grasp the issue with Jordan P as you see it?
Terry, oh where to start?? Here, for instance: https://donaldrobertson.substack.com/p/could-jordan-peterson-be-the-messiah
this essay is paywalled, however, I've discovered other essays by Donald on JP. I can't evaluate his criticism of the books - I have 12 rules and its sequel but have been unable to get through them. Meanwhile his conversations favourably reveal his moral compass for example with Abigail Shrier, Nigel Biggar, Lee Jussim, Andrew Doyle, Gad Saad, Katherine Birbalsingh, Jonathan Haidt, Greg Lukianoff ... I pass by the religious ones. I have never thought of JP as a Stoic - some people do apparently.