The Unz Review - Mobile
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
 BlogviewJames Thompson Archive
Correlation Is Not Causation, But It’s the Way to Bet.
🔊 Listen RSS
Email This Page to Someone

 Remember My Information



=>

Bookmark Toggle AllToCAdd to LibraryRemove from Library • BShow CommentNext New CommentNext New ReplyRead More
ReplyAgree/Disagree/Etc. More... This Commenter This Thread Hide Thread Display All Comments
AgreeDisagreeLOLTroll
These buttons register your public Agreement, Disagreement, Troll, or LOL with the selected comment. They are ONLY available to recent, frequent commenters who have saved their Name+Email using the 'Remember My Information' checkbox, and may also ONLY be used once per hour.
Ignore Commenter Follow Commenter
Search Text Case Sensitive  Exact Words  Include Comments
List of Bookmarks

Some things are associated with others. Some things you eat make you ill. Some animals attack you. Some places are dangerous, some people likewise. On a brighter note, some foods are tasty and healthy. Some animals can be domesticated, or at least are easy to hunt or trap. Some places are safe, and some people likewise.

Correlation is not causation, but it’s the way to bet. Your life may depend upon it. Under-predict dangers and you could end up dead. Better to be safe than sorry. Better to be sorry that you have missed some opportunities than to be dead. It is sensible to worry about what may happen. Stereotypes are your friend. They are preliminary observations about life. Improve them as you learn more. Some must be discarded, but many more can be sharpened up and refined.

Life is a dilemma. When searching for a meal you must avoid ending up as a meal. Be careful, but don’t worry so much that you cannot forage for food. Hunger will make you adventurous, and then you are at risk again.

Ideally, we would never calculate correlations coefficients, but would just look at the data properly plotted out, ideally over a long period, and judge things by eye. The shape of the distribution matters. Intellectual and scholastic tests need not be a perfect bell curve, though they can be pretty close to one.

Sometimes an unknown force distorts the distribution, as when illness and infections sap the wits of poor citizens living in bad circumstances. More mysteriously, sometimes distributions are almost normal, but pinched into a narrower range, as if bound by a tighter central limit. Why are some groups narrower than others? Women, for example? African Americans, for another? Easy to see how systematic disadvantage could shift a mean downwards, less easy to see how those forces could both encourage low scorers and discourage high scorers.

A correlation coefficient is a straight-line simplification. Useful, though. It captures a lot in a little number. Standard deviations are also very informative.

It is no disproof of a correlation that it is not unity. Most real-life correlations are far less than perfect, but will be much better than guessing, even though there will always be outliers. Adding up those outliers in terms of residuals (errors of prediction) is a useful way of understanding the power of predictions based on correlations. For example, if you have to predict the height of an unknown person, your best bet (least error prone) is to predict that they are of average height. If you are asked to predict the height of 100 people, betting that everyone of them is of average height results in your error of prediction being the same as the standard deviation of the height of the general population.

If you have extra knowledge, such as being told the height of the individual’s parents, then you can improve your prediction by taking that into account. You will have reduced your error of prediction, and can compare how much it improves your bets by comparing your reduced residual with that of the standard deviation of the population.

Some people really believe they have invalidated a correlation by drawing attention to a particular outlier. If you conceive of a correlation as an ellipse rather than a straight line you can see that the highest scorer on one variable will not be the highest scorer on the other variable. That only happens with perfect correlation. Steve Hsu explains the issue here:

https://infoproc.blogspot.com/2014/07/success-ability-and-all-that.html

Correlation is not causation, but you are more likely to find a cause in a correlated variable than in an uncorrelated one. Search where there is at least a trace of a putative connective tissue. If you think it was the tomato that upset your digestion, start your controlled trial on tomatoes.

Correlation is not causation, but sometimes a finding is suggestive, like a trout in the milk. It does not prove that the milk was watered, but it makes you suspicious.

The “correlation is not causation” mantra is true as far as it goes, but it tends to be used so as to argue that, despite many correlations linking A with B being found in different circumstances, these will somehow never suffice to strongly suggest a causal link between A and B. On the contrary, correlation is a necessary feature of causation, but not a sufficient proof. Correlation is not always causation, but it helps find causes. Correlation is a pre-condition of causality.

Michael Woodley has set a challenge: “Sure, correlation does not equal causation, but find me just one single instance of a causal relationship where there is no correlation (just one would suffice).”

http://www.unz.com/jthompson/the-woodley-challenge/

http://www.unz.com/jthompson/woodley-elaborates/

Whilst it is true that correlation does not necessarily equate to causation, all causally related variables will be correlated. Thus correlation is always necessary (but not in and of itself sufficient) for establishing causation.

Woodley continues:

The claim that ‘correlation does not equal causation’ is therefore meaningless when used to counter the results of correlative studies in which specific causal inferences are being made, as the inferred pattern of causation necessarily supervenes upon correlation amongst variables. Whether the variables being considered are in actuality causally associated as per the inference is another matter entirely.

The correct critique of such findings therefore is from mediation, i.e. the idea that a given correlation might be spurious owing to the presence of ‘hidden’ variables that are generating the apparent correlation. A famous example is yam production and national IQ, which across countries correlate negatively. It would be wrong to say that yam production somehow inhibits IQ, as the association will in fact turn out to be mediated by something like temperature and latitude. These variables are in turn proxies for historical and ecological trends that make the sort of countries that yield fewer yams the sort of countries that are typically populated by higher ability people, and vice versa. The causation in this case is via additional variables, which cause the covariance between the two variables of interest, without there being a direct effect of one on the other.

Properly constructed multivariate models can use these patterns of mediation to infer the likelihood of causation going in one direction or another. Thus, it is possible to actually test causal inference amongst a population of correlated variables. By far the best way of doing this is to compare the fits of models containing specific theoretically prescribed patterns of causal inference against (preferably many) alternative theoretically plausible models, in which alternative patterns of causation are inferred (Figueredo & Gorsuch, 2007).

Sir William Gemmell Cochran termed this “Fisher’s Dictum‟:

“About 20 years ago, when asked in a meeting what can be done in observational studies to clarify the step from association to causation, Sir Ronald Fisher replied; `Make your theories elaborate.’ The reply puzzled me at first, since by Occam’s razor, the advice usually given is to make theories as simple as is consistent with known data. What Sir Ronald meant, as subsequent discussion showed, was that when constructing a causal hypothesis one should envisage as many different consequences of its truth as possible, and plan observational studies to discover whether each of these consequences is found to hold. (Cochran, 1965, §5).

Ref.

Cochran, W. G. (1965). The planning of observational studies of human populations (with Discussion). Journal of the Royal Statistical Society. Series A, 128, 134–155.

Figueredo, A. J., & Gorsuch, R. L. (2007). Assortative mating in the jewel wasp. 2.
Sequential cononical analysis as an exploratory form of path analysis. Journal of the Arizona-Nevada Academy of Science, 39, 59-64.

Following on from Woodley it is interesting to calculate how many replications of a correlation would be needed to strongly suggest a causal relationship, including the possibility that it was caused by an ever-present hidden variable yet to be identified. Would three replications be enough if one was in a hurry?


Faced with the 1854 cholera outbreak in London, John Snow had no idea what mechanism caused cholera, and his instruments could not reliably identify the contaminants in water supplies, but he noted what we would now call correlations: some water companies had more of their clients die than others, even though all of them supplied to rich and poor households alike. South of the river companies were more deadly, and they drew more contaminated water from the river rather than other sources, and filtered it less than other companies. Some neighbourhood pumps had more deaths nearby than did others. This was a geographic form of correlation (now called a Voronoi diagram) and it was on that correlational basis, without knowledge of the real mechanism, that he took the handle off the Broad Street pump, and stopped the epidemic.

That is the way we tell the story now, but Snow was a careful and clever man, and pointed out another explanation: the cholera outbreak was coming to an end anyway, as people ran away from areas where there were many deaths. The common folk who believed that correlation implied causation ran for their lives and lived to see another day.

Snow also had to cope with a major anomaly in his geographic correlational investigations. None of the brewery workers right next to the Broad Street water pump fell ill with cholera. It turned out that they received free beer, and the water for the beer was boiled so as to release the flavour of hops, thus inadvertently killing off the water-borne organisms.

Snow jumped to a conclusion because his mind was prepared to interpret associations in a particular way, intially by his doubts about the air transmission miasma theory and later by his own hypothesis of water-borne transmission. He jumped to the right conclusion, without proofs of the causal mechanism which were only available years later.

Correlation is not causation, but it’s the way to bet.

 
• Category: Science • Tags: Correlation, IQ, Statistics 
Hide 88 CommentsLeave a Comment
Commenters to Ignore...to FollowEndorsed Only
Trim Comments?
    []
  1. Sean says:

    http://www.unz.com/pfrost/what-caused-rickets-epidemic/
    By the mid-20th century, rickets was again rare, thus vindicating not only UV therapy but also the view that lack of sun had been the cause (Harrison, 1966). This view nonetheless remains unproven. Although vitamin D does help the body absorb more calcium and phosphorus, no one knows for sure whether the epidemic was due to low levels of this vitamin. […]
    At the height of the epidemic, one physician did suggest that something was immobilizing phosphorus in people with rickets. Dr. John Snow (1857) observed that the illness was most frequent in London and the south of England where industrial bakeries used alum to make bread look whiter. It was rare in the north where bread was normally home-baked. He reasoned that this alum combined with phosphorus in the body to form insoluble aluminum phosphate, thus depleting the reserves of phosphorus needed for strong bones

    .

  2. Alfa158 says:

    That saying has never been correct. It should be “correlation is causation, but not necessarily the cause you think”. Two new roommates Dick, who is a sock thief, and Tom move in with me and suddenly my socks start disappearing. The sock thefts correlate with the presence of Tom and also correlate with the presence of Dick. One is correlation is causation, the other isn’t.

    • Replies: @Dieter Kief
    , @c matt
  3. Sean says:

    Some people really believe they have invalidated a correlation by drawing attention to a particular outlier. If you conceive of a correlation as an ellipse rather than a straight line you can see that the highest scorer on one variable will not be the highest scorer on the other variable. That only happens with perfect correlation. Steve Hsu explains the issue here:

    It is a bit more simple that that in practice I think

    https://en.wikipedia.org/wiki/Jeanne_Calment

    Demographers have highlighted that Calment having lived multiple years longer than the next oldest person ever documented would make her a disproportionately extreme outlier.[18][19] Natalya Gavrilova and Leonid Gavrilov argued in 2000 that this anomaly casts doubt on the authenticity of her age, and that the evidence for her being genuine was inadequate to overcome this doubt.[19] Gerontologist Tom Kirkwood in 2001 wrote, “could she be a fraud, it is hard to see how unless it was the mother not the daughter who died in 1934, the daughter assuming the identity of her mother.”[20]

  4. @Alfa158

    Correlation might not tell us what (kinds of) causes there are, and not even if there are any causes at work between the observed phenomena at all (could well have been neither Dick nor Tom who stole the socks).

    I hesitate only a little bit because my thoughts don’t go along perfectly well with Friedrich Schlegels (quite romantic) idea, that the “darn things hang all together”:
    Eins oder das andere ist auch ein Mitglied einer Masse, die sich nicht trennen läßt. Ueberhaupt hängen die verdammten Dinger so zusammen.” – – – In the end, everything might well be caused by everything – or at least anything. All we lack then is the degree, up to which the – ultimate I might say – causation might be one at all. Now I think of the wing of the butterfly in – – – Norway, let’s assume – – -, which causes a tropical thunderstorm…
    (The funny thing is, that seen from such a perspective, cosmological irony and the hard sciences are hard to tell apart. The romantics discovered this now a quite popular idea, and the virtuoso of these kinds of musings is Jean Paul. The master of the universal, but never unrational (!) irony (the first Kantian who understood, that the three Critiques could be quite easily synchronized with laughter… – and be taken seriously all the same).

    (Friedrich Schlegel an August Wlhelm von Schlegel https://august-wilhelm-schlegel.de/briefedigital/letters/view/2763?left=text&right=manuscript&query_id=5ba0e4d0665a9

  5. If by “bet,” you mean formulate a theory and test it, ok.

    If you mean, “destroy an economy,” not so much.

    The problem arises when some people take it upon themselves to place bets for others.

    Also when the causation link is there, but the analysis of it has it reversed. An example might be deflation during a crash following an inflationary bubble. Would you “bet” the deflation caused the crash?

  6. dearieme says:

    “the water for the beer was boiled so as to release the flavour of hops, thus inadvertently killing off the water-borne organisms.” Wouldn’t the alcohol have done the trick anyway, even if the water hadn’t been boiled? After all, wine was safe to drink without anyone having boiled its water.

    Oh you sly fellow, doc. You wanted someone to point out that the boiling water/survival link was mere correlation and that the cause was the alcohol. Hats off, sir, for such subtlety.

    By the by, isn’t it customary in such discussions for someone to allude to Hill’s recommendations about taking correlations seriously? Though it’s daft to elevate them to doctrine, they do come across as the heightened common sense of an intelligent man who has reflected carefully on his experience on the matter.

    https://en.wikipedia.org/wiki/Bradford_Hill_criteria

    • Replies: @James Thompson
  7. res says:

    Revisiting the latter of the two Woodley posts I think these comments were very helpful in thinking his point through.

    Michael Woodley: http://www.unz.com/jthompson/woodley-elaborates/#comment-1683644

    Cryptography does not count for the reasons articulated by the previous commenter. To reiterate, correlation in this instance is simply being deliberately hidden via encryption. The clue here is actually in the word crypt, which stems from the Greek kryptós meaning hidden or secret.

    I will say it one more time – all causal relationships are necessarily correlated. Whether the correlation exists in plain sight, or is known only to the ‘Mind of God’ (as in the case of encryption) has no bearing on the validity of this axiom.

    The second paragraph clarifies greatly what was meant, but at the same time renders the assertion almost meaningless IMO. Practically speaking the correlation coefficient is what we are generally talking about. The squared variable and donut cases mentioned earlier in the thread are very much on point here. Whether or not any necessary variable transformations are easy to find and implement is important (more so from a practical than philosophical POV).

    I think refining the point away from the dichotomous zero or not correlation/causation to a gross mismatch (in either direction, how to define this though?) between the correlation observed and the actual strength of the causation is worthwhile.

    Timothy Bates: http://www.unz.com/jthompson/woodley-elaborates/#comment-1683645

    The best description I’ve seen is that correlation is “an unexplained causal nexus”.

    This acknowledges that correlations always reflect causes; does not make the (false) counterclaim that lack of correlation proves absence of causal relations (see above); and implies (correctly) that correlation alone can’t elucidate these causes for us.

    To say that causes always result in correlations and then add that by correlation is meant “what God sees’ just confuses the situation.

    Learn your Pearl is great advice 🙂 Eysenck (and of course Fisher) are also worth reading in addition to Pearl on this topic. As are a couple of others – for instance Chang’s excellent “Inventing Temperature”. People long knew, for instance that exposure to fire raised temperature, but the theories (i.e., models of causality) were terribly varied (and often terribly wrong) until the kinetic model. Which, of course, may also be supplanted. Correlation all along. Causality all along. Good understanding of it, effortful, deliberate, inspired, and fallible.

    http://www.amazon.com/Inventing-Temperature-Measurement-Scientific-Philosophy/dp/0195337387/ref=sr_1_1?s=books&ie=UTF8&qid=undefined&sr=1-1&keywords=history+of+temperature

    Well said. I wish Timothy Bates would comment here more often. Does he have a blog? He is active on twitter, but I don’t find that as good as longer form posts: https://twitter.com/timothycbates

    He offers insightful commentary at Steve Hsu’s blog from time to time as well: https://disqus.com/by/TimothyBates/

    P.S. This tweet is pretty funny and on topic for other recent posts if not this one:

    • Replies: @James Thompson
    , @Sean
  8. Why are some groups narrower than others? Women, for example? African Americans, for another? Easy to see how systematic disadvantage could shift a mean downwards, less easy to see how those forces could both encourage low scorers and discourage high scorers.

    The putative explanation for the lower variance of females is that they have two X chromosomes with random inactivation of most of one X in each cell leading to averaging of the effect.

    However, what explains the lower variance of blacks? Could it be that a certain number of IQ-affecting loci are fixed at the lower-IQ producing allele within lower-average-IQ populations. There are good selection arguments for thinking so.

    That also suggests that in mixed-population (ie, mixed-race) offspring, males will regress to the IQ of the mother’s population while females would regress to something like the mid-point of the parental populations.

    I guess these could all be tested to see whether I am talking shit or not.

    • Replies: @James Thompson
  9. @dearieme

    It is years since I brewed beer. You are right that full fermentation should be enough to produce a drinkable and thus safe brew.

    Leaving out Hill criteria was a lapse on my part. Very useful and sensible list, including dose-response relationships and the possibility of a fairly uniform time lapse between cause and effect. Thank you.

    One thing I did read for ages was the Communicable Diseases Register, which taught me to be very careful about the food served at wedding receptions, particularly when the service ran late. Few detective stories are as interesting.

    • Replies: @dearieme
  10. @res

    Tim Bates is a thoroughly good egg.

  11. @Peripatetic Commenter

    Agree with you that hypotheses of that sort should be tested. Need to check that the narrower standard deviations are still the case, because a consistent change in them would suggest recent secular changes. Don’t expect that to be the case, but worth checking, and fun if it turned out to be the case.

    • Replies: @Peripatetic Commenter
  12. @James Thompson

    There is one other case I forgot to mention.

    If the explanation for the narrower female variance is correct, I would expert black-female variance to be narrower than black-male variance.

    That could be tested also.

    There are some hints of narrower variance among East Asians.

    In that case a possible explanation is some number of loci fixed at the higher-IQ causing alleles.

    • Replies: @Wizard of Oz
  13. Dr. Thompson:

    Here’s an excellent little article at Vox Day about correlation and causation that has real world value on the outbreak of Cholera:

    http://voxday.blogspot.com/2019/01/heuristics-are-not-proof.html

    Perhaps this too should be sent to Taleb. Even primitive humans like Indigenous Natives knew that if they found a dead animal at the waterhole maybe they shouldn’t drink there. They didn’t need to be a microbiologist to figure out which microbes were in the water.

    • Replies: @James Thompson
  14. @niteranger

    Er, that’s my article, isn’t it?

    • LOL: Hail
    • Replies: @niteranger
    , @Sean
  15. @James Thompson

    Of course…..I was just checking if you’re reading because Vox really didn’t give you credit up front.

  16. @niteranger

    Thanks for letting me know about it.

  17. dearieme says:
    @James Thompson

    Wedding receptions, eh? Are wakes safer? Christening parties? Silver weddings?

    Over the years my fondness for sausage rolls has been moderated by the tummy upsets that can follow them. Biltong has caused problems too.

    I never have any trouble with pork pies (preferably Melton Mowbray) or with tinned corned beef (from your native land, perhaps?).

    Anyway, these correlations are consistent enough to suggest causation. I now eat sausage roll only if I’m the fellow who has bunged it in the oven, and I eat it only piping hot.

  18. j2 says:

    “Michael Woodley has set a challenge: “Sure, correlation does not equal causation, but find me just one single instance of a causal relationship where there is no correlation (just one would suffice).”

    Is there any prize on solving this? I mean it is easy. The period of ice ages between integalaciation, about 100,000 years +-25,000 years. Everybody agrees that the reason is some variation in the radiation on the Earth, because what else could it be. That is causation, it is accepted and most probably true. However, there is no suitable period related to the Earth’s orbit, that is, there is no good enough correlation with any period. Causation, but no correlation.

  19. MarkU says:
    @j2

    Is there any prize on solving this? I mean it is easy. The period of ice ages between integalaciation, about 100,000 years +-25,000 years. Everybody agrees that the reason is some variation in the radiation on the Earth, because what else could it be. That is causation, it is accepted and most probably true. However, there is no suitable period related to the Earth’s orbit, that is, there is no good enough correlation with any period. Causation, but no correlation.

    Sorry but no prize for you. If you (or the imaginary ‘everybody’ that you cited) believe that the only factor affecting the temperature of the earth is solar radiation then you are very much mistaken.

    • Replies: @j2
    , @Peripatetic Commenter
    , @j2
  20. j2 says:
    @MarkU

    OK, so another example: plaintext is the causal reason for cryptotext, but they do not correlate. Good cryptotext has no correlations.

    • Replies: @res
  21. @MarkU

    Of course. As we all know, Carbon Dioxide is the most powerful force in the universe.

  22. j2 says:
    @MarkU

    Take a function f that
    cannot be inverted in an efficient way and does not show any measurable correlations,
    let y=f(x), you ask
    give at least one example where y and x do not correlate. There are relations we know exist, yet they do not show up as correlations.

  23. res says:
    @j2

    So comment 7 above and the linked comment thread for Woodley’s thoughts on cryptography.

    • Replies: @j2
  24. Some things are associated with others. Some things you eat make you ill. Some animals attack you. Some places are dangerous, some people likewise. On a brighter note, some foods are tasty and healthy. Some animals can be domesticated, or at least are easy to hunt or trap. Some places are safe, and some people likewise.

    Already stupid comparisons this dying old trash is doing…

    And and and

    NOTHING NEW ON UNS

    • Replies: @ploni almoni
  25. j2 says:
    @res

    “So comment 7 above and the linked comment thread for Woodley’s thoughts on cryptography.”

    Yes. I missed that comment. It is basically correct.
    A causal relation can be taken as a function y(t)=f(x(t)) and if you want
    x(t) only to partially determine y(t) you add some n(t): y(t)=f(x(t),n(t)). To say that this is correlation in the eyes of God is nonsense. It is correlation if y(t) and x(t) correlate as functions.
    Correlation is a function. It does not mean the existence of a relation.

    We can still call it correlation if we figure out f and can calculate either an estimate of x from y or an estimate of y from x and show that they correlate. But this cannot be always done for theoretical reasons (pseudorandom one-way functions show it), so there are causal relations that do not show as correlation.

    I gave the timing of the ice ages as an example. They have had a rough period of 100,000 years lately. Finally all that can effect the period goes down to the radiation from the sun and some other input, like cooling of the magma inside the earth. But the function is terribly complicated with biomass developing and producing CO2 and other gases and the seas and glaciers and winds and volcanoes and all this. So, everybody, or at least me, I hope also you, admits that in the very end radiation from the sun causes the periodicity of the ice ages, but through a function we do not know and this function involves all those aspects you were thinking of, but they also depend on the sun, so there is causation but we cannot figure out the correlation.

  26. Sean says:
    @James Thompson

    But Snow also came up with a theory about rickets that seemed unlikely in the light of later knowledge (though not Darwin’s sexual selection book) about vitamin D, and light skin being found at northern latitudes of Europe. Lo and behold, it turned out that stone age Britons’ DNA shows they were black. So we should not be dogmatic, in theory.

    It seems to me that accepting ‘the way to bet’ on IQ and race was on correlation is predicated on no reason not to, apart from the odds of being right . But there are reasons. It is the high IQ professionals of the ruling class who it suits to pretend that there is already a classless meritocratic system in place, though unfortunately being stymied by ignorance and prejudice which racial preferences are there to break down. Immigration isthe importation of allies by a technocratic ruling class against their only dangerous competitor the working class majority.

    It is conceivable that a feel-good one-nation conservatism could come to power. However being feel-good it could not be elected while explaining the facts of life in relation to genes and intelligence to the population. Hence racial preferences would have to be abolished on ceasing-to–discriminate grounds in the unspoken expectation that there will be a equality of result when a efficient welfare and education system upgrades the people irrespective of race. Hence any such government is going to run into terrible trouble as soon as it becomes obvious that the millions of people of other races in the UK are unable to compete on equal terms in a genuine meritocracy.

    Bio-logic is not compatible with individualist-egalitarianism, and for the latter to be discredited the country as a whole would need to face an external threat. The lower orders cannot win against massive wealth backed by expertise in law and politics, media, universities and nonprofit organisations.. For them to lose their grip it would take cataclysmic events that no one would wish for.

  27. Sean says:
    @res

    “How to avoid facts and influence people”.

    That is very profound. Robin Dunbar the eminent anthropologist has a theory that the development of human intelligence was partly driven by a need to navigating the social landscape of large groups. More pertinently he says that conversation is for influencing people by social grooming in which actual factual information (apart from gossip about other people) is being given.

    Quite the opposite, because according to Dunbar people who want to impart information are the classic cocktail party bore. Taleb says something very similar in his latest book about how the classical Art of the Courtier book written in Renaissance Italy wisely advocated an balanced equality between the participants as what to aim for in a conversation. This is the way to influence.

  28. Dutch Boy says:

    “Correlation is a pre-condition of causality.” This cannot be emphasized enough to the nitwits who chant the “correlation does not prove causation” mantra as though correlation disproves causation.

  29. @niteranger

    I like Vox’s comparing correlation to a clue. Maybe that clue proves someone guilty, or maybe it points you in the right direction to finding additional evidence that helps you solve the case, or maybe it’s a false flag. But two out of three ain’t bad, certainly better than the odds that you had before finding the correlation.

  30. dearieme says:

    I understand that there have been lots of observational studies in medicine that have thrown up correlations that have not survived scrutiny in controlled experiments i.e. in clinical trials.

    Maybe somebody should write a book about it.

    (By which I mean, surely somebody must have written a book about it.)

  31. On the contrary, correlation is a necessary feature of causation, but not a sufficient proof.

    Sociology-Inumeracy Alert.

    True iff the relationship is linear and contemporaneous.

    I also note that you made absolutely no reference to the fact that the correlations must be of variables that are both de-trended – because the number 1 source of spurious correlations is when the relationship is not between I(0) variables (variables that are integrated of order zero); that is, they have no trend and are not both random walks (since a random walk is not I(0).

    It is trivial to construct a series where Y[t] is a deterministic linear function of some distributed-lag mechanism in X[t-i] , {0<i}, but where Y[t] and X[t] have zero (or near-zero) correlation.

    It's also trivial to construct series in which Y[t] is a completely deterministic non-linear function of X[t], but ρ[X,Y] is zero (or is statistically not different from zero).

    I used to generate such series all the time as tutorial examples to illustrate why the Gauss-Markov Conditions (for OLS regression) were so important, and why (contemporaneous) correlation is absolutely not a necessary condition of causation.

    You spent some number of words trying to exculpate the shitful statistical drivel that is the stock-in-trade of the sociologist and the psychosophaster, and just managed to add to the evidence. Stop digging.

  32. mikesmith says:
    @j2

    “The period of ice ages between integalaciation, about 100,000 years +-25,000 years. Everybody agrees that the reason is some variation in the radiation on the Earth, because what else could it be. That is causation, it is accepted and most probably true. However, there is no suitable period related to the Earth’s orbit, that is, there is no good enough correlation with any period. Causation, but no correlation.”

    Sure there is: the obliquity cycle of 41,000 years. Earlier in the Pleistocene, we used to have one interglacial in each 41k years long obliquity cycle, but for the last million years it has been only one interglacial for a series of either 2 or 3 obliquity cycles (82k or 142K years, all together averaging out about 100k years). Less frequent interglacial episodes might indicate that the earth is in a secular cooling trend. The last glacial cycle lasted for 3 obliquity cycles, and the current obliquity cycle indicates that our interglacial will terminate some time during the fourth millennium AD, although we are so close that I would not totally rule out the possibility of an earlier termination. On a positive note, even if the Little Ice Age period (AD 1300 to 1870) is an indication that the cooling trend that began about 6000 years ago is steepening, it is not a straight line down, and we now know that previous glacial inceptions were preceded by warming periods of similar duration and intensity to what we experienced in the last quarter of the 20th century (when the sun was more active than in 8000 years, back in the Holocene Thermal Maximum), and sometimes more than one such episode, so mankind may be lucky enough to again experience one or more blessed but brief episodes of global warming before our long interglacial summer ends. Not us, though. Atmospheric temperatures have been declining for over a decade, and, more deterministically, ocean temperatures since the 1990’s, so colder winters are in our future, and more erratic weather as the temperature gradient between the tropics and the temperate zones widens. Prepare accordingly.

    • Replies: @j2
    , @j2
  33. Sean says:

    Society decides what is good science. It has to be that way, and probably is not that way nearly enough.

    • Replies: @jilles dykstra
  34. Anon[301] • Disclaimer says:

    I used to agree with the article’s claim, but in the field of genetics we now have “The Fifth Law of Behavioral Genetics,” from Emil Kirkegaard, also known as “Finding 4” from Robert Plomin’s paper “Top 10 Replicated Findings From Behavioral Genetics.”

    All phenotypic relationships are to some degree genetically mediated or confounded.
    http://emilkirkegaard.dk/en/?p=6636

    Phenotypic correlations between psychological traits show significant and substantial genetic mediation.
    https://www.gwern.net/docs/genetics/2016-plomin.pdf

    JayMan on the Fifth Law:

    Whenever there is an association between two phenotypes (such as poverty and crime), there will be a genetic association driving both.

    This was drawn from studies like those of Amir Sariaslan’s and others showing the confounded nature of phenotypical associations (even extended phenotypes like social circumstances).

    This essentially strikes at the heart of modern social science (and for that matter, medical science), which assumes, wrongly, that association between social and/or behavioral factors is an indication that one causes the other. In reality, genetic forces cause both.

    Indeed, we see this with health and lifestyle: people who exercise more have fewer/later health problems and live longer, so naturally conventional wisdom interprets this to mean that exercise leads to health and longer life, when in reality healthy people are driven to exercise and have better health due to their genes.

    http://www.unz.com/jman/the-five-laws-of-behavioral-genetics/

    Another specific example of this is body weight, extensively discussed by Plomin in his book Blueprint, as a proxy for more dangerous topics like race and intelligence. Body weight is 70 percent heritable (80 percent in adults), and correlates with a number of traits and activities, but … wait for it … [I’ll leave this for the student to review at home.]

  35. j2 says:
    @mikesmith

    “Sure there is: the obliquity cycle of 41,000 years.”

    You mix up two things: 1) is there an orbital cycle that we may with justification consider as a probable cause of ice ages, and 2) is there a cycle that correlates significantly with the start of ice ages.

    See the situation e.g. from the always as reliable Wikipedia:
    https://en.wikipedia.org/wiki/Milankovitch_cycles
    and notice this text in it
    “Some have argued that the length of the climate record is insufficient to establish a statistically significant relationship between climate and eccentricity variations”

    The expression “some have argued” apparently means that more than one statisticians has pointed out a fact that the correlation is not statistically significant. Though is only says of eccentricity, it is true also of obliquity. The time series is too short.

    Thus, 1) is true and the cycle is either the obliquity cycle with its multiples, or the a subcycle of the eccentricity cycle. Neither one of them gives the 100,000 year cycle, but inventive scientists can always explain these things with a strange combination. Just see the figure of the total insolation from the wiki page (the last figure, insolation of 65 N), the correlation of the combined effect with the weather is rather weak and not really statistically significant. 2) is not filled.

    What your comment however makes clear is that there is good reason to expect that the cause is
    Milankowitch cycles, but we just do not agree if it is obliquity with 41,000 period or eccentricity with 413,000, 125,000 and 95,000 period that is the reason for 100,000 year period. Mainly because however you do it, the result seems to be before the cause in some point of the graph, and that is not so good for a causal reason.

    • Replies: @jilles dykstra
  36. j2 says:
    @mikesmith

    “Atmospheric temperatures have been declining for over a decade, and, more deterministically, ocean temperatures since the 1990’s, so colder winters are in our future, and more erratic weather as the temperature gradient between the tropics and the temperate zones widens. ”

    Otherwise, global weather and its change and the reasons for it is an interesting topic, but a bit off-line from this thread. Maybe UNZ will publish some article of it sometimes, then we can comment what the reasons and the future might be, like why the natural feed-back mechanisms do not keep CO2 in limits as they should.

  37. dearieme says:
    @James Thompson

    The Cochrane Collaboration? Good idea; could be. The one I’ve just found is

    Critical Appraisal of Epidemiological Studies and Clinical Trials, Mark Elwood

    Paperback – 2 Feb 2017

  38. utu says:

    “Sure, correlation does not equal causation, but find me just one single instance of a causal relationship where there is no correlation (just one would suffice).”

    A periodic function. A causation with zero correlation.

    • Replies: @res
  39. Sparkon says:

    No. Milankovitch Cycles progress far too slowly to account for the relatively rapid (in geological terms) observed climate change on Earth just over the last 2000 years, when the Roman Warm Period was followed by the Dark Ages Cold Period followed by the Medieval Warm Period (When the Vikings farmed Greenland) followed by the Little Ice Age (when panicky, superstitious humans blamed witches(!) for the poor weather, and burned many) followed by the Modern Warm Period, where some humans are blaming the beneficial trace gas CO₂, so there is no correlation with any MC.

    Additionally, a fatal blow to the Milankovitch Cycle theory is that recent climate change like the Little Ice Age was observed in both Northern and Southern Hemispheres simultaneously, and you can’t do that with MCs.

  40. A long story to explain the obvious, in Sweden storks bring the babies, according to statistical high correlation between the returning of the storks, and the births.
    Of course we have the same problem with average earth temperature and CO2 ppm.
    We live in an era where facts no longer matter, as Lasch already wrote, and Sarrazin explains today.
    Christopher Lasch, ‘The Culture of Narcissism, American Life in an Age of Diminishing Expectations’, 1979, 1980, London
    https://www.achgut.com/artikel/zwei_parteiausschlussverfahren_und_drei_faelle_von_gezielter_taeuschung
    Twice it was tried to take his party membership from him, twice it failed, he had simply stated the unwelcome facts.

  41. @j2

    What your comment however makes clear is that there is good reason to expect that the cause is Milankowitch cycles

    It has been demonstrated that these cycles cannot explain the ice ages.

    • Replies: @j2
  42. @Sean

    Society can decide anything it wants, but the decision that ripe apples fall towards heaven will not change the behaviour of ripe apples.

    • Replies: @Sean
  43. Bruno says:

    If’you find correlation among DNA and a trait, it’s causation. The warning « correlation isn’t causation », however mostly true, doesn’t held in some instances.

  44. Michael Woodley has set a challenge: “Sure, correlation does not equal causation, but find me just one single instance of a causal relationship where there is no correlation (just one would suffice).”

    The simultaneous occurrence of two events is sufficient to preclude a non-zero correlation between the two, so the challenge is a nullity that cannot be tested.

  45. @j2

    Isn’t the classic example, y = x^2 with x a unit normal?

    • Replies: @j2
  46. @Santoculto

    This correlates to a troll.

    • Replies: @Counterinsurgency
  47. Sean says:

    Snow jumped to a conclusion because his mind was prepared to interpret associations in a particular way, intially by his doubts about the air transmission miasma theory and later by his own hypothesis of water-borne transmission. He jumped to the right conclusion, without proofs of the causal mechanism which were only available years later.

    Correlation is not causation, but it’s the way to bet.

    But which correlation? In the case of rickets he was proved to have selected the wrong one, or was he

  48. @Peripatetic Commenter

    The idea that some alleles for higher IQ have become fixed in some populations and not others is an interesting one although the logical possibility that lower IQ fixation have occurred in some populations seems bizarre. And of course things are bound to be more complicated with odd epigenetic effects and no doubt interactions yet to be discovered.

    I see the problem with taking your
    (and others’) restricted variance in females theory very far as being its requirement that a significant number of what might be called IQ alleles need to be on the X chromosome.

    • Replies: @j2
  49. res says:
    @utu

    See comment 7 above and the linked thread.

    I tend to agree with you though, and think Woodley has watered his assertion down until it is worthless in a practical sense.

    • Agree: utu
  50. @Sean

    “allies against the working class majority”.

    You neglect “more net taxpayers to share support of the tax eaters”.

    In Australia I hope we are building the smart fraction through immigration in the face of dysgenic breeding by the natives because the comfort provided by China (mostly) buying our iron ore and coal won’t last forever. For the US it is almost infinitely more complicated and I don’t see a ruling class which can solve the problems by intelligence.

    • Replies: @YetAnotherAnon
  51. Every spring women start wearing less clothing and plants turn green. But I’ve been unable to determine if the decreased clothing is causing increased greenery, or if the increased greenery is causing decreased clothing.

  52. The claim that ‘correlation does not equal causation’ is therefore meaningless when used to counter the results of correlative studies in which specific causal inferences are being made, as the inferred pattern of causation necessarily supervenes upon correlation amongst variables.

    Never have I felt more like the affirmative action student in a post-graduate class.

    • Replies: @Wizard of Oz
  53. Sean says:
    @jilles dykstra

    Perhaps, but if something was was deemed true by society and government that scientists objected to, then scientists would cease to listen to either government or society. Some scientists may be about to meddle in things far too fateful for them to play Lone Ranger. We may be on the verge of an invention that will have a mind to cancel our series.

    For a long time scientists were predicting artificial general intelligence was imminent, yet said nothing about risks. It was only in scifi that things like Colossus or Sky-net were predicted, and as Bostrom notes this is the clearest possible indication that scientists will not pay sufficient attention to the risks of AI, if operating within their own sub-cultural assumptions.The most brilliant scientist are quite often viscerally opposed to saying , for example, that intelligence is genetic, let alone that it is in a way that can identify disadvantaged groups as genetically disadvantaged.

  54. j2 says:
    @jilles dykstra

    “”What your comment however makes clear is that there is good reason to expect that the cause is Milankowitch cycles”

    It has been demonstrated that these cycles cannot explain the ice ages.”

    You are right, but what I mean is also right. Milankowitch cycles alone cannot explain the ice ages,
    but added with other mechanims they are a causal reason for ice ages. Or let us say: radiation from the sun is a causal reason for the ice age period, because if there is no radiation from the sun, ice does not melt and if ice does not melt, there is no ice age period as period implies melting and freezing.

    There are other reasons, like the Younger Dryas apparently were caused by a comet impact 12.800 BC,
    and there has been a warming cycle of 2000 years, which most probably is internal to the Earth and not dependent on insolation. It may be a biosphere cycle, or an atmospheric cycle with glaciation and CO2, but it is not a Milankowitch cycle. Milankowitch cycles may explain the 100,000 period, but only if you add other mechanism that depend on the sun.

    So, variation in insolation + lots of Earth related mechanims that also depend on the sun – give in some complicated way a rough period of 100,000 years, which earlier was 41,000 years, but the time series of temperatures there is from glaciers is too short to get a working model and show that there is a correlation.

    I consider radiation from the sun to be the causal reason of ice age period to be obvious for logical reasons (no sun-> no melting of ice-> no ice age period), so it is true, but a causal correlation has not been shown to exist and it does not need to exist, even though there is causation.

    • Replies: @jilles dykstra
  55. Anonymous[375] • Disclaimer says:

    The “correlation is not causation” mantra is true as far as it goes, but it tends to be used so as to argue that, despite many correlations linking A with B being found in different circumstances, these will somehow never suffice to strongly suggest a causal link between A and B. On the contrary, correlation is a necessary feature of causation, but not a sufficient proof. Correlation is not always causation, but it helps find causes. Correlation is a pre-condition of causality.

    At this point, why not just drop the notion of causation altogether? It adds nothing. There are only patterns of correlation, and causation is nothing but correlation. Bertrand Russell and Quine regarded causality as a relic of animism:

    “The law of causation,… is a relic of a bygone age, surviving, like the monarchy, only because it is erroneously supposed to do no harm.”

    “In the motions of mutually gravitating bodies, there is nothing that can be called a cause, and nothing that can be called an effect; there is merely a formula.”

    • Agree: AaronB
    • Replies: @James Thompson
  56. j2 says:
    @Beefcake the Mighty

    “Isn’t the classic example, y = x^2 with x a unit normal?”

    You mean y and x are real valued time functions and x has expectation zero and is stationary, and the correlation function (as calculated e.g in the signal theory) is zero:
    C(s)=int y(t)x(t-s) dt= int x^2(t) x(t-s) dt= int_x(t-s)>0 x^2(t) x(t-s) dt – int_x(t-s)<0 x^2(t) x(t-s) dt
    =0
    but these IQ people take a correlation coefficient that uses the covariance cov(x(t-s),y(t)) and normalize it with standard deviations. In this case calculating the covariance you have to subtract the expectation value of y, so it also varies around the zero. But it does not change anything. You can move the expectation value of y outside, so the covariance and thus the correlation coefficient is zero.

    Yes, you are correct and you should get the prize if there is any, but they will say you failed as they would just divide x to positive and negative parts and take a square root of y and get a perfect correlation and say that that is what they meant. You see, they always cheat on everything and I do not mean (((they))) but all they.

  57. @Anonymous

    As a sometime billiards player, I find your argument hard to follow.

  58. Muse says:

    I really love this video by Daniella Witten that explains statistical methods used to infer direct relationships between different variables in complex biological networks. There is a lot of detail beyond causation and correlation, but if you can follow her discussion, you will have cause and correlation down cold.

    Sure some have seen it. Believe Steve Hsu originally posted it on his blog.

  59. j2 says:
    @Wizard of Oz

    “being its requirement that a significant number of what might be called IQ alleles need to be on the X chromosome.”

    there are many alleles that cause reduced intelligence in X and they do widen men’s IQ scale, as mildly retarded cannot be excluded.

    “the logical possibility that lower IQ fixation have occurred in some populations seems bizarre.”

    The higher IQ allele did not arrive to all populations, so they are limited to what they have. If it is
    one allele, the allele is fixed.

    Schools are one important factor reducing the standard deviation. In Finland you see SD more like 3.5 instead of 4-5 as in the USA in WAIS standardization sample. It is between these figures in Scandinavia. The likely reason is equal teaching level in all schools. You can artificially increase (or lower) the number of high (low) achievers by teaching them. About the same dirty trick as is used in sport: some
    crooked people excercise more and appear better in all tests and competitions.

    • Replies: @Counterinsurgency
  60. @j2

    but added with other mechanims they are a causal reason for ice ages.

    Alas as far as I know these other mechanisms have never been identfied, there is just speculation.

  61. Where’s the usual “low IQ = something or other” correlation you guys are always whinging on about? You forgot it.

    • Replies: @Wally
  62. As we’ve expanded our universities and increasingly funded research into areas that would have gotten a belly laugh 2 generations ago, we will get more and more grad students in need of more and more untilled fields to plow through.

    The correlation-causation gap will widen until causation becomes meaningless. Feelings not Facts are the future.

  63. @Nancy Pelosi's Latina Maid

    No, that was the affirmative action student getting his rehearsed lines a bit muddled at the end 😉

  64. @Wizard of Oz

    “in Australia I hope we are building the smart fraction through immigration”

    If the immigrants are from non-Brit groups whose primary loyalty is to their own group, you may be storing up future trouble. The last thing you want is high-IQ, wealthy outsiders* funding your major parties and controlling your major businesses.

    What are levels of Chinese immigration like? You only have to look at the map to see how isolated Oz and NZ are.

    * i.e. people whose ethnic interests aren’t those of your historic population.

    • Replies: @Wizard of Oz
  65. utu says:

    A belief that correlation always is an indication of causation (as Woodley seems to suggest) can lead one seriously astray particularly when dealing with undetermined system such as genotype vs. trait when samples are small.

    This was the case of Davide Piffer who selected 8 or 9 SNPs to calculate a polygenic score that correlated very well with Lynn’s IQ’s for 20 plus countries. While estimating the P-value he found that with probability of 1/100 (iirc) he could find other sets of SNPs that produced even higher correlations. This however did not alert him the correlation he found was spurious.

    Among 10 millions of SNPs there is, for all practical reasons, infinite number of combinations of 8-long sequences so virtually for any sequence of 20 plus radom there is a polygenic score that will correlate with it.

    The problem of over-fitting is a serious problem when trying to find a predictor function for a trait like IQ score because only the existence of an independent validation sample can guard against the over-fitting. If however the validation sample will be lost by being absorbed into the exploratory sample there will be no way of verifying the validity of a fit. So once all 8 billions people are genotypes and IQ tested and some evil genius will fit their IQ’s with a polygenic score using a subset from 10 millions of SNPs there will be no way of validating this fit until new people are born.

  66. George says:

    IQ and human genetics are not as stable as you assume. South African President Jacob Zuma seems like a pretty cable, high or higher IQ sort of person. I don’t know the details of his 6 wives but I am going to assume they are as smart or smarter than he as the few pictures I have seen seem to imply he went for quality. So his 20 children will likely also be high or higher IQ according to your genetic theories.

    In Europe 6 women as smart as smart as Zuma’s wives might have 5 children while Zuma’s 6 wives produced 20. So the IQ genetic process you imagine as continuing generation after generation might be quickly reversed due to very low Euro fertility and African polygamy.

    Who ya gonna bet on in 2100 Europe or Africa?

  67. @YetAnotherAnon

    I can’t give anything like a statistically valid response though I think highly mixed Chinese and Indian descended approx 5 per cent post 1966 populations are likely to fit in well with existing middle middle and upper middle class people including the core Anglo Celtic who came to see the post WW2 East European, Italian and Greek immigration as a good thing. So many will have come as students, many at primary or secondary school level that they won’t have a strong ethnic identification and will I trust be multiracial defenders of capitalism and anti taxeater immigration policies.

    The sort of details that encourage me are the obvious differences in background and culture of SE Asian Chinese and HK compared with mainland Chinese (and each other), the frequent phenomenon of wife being in Oz with the school kids while husband father is away doing business, the likelihood that they have reason to be wary of the CCP and Chinese government, even the possibility that they are making it easier for their kids to get into a good university (and Australia has plenty that are worth attending) rather than face the intense narrow meritocratic competition in China (after all Australia prima facie benefits if that brings IQ 125 kids to grow up in Australia).

    Amongst my half East Asian young relatives I see that they are all bright and popular. The fact that they are all strikingly talented at sports and generally seems to be a plus rather than a source of envy and snide bitching. You don’t get elected to be captain of your surf club in Oz if your contemporaries harbour racist feelings.

    By contrast, my charming former Yugoslav bookkeeper who has been living happily in an apartment in a bustling multi everything wayside suburb tells me of suddenly feeling unsafe, especially for her daughter, because of large crowds of blacks of African appearance behaving threateningly, and worse, on the beach. As there are only 8000 African origin people in a city of 5 million I trust the state government and police can find the gumption to fix this and justify my long held view that the idiocy that allowed the Africans in will pay off as inoculation….

    • Replies: @YetAnotherAnon
  68. You sound like you are hoping so bad that it’s true. The reality is that even with negroes that have higher IQ’s than average (Bill Cosby et al), they seem to have this built -in gene for self-destructive behavior that always makes them take two steps backward for every one forward.

  69. Forbes says:

    Amongst adolescents (pre-teens) shoe size is correlated with academic achievement, but surely not the cause. Shoe size is merely a proxy for age, and older students have experienced more schooling and hence, will demonstrate greater academic achievement.

    There are many more examples like this. The article seems a bit overdone to demonstrate correlation is causation, when often cause/effect is inverted, or only slightly related. And yes, often correlation is the statistical demonstration of cause and effect.

    The caution that correlation is not causation is prudent against the misuse of “proving” a relationship effect that doesn’t exist–or is merely a curiosity. If cause and effect is demonstrable, then correlation is confirmation.

  70. RJJCDA says:

    When quoted that “correlation is not causation,” usually by recent college grads, I respond: YES, BUT ITS A CLUE!

    • Agree: utu
  71. @Sean

    I’ve spent decades working with elite outwash. “High IQ” isn’t the way they look up close, or at a distance once you see through the mask. The current US elite classes are similar to but clearly worse than the US elite class during the Gilded Age. They have the same corrupt Tammany Hall politics, but they destroy rather than build. They are on the way out, although I can’t say whether they’ll be replaced by something better or worse. Trump should have moved by now, for example. I can’t predict outcome even for the present super-obvious situation, hence can’t say better or worse for the long term.

    But:
    We don’t usually get what we wish for, good or bad.
    “Plant a carrot, get a carrot,
    not a Brussels Sprout.
    That’s why I like Vegetables.
    You know what you’re about.”
    “The Fantastiks”
    Go beyond causation like the above, and wishes don’t come true.

    Elite classes can die, as corporations die (30 year lifespan?), as nations can die (Anglo Saxon America, Communist USSR, Aristocratic UK), as the entire West dies (Religious West, Royalist West, etc., with noticeable cycles of about 80 years on average, used to be). Elite death happens, and a lot more often than people notice. The current set of elites is walking dead.
    Which means that replacement elite classes, corporations, nations, are born a lot more than people notice. The next set of elites is walking and alive, but (as with the late 1960s) nobody notices them even when everybody is thinking about them. It’s the way of the world.

    Counterinsurgency

    • Replies: @Wizard of Oz
  72. Wally says:
    @obwandiyag

    The focus is other than the very low African, Latin American IQs …. which everyone already knows about.

  73. @Counterinsurgency

    Your case would sound more convincing if you didn’t convey such a vague notion of what you think are or were elites. The idea that the “US elite class during the Gilded Age” had anything to do with the New York thugs of Tammany Hall who controlled Democrat patronage for quite a long period (mostly after the Gilded Age) is ludicrous. Even the Californian parochial Alden (now [Anon257] ) would know that.

    • Replies: @Counterinsurgency
  74. @Wizard of Oz

    “as there are only 8000 African origin people in a city of 5 million I trust the state government and police can find the gumption to fix this and justify my long held view that the idiocy that allowed the Africans in will pay off as inoculation”

    It didn’t pay off as inoculation in 1950s and early 60s London, and Britain was a lot less pozzed in 1960 than Australia is now. I can’t think of any Anglo nation in which poor behaviour by an immigrant group has resulted in restrictions on further immigration.

    What you’re more likely to find is that the assaults won’t be publicised, and anyone pointing them out will be the problem.

    • Replies: @Wizard of Oz
  75. @YetAnotherAnon

    The difference is that Australia doesn’t have anything like the immigration policy of the UK in the 50s and 60s. And even the UK’s indiscriminate allowing of Commonwealth immigrants with no skills ended in the late 60s after Enoch Powell’s famous speech (and was applied to the old white Commonwealth citizens equally, which was unfortunate). The South Sudanese only got into Australia as part of the refugee quota and that’s not likely to be repeated.

  76. c matt says:
    @Alfa158

    I would phrase it as correlation is necessary for causation, but not sufficient. If two things are not correlated, one cannot be the cause of the other. If two things are correlated, one could be the cause of the other.

    Correlation also provides guidance, even if there is no cause/effect (e.g., staying out late at night in questionable areas of town may not cause your mugging, but it increases the probability of it occurring).

    Another faux-fallacy I always found annoying was the anti-slippery slope argument. Again, it is only a fallacy in the sense that if you give an inch, the other party is not required to take a mile. But history proves 99% of the time, they will.

  77. ”IQ correlates with better socioikomomic outcums”

    Correlates is when a factor is partly responsibly by another factor.

    When a factor essentially or is foundationally responsible for another so weak’all causalitet.

    Thanks self-mister Ozzi..

  78. Co-relation is not the same to coincidence of relation… my vil..

    Finnish people love ice cream. They have excellent socio-economuc outcum = coincidence of relation and not co-relation. Two factors there are in the same geographical space but it doesn’t mean they interact in cooperation.

  79. @Wizard of Oz

    It’s true that Tammany and the Anglo Saxon elite were on opposite sides during the Gilded Age. However.

    The Anglo Saxon elite, to which I was referring, was responsible for the industrialization of the United States. That elite was dominant back then (c.a. AD 1870, post-Civil War), and set large parts of national policy. It was clearly responsible for the flow of immigration, which it regarded as essential for rapid industrialization and the large projects that Steam Age equipment made possible.

    Tammany Hall, and similar organizations (or “political machines”, if you like) ruled the large cities, one organization per city. This was direct consequence of a large flow of polyglot immigrants, who had no history of or experience with self government. Anglo Saxon governmental forms of self government failed badly. In particular, self-government failed badly: office holders from strong man / clan societies used political organizations for their enrichment, not the enrichment of the city.
    The United States Federal Government could have declared martial law, and attempted to rule the cities by force. It did not do that, but it did insist on democratic rule within the cities, and there was a substantial remnant population of Anglo Saxons, quite a few of whom had economic power.
    Tammany was one of the organizations that emerged and, in effect, limited the damage of the mass migration to the US. Tammany (like its contemporary organizations, and its descendants contemporary to us) managed to rule and gradually pacify the cities, mainly by establishing a political base that found Tammany rule to its monetary advantage. In fact, they pacified the cities a bit _too_ much. As the US urbanized, Tammany and its allies transformed itself (that is, wrote the enabling legislation for) the contemporary welfare state. This welfare state wrote Tammany methods of getting and keeping political power into a bureaucracy that insulated political actors from the exhausting life of a ward boss. Insulated them a bit _too_ much, in fact — our present rulers appear to have forgotten that they have a political base outside of the check writing bureaucracy, and are spending their time arguing about (metaphorically) angels and heads of pins.

    OK, two sides: Tammany, Anglo Saxon establishment. Both sides hated the other, but they did interact [1]. Usually fought like dogs and cats, at least on the surface. Anglo Saxon victories were significant (preservation of at least the form of voting, no sanction under law for absolute rule in cities, Tammany under constant threat of the criminal code), but ultimately transitory.
    During the fight, however, (say, AD 1840-1942), the two sides not only collaborated (while maintaining mutual contempt and hatred), but did so with great success. The Anglo Saxons provided the jobs for immigrants, and Tammany Hall (and similar) provided the government that prevented actual secession of the larger cities from the US Federal Government, or open warfare between ethnic groups (no active IRA in the US, for example). There was nothing in the Gilded Age and following like the current partial secession of California and the developing secession of New York.
    So, if you wanted to, you could say that Tammany (and similar) was co-opted by the Anglo Saxon establishment, and kept order among the immigrant population. Or, you could say that Tammany (and similar) bided their time until they could take over the Federal government through simple demographics. Both statements could be correct, and both suggest _both_ mutual dislike and mutual cooperation — a rather amazing American success when you consider Europe from 1840-1914.

    Counterinsurgency

    1] Theodore Roosevelt.
    “Theodore Roosevelt: How we overthrew corrupt Tammany Hall”.
    https://internauta-online.com/2014/06/theodore-roosevelt-how-we-overthrew-corrupt-tammany-hall/

    • Replies: @Wizard of Oz
  80. “Correlation is not causation” is true enough as far as it goes.

    It it’s interpreted as “Correlation is not always causation”, or “causation implies correlation”, then it’s true.

    If it is interpreted as “If I take a data set, calculate the correlation coefficient, and it’s greater than some value, then I’ve found causation”, that’s false. You may have found a promising subject for further research, but that’s it. You need some kind of demonstration that varying the causative measurement actively results in a corresponding change in the caused measurement.

    If it is interpreted as per David Hume, “There is no observable reality, only constant conjunction of cause and effect, so causation does not exist”, or Kant’s extension of same “Measurements depend on instruments and decisions by people using them, so instruments don’t measure reality, they measure their designers”, or the current Postmodernist “There are no individuals, only groups. Groups oppress each other, and that’s all they do. Therefore, instruments do not measure reality, they measure nothing, and are tools of oppression. An science, if it existed, would describe reality quite differently, and be just as valid as any other science since would validate it socially, and there is no validation except social validation.” Or, in most basic terms “Let’s end all this endless talk and fight, see who ends up oppressing whom.”
    Well, if that’s the interpretation, and it’s the dominant one today, socially imposed and socially validated, that interpretation is false in that it doesn’t lead to successful prediction.

    Counterinsurgency

  81. @j2

    The higher IQ allele did not arrive to all populations, so they are limited to what they have.

    That can be a good thing.
    Intelligence demonstrably doesn’t always mean survival. Consider the expulsion of Europeans from Africa, even from South Africa, where the Europeans arrived before the Bantu expansion arrived, but eventually lost to it. Intelligence seems linked to limited fertility; I’ve seen references to that in non-human organisms (Fish, if I remember correctly), although I don’t know that results have been replicated. High IQ members of the Australian aboriginal population seem to have been displaced by low IQ but better physically adapted members, possibly after sparse populations made retention of complex tool making non-adaptive.

    “If it works and it’s stupid, then it isn’t stupid” is another way to put the above.

    Counterinsurgency

  82. @ploni almoni

    Yep. Note the viciousness of the comment. On Unz, at least, Troll correlates to Postmodernist SJW. Vicious attacks of various sorts are about the only way left for Postmodernists to interact with others.

    In the artificial massive multi-player dungeons there are always people who simply show up, kill every other character they can reach, then vanish. In one account of this they managed to “capture” the killing character, and attempted to talk to it. The capturing group were of a universalist / social constructionist sort, and offered the person full membership in their group if he would simply stop killing them. The person bluntly refused, and logged off for good. The account’s author hypothesized that the captured person was incapable of social interaction, but needed it, and was killing other characters _because killing is the simplest form of social interaction_, all the killing character’s mind could support.

    I suppose that, given an increasingly isolated and anomic population, the extremely simple social interactions that characterized Postmodernists / leftists should be expected to increase.

    Counterinsurgency

    • Replies: @Santoculto
    , @Santoculto
  83. @Counterinsurgency

    Some interesting meat in that now that you have focused on that era/those eras.

    • Replies: @Counterinsurgency
  84. JLK says:

    On Unz, at least, Troll correlates to Postmodernist SJW.

    The trolls here can be very creative. There’s not much than can done about the paid ones. I hope that some of the others will eventually see the benefits of unrestrained exchanges of ideas. As Ron Paul said once, libertarians have to be prepared to tolerate a few things that they don’t like.

    I used to really enjoy posting at another forum. It had high level people from a number of different disciplines, and we thoroughly got into various world events. Something apparently touched someone’s nerve, and it was spammed out of existence. Hundreds of nonsense posts per hour, faster than the mods could delete.

    So some of us are here putting up with the dreck because we were herded away from greener pastures.

  85. @Wizard of Oz

    Thanks!
    There is another book, an entertaining one that’s a primary source by Riorden [1] from the late AD 1800s. It is generally considered that Riorden, a newspaper man back then, was effectively a ghost writer for George Washington Plunkett, a Tammany ward boss. It presents Tammany Hall as it wished it could be.

    Plunkett could be used to support a case that Tammany Hall was better than the current system. At least the old ward bosses had _some_ contact with the inhabitants of their territory. The present system is epitomized by the argument under L. B. Johnson that Black families should divorce so that their children would get more money. I’ve seen that argument used _by the woman’s mother_ to successfully prevent marriage to an employed father who wanted to marry a women who was having his child. Bad system.

    Anyway, enjoyable book, has more on that era. The whole story, from start of mass Irish immigration (AD 1845) to immigration’s end in AD 1924, is fascinating.

    Counterinsurgency

    1] William L. Riorden.
    _Plunkitt of Tammany Hall: A series of very plain talks about very practical politics_.
    Various publishers, also available from Project Gutenberg.
    It is generally considered that Riorden, a newspaper man back then, was effectively a ghost writer for George Washington Plunkett, a Tammany ward boss. It presents Tammany Hall as it wished it could be. Plunkett supports case that Tammany Hall was better than the current system.

  86. @Counterinsurgency

    Demons want to put the fallacy

    ”good is to be coward”

    ”goody is to be good with evil”

    Most writers here are sociopaths… as this old dying tra$$h..

    ”Those who don’t have mercy don’t deserve it”

    The world would be fantastic with this simple sentence was the common practices.

  87. @Counterinsurgency

    These DEMON conserFs manipulate every peace of reality…

    MOST trolls are right-wingers!!!

    Most snowflakes, those who suffer massive discrimination and became over-upset are left-wingers..

    To be intolerant with the intolerant is nothing bad at all, sound homeopathic but trutful.

    A lot of leftists are intolerant but many [not all] of this intolerance IS ABOUT right winger anachronic intolerants…

Current Commenter
says:

Leave a Reply -


 Remember My InformationWhy?
 Email Replies to my Comment
Submitted comments become the property of The Unz Review and may be republished elsewhere at the sole discretion of the latter
Subscribe to This Comment Thread via RSS Subscribe to All James Thompson Comments via RSS