My story about how I recently left three cults
I will admit something: I recently left three different Internet groups
which seem to frequently engage in cult-like reasoning. Those are
anarcho-capitalists, stereotypical Internet vegans, and the people who
believe they are scientifically studying the names of places. Some of
the events that led to me to leave them happened in my head, most of
them have to do with discussions I had on the Internet forums.
Then some very smart guy told me: "Look, I agree with you that lockdowns and mask mandates are relatively weak arguments for the existence of government, but not all public-health arguments for the existence of government are that weak. What do you think about superbacteria?".
I responded like anarchists usually respond: "Superbacteria are a very temporary problem. It's mostly caused by animal agriculture, since around 80% of antibiotics go to farm animals. Lab-grown animal products will soon mostly-solve that problem. And you can do your part even now by being a vegetarian.".
That person told me: "But, see, lab-grown animal products will not solve that problem any time soon. Most of the antibiotics used in agriculture go to the egg industry, and we will not have lab-grown eggs any time soon. We sturggle to even produce muscle meat in laboratories, and eggs are far more complicated than muscles. And, for the exact same reason, many people going vegetarian would also not address the problem. For all we know, it might even have a negative effect because, well, if meat is produced in a lab or not not eaten at all, that removes the incentive to kill old laying hens, so they might end up being pumped with antibiotics for longer than they are now.".
I did not believe him. I was under a strong impression that most of the antibiotics these days go to cows and pigs.
That person told me: "Well, the estimate cited on the Wikipedia article about AMR says that around 70% of natibiotics are used in the egg industry. And other estimates are not a lot lower. You need to take into account that 45% of antibiotics we use today are ionophores, which are antibiotics effective in birds, but not in mammals.".
I admitted my mistake.
And that shaked me both out of my anarchism and out of my delusion that vegetarianism would solve most of the world's problems. I went on to discuss that fact within my cults. The responses I got there were absolutely pathetic. When I tried discussing superbacteria on an Internet forum about anarchism, I got a lot of responses that were the likes of: "Governments have murdered over a hundred million people in the 20th century, but, according to a brainwashed statist like yourself, this is all justified because muh superbacteria!". Look, I think the denial that the massive use of antibiotics in the egg industry leads to superbacteria is even more dangerous than Lysenkoism (the pseudoscience that killed tens of millions of people by causing ecological catastrophies) because Lysenko could indeed defend himself with "I didn't know.", but everybody who understands elementary school biology realizes that abuse of antibiotics has to lead to superbacteria. And on vegan forums, I got hated for saying that superbacteria are more important than global warming and that eggs are more of a threat than meat. Why? I assume that's because being against eggs sounds like you are being a right-winger that is also against abortion. Vegans are mostly left-wing. And hardly anybody is rational enough to realize that being against eggs and being against abortion is not remotely similar. I made a meme about that:
I thought almost everybody except me was disconnected from the reality
when it comes to superbacteria, because almost everybody believes that
most antibiotics go to humans. The truth is, what I believed, that
lab-grown meat will soon resolve the problem of superbacteria, is only
slightly less disconnected from the reality than what most people
believe.
When libertarians and anarchists say they will tackle the problem of superbacteria with lab-grown animal products, that's every bit of an empty and potentially dangerous rhetoric as when the greens and the liberals say they will tackle global warming by powering the national grid using solar power and wind power. I've made a list of such political rhetorics which are empty and potentially dangerous because of engineering problems. I think it's one of the biggest problems when discussing politics.
Look, I still agree with anarcho-capitalists that the vast majority of the laws we have are either unnecessary or harmful. It's difficult to imagine that 80'000 pages of regulation (which is how many federal regulations there are in the US) are not too much. But that doesn't mean that replacing the government with some vaguely imagined unicorns which have no obvious way of enforcing actually good laws (such as laws against the abuse of antibiotics in the egg industry or the laws against open DNS servers) does any good.
I've noticed that, in various debates on Internet forums, people are using relatively good arguments, but, when asked to clarify them, they are unable to present them properly. That has misled me quite a few times into thinking those arguments are invalid, when they are in fact excellent. For example, using the Internet as an argument against anarchism. Or, back when I was confused about the shape of the Earth, using GPS as evidence that the Earth is round and that space travel is possible. Quite a few people understand that GPS as we know it couldn't work if the Earth was flat, but they are incapable of explaining why. In reality, there is a simple experiment one can do to prove that GPS does not work like Flat-Earthers claim it works: find a spot with only 3 GPS signals available and ask the device to approximate your location. It will be able to (not as accurately as with 4 signals, but still reasonably accurate). But if the Earth was flat and the GPS devices were receiving signals from some secret emitters on the ground, it would not be able to. When you know your distance from 3 points in a 3D space, you can calculate 2 very-far-away points where you might be. Because the Earth is round and emitters are very high up in the sky, the GPS device can elliminate the point that's above the satellites as impossible. If the emitters were on the ground, it would not be able to do that: how could it know if it's above or below them? Another example of such a situation on the Internet forums is with anarchism and superbacteria: plenty of people say that sane government regulation is necessary to prevent superbacteria, but, if you ask them the obvious question "Isn't investing in lab-grown animal products a better solution?", they will not respond sensibly (many people will insist that antibiotics given to animals are somehow a lot less important than antibiotics given to humans when it comes to superbacteria, even if science is quite clear that the biggest cause of superbacteria in humans are trace amounts of antibiotics found in the eggs). I don't know what's the appropriate response if somebody on an Internet forum is using what might be a good argument, but is unable to present it properly.
So I went to discuss vegetarianism on an Internet forum. And, of course, I complained about factory farmed cows causing superbacteria.
Somebody told me: "So, why not switch to grass-fed cows? They require a lot fewer antibiotics."
I responded with: "Because grass-fed cows emit 3 times as much methane per a litre of milk than grain-fed cows do. Methane emissions of a cow are approximately proportional to her cellulose intake. Grass contains very little starch and a lot of cellulose. And the feed-conversion-ratio is a lot higher for grass-fed cows than for grain-fed cows.".
He told me: "Grass-fed cows emit more methane than grain-fed cows, but the difference cannot be anywhere near three times. Cows emit methane also when digesting starch, just less methane than when they digest cellulose. And, furthermore, the methanotrophic bacteria on pastures absorb a lot of methane that a cow emits before it goes up into the atmosphere.".
I responded with: "I think that's the only explanation, because you can see from this diagram that our methane emissions reached their peak in the 1980s (right when there were the most grass-fed cows around the world), and have been decreasing ever since. See how methane concentrations were growing a lot faster in the 1980s than they are growing now?".
Now, in my experience, that argument makes the vast majority of people
shut up.
But, thankfully, not this person. He told me: "Well, to me, this diagram looks almost exactly like what we would expect to see if our methane emissions haven't changed since the 1980s. Try doing a computer simulation, I think you will quickly understand why. All you need to know is that the half-life of the methane in the atmosphere is somewhere between 9 years and 12 years.".
I thought he was going crazy. I responded with: "What? This is an IT1-type system, do we agree? You know what the step response of a IT1-type system looks like? Not at all like the diagram I just showed you.".
Then he told me: "No, this is not an IT1-type system. Try deriving the
transfer function. You know what the impulse response is, right? Well,
you can derive a transfer function from the impulse response, as a
Laplace transform of the impulse response.".
And, indeed, after a bit of calculating, I got that the transfer function is:
. That's not a transfer function of an IT1-type system, and
the step response (printed by Octave) looks very similar to that diagram
of empirically measured concentrations of methane in the atmosphere over
time.
I admitted my mistake.
I decided to do a more complicated computer simulation, just to be sure. What might our methane emissions have been? Well, if you assume the half-life of the methane is 12 years, you get this diagram:
Only a slight downward trend, right?
And if you assume it is 9 years, you get this:
No discernable trend here, right?
So I went to discuss with vegans on another Internet forum, one about veganism. The end result? I got banned from that forum for a year and a half.
Look, I still think it's important to protect the environment, and that eliminating animal products from our diet, especially eggs, is a way to go. But the veganism on the Internet is not at all about it. It has grown into a cult that discourages rational conversation. And which, paradoxically, gives credit to factory farming that it doesn't deserve.
I think one of the reasons why people find that argument with superbacteria and this argument with factory farming and methane appealing is because they make the world's problems look like they are self-solving. "You don't need to do anything about superbacteria, and especially the government doesn't have to. Lab-grown meat will soon solve that problem.". Or "You or even the government doesn't need to do anything about methane in the atmosphere. The invisible hand of the market has basically already solved the problem via the switch from grass-fed cows to grain-fed cows.". Unfortunately (Or is it fortunately? I would not really like if the idea that torturing cows in factory farms saved us from global warming were true.), a little more research shows those contentions are not remotely accurate.
People are often using global warming denialism as an example of a harmful pseudoscience. The truth is, the belief that global warming is caused mostly by grass-fed cows (and that therefore factory farming is the only realistic solution to global warming) is even more harmful. Partly because talking about how the temperature data has been manipulated to show warming where there is no real warming triggers almost everyone's BS detectors, whereas implying that factory farming saved us from global warming does not trigger everybody's BS detectors.
A year ago, when I wrote a blog-post on what not to do when doing vegegelicism, I was thinking that the very worst thing you can do while doing vegegelicism is to yell unrelated fringe claims such as Flat-Earthism, radical anarchism, or global warming denialism, in other words, things that trigger the BS detectors in most of the people. And I still think that's a wrong thing to do. But now I think there is something way worse: spreading misinformation which does not trigger the BS detectors of most people, such as implying that factory farming saved us from global warming.
So, there seem to be two basic methodologies of studying the names of places competing with one another. One is Antun Mayer's methodology and the other is the Hans Krahe's methodology. Antun Mayer's methodology is based on the assumption that the etymologies from languages we know a lot about are somehow more probable than etymologies from languages we know little about. For example that, other things being equal, the commonly cited etymology that the river name Karašica comes from Latin fish name carassius is a lot more probable than my etymology that Karašica comes from Illyrian *Kurr-urr-issia (flow-water-suffix), because we know a lot about Latin and very little about Illyrian. Hans Krake's methodology is based on the assumption that repetition of same or similar elements in some meaning is evidence that the names of places come from the same language. For example, where Scythians lived in antiquity, many river names start with consonants 'd' and 'n' (sometimes separated by a vowel): Danube, Don, Dniester, and Dnieper. That, according to Krahe, strongly suggests that this d-n was the Scythian word for "to flow". Or, putting that methodology to the extreme (I don't know what Krahe would have thought about that), the fact that many river names in Croatia start with consonants 'k' and 'r' suggests that this k-r was the Illyrian word for "to flow": Krka, Krbavica, Kravarščica, Krapina, Korana, and two rivers named Karašica. The Mayer's methodology is, for a reason that escapes me, a lot more mainstream.
I had known for a long time that there seems to be one quirky mathematical problem with the Krahe's methodology: Krahe seems to regularly give Indo-European etymologies to the patterns he finds in the European river names. But a simple calculation that takes into account how stable river names are strongly suggests most of the river names in Europe date back to before 8'000 BC. That probably excludes the possibility of them being Indo-European in origin, since there were probably no Indo-European languages spoken in Europe at the time. But that fact didn't bother me too much.
After I passed my Information Theory course at the university, an idea occurred to me: "Hang on a second. The Mayers's methodology does not appear to be based on mathematics, but the Krahe's methodology appears to be. The patterns in names of places such as the k-r pattern in the Croatian river names have p-values, and that p-value can be calculated using the Collision Entropy and Birthday Paradox.". So I did the collision entropy measurements of the Aspell word-list for the Croatian language, and the Monte Carlo calculations which take those measurements and the Birthday Paradox into account, and I came to the conclusion that the p-value of that k-r pattern is somewhere between 1/300 and 1/17. And I published a paper about it, called "Etimologija Karašica", in two peer-reviewed (well, nominally peer-reviewed, at least) journals: Valpovački Godišnjak and Regionalne Studije.
I went on to discuss my findings on Internet forums where the toponymy enthusiasts meet. And I was hated there. They were giving me all kinds of awkward responses. For example, two people independent of one another told me: "What if nouns have a significantly lower collision entropy than the rest of the words in the Aspell word-list? River names are nouns.". When I asked them: "How could that even work? I can see how that might work in Swahili, where nouns can only start with one of 18 prefixes called noun classes, while verbs can start with however the phonotactics allows, but Croatian grammar is not remotely like that.", they pretended (in my opinion) to be unable to understand my response. One person told me: "Your supposed Proto-Slavic form of the name Karašica, *Kъrъrьsьja, is impossible because Proto-Slavic phonotactics didn't allow four consecutive syllables with yers. Most likely, *Kurr-urr-issia would have been borrowed as *Kъryrьsьja, which would give *Kriraša in modern Croatian.". When I asked him to give me some source making that extraordinary claim about Proto-Slavic phonotactics (that doesn't seem to be how phonotactics in general works, phonotactics of many languages does vowel harmony, which is the opposite of what that person suggested for Proto-Slavic), they couldn't do that, but that didn't convince them they were wrong. Not even asking them to explain the phonological evolution from civitate to Cavtat, which appears to involve a Proto-Slavic form with four consecutive yers, convinced them. In fact, when I asked that question about Cavtat, I got banned from the forum.
About a year later, I made a YouTube video about the Croatian river names in Latin (In case you cannot open it, try downloading this MP4 file and opening it in VLC or something similar), and I went on to discuss it on another Internet forum. And somebody told me: "To me it seems like your computer simulations are assuming that pairs of consonants which are common in the middle and at the end of a word are also common at the beginning of a word. That's not true, and, for the languages such as Croatian or English, that is especially not true. Croatian and English allow for many consonant clusters word-initially and word-finally, and those consonant clusters are governed, to a large extent, by the Law of Sonority. The Law of Sonority says, among other things, that consonant clusters which are common word-initially are rare word-finally and vice versa. Try to control for that, I assume you will get a p-value significantly higher than 1/17.". And, indeed, when I tried to control for that fact, I got that the p-value of the k-r pattern in the Croatian river names is around 85%, because consonant pairs at the beginning of a word have around 0.9 bits per consonant pair lower collision entropy than all the consonant pairs in the Aspell word-list.
So, it took me 9 years of studying toponyms to understand the complexity of the problem. That the Antun Mayer's methodology doesn't hold up to even elementary scrutiny (What would be the mathematical basis for that methodology? I don't see it.), and that the Hans Krahe's methodology appears to be based on information theory, so much so that a surface analysis suggests that information theory gives precise numbers (for example, that the p-value of that k-r pattern in the Croatian river names is between 1/300 and 1/17)... but a deeper analysis shows that it really isn't. Mayer's methodology is the linguistic equivalent of Flat-Earthism, whereas the Krahe's methodology (as well as the nearly-equivalent methodology I was using in the paper "Etimologija Karašica") is the linguistic equivalent of Creationism. Or, better say, the Mayer's methodology does not even work on paper, while the Krahe's methodology works on paper but not in reality.
It took me 9 years of studying names of places to realize what is obvious to most laymen: that there is no good scientific methodology for studying names of places. And, indeed, there is very little scientific consensus among those who study Illyrian toponyms. There are only a handful of toponyms which all the Illyrologists agree are Illyrian in origin and what they meant: that Serapia (ancient name for Bednja) was Illyrian for "flowing water", that Salapia (ancient name for the salt marsh Salpi in Apulia) was Illyrian for "salty water"... And those few toponyms that all Illyrologists agree on what they meant, they agree on them not because of some systematic study of the toponyms, but because of the harder science that is comparative Indo-European linguistics. Yet, the study of the toponyms was, in my mind (and in minds of quite a few people), highly dignified as science.
UPDATE on 13/07/2025: You know, I am no longer sure that comparative historical linguistics is a much harder science than the study of the names of places is. What really hurt me is this. In Latin, there were two words meaning "tear (of an eye)": dacrima and lacrima. The word "lacrima" prevailed, and its descendants are the words for tears in modern Romance languages. And there were two words meaning "tongue (the body part)" and "language": dingua and lingua. The word lingua prevailed. That is called Sabine L, and is usually explained as initial 'd' changing to 'l' in some dialects of Latin. Now, if you ask linguists: "How do you know it is not precisely the opposite? What if 'lingua' and 'lacrima' were older forms?", they will respond with: "We have comparative evidence. The cognates to Latin 'lacrima'/'dacrima' are English 'tear' and Greek 'δᾰ́κρῠ', and a cognate to Latin 'lingua'/'dingua' is English 'tongue'. That's how we know the forms starting with 'd' are older.". But there is an obvious question here: If English tear comes from the same Indo-European root as lacrima, why is it not spelt something like *teighr? The 'k' in the middle of a word in Latin and Greek clearly corresponds to 'gh' in English, as in the number "eight" (compare Latin "octo" and Greek "ὀκτώ") and the word "night" (Latin "nox" and Greek "νῠ́ξ"), so why wouldn't it do that in the word "tear"? Well, I asked that question on Latin StackExchange, where it received two downvotes and was closed by the moderator. The reason? It's off-topic since the answer obviously lies in the English historical phonology, rather than in Latin or Greek. OK, let's ask that same question on English StackExchange. The result? They are pretending (in my opinion) to be unable to understand my question and have closed it. Why do people react to such questions that way? Is it because the contention that "tear" and "lacrima" are cognates is considered to be a "basic fact" and that therefore doubting it is considered unscientific and anti-intellectual? Are such supposed facts based on evidence or on nothing more than groupthink? You be the judge. One moderator of the Latin Language StackExchange agrees with me that this is a relatively good question and he suggested me to post it on Linguistics StackExchange. I did, but I am not at all sure that's the appropriate response. Cross-posting onto another StackExchange is almost never a good idea.
You can avoid many cults by asking yourself: "Are the arguments they are using not-even-wrong?", and while this works against Flat-Earthism (most of the commonly-used arguments for Flat-Earthism, such as "The horizon appears to be rising with the observer as he climbs.", are not-even-wrong: it's not at all obvious how the Earth being flat and there being magical curvings of light explains that phaenomenon), it doesn't really work against the cults I was until recently a part of (asking "How it is that our methane emissions have been decreasing over time?" with the implication it's because of the number of grass-fed cows decreasing is hardly not-even-wrong).
Not even asking experts always works: many of my professors at the Computer Engineering programme, who knew far more about information theory than I did, told me that the arguments I presented in my paper "Etimologija Karašica" sound compelling to them. It's just that there is a technical detail in the field that connects information theory and linguistics that makes my calculations invalid, and not even my professors happened to know that tehnical detail.
I have long realized that the knowledge which is taught in middle school and high school was counter-productive in detecting pseudoscience, but I hoped the knowledge taught at the university would be better. In reality, it's every bit as useless and counter-productive. In our cybernetics classes we are taught that if some system is "slightly less than an integral", it's probably an IT1-type system, and that piece of knowledge led me astray with the methane. And this information theory... give me a break. It gives precise numbers (for example, that the p-value of that k-r pattern is between 1/300 and 1/17), but those numbers are wildly wrong. Hardly any piece of "knowledge" is that misleading. You can argue that computer engineeering is not building those flawed intuitions (I believed methane emissions were decreasing long before I learned basic cybernetics.), but it's undeniable that it gives people the ability to appear to mathematically justify those flawed intuitions.
People say that debating is a cure for irrationality. The truth is, if you want to be sure some argument is correct, you probably need to have the same tired old debate on many different Internet forums, to make sure that really nobody notices the fallacy. On most Internet forums, the vast majority of people, when presented with that argument "This diagram shows our methane emissions have been decreasing." or "Lab-grown meat will soon solve the problem of superbacteria.", will either shut up or respond with some incomprehensible word-salad. That will probably mislead you into thinking that your argument is correct. It goes without saying that an Internet debate going so well like the ones I described above is an exceedingly rare event. Usually debates turn into insults. And sometimes you need to be creative to provoke a productive discussion. It seems logical that a way to provoke a productive discussion about Croatian toponyms is to write papers and make YouTube videos in the Croatian language. But I've seen time and time again that didn't work. It's only after I made a video about Croatian toponyms in the Latin language that a productive discussion started. Perhaps it's sometimes necessary to talk about dead languages using dead languages (for example, talking about Illyrian in Latin), so that you attract people who are actually knowledgeable about the topic.
It should go without saying that quite often debates are not productive because both sides of the debate are incredibly disconnected from the reality. The debates about superbacteria are unlikely to be productive because most people wrongly believe most antibiotics go to humans, and that the main reason for superbacteria are physicians misprescribing antibiotics, and those who realize that's not true (exempli gratia, most vegetarians) mostly wrongly believe most antibiotics go to cows and pigs, and it seems plausible to them that lab-grown meat will soon solve the problem. Or like in the debates about Croatian toponyms, those who believe the Krahe's methodology is justified by the basic information theory (like I used to believe until recently) are only slightly less disconnected from the reality than those who think the Mayer's methodology is good. Debates in that case are almost guaranteed to be counter-productive.
If you are a Flat-Earther using not-even-wrong arguments such as "Why does the horizon appear to rise with the observer if he climbs? If the Earth was round, wouldn't we expect it to be falling?", once you get outside of your echo-chamber, you will actually probably quickly get correct responses such as "Until the Flat-Earth Theory gives us the math which predicts those phaenomena, we do not owe you that answer." (which is actually unlikely to convince you) or even an explanation of how basic trigonometry predicts that the angle at which you see the horizon is very small and how human perception doesn't work that way to be able to detect it (and that explanation is likely to convince you). But once you try debating the arguments such as "With the methods one learns in Information Theory course at the university, one can estimate that the p-value of that k-r pattern in the Croatian river names is somewhere between 1/300 and 1/17.", don't be surprised if you have to discuss it with dozens of different people on various forums before you get a useful response (that those methods fail because of the Sonority Sequencing Principle).
Maybe my mental illness has something to do with me not seeing how those arguments used to attract people into those cults ("Lab-grown meat will soon solve the problem of superbacteria, and you can do your part already by going vegetarian.", "Why have our methane emissions decreased, even though we are using more methane than ever? It has to be because most of our methane emissions come from grass-fed cows, which emit 3 times as much methane as grain-fed cows.", "Basic information theory says that the p-value of that k-r pattern in the Croatian river names is somewhere between 1/300 and 1/17. It is statistically significant. It's only logical that there was a word *karr~kurr in Illyrian meaning to flow, and therefore that Karašica comes from Illyrian *Kurr-urr-issia (flow-water-suffix)."...) are faulty. I have psychosis and I need to take Risperidone, Biperiden, and Alprazolam.
Table of Content
- Realization that lab-grown meat and vegetarianism do not address superbacteria
- Realization that the Internet needs a government to work
- Realization that methane emissions haven't decreased and that factory farming didn't save us from global warming
- Realization that the study of names of places is a soft science at best
- Conclusion
Realization that lab-grown meat and vegetarianism do not address superbacteria
One discussion on an Internet forum that made me question both my anarcho-capitalism and my stereotypical Internet veganism is this. It was year 2021, and I was advocating radical anarchism on an Internet forum. Of course, since it was the height of the COVID-19 pandemic, the conversation quickly devolved into a discussion about lockdowns and mask mandates. Most of the people were telling me that the government is necessary in order to enforce lockdowns and mask mandates during a pandemic, while I was citing statistics supposedly showing lockdowns and mask mandates do not work.Then some very smart guy told me: "Look, I agree with you that lockdowns and mask mandates are relatively weak arguments for the existence of government, but not all public-health arguments for the existence of government are that weak. What do you think about superbacteria?".
I responded like anarchists usually respond: "Superbacteria are a very temporary problem. It's mostly caused by animal agriculture, since around 80% of antibiotics go to farm animals. Lab-grown animal products will soon mostly-solve that problem. And you can do your part even now by being a vegetarian.".
That person told me: "But, see, lab-grown animal products will not solve that problem any time soon. Most of the antibiotics used in agriculture go to the egg industry, and we will not have lab-grown eggs any time soon. We sturggle to even produce muscle meat in laboratories, and eggs are far more complicated than muscles. And, for the exact same reason, many people going vegetarian would also not address the problem. For all we know, it might even have a negative effect because, well, if meat is produced in a lab or not not eaten at all, that removes the incentive to kill old laying hens, so they might end up being pumped with antibiotics for longer than they are now.".
I did not believe him. I was under a strong impression that most of the antibiotics these days go to cows and pigs.
That person told me: "Well, the estimate cited on the Wikipedia article about AMR says that around 70% of natibiotics are used in the egg industry. And other estimates are not a lot lower. You need to take into account that 45% of antibiotics we use today are ionophores, which are antibiotics effective in birds, but not in mammals.".
I admitted my mistake.
And that shaked me both out of my anarchism and out of my delusion that vegetarianism would solve most of the world's problems. I went on to discuss that fact within my cults. The responses I got there were absolutely pathetic. When I tried discussing superbacteria on an Internet forum about anarchism, I got a lot of responses that were the likes of: "Governments have murdered over a hundred million people in the 20th century, but, according to a brainwashed statist like yourself, this is all justified because muh superbacteria!". Look, I think the denial that the massive use of antibiotics in the egg industry leads to superbacteria is even more dangerous than Lysenkoism (the pseudoscience that killed tens of millions of people by causing ecological catastrophies) because Lysenko could indeed defend himself with "I didn't know.", but everybody who understands elementary school biology realizes that abuse of antibiotics has to lead to superbacteria. And on vegan forums, I got hated for saying that superbacteria are more important than global warming and that eggs are more of a threat than meat. Why? I assume that's because being against eggs sounds like you are being a right-winger that is also against abortion. Vegans are mostly left-wing. And hardly anybody is rational enough to realize that being against eggs and being against abortion is not remotely similar. I made a meme about that:
I thought almost everybody except me was disconnected from the reality
when it comes to superbacteria, because almost everybody believes that
most antibiotics go to humans. The truth is, what I believed, that
lab-grown meat will soon resolve the problem of superbacteria, is only
slightly less disconnected from the reality than what most people
believe.When libertarians and anarchists say they will tackle the problem of superbacteria with lab-grown animal products, that's every bit of an empty and potentially dangerous rhetoric as when the greens and the liberals say they will tackle global warming by powering the national grid using solar power and wind power. I've made a list of such political rhetorics which are empty and potentially dangerous because of engineering problems. I think it's one of the biggest problems when discussing politics.
Realization that the Internet needs a government to work
Back when I was an anarchist, many people were telling me the Internet needs a government to work. However, they were not able to present that argument properly. If you asked them how exactly the Internet relies on there being a government to work, they would give you some incomprehensible word-salad. Then, after I passed the course of Computer Networks on the university, I realized: "Hang on a second, what would happen if there were no laws requiring ISPs that, if they set up an unencrypted DNS server, they make it respond only to the IP addresses it is supposed to serve, rather than to requests from all IP addresses? Well, the answer seems obvious: some ISPs would set up their DNS servers improperly, and, since DNS servers sometimes respond to short queries with long responses (up to 4 UDP packets to one-packet-long DNS requests), the hackers would paralyze the Internet by spoofing their IP addresses and making those long-response-causing DNS requests to the improperly configured DNS servers.". That's but one of the reasons the Internet needs a government. I went on to discuss it on an anarchist forum, and most of the people there were convinced that your ISP can somehow protect you against denial-of-service attacks caused by DNS reflection. When I asked them how exactly might your ISP do that, they could not give a coherent response.Look, I still agree with anarcho-capitalists that the vast majority of the laws we have are either unnecessary or harmful. It's difficult to imagine that 80'000 pages of regulation (which is how many federal regulations there are in the US) are not too much. But that doesn't mean that replacing the government with some vaguely imagined unicorns which have no obvious way of enforcing actually good laws (such as laws against the abuse of antibiotics in the egg industry or the laws against open DNS servers) does any good.
I've noticed that, in various debates on Internet forums, people are using relatively good arguments, but, when asked to clarify them, they are unable to present them properly. That has misled me quite a few times into thinking those arguments are invalid, when they are in fact excellent. For example, using the Internet as an argument against anarchism. Or, back when I was confused about the shape of the Earth, using GPS as evidence that the Earth is round and that space travel is possible. Quite a few people understand that GPS as we know it couldn't work if the Earth was flat, but they are incapable of explaining why. In reality, there is a simple experiment one can do to prove that GPS does not work like Flat-Earthers claim it works: find a spot with only 3 GPS signals available and ask the device to approximate your location. It will be able to (not as accurately as with 4 signals, but still reasonably accurate). But if the Earth was flat and the GPS devices were receiving signals from some secret emitters on the ground, it would not be able to. When you know your distance from 3 points in a 3D space, you can calculate 2 very-far-away points where you might be. Because the Earth is round and emitters are very high up in the sky, the GPS device can elliminate the point that's above the satellites as impossible. If the emitters were on the ground, it would not be able to do that: how could it know if it's above or below them? Another example of such a situation on the Internet forums is with anarchism and superbacteria: plenty of people say that sane government regulation is necessary to prevent superbacteria, but, if you ask them the obvious question "Isn't investing in lab-grown animal products a better solution?", they will not respond sensibly (many people will insist that antibiotics given to animals are somehow a lot less important than antibiotics given to humans when it comes to superbacteria, even if science is quite clear that the biggest cause of superbacteria in humans are trace amounts of antibiotics found in the eggs). I don't know what's the appropriate response if somebody on an Internet forum is using what might be a good argument, but is unable to present it properly.
Realization that our methane emissions didn't decrease
So, stereotypical Internet vegans believe that a switch from grain-fed cows to grass-fed cows is not feasible because grass-fed cows supposedly emit 3 times as much methane per a litre of milk, and that methane is a more important greenhouse gas than CO2. Furthermore, they believe that our methane emissions reached their peak somewhen in the 1980s and have been decreasing ever since, primarily because today we are using more grain-fed cows and less grass-fed cows. In other words, though they don't say that explicitly, that factory farming saved us from global warming.So I went to discuss vegetarianism on an Internet forum. And, of course, I complained about factory farmed cows causing superbacteria.
Somebody told me: "So, why not switch to grass-fed cows? They require a lot fewer antibiotics."
I responded with: "Because grass-fed cows emit 3 times as much methane per a litre of milk than grain-fed cows do. Methane emissions of a cow are approximately proportional to her cellulose intake. Grass contains very little starch and a lot of cellulose. And the feed-conversion-ratio is a lot higher for grass-fed cows than for grain-fed cows.".
He told me: "Grass-fed cows emit more methane than grain-fed cows, but the difference cannot be anywhere near three times. Cows emit methane also when digesting starch, just less methane than when they digest cellulose. And, furthermore, the methanotrophic bacteria on pastures absorb a lot of methane that a cow emits before it goes up into the atmosphere.".
I responded with: "I think that's the only explanation, because you can see from this diagram that our methane emissions reached their peak in the 1980s (right when there were the most grass-fed cows around the world), and have been decreasing ever since. See how methane concentrations were growing a lot faster in the 1980s than they are growing now?".
Now, in my experience, that argument makes the vast majority of people
shut up.But, thankfully, not this person. He told me: "Well, to me, this diagram looks almost exactly like what we would expect to see if our methane emissions haven't changed since the 1980s. Try doing a computer simulation, I think you will quickly understand why. All you need to know is that the half-life of the methane in the atmosphere is somewhere between 9 years and 12 years.".
I thought he was going crazy. I responded with: "What? This is an IT1-type system, do we agree? You know what the step response of a IT1-type system looks like? Not at all like the diagram I just showed you.".
Then he told me: "No, this is not an IT1-type system. Try deriving the
transfer function. You know what the impulse response is, right? Well,
you can derive a transfer function from the impulse response, as a
Laplace transform of the impulse response.".And, indeed, after a bit of calculating, I got that the transfer function is:
. That's not a transfer function of an IT1-type system, and
the step response (printed by Octave) looks very similar to that diagram
of empirically measured concentrations of methane in the atmosphere over
time.
I admitted my mistake.I decided to do a more complicated computer simulation, just to be sure. What might our methane emissions have been? Well, if you assume the half-life of the methane is 12 years, you get this diagram:
Only a slight downward trend, right?And if you assume it is 9 years, you get this:
No discernable trend here, right?So I went to discuss with vegans on another Internet forum, one about veganism. The end result? I got banned from that forum for a year and a half.
Look, I still think it's important to protect the environment, and that eliminating animal products from our diet, especially eggs, is a way to go. But the veganism on the Internet is not at all about it. It has grown into a cult that discourages rational conversation. And which, paradoxically, gives credit to factory farming that it doesn't deserve.
I think one of the reasons why people find that argument with superbacteria and this argument with factory farming and methane appealing is because they make the world's problems look like they are self-solving. "You don't need to do anything about superbacteria, and especially the government doesn't have to. Lab-grown meat will soon solve that problem.". Or "You or even the government doesn't need to do anything about methane in the atmosphere. The invisible hand of the market has basically already solved the problem via the switch from grass-fed cows to grain-fed cows.". Unfortunately (Or is it fortunately? I would not really like if the idea that torturing cows in factory farms saved us from global warming were true.), a little more research shows those contentions are not remotely accurate.
People are often using global warming denialism as an example of a harmful pseudoscience. The truth is, the belief that global warming is caused mostly by grass-fed cows (and that therefore factory farming is the only realistic solution to global warming) is even more harmful. Partly because talking about how the temperature data has been manipulated to show warming where there is no real warming triggers almost everyone's BS detectors, whereas implying that factory farming saved us from global warming does not trigger everybody's BS detectors.
A year ago, when I wrote a blog-post on what not to do when doing vegegelicism, I was thinking that the very worst thing you can do while doing vegegelicism is to yell unrelated fringe claims such as Flat-Earthism, radical anarchism, or global warming denialism, in other words, things that trigger the BS detectors in most of the people. And I still think that's a wrong thing to do. But now I think there is something way worse: spreading misinformation which does not trigger the BS detectors of most people, such as implying that factory farming saved us from global warming.
Realization that the study of names of places is a soft science at best
I used to believe with a rather cult-like mentality that the study of the names of places is a respectable science. And, unlike those other cults I became a part of, this one did not come from the Internet, but from my father. When I had almost finished writing my book "Jezici za gimnazijalce", my father suggested me to finish the linguistics part of that book with an overview of the Croatian names of places. I was immediately skeptical. My intuition told me that the names of places cannot really be studied scientifically. But I couldn't justify that intuition, so I gave in.So, there seem to be two basic methodologies of studying the names of places competing with one another. One is Antun Mayer's methodology and the other is the Hans Krahe's methodology. Antun Mayer's methodology is based on the assumption that the etymologies from languages we know a lot about are somehow more probable than etymologies from languages we know little about. For example that, other things being equal, the commonly cited etymology that the river name Karašica comes from Latin fish name carassius is a lot more probable than my etymology that Karašica comes from Illyrian *Kurr-urr-issia (flow-water-suffix), because we know a lot about Latin and very little about Illyrian. Hans Krake's methodology is based on the assumption that repetition of same or similar elements in some meaning is evidence that the names of places come from the same language. For example, where Scythians lived in antiquity, many river names start with consonants 'd' and 'n' (sometimes separated by a vowel): Danube, Don, Dniester, and Dnieper. That, according to Krahe, strongly suggests that this d-n was the Scythian word for "to flow". Or, putting that methodology to the extreme (I don't know what Krahe would have thought about that), the fact that many river names in Croatia start with consonants 'k' and 'r' suggests that this k-r was the Illyrian word for "to flow": Krka, Krbavica, Kravarščica, Krapina, Korana, and two rivers named Karašica. The Mayer's methodology is, for a reason that escapes me, a lot more mainstream.
I had known for a long time that there seems to be one quirky mathematical problem with the Krahe's methodology: Krahe seems to regularly give Indo-European etymologies to the patterns he finds in the European river names. But a simple calculation that takes into account how stable river names are strongly suggests most of the river names in Europe date back to before 8'000 BC. That probably excludes the possibility of them being Indo-European in origin, since there were probably no Indo-European languages spoken in Europe at the time. But that fact didn't bother me too much.
After I passed my Information Theory course at the university, an idea occurred to me: "Hang on a second. The Mayers's methodology does not appear to be based on mathematics, but the Krahe's methodology appears to be. The patterns in names of places such as the k-r pattern in the Croatian river names have p-values, and that p-value can be calculated using the Collision Entropy and Birthday Paradox.". So I did the collision entropy measurements of the Aspell word-list for the Croatian language, and the Monte Carlo calculations which take those measurements and the Birthday Paradox into account, and I came to the conclusion that the p-value of that k-r pattern is somewhere between 1/300 and 1/17. And I published a paper about it, called "Etimologija Karašica", in two peer-reviewed (well, nominally peer-reviewed, at least) journals: Valpovački Godišnjak and Regionalne Studije.
I went on to discuss my findings on Internet forums where the toponymy enthusiasts meet. And I was hated there. They were giving me all kinds of awkward responses. For example, two people independent of one another told me: "What if nouns have a significantly lower collision entropy than the rest of the words in the Aspell word-list? River names are nouns.". When I asked them: "How could that even work? I can see how that might work in Swahili, where nouns can only start with one of 18 prefixes called noun classes, while verbs can start with however the phonotactics allows, but Croatian grammar is not remotely like that.", they pretended (in my opinion) to be unable to understand my response. One person told me: "Your supposed Proto-Slavic form of the name Karašica, *Kъrъrьsьja, is impossible because Proto-Slavic phonotactics didn't allow four consecutive syllables with yers. Most likely, *Kurr-urr-issia would have been borrowed as *Kъryrьsьja, which would give *Kriraša in modern Croatian.". When I asked him to give me some source making that extraordinary claim about Proto-Slavic phonotactics (that doesn't seem to be how phonotactics in general works, phonotactics of many languages does vowel harmony, which is the opposite of what that person suggested for Proto-Slavic), they couldn't do that, but that didn't convince them they were wrong. Not even asking them to explain the phonological evolution from civitate to Cavtat, which appears to involve a Proto-Slavic form with four consecutive yers, convinced them. In fact, when I asked that question about Cavtat, I got banned from the forum.
About a year later, I made a YouTube video about the Croatian river names in Latin (In case you cannot open it, try downloading this MP4 file and opening it in VLC or something similar), and I went on to discuss it on another Internet forum. And somebody told me: "To me it seems like your computer simulations are assuming that pairs of consonants which are common in the middle and at the end of a word are also common at the beginning of a word. That's not true, and, for the languages such as Croatian or English, that is especially not true. Croatian and English allow for many consonant clusters word-initially and word-finally, and those consonant clusters are governed, to a large extent, by the Law of Sonority. The Law of Sonority says, among other things, that consonant clusters which are common word-initially are rare word-finally and vice versa. Try to control for that, I assume you will get a p-value significantly higher than 1/17.". And, indeed, when I tried to control for that fact, I got that the p-value of the k-r pattern in the Croatian river names is around 85%, because consonant pairs at the beginning of a word have around 0.9 bits per consonant pair lower collision entropy than all the consonant pairs in the Aspell word-list.
So, it took me 9 years of studying toponyms to understand the complexity of the problem. That the Antun Mayer's methodology doesn't hold up to even elementary scrutiny (What would be the mathematical basis for that methodology? I don't see it.), and that the Hans Krahe's methodology appears to be based on information theory, so much so that a surface analysis suggests that information theory gives precise numbers (for example, that the p-value of that k-r pattern in the Croatian river names is between 1/300 and 1/17)... but a deeper analysis shows that it really isn't. Mayer's methodology is the linguistic equivalent of Flat-Earthism, whereas the Krahe's methodology (as well as the nearly-equivalent methodology I was using in the paper "Etimologija Karašica") is the linguistic equivalent of Creationism. Or, better say, the Mayer's methodology does not even work on paper, while the Krahe's methodology works on paper but not in reality.
It took me 9 years of studying names of places to realize what is obvious to most laymen: that there is no good scientific methodology for studying names of places. And, indeed, there is very little scientific consensus among those who study Illyrian toponyms. There are only a handful of toponyms which all the Illyrologists agree are Illyrian in origin and what they meant: that Serapia (ancient name for Bednja) was Illyrian for "flowing water", that Salapia (ancient name for the salt marsh Salpi in Apulia) was Illyrian for "salty water"... And those few toponyms that all Illyrologists agree on what they meant, they agree on them not because of some systematic study of the toponyms, but because of the harder science that is comparative Indo-European linguistics. Yet, the study of the toponyms was, in my mind (and in minds of quite a few people), highly dignified as science.
UPDATE on 13/07/2025: You know, I am no longer sure that comparative historical linguistics is a much harder science than the study of the names of places is. What really hurt me is this. In Latin, there were two words meaning "tear (of an eye)": dacrima and lacrima. The word "lacrima" prevailed, and its descendants are the words for tears in modern Romance languages. And there were two words meaning "tongue (the body part)" and "language": dingua and lingua. The word lingua prevailed. That is called Sabine L, and is usually explained as initial 'd' changing to 'l' in some dialects of Latin. Now, if you ask linguists: "How do you know it is not precisely the opposite? What if 'lingua' and 'lacrima' were older forms?", they will respond with: "We have comparative evidence. The cognates to Latin 'lacrima'/'dacrima' are English 'tear' and Greek 'δᾰ́κρῠ', and a cognate to Latin 'lingua'/'dingua' is English 'tongue'. That's how we know the forms starting with 'd' are older.". But there is an obvious question here: If English tear comes from the same Indo-European root as lacrima, why is it not spelt something like *teighr? The 'k' in the middle of a word in Latin and Greek clearly corresponds to 'gh' in English, as in the number "eight" (compare Latin "octo" and Greek "ὀκτώ") and the word "night" (Latin "nox" and Greek "νῠ́ξ"), so why wouldn't it do that in the word "tear"? Well, I asked that question on Latin StackExchange, where it received two downvotes and was closed by the moderator. The reason? It's off-topic since the answer obviously lies in the English historical phonology, rather than in Latin or Greek. OK, let's ask that same question on English StackExchange. The result? They are pretending (in my opinion) to be unable to understand my question and have closed it. Why do people react to such questions that way? Is it because the contention that "tear" and "lacrima" are cognates is considered to be a "basic fact" and that therefore doubting it is considered unscientific and anti-intellectual? Are such supposed facts based on evidence or on nothing more than groupthink? You be the judge. One moderator of the Latin Language StackExchange agrees with me that this is a relatively good question and he suggested me to post it on Linguistics StackExchange. I did, but I am not at all sure that's the appropriate response. Cross-posting onto another StackExchange is almost never a good idea.
Conclusion
I was reasoned into those beliefs, so I was able to be reasoned out of them. The real question is how not to get reasoned into following a cult.You can avoid many cults by asking yourself: "Are the arguments they are using not-even-wrong?", and while this works against Flat-Earthism (most of the commonly-used arguments for Flat-Earthism, such as "The horizon appears to be rising with the observer as he climbs.", are not-even-wrong: it's not at all obvious how the Earth being flat and there being magical curvings of light explains that phaenomenon), it doesn't really work against the cults I was until recently a part of (asking "How it is that our methane emissions have been decreasing over time?" with the implication it's because of the number of grass-fed cows decreasing is hardly not-even-wrong).
Not even asking experts always works: many of my professors at the Computer Engineering programme, who knew far more about information theory than I did, told me that the arguments I presented in my paper "Etimologija Karašica" sound compelling to them. It's just that there is a technical detail in the field that connects information theory and linguistics that makes my calculations invalid, and not even my professors happened to know that tehnical detail.
I have long realized that the knowledge which is taught in middle school and high school was counter-productive in detecting pseudoscience, but I hoped the knowledge taught at the university would be better. In reality, it's every bit as useless and counter-productive. In our cybernetics classes we are taught that if some system is "slightly less than an integral", it's probably an IT1-type system, and that piece of knowledge led me astray with the methane. And this information theory... give me a break. It gives precise numbers (for example, that the p-value of that k-r pattern is between 1/300 and 1/17), but those numbers are wildly wrong. Hardly any piece of "knowledge" is that misleading. You can argue that computer engineeering is not building those flawed intuitions (I believed methane emissions were decreasing long before I learned basic cybernetics.), but it's undeniable that it gives people the ability to appear to mathematically justify those flawed intuitions.
People say that debating is a cure for irrationality. The truth is, if you want to be sure some argument is correct, you probably need to have the same tired old debate on many different Internet forums, to make sure that really nobody notices the fallacy. On most Internet forums, the vast majority of people, when presented with that argument "This diagram shows our methane emissions have been decreasing." or "Lab-grown meat will soon solve the problem of superbacteria.", will either shut up or respond with some incomprehensible word-salad. That will probably mislead you into thinking that your argument is correct. It goes without saying that an Internet debate going so well like the ones I described above is an exceedingly rare event. Usually debates turn into insults. And sometimes you need to be creative to provoke a productive discussion. It seems logical that a way to provoke a productive discussion about Croatian toponyms is to write papers and make YouTube videos in the Croatian language. But I've seen time and time again that didn't work. It's only after I made a video about Croatian toponyms in the Latin language that a productive discussion started. Perhaps it's sometimes necessary to talk about dead languages using dead languages (for example, talking about Illyrian in Latin), so that you attract people who are actually knowledgeable about the topic.
It should go without saying that quite often debates are not productive because both sides of the debate are incredibly disconnected from the reality. The debates about superbacteria are unlikely to be productive because most people wrongly believe most antibiotics go to humans, and that the main reason for superbacteria are physicians misprescribing antibiotics, and those who realize that's not true (exempli gratia, most vegetarians) mostly wrongly believe most antibiotics go to cows and pigs, and it seems plausible to them that lab-grown meat will soon solve the problem. Or like in the debates about Croatian toponyms, those who believe the Krahe's methodology is justified by the basic information theory (like I used to believe until recently) are only slightly less disconnected from the reality than those who think the Mayer's methodology is good. Debates in that case are almost guaranteed to be counter-productive.
If you are a Flat-Earther using not-even-wrong arguments such as "Why does the horizon appear to rise with the observer if he climbs? If the Earth was round, wouldn't we expect it to be falling?", once you get outside of your echo-chamber, you will actually probably quickly get correct responses such as "Until the Flat-Earth Theory gives us the math which predicts those phaenomena, we do not owe you that answer." (which is actually unlikely to convince you) or even an explanation of how basic trigonometry predicts that the angle at which you see the horizon is very small and how human perception doesn't work that way to be able to detect it (and that explanation is likely to convince you). But once you try debating the arguments such as "With the methods one learns in Information Theory course at the university, one can estimate that the p-value of that k-r pattern in the Croatian river names is somewhere between 1/300 and 1/17.", don't be surprised if you have to discuss it with dozens of different people on various forums before you get a useful response (that those methods fail because of the Sonority Sequencing Principle).
Maybe my mental illness has something to do with me not seeing how those arguments used to attract people into those cults ("Lab-grown meat will soon solve the problem of superbacteria, and you can do your part already by going vegetarian.", "Why have our methane emissions decreased, even though we are using more methane than ever? It has to be because most of our methane emissions come from grass-fed cows, which emit 3 times as much methane as grain-fed cows.", "Basic information theory says that the p-value of that k-r pattern in the Croatian river names is somewhere between 1/300 and 1/17. It is statistically significant. It's only logical that there was a word *karr~kurr in Illyrian meaning to flow, and therefore that Karašica comes from Illyrian *Kurr-urr-issia (flow-water-suffix)."...) are faulty. I have psychosis and I need to take Risperidone, Biperiden, and Alprazolam.