Somewhat related, a hypothesis of mine that I could never prove but believe in is that the Latin Alphabet is responsible for why the industrial revolution happened in Europe.
I often see people say, "It is unclear why the Industrial revolution occured in Europe, and not China which tended to have a much larger economy and stable states, or anywhere else on Earth". I think the reason is that Europe was experiencing much more scientific advancement over the years ~1450-1800 when the industrial revolution was beginning, inventing stuff like calculus, and all sorts of advances in chemistry and physics. It should be obvious why stuff like improved engineering, chemistry, and physics are very useful for building engines and other industrial machines.
But why was Europe having more scientific advancement? I think it's clearly the invention of the printing press. There is a jump in the amount and quality of science being done after the invention of the Gutenburg Printing press in 1440. And it makes very intuitive sense why there would be- the printing press rapidly increases the ease of distributing science textbooks and pamphlets about the latest scientific discoveries, and science is an exponential field, where each discovery makes more discoveries possible. At least in 1440 anyway, when there were tons of low hanging fruit for people to pick once they had a grounding in the basics and started to look.
But why did Europe have such success with the printing press and nowhere else? I think it's because of the nature of the latin alphabet, where each letter takes up one distinct space on a line. It lets someone with a printing press and some sets of 26 letter blocks rearrange those blocks into any combination of words. Whereas someone using written Chinese would need a seperate block for each word, likely multiple blocks since some words would appear multiple times on a page. In Arabic, it's more feasible, but it's a cursive script, which doesn't print as well and would inhibit the widespread use of printing presses.
I don't know much about Korean history, but trying to read up on it, it sounds like they did have a scientific renaissance after their invention of the printing press. But in 1637 they became a tributary of China, and that may have killed their potential for further scientific discoveries. It's hard to say. But there probably could be more investigation into whether my theory of alphabet -> science has any truth to it by looking more at Korea and comparing it to China and Japan.
I think that is a result of Europe's geography. I don't think any other land area has the same mix of good rivers and coastlines facilitating trade and communication, and mountains and other obstacles to huge empires. The earlier Euro-area civilizations were the Tigris-Euphrates and Nile rivers, both great for forming civilizations, surrounded by deserts to deter attackers. Greece is full is close islands, rugged coasts, and mountains, encouraging trade and defensive isolation. Italy is similar. Northern Europe with the Baltic and British island has a lot of that. Look at all the rivers -- Rhine and Danube headwaters are only (sez memory!) 1-200 km apart, and the same applies to a bunch of Russian rivers going north to the Arctic and south to the Black and Caspian Seas.
All that is fertile breeding grounds for cooperation and competition, a vast variety of languages and cultures.
Africa has few mountains, a very simple coast with few good harbors, and few rivers partly from the lack of mountains.
China is a huge single territory. It has two or three main rivers which don't encourage interior communications and trade like Europe, and few interior mountains or coastal features to isolate different cultures. It is ready made for a huge single culture and little competition. Japan is too resource-poor and small to make up for that.
The Americas have three or four major mountain chains (Sierras, Rockies, Appalachians, Andes) which don't have the same utility as all those small European mountains, and a few big rivers which are too isolated to aid communication and trade. The Ohio-Mississippi route was really good for an expanding US, but the southeast has no useful rivers for that, and the western rivers are too short and scattered.
I have always thought that was why Europe turned into the leading technologists and philosophers of the world. And all that competition is why Europe invented alphabets -- all those languages encouraged developing some common denominator.
If you can locate Publishing, Culture, and Power in Early Modern China, by Chow Kai-wing, it provides information on early Chinese ventures in printing, using not movable type (for the most part) but single carved wooden blocks for entire pages. Apparently such a block could produce 1500 copies in a day. I believe the early modern era was when the four classic Chinese novels were written and published, so there may have been a movement to a larger audience for written material.
I think Chinese publishing could do large scale printings of single books just as well or maybe even better. But the trouble would be in quickly switching to small and medium scale publishing of many different books
Japanese raises interesting questions. The system of hiragana (a syllabary used for native Japanese words) and katakana (a syllabary used largely for loan words) is not an alphabet, but is hardly less efficient for a language with restrictive phonotactics such as Japanese; and any Japanese word can be spelled using syllabic characters. But the Japanese have not gone entirely over to kana; they retain the use of kanji for many words, indeed most—but they seem to have a high level of literacy.
My memory is that when my father asked about the issue during a long ago trip to Japan, he was told that Japanese students were in about sixth grade before they were able to read a newspaper, which is some evidence of the greater difficulty of learning to read kanji.
On one hand, that's consistent with what I've read. But on the other, it raises the question of whether they gain anything from initially learning kana. My wife has a neat little book that goes through the hiragana, and provides for each of them a drawing of an animal whose Japanese name begins with the syllable that character represents; this seems to be an analog of an English language alphabet book. Is this a helpful step toward literacy?
This might be an interesting experiment to look at: What is the effect of discriminatory practices by western colonial powers on a country's development? Case in point: Haiti vs Dominican Republic. It might be hard to argue that Haiti's condition is its own fault, giving the price of reparations that Haiti was forced to pay to France and the hostility of states in the soon-to-be Confederate States of America to a black-led government.
I almost always hear colonialism as unequivocal bad. I would think that one major benefit of colonialism is the introduction of new ideas, especially technological advances. But I suspect this is not the case for India, if they already had 'communication' with the outside world before being colonized.
But I wanted to add some potential different views on the subject.
(I am commenting as a non-economist and without any significant knowledge on the subject, and I have not read the full article)
Lakshmi Iyer writes in the Introduction of "Direct versus Indirect Colonial Rule in India: Long-term Consequences" that:
"Third, the fact that the British retained the right to depose native state rulers in cases of “misrule” appears to play an important role. This right was exercised quite often and thus constituted a credible threat. I find that areas where the British deposed the rulers were indeed the worst-performing ones in the sense that these areas do not show any significant advantage compared to the directly-ruled
British areas"
First of all, this may also give a selection bias towards British rule being worse performing (despite them initially ruling the areas of highest agricultural potential). I do not know whether/how Iyer took this into account. Secondly, this gives quite different incentives, besides Indians knowing Indians better, then Indian rulers had a much larger incentive to rule their area well, since else they will lose power and their (probably large) source of income. While British rulers did not have this consequence in the case of misruling, as Iver says in the conclusion: "The key difference is that [Indian] kings were explicitly subject to being removed in cases of gross misrule, while [British] landlords did not have this institutional constraint." So the incentives are potentially better for the Indian rulers, which may also lead to better results.
Moreover, leading and managing is two different things, where leading is setting the overall direction, while managing is organizing people to do the job. I have no idea whether the British have been worse or better at leading (e.g. based on ideological ideas, like capitalism or socialism), than the Indians would have been. The study might more go to show that the Indian rulers are better at managing, and does not say anything about leading.
Finally, I do not regard
"lower availability of public goods such as schools, health centers and roads" or "more equal" as important measures of performance in the sense that equality says nothing about how good life people are living (e.g. rich or poor by relative or more so: absolute measure). And public schools, health centers and to lesser extend roads, do not say anything about the total quantity and quality of schooling and health care, since it is private goods, and can be provided by the private sector as well.
It reminds me of a study which measured the quality of health care, by "how large the public health care system is" and "How many people have acces to health care at very low/no cost". This of course ranks USA low, since the health care system is more private (and expensive). But better measures as the survival rate of deadly diseases would, as I understand it, rank USA very high.
But Iyer might not have made any mistake by e.g. considering pupils in school instead of pupils in public schools (I have not read the entire article).
But the measures for poverty and infant mortality still points in the same direction, to give his conclusion.
I think we will have some solid data points about this in 10 year or so because of India. Jio a 4G network provider made the cost of internet near negligible in rural India (which is otherwise also a cesspit of sexual violence and other acts of degeneracy). This has resulted in massive consumption of porn as well as creation of lots of amateur porn. (Both are illegal in India but law has little tooth here).
We will soon find out if this results into lower sexual violence. Since in these areas even middle aged people live with their parents, I think age groups too will be irrelevant.
Compulsory licensing applied to a past patent is evidence that future patents are more likely to have their values reduced as well via compulsory licensing, which revelation should reduce the expected profitability of future research, which should reduce the incentives and amount of that research.
The trouble is that you *still* can't use that to run any kind of robust study. The future expected value reduction applies equally to companies who were vs who weren't hit by the present compulsory licensing, so you can't compare them against each other, and there are too many other factors correlated with the passage of time so you can't compare before-vs-after. You can't even be sure when compulsory licensing expectations priced in by the market - if rumors lead the market to expect the government to take over 6 patents, but then it only takes over 2, research funding availability should go *up* after the taking, despite 2 takings reducing future research compared to the counterfactual of 0.
If I correctly understood the talk, these were companies that got compulsory licensing imposed for special reasons, such as as a condition of permitting a merger. So they had no more reason than other companies to expect compulsory licensing requirements in the future. It wasn't a case of the government simply taking over patents.
The exception were the two companies that agreed to licensing restrictions on their future patents.
To prove causality you need a ceteris paribus condition. In the real world there is no ceteris paribus, but everything is always changing. So you cannot observe causality (isolated cause-effect relationship), you can only impute causality. There is always room for error and doubt. It is Hume's induction problem. But it is not all a problem. This room of doubt is where you can choose and decide and take a different path. For example, the decision to be vaccinated with a certain substance or not. No experiment could completely eliminate the need for such a decision.
Natural experiments are situations that do give you evidence, not proof, of causality, because you can compare otherwise similar situations with and without the cause — states where Internet became available early vs states where it did not.
Well, proof is impossible on an empirical level anyway. It is certainly about evidence. But the natural experiment is the furthest thing away from proof or evidence of causality. Because many other influencing factors cannot be ruled out here. The best would be the double blind study. But even then, there would be room for error and for alternative evidence.
There is an interesting passage in Karl Popper's "The Logic of Scientific Discovery":
"Generally, similarity, and with it repetition, always presuppose the adoption of a point of view: some similarities or repetitions will strike us if we are interested in one problem, and others if we are interested in another problem. But if similarity and repetition presuppose the adaption of a point of view, or an interest, or an expectation, it is logically necessary that points of view, or interests, or expectations, are logically prior, as well as temporally (or causally or psychologically) prior, to repetion. But this result destroys both the doctrines of the logical and of the temporal primacy of repetitions.
[...] for any given finite group or set of things, however variously they may be chosen, we can, with a little ingenuity, find always points of view such that all the things belonging to that set are similar (or partially equal) if considered from one of these points of view; which means that anything can be said to be a 'repetition' of anything, if only we adopt the appropriate point of view. This shows how naive it is to look upon repetion as something ultimate, or given."
So we can look from different directions and perhaps find evidence of different causality, leading to conflicting conclusions. To stay with the example of a vaccine, you can look at a very narrow setting and potentially miss long-term effects, effects that then lead to the conclusion that the vaccine should be rejected. But that is likely to lead to bias. What is more important? What is preferred? Every test, every experiment depends on an interest, and so do the results and conclusions. But is this a problem to solve? I think it creates an openness, which is the whole reason for freedom, to choose different paths.
edit:
To put it this way:
What is the ultimate goal of an experiment or test? To find evidence?
If you could objectively measure evidence, and decide which test or experiment produced "more" evidence, then there would be an undeniable rationale for choosing the result with the highest objective evidence score. It would be "irrational" to choose the conclusion with a significantly lower score.
In the field of economics, for example, you could start "testing" the "right price" for something. In the end, socialism would be the way to go, the planned economy, because if you could "test" everything, you could weed out all the "wrong" things and only do the "right" things. That is exactly the dream of socialism. But otherwise there will be evolving markets or "anarchy", i.e. alternative ways.
Consider the evidence I reported that porn is a substitute for rape, not a complement. From what plausible viewpoint is that not a reason to raise your estimate of the probability that that is true, which I take as what "is evidence for" means?
There are two parts whose connection is in question:
1. an explanation, a theory, an idea about a cause-and-effect relationship
2. empirical data collected to prove that idea. The hoped-for result is a sufficiently good correlation.
But there may be unknown, unexplored causes for the observed effect, causes that lead to the observed effect. I cannot talk about the unknown. Or there may be (side) effects that occur much later, outside the time frame of the data collection, that counteract the hoped-for effect. Or there may be effects of internet porn other than rape that are considered bad. It is a value judgment as to which effect is important in judging whether internet porn is a bad or good thing. It is our interest or preference that leads us to observe a particular part of the world in order to prove our expectations and ideas about the world. Our inductive reasoning is never complete. Only in the logical sphere there can be complete induction. The connection between logic and the real world is in question. Nietzsche: "logic is bound to the condition: assume there are identical cases." But in reality there is similarity, but not complete identity.
We never know the whole story. In the end, we believe and judge. Even in the hardest science. Sure, I believe that internet porn can reduce rape.
The question was not whether internet porn can reduce rape. It was not whether internet porn is good or bad. It was whether the research I described is evidence that internet porn does reduce rape and, if not, what alternative plausible explanation for the evidence you could offer.
The rest of your response seems like a very lengthy way of saying that evidence is not proof.
Empirical tests aim to provide correlations, correlations can be taken as evidence of causality. Sure. The research you mention is evidence that Internet access reduces rape. What I am trying to say with Popper is that evidence is inevitably perspectival, and as such deniable.
In the end, it is counting cases of the past within a chosen frame.
But only the single unique case is the real case. A class of cases is always an artifact. It is all right to create classes and artifacts. The whole human language is nothing but artifacts. But there is always an escape. I like such escapes because I am an epistomological anarchist and I believe that any anarchism has to start at least at the epistomological level.
The adoption of Latin script, phonetic written Turkish in 1928 over the previous unphonetic written Turkish Arabic might also be a useful example. Modern written Turkish was later adopted for multiple Turkic languages across central Asia, some quite recently.
One interesting thing about this example is that the word `fish` is spelled perfectly phonetically (if we allow for digraphs). Compare this to Japanese 魚 (sakana 'fish'), where there is 0 relationship between sound and spelling. Ottoman Turkish بالق (balık 'fish') is in an intermediate position, but much closer to English than to Japanese.
Sakana (fish) in Japanese can also be written phonetically as さかな in the hiragana syllabary, which is more phonetically pronounceable than in English. The three characters are さ ‘sa’, か ‘ka’ and な ‘na’. The kanji 魚is a borrowed Chinese ideograph for fish, which when used in Japanese is also pronounced as SAKANA, but may also be pronounced as GYO if used in compound words, making it ideographic not phonetic.
A children’s book would use さかな, since kanji are taught after the phonetic and easier to teach hiragana. Hiragana (or the alternate form katakana) is perfectly phonetic with fewer exceptions than English or other alphabets. Kanji was added as a way to make longer texts more compact and to disambiguate homophones.
Saying that Kanji was added seems misleading. Kanji were the original way to write Japanese, before kana were invented. It is not quite correct to say that kana spelling is perfectly phonetic – ぢ/じ and づ/ず have merged in standard Japanese, and the particles へ, は and を are spelled irregularly, and pitch accent is not indicated at all, meaning that non-homophones such as 換える (kàérú 'to exchange') and 返る (káèrù 'to return') are homographs in kana (かえる).
In Katherine Mayo's Mother India, written I think around 1920, there is a discussion of the Indian custom of marrying prepubescent girls to mature men, and the British efforts to do away with it (the famous poet Rabindranath Tagore apparently wrote an essay denouncing those efforts). There was a theological rationale, but it may also have been that men wanted to be sure of exclusive access to their wives' fertility. I've seen the argument that if women are monogamous, the optimal age for men to be attracted to is the age of puberty (they get 100% of the reproductive opportunities), but if they are not, the optimal age is peak fertility, say, 18-25 (giving the maximum chance of fathering a child in any one sexual act, when you can't count on getting another); the Indian custom is an exaggeration of the former strategy. And Mayo wrote of some quite young girls giving birth, though of course it was much riskier for them.
Yes. I use "pedophilia" in that same sense. There is a technical term "ephebephilia," but I find it dubious; when I was 14 I was attracted to girls my own age (would it have been better for me to be attracted to young woman of 18 or 25?), and it seems to me that such attraction is a normal human experience.
As for consummating marriages with prepubescent girls, Mayo's book gives explicit accounts of such consummation taking place in India.
I don't think we should assume that sexual activity with prepubescents is selected against because it doesn't usually produce children. In the first place, there are a lot of forms of human sexual behavior that have the same liability, starting with same-sex attraction. In the second, a man who marries a prepubescent girl presumably expects to remain married to her until she eventually becomes fertile, so he doesn't decrease his reproductive success by starting early. Indeed, doing so gives him a chance of impregnating her as soon as she becomes fertile, whereas waiting might cause him to miss such an opportunity.
Well, it's true that there's a distinction between being capable of sexually desiring prepubescent children, and ONLY being attracted to prepubescent children. Though there are surely intermediate cases, just as there used to be many men whose primary attraction was to other men, or boys, but who got married, consummated the marriages, and even fathered children (and there still are some, but legal protection of homosexual acts and relationships has reduced the incentive to form such relationships).
Is it really the case that prepubescent girls cannot bear children? That's the general assumption, but I've read more than one account of girls of such an age that one would assume they were not yet pubescent being impregnated and giving birth. I'm not certain of the biology here. Though I'll grant that even if it does happen, it's not a strategy with a high payoff genetically.
>The Value of an Alphabet
Somewhat related, a hypothesis of mine that I could never prove but believe in is that the Latin Alphabet is responsible for why the industrial revolution happened in Europe.
I often see people say, "It is unclear why the Industrial revolution occured in Europe, and not China which tended to have a much larger economy and stable states, or anywhere else on Earth". I think the reason is that Europe was experiencing much more scientific advancement over the years ~1450-1800 when the industrial revolution was beginning, inventing stuff like calculus, and all sorts of advances in chemistry and physics. It should be obvious why stuff like improved engineering, chemistry, and physics are very useful for building engines and other industrial machines.
But why was Europe having more scientific advancement? I think it's clearly the invention of the printing press. There is a jump in the amount and quality of science being done after the invention of the Gutenburg Printing press in 1440. And it makes very intuitive sense why there would be- the printing press rapidly increases the ease of distributing science textbooks and pamphlets about the latest scientific discoveries, and science is an exponential field, where each discovery makes more discoveries possible. At least in 1440 anyway, when there were tons of low hanging fruit for people to pick once they had a grounding in the basics and started to look.
But why did Europe have such success with the printing press and nowhere else? I think it's because of the nature of the latin alphabet, where each letter takes up one distinct space on a line. It lets someone with a printing press and some sets of 26 letter blocks rearrange those blocks into any combination of words. Whereas someone using written Chinese would need a seperate block for each word, likely multiple blocks since some words would appear multiple times on a page. In Arabic, it's more feasible, but it's a cursive script, which doesn't print as well and would inhibit the widespread use of printing presses.
I don't know much about Korean history, but trying to read up on it, it sounds like they did have a scientific renaissance after their invention of the printing press. But in 1637 they became a tributary of China, and that may have killed their potential for further scientific discoveries. It's hard to say. But there probably could be more investigation into whether my theory of alphabet -> science has any truth to it by looking more at Korea and comparing it to China and Japan.
I think that is a result of Europe's geography. I don't think any other land area has the same mix of good rivers and coastlines facilitating trade and communication, and mountains and other obstacles to huge empires. The earlier Euro-area civilizations were the Tigris-Euphrates and Nile rivers, both great for forming civilizations, surrounded by deserts to deter attackers. Greece is full is close islands, rugged coasts, and mountains, encouraging trade and defensive isolation. Italy is similar. Northern Europe with the Baltic and British island has a lot of that. Look at all the rivers -- Rhine and Danube headwaters are only (sez memory!) 1-200 km apart, and the same applies to a bunch of Russian rivers going north to the Arctic and south to the Black and Caspian Seas.
All that is fertile breeding grounds for cooperation and competition, a vast variety of languages and cultures.
Africa has few mountains, a very simple coast with few good harbors, and few rivers partly from the lack of mountains.
China is a huge single territory. It has two or three main rivers which don't encourage interior communications and trade like Europe, and few interior mountains or coastal features to isolate different cultures. It is ready made for a huge single culture and little competition. Japan is too resource-poor and small to make up for that.
The Americas have three or four major mountain chains (Sierras, Rockies, Appalachians, Andes) which don't have the same utility as all those small European mountains, and a few big rivers which are too isolated to aid communication and trade. The Ohio-Mississippi route was really good for an expanding US, but the southeast has no useful rivers for that, and the western rivers are too short and scattered.
I have always thought that was why Europe turned into the leading technologists and philosophers of the world. And all that competition is why Europe invented alphabets -- all those languages encouraged developing some common denominator.
You don't think Shanxi, Shaanxi, Sechuan and Yunnan count as deeply mountainous?
Not like the eastern Mediterranean.
If you can locate Publishing, Culture, and Power in Early Modern China, by Chow Kai-wing, it provides information on early Chinese ventures in printing, using not movable type (for the most part) but single carved wooden blocks for entire pages. Apparently such a block could produce 1500 copies in a day. I believe the early modern era was when the four classic Chinese novels were written and published, so there may have been a movement to a larger audience for written material.
I think Chinese publishing could do large scale printings of single books just as well or maybe even better. But the trouble would be in quickly switching to small and medium scale publishing of many different books
Japanese raises interesting questions. The system of hiragana (a syllabary used for native Japanese words) and katakana (a syllabary used largely for loan words) is not an alphabet, but is hardly less efficient for a language with restrictive phonotactics such as Japanese; and any Japanese word can be spelled using syllabic characters. But the Japanese have not gone entirely over to kana; they retain the use of kanji for many words, indeed most—but they seem to have a high level of literacy.
My memory is that when my father asked about the issue during a long ago trip to Japan, he was told that Japanese students were in about sixth grade before they were able to read a newspaper, which is some evidence of the greater difficulty of learning to read kanji.
On one hand, that's consistent with what I've read. But on the other, it raises the question of whether they gain anything from initially learning kana. My wife has a neat little book that goes through the hiragana, and provides for each of them a drawing of an animal whose Japanese name begins with the syllable that character represents; this seems to be an analog of an English language alphabet book. Is this a helpful step toward literacy?
This might be an interesting experiment to look at: What is the effect of discriminatory practices by western colonial powers on a country's development? Case in point: Haiti vs Dominican Republic. It might be hard to argue that Haiti's condition is its own fault, giving the price of reparations that Haiti was forced to pay to France and the hostility of states in the soon-to-be Confederate States of America to a black-led government.
> Colonialism and Industrial Development
I almost always hear colonialism as unequivocal bad. I would think that one major benefit of colonialism is the introduction of new ideas, especially technological advances. But I suspect this is not the case for India, if they already had 'communication' with the outside world before being colonized.
But I wanted to add some potential different views on the subject.
(I am commenting as a non-economist and without any significant knowledge on the subject, and I have not read the full article)
Lakshmi Iyer writes in the Introduction of "Direct versus Indirect Colonial Rule in India: Long-term Consequences" that:
"Third, the fact that the British retained the right to depose native state rulers in cases of “misrule” appears to play an important role. This right was exercised quite often and thus constituted a credible threat. I find that areas where the British deposed the rulers were indeed the worst-performing ones in the sense that these areas do not show any significant advantage compared to the directly-ruled
British areas"
First of all, this may also give a selection bias towards British rule being worse performing (despite them initially ruling the areas of highest agricultural potential). I do not know whether/how Iyer took this into account. Secondly, this gives quite different incentives, besides Indians knowing Indians better, then Indian rulers had a much larger incentive to rule their area well, since else they will lose power and their (probably large) source of income. While British rulers did not have this consequence in the case of misruling, as Iver says in the conclusion: "The key difference is that [Indian] kings were explicitly subject to being removed in cases of gross misrule, while [British] landlords did not have this institutional constraint." So the incentives are potentially better for the Indian rulers, which may also lead to better results.
Moreover, leading and managing is two different things, where leading is setting the overall direction, while managing is organizing people to do the job. I have no idea whether the British have been worse or better at leading (e.g. based on ideological ideas, like capitalism or socialism), than the Indians would have been. The study might more go to show that the Indian rulers are better at managing, and does not say anything about leading.
Finally, I do not regard
"lower availability of public goods such as schools, health centers and roads" or "more equal" as important measures of performance in the sense that equality says nothing about how good life people are living (e.g. rich or poor by relative or more so: absolute measure). And public schools, health centers and to lesser extend roads, do not say anything about the total quantity and quality of schooling and health care, since it is private goods, and can be provided by the private sector as well.
It reminds me of a study which measured the quality of health care, by "how large the public health care system is" and "How many people have acces to health care at very low/no cost". This of course ranks USA low, since the health care system is more private (and expensive). But better measures as the survival rate of deadly diseases would, as I understand it, rank USA very high.
But Iyer might not have made any mistake by e.g. considering pupils in school instead of pupils in public schools (I have not read the entire article).
But the measures for poverty and infant mortality still points in the same direction, to give his conclusion.
Article online at:
https://www.hbs.edu/ris/Publication%20Files/05-041_1feff996-f50e-4e5a-b057-e0119cd19a62.pdf
> Porn and Sex: Substitutes or Complements?
I think we will have some solid data points about this in 10 year or so because of India. Jio a 4G network provider made the cost of internet near negligible in rural India (which is otherwise also a cesspit of sexual violence and other acts of degeneracy). This has resulted in massive consumption of porn as well as creation of lots of amateur porn. (Both are illegal in India but law has little tooth here).
We will soon find out if this results into lower sexual violence. Since in these areas even middle aged people live with their parents, I think age groups too will be irrelevant.
Compulsory licensing applied to a past patent is evidence that future patents are more likely to have their values reduced as well via compulsory licensing, which revelation should reduce the expected profitability of future research, which should reduce the incentives and amount of that research.
The trouble is that you *still* can't use that to run any kind of robust study. The future expected value reduction applies equally to companies who were vs who weren't hit by the present compulsory licensing, so you can't compare them against each other, and there are too many other factors correlated with the passage of time so you can't compare before-vs-after. You can't even be sure when compulsory licensing expectations priced in by the market - if rumors lead the market to expect the government to take over 6 patents, but then it only takes over 2, research funding availability should go *up* after the taking, despite 2 takings reducing future research compared to the counterfactual of 0.
If I correctly understood the talk, these were companies that got compulsory licensing imposed for special reasons, such as as a condition of permitting a merger. So they had no more reason than other companies to expect compulsory licensing requirements in the future. It wasn't a case of the government simply taking over patents.
The exception were the two companies that agreed to licensing restrictions on their future patents.
To prove causality you need a ceteris paribus condition. In the real world there is no ceteris paribus, but everything is always changing. So you cannot observe causality (isolated cause-effect relationship), you can only impute causality. There is always room for error and doubt. It is Hume's induction problem. But it is not all a problem. This room of doubt is where you can choose and decide and take a different path. For example, the decision to be vaccinated with a certain substance or not. No experiment could completely eliminate the need for such a decision.
Natural experiments are situations that do give you evidence, not proof, of causality, because you can compare otherwise similar situations with and without the cause — states where Internet became available early vs states where it did not.
Well, proof is impossible on an empirical level anyway. It is certainly about evidence. But the natural experiment is the furthest thing away from proof or evidence of causality. Because many other influencing factors cannot be ruled out here. The best would be the double blind study. But even then, there would be room for error and for alternative evidence.
There is an interesting passage in Karl Popper's "The Logic of Scientific Discovery":
"Generally, similarity, and with it repetition, always presuppose the adoption of a point of view: some similarities or repetitions will strike us if we are interested in one problem, and others if we are interested in another problem. But if similarity and repetition presuppose the adaption of a point of view, or an interest, or an expectation, it is logically necessary that points of view, or interests, or expectations, are logically prior, as well as temporally (or causally or psychologically) prior, to repetion. But this result destroys both the doctrines of the logical and of the temporal primacy of repetitions.
[...] for any given finite group or set of things, however variously they may be chosen, we can, with a little ingenuity, find always points of view such that all the things belonging to that set are similar (or partially equal) if considered from one of these points of view; which means that anything can be said to be a 'repetition' of anything, if only we adopt the appropriate point of view. This shows how naive it is to look upon repetion as something ultimate, or given."
So we can look from different directions and perhaps find evidence of different causality, leading to conflicting conclusions. To stay with the example of a vaccine, you can look at a very narrow setting and potentially miss long-term effects, effects that then lead to the conclusion that the vaccine should be rejected. But that is likely to lead to bias. What is more important? What is preferred? Every test, every experiment depends on an interest, and so do the results and conclusions. But is this a problem to solve? I think it creates an openness, which is the whole reason for freedom, to choose different paths.
edit:
To put it this way:
What is the ultimate goal of an experiment or test? To find evidence?
If you could objectively measure evidence, and decide which test or experiment produced "more" evidence, then there would be an undeniable rationale for choosing the result with the highest objective evidence score. It would be "irrational" to choose the conclusion with a significantly lower score.
In the field of economics, for example, you could start "testing" the "right price" for something. In the end, socialism would be the way to go, the planned economy, because if you could "test" everything, you could weed out all the "wrong" things and only do the "right" things. That is exactly the dream of socialism. But otherwise there will be evolving markets or "anarchy", i.e. alternative ways.
Consider the evidence I reported that porn is a substitute for rape, not a complement. From what plausible viewpoint is that not a reason to raise your estimate of the probability that that is true, which I take as what "is evidence for" means?
There are two parts whose connection is in question:
1. an explanation, a theory, an idea about a cause-and-effect relationship
2. empirical data collected to prove that idea. The hoped-for result is a sufficiently good correlation.
But there may be unknown, unexplored causes for the observed effect, causes that lead to the observed effect. I cannot talk about the unknown. Or there may be (side) effects that occur much later, outside the time frame of the data collection, that counteract the hoped-for effect. Or there may be effects of internet porn other than rape that are considered bad. It is a value judgment as to which effect is important in judging whether internet porn is a bad or good thing. It is our interest or preference that leads us to observe a particular part of the world in order to prove our expectations and ideas about the world. Our inductive reasoning is never complete. Only in the logical sphere there can be complete induction. The connection between logic and the real world is in question. Nietzsche: "logic is bound to the condition: assume there are identical cases." But in reality there is similarity, but not complete identity.
We never know the whole story. In the end, we believe and judge. Even in the hardest science. Sure, I believe that internet porn can reduce rape.
The question was not whether internet porn can reduce rape. It was not whether internet porn is good or bad. It was whether the research I described is evidence that internet porn does reduce rape and, if not, what alternative plausible explanation for the evidence you could offer.
The rest of your response seems like a very lengthy way of saying that evidence is not proof.
Empirical tests aim to provide correlations, correlations can be taken as evidence of causality. Sure. The research you mention is evidence that Internet access reduces rape. What I am trying to say with Popper is that evidence is inevitably perspectival, and as such deniable.
In the end, it is counting cases of the past within a chosen frame.
But only the single unique case is the real case. A class of cases is always an artifact. It is all right to create classes and artifacts. The whole human language is nothing but artifacts. But there is always an escape. I like such escapes because I am an epistomological anarchist and I believe that any anarchism has to start at least at the epistomological level.
Interesting read for this non-economist…
The adoption of Latin script, phonetic written Turkish in 1928 over the previous unphonetic written Turkish Arabic might also be a useful example. Modern written Turkish was later adopted for multiple Turkic languages across central Asia, some quite recently.
Ottoman Turkish writing seems pretty phonetic to me, about on par with modern English, I'd say.
I think that is damning with faint praise.
Consider Shaw's spelling "fish" as "ghoti."
"gh" as in "tough," "o" as in "women," "ti" as in "nation."
One interesting thing about this example is that the word `fish` is spelled perfectly phonetically (if we allow for digraphs). Compare this to Japanese 魚 (sakana 'fish'), where there is 0 relationship between sound and spelling. Ottoman Turkish بالق (balık 'fish') is in an intermediate position, but much closer to English than to Japanese.
Sakana (fish) in Japanese can also be written phonetically as さかな in the hiragana syllabary, which is more phonetically pronounceable than in English. The three characters are さ ‘sa’, か ‘ka’ and な ‘na’. The kanji 魚is a borrowed Chinese ideograph for fish, which when used in Japanese is also pronounced as SAKANA, but may also be pronounced as GYO if used in compound words, making it ideographic not phonetic.
A children’s book would use さかな, since kanji are taught after the phonetic and easier to teach hiragana. Hiragana (or the alternate form katakana) is perfectly phonetic with fewer exceptions than English or other alphabets. Kanji was added as a way to make longer texts more compact and to disambiguate homophones.
Saying that Kanji was added seems misleading. Kanji were the original way to write Japanese, before kana were invented. It is not quite correct to say that kana spelling is perfectly phonetic – ぢ/じ and づ/ず have merged in standard Japanese, and the particles へ, は and を are spelled irregularly, and pitch accent is not indicated at all, meaning that non-homophones such as 換える (kàérú 'to exchange') and 返る (káèrù 'to return') are homographs in kana (かえる).
Lots of things are innate that reduce fitness — hemophilia, Down Syndrome, ...
In Katherine Mayo's Mother India, written I think around 1920, there is a discussion of the Indian custom of marrying prepubescent girls to mature men, and the British efforts to do away with it (the famous poet Rabindranath Tagore apparently wrote an essay denouncing those efforts). There was a theological rationale, but it may also have been that men wanted to be sure of exclusive access to their wives' fertility. I've seen the argument that if women are monogamous, the optimal age for men to be attracted to is the age of puberty (they get 100% of the reproductive opportunities), but if they are not, the optimal age is peak fertility, say, 18-25 (giving the maximum chance of fathering a child in any one sexual act, when you can't count on getting another); the Indian custom is an exaggeration of the former strategy. And Mayo wrote of some quite young girls giving birth, though of course it was much riskier for them.
Yes. I use "pedophilia" in that same sense. There is a technical term "ephebephilia," but I find it dubious; when I was 14 I was attracted to girls my own age (would it have been better for me to be attracted to young woman of 18 or 25?), and it seems to me that such attraction is a normal human experience.
As for consummating marriages with prepubescent girls, Mayo's book gives explicit accounts of such consummation taking place in India.
I don't think we should assume that sexual activity with prepubescents is selected against because it doesn't usually produce children. In the first place, there are a lot of forms of human sexual behavior that have the same liability, starting with same-sex attraction. In the second, a man who marries a prepubescent girl presumably expects to remain married to her until she eventually becomes fertile, so he doesn't decrease his reproductive success by starting early. Indeed, doing so gives him a chance of impregnating her as soon as she becomes fertile, whereas waiting might cause him to miss such an opportunity.
Well, it's true that there's a distinction between being capable of sexually desiring prepubescent children, and ONLY being attracted to prepubescent children. Though there are surely intermediate cases, just as there used to be many men whose primary attraction was to other men, or boys, but who got married, consummated the marriages, and even fathered children (and there still are some, but legal protection of homosexual acts and relationships has reduced the incentive to form such relationships).
Is it really the case that prepubescent girls cannot bear children? That's the general assumption, but I've read more than one account of girls of such an age that one would assume they were not yet pubescent being impregnated and giving birth. I'm not certain of the biology here. Though I'll grant that even if it does happen, it's not a strategy with a high payoff genetically.