Hear Briggs Speak Free! Where by “Free” I mean you pay my expenses. Here are the details.
Hear the One
William M. Briggs
The Statisticians to the Stars!
speak on the Astounding Subject
Why You’re Doing Statistics Wrong.
Listen and learn…Why hypothesis tests should on no account be used for any purpose…Why p-values should be tossed outside the gate where there is weeping and gnashing of teeth…Why model parameters are no use to man or beast…Why probability models cannot determine causal relationships…Why regression is not what you think…Why you should be making predictions and not over-examining model fit…Why some models can be deduced and not appealed to ad hoc…Why some models are better than other and how to tell…Why you should learn the Third Way of Probability & Statistics…
And more importantly…Why you are making faulty, sub-optimal, and wrong decisions..and much more!
If you’d like examples of what steps the Dancing Briggs makes, you can watch these videos (many will be reminded of Fred Astaire):
- The Crisis Of Evidence, Or, Why Probability & Statistics Cannot Discover Cause
- The Need To Believe In “The Solution” To Global Warming
- The Mysticism of Randomness
- Probability Is Logic
- Statistical Follies and Epidemiology
I don’t have anything to sell except these ideas. I’m anxious we do a better job and reduce the pandemic of over-certainty the old ways have caused.
If you can’t have me come and speak, but still want to contribute—don’t forget I’m independent and free of all regular sources of employment—you can! By clicking here.
Categories: Fun, Statistics
I’ll have to wait for the book. I can’t do YouTube videos (as seen by my giving up on my last online course because they used YouTube videos and no text). Sorry.
I don’t know of any where to suggest you go for speaking. The one university in Wyoming is overrun with liberals who hate oil and gas but love their paychecks. It’s some form of mental illness and it may be catching. I’ll drop on over and pick up one of your classy mugs at Zazzle.
Keep up the good work. It’s frustrating, I’m sure, but someone has to do it.
Actually, Zazzle requires me to make an account, so I won’t be ordering your mug unless you have a different avenue for purchase.
The Crisis Of Evidence, Or, Why Probability & Statistics Cannot Discover Cause
This title is misleading. You don’t actually show (except for a single methodology — namely hypothesis testing) that statistics cannot. You should change the title by altering “Statistics” to “Hypothesis Testing”.
DAV: How does statistics show cause other then hypothesis testing? I can’t really think of any examples. Do have one or two?
The simplest example is when faced with three correlated variables A, B and C and A and B are independent given C then C can be labelled a cause at least for predicting A and B which is what people really mean when they say X causes Y.
Isn’t what you are doing confirming the hypothesis that C is the cause?
Perhaps but not with what is called hypothesis testing which doesn’t really do anything but affirm what you already know, if that. In addition, note that in my example there were three variable s and not two. Standard hypothesis testing doesn’t make this distinction and is often applied to two only two variable. Three variables is the minimum required and, even then, all of the conditions in the example must hold or you can’t assign a causality label.
DAV: I was not aware that hypothesis testing was limited to two variables. I have seen examples with three variables and these were called hypothesis testing. There are also multivariable hypotheses, correct? I don’t understand how your example is different in a way that does not affirm what you already know. Are you referring to just taking data and throwing statistics at it to look for causes? That would be different than hypothesis testing so probably not? I just don’t see the difference.
See the video on cause linked above. Hypothesis testing is based on a false dichotomy. Also, statistics per se cannot discover any cause.
Hmmmm….. the normal use of Hypothesis Testing is with two variables, e.g., does taking Unhirsote(TM) prevent moutstache growth? In these cases one starts with data (perhaps) that clearly shows that increasing doses of Unhirsote(TM) lip gloss is correlated to lip hair loss. All the hypothesis test can do is say “Hey! It really does (for these data)” or “Uhhh, maybe not”. IOW: it doesn’t tell you anymore than what you have seen (or should have) yourself and doesn’t answer the question if the hypothesis is valid. It’s a pointless ritual.
The multivariate test is really a form of regression (e.g, Ax+By=z) and what the p-value tells you is how nice the coefficients are. which tells you nothing about the hypothesis of how z is related to x and y.
In my example, I can start with a data set containing three variables and test if the conditions hold for a given configuration of the (A,B,C) tuple and if so then I’ve found a causal relationship that I may not even be able to discern by inspection. It’s not a confirmation. But, yes, you DO have to start with the proposition that one variable causes the other two then test it and, if it does, what falls out (mostly because you needed to determine it to apply the tests) is Pr(A|C) and Pr(B|C) albeit estimates. IOW: your uncertainties in model predictions. You get none of this from so called “Hypothesis Testing”.
This example is the simplest case and rarely is found by itself in real life. However, it forms a stepping stone for much harder and far more obtuse problems involving many more variables. The net result would be a causal network.
statistics per se cannot discover any cause.
Perhaps it depends on what you mean by per se. The algorithms developed by Pearl and Glymour, Scheines and Spirtes can be quite good at structure recovery when presented only data.
Lately I’ve been using Tetrad. The algorithms have been tested by generating random causal networks; sampled; and the data passed to one of the Tetrad algoritms. As I said before, the reconstruction of the network from the data can be quite good although not perfect. It’s been a big help on my project.
Briggs: I’ll try the videos since they have closed captioning. Maybe that will work. I really am not comfortable with video. I need text.
I understand statistics cannot discover any cause per se, which is part of why I did not understand what DAV meant.
DAV: The normal use? Okay, I’ll go with it but I still cannot understand the difference. Your further explanation helps some, though you are starting with an hypothesis that one variable is causal, you just don’t know which one. If you are not testing an hypothesis that one of the variable causes the phenomena, what are you doing? The end result is not verification of the hypothesis or the null hypothesis, but I still see it as confirming your original hypothesis that one variable is causal. However, I will for the time being just study the idea and see if I can eventually understand the difference. (A causal network makes some sense–I suppose in that case you are technically not decreeing cause versus null, but you’re still building on cause combined with uncertainty. I’ll think on it more.)
I still cannot understand the difference.
The difference is that “Hypothesis Testing” as normally defined doesn’t actually test the hypothesis.
The end result is not verification of the hypothesis or the null hypothesis, but I still see it as confirming your original hypothesis that one variable is causal.
A rather simple examples is growing legs of its own. 🙂 What I listed were the conditions that must be true if C caused A and B. It wasn’t meant as an hypothesis but how a what if question might be answered. You originally asked how one goes about determining causality with statistics but without using the very specifically defined term Hypothesis Testing, or so I thought.
With three variables, there are 12 ways to link them. It’s quite possible to determine the most likely linking configuration. It is NOT confirming an original hypothesis. It is instead asking, what if the variables were linked this particular way. Normally, one doesn’t just stop with just one configuration and would examine all of them picking the most probable. This, however, becomes quite difficult as the number of combinations grows with N, namely, 2*N*(N-1). With many data rows, this can take a very long time or simply be completely impractical. So some heuristics are almost always applied as well.
The algorithms I talked about earlier do this.
“Why probability models cannot determine causal relationships…”
What method determines causal relationships?
Stove argues that induction is evidence for causal relationships. Dr Brigg’s claims there is no “problem of induction”. Why then turn around and state the opposite?
“Hypothesis testing is based on a false dichotomy. Also, statistics per se cannot discover any cause.”
It isn’t. Although it would be correct to argue that hypothesis testing is routinely misused such that a false dichotomy arises. That is to say, the hypothesis test can only falsify your hypothesis, it does not imply that, therefore, the opposite of the hypothesis must therefore be true. It usually is not even clear what the opposite of the hypothesis is.
Final comment for now:
Does smoking cause cancer or not?
Because if you say probability models cannot be used to establish that fact, then you’re arguing that there doesn’t exist evidence that smoking causes cancer. Surely not?
But what I suspect is going on here is a subtle shift of meaning. Here are two sentences that sound similar but are radically different claims in relation to their epistemology.
Probability models prove that smoking causes cancer.
Probability models furnish evidence that smoking causes cancer, beyond reasonable doubt.
It is a false dichotomy, which I’ve elsewhere proved (and in my book, too). And you’re quite, quite wrong that a hypothesis test can falsify anything.
Probability models are incapable of proving cause. Cause is a different kind of knowledge.
But I won’t argue with you, because that is futile, no? If you want proofs, see my paper—or buy the book! (When it will come out, who knows.)
Yes I read what you wrote, but I found it unconvincing and your rebuttal was to refer me back to the argument you originally presented. Which obviously doesn’t help. It would be more constructive to show me why I am mistaken in my argument. Sometimes theses differences arise not due to the logic of the argument but its wording.
Your last sentence does not address what I was pointing out but steam rolled over it. A hypothesis does not prove or falsify anything. It’s evidence for, or against, something. The argument you present is misleading because there is no method to prove cause or even disprove cause. You only have more or less, good or bad, evidence. Now if you’re phrasing things that way as a rhetorical argument to get people interested in what else you have to say, that seems to me to be a perfectly legitimate form of intellectual salesmanship. But if you think it points to something logically deeper, I and most others who are philosophically minded, would have an issue with that, or at least, would like to better understand what you’re getting at.
It would certainly be more constructive, sure. But it would take too long in comments boxes, especially since keeping on track isn’t so easy. I’ll simply point out that if you think a hypothesis test can falsify anything, your view of probability is fundamentally flawed. And to correct that kind of error would take more than a sentence or two in a combox.
The point I’m making is *no method* proves anything and *no method* disproves anything. So isn’t what you’re writing a tautology? But if you can’t clarify that point in a paragraph or two, then leave it for now.
So you’re admitting hypothesis tests cannot falsify anything? And that, it follows, they cannot prove anything? One step at a time.
(I’ll be signing off for today, incidentally.)
Will: Smoking can only contribute to cancer. If it caused it, virtually all people who smoke would get cancer. Viruses cause measles and virtually everyone exposed gets measles. No one gets measles without the virus being present, but people get cancer all the time without smoking. There are multiple variables at play.
@ Dr Briggs,
Again ambiguity of language at play that I fear you’re failing to clarify hence why I’m seeking clarification. Strictly speaking the only propositions that can be deduced as true or false are deductive in nature, hence tautological. Mathematical proofs take this form, few other things in life do. I can’t prove that the sun will rise tomorrow, but I can give you very strong evidence for why that is extremely likely. On the other hand, the Earth might get eaten up by a wandering black hole tomorrow in which case there will not be a sun rise. The probability of that happening is low, but not zero.
The problem is that there is a world of difference between the claim that I have proof of X versus I have (strong) evidence of X, although to the non philosopher they may sound like you are making the same claim.
No, smoking can cause cancer. Not merely ‘contribute’ unless by ‘contribute’ you mean increase the probability. But other things can cause cancer too. Hence if you have lung cancer and you smoke, that doesn’t mean that the smoking caused cancer in your individual case. You can only look at the probability of those who get cancer who don’t smoke versus those who do, and say the probability in your case was X relative to some population. Which I think is meant you meant to write.
Will: If something causes a disease, it must do so in all cases. Otherwise, it can contribute but is not the cause. For something to be causal, if A causes B, then B must occur every time A does. It requires 100% correlation. No one gets measles in the absence of the measles virus.
I think that a general proof that statistics does not pertain to (efficient) cause would begin by acknowledging that mathematics as such abstracts from motion and considers only formal and material cause and then only in the limited way of considering magnitude.
The paradigmatic example of an equation of causation would be F=ma. However, the equation as such, that is mathematics as such, cannot tell us which causes which or even if there is a cause in there. It is from outside of the mathematics that we find the force to be the cause of the acceleration of a mass. dc
The sun will rise tomorrow unless something prevents it. The fact that you postulate a black hole (or something else) implies that you recognize this sort of necessity. The sun will not rise for no reason whatsoever, something must cause it not to.
Some things will never change; here’s one:
“When we wish to correct with advantage, and to show another that he errs, we must notice from what side he views the matter, for on that side it is usually true, and admit that truth to him, but reveal to him the side on which it is false. He is satisfied with that, for he sees that he was not mistaken, and that he only failed to see all sides. Now, no one is offended at not seeing everything; but one does not like to be mistaken, and that perhaps arises from the fact that man naturally cannot see everything, and that naturally he cannot err in the side he looks at, since the perceptions of our senses are always true.”
Blaise Pascal, Pensées (1670)
“Hoc opus, hic labor”
(this is the hard work, this is the toil)
Dale Carnegie said the same thing more succinctly.
And in another 300 years someone else will again.
Meantime, scores of people & would-be consultants will fail to grasp this basic fact of human nature, and prosper accordingly….
There are many very successful people who have no problem telling others they are wrong–see Larry Winget, Ann Coulter or any liberal politician. At some point being “nice” may not work anymore if people cannot grasp that another view exists. As in global warming or Obamacare. No amount of alternative theories or niceness works when the audience simply does not care about the truth.
Briggs, would you want to join the brouhaha over at WUWT? I think your two cents would add greatly to the discussion over there.
Chuck L: That’s an interesting brouhaha.
“Will: If something causes a disease, it must do so in all cases.”
I’m not sure where you are getting such strange ideas from. Cancers are caused by random mutations so there is a probabilistic element to consider. Tobacco smoke raises the probability, genetic factors raise the probability, and so on. You can’t say that one risk factor caused the disease when another risk factor did not.
“The sun will rise tomorrow unless something prevents it… The sun will not rise for no reason whatsoever, something must cause it not to.”
Correct, which is why I’m not a big fan of Cause & Effect as explanatory concepts in science, but we must be pragmatic and work with what we have. We can hardly get rid of them until we understand the natural world a lot better.
Although I will make the observation that when I asked a fairly routine question, i.e., does Dr Brigg’s believe smoking causes cancer? The response from Dr Brigg’s was mumble mumble. Not an inspiring response. I am not being contrarian for the sake of being contrarian. Dr Brigg’s is welcome to being the biggest contrarian in the room. I’m just trying to understand Dr Brigg’s position at this point, and not being able to answer such a straightforward question without directing me to read his entire book, is a little worrying. He is certainly going to have an issue on the chat show circuit.
Will: I think I’m getting those “strange ideas” from reality. I don’t have any idea where you picked up yours. Causality requires 100% correlation. If it’s not 100%, then it’s not causality. (Only about 15% of smokers get lung cancer—85% of more do not. This was the highest percentage I could find for how many smokers get lung cancer. This is not causality.)
Cancers are said to be caused by “random” mutations as far as science currently understands the phenomena. I did not say one risk factor causes the disease and other does not. We do not know the causes of cancer because no factor or group of factors causes it in every instance the factor exists. Without 100% correlation, there is no cause. Probability may help predict who will get cancer, but people get cancer all the time with no know risk factors, so probability is pretty much useless as far as predictions of who will and will not get cancer (women who don’t smoke are increasing being diagnosed with lung cancer). If it can’t predict and it can’t show cause, who cares about the probability thing. The only thing risk factors may do is result in less aggressive treatment for some cancers, the assumption being removing the risk factor will help prevent recurrence (which is scary if you think about it).
Sheri: There are a quite a few liberals at University of Wyoming, not all of whom are stupid. Wyoming is actually quite pragmatic. We are about ten years behind the world, and are in a good position to see how new ideas go wrong before we implement them. As far as the dopey liberals go, well, folks unemployable at any useful job tend to hang around a university.
The question of whether Briggs could get a gig talking about his third way here in Laramie I can answer in the negative. First, if Briggs is on to something new, Wyoming is about ten years behind and will not take the risk of inviting him here and finding itself on a frontier. Then we have a very traditional Statistics Department, separate from the Mathematics Department, who will disagree with any paid visit. Finally, with declines in oil, coal, and gas sales; we are actually getting to the point of belt-tightening.
Sheri: Shoot. I meant to speak up about your thoughts on causality, and in discussing Laramie and Briggs’ odds of getting an invitation here, I forgot what I was doing.
You are describing a part of what I learned as Koch’s postulates; did you mention this? They form a very rigorous framework for proving causation. Statistical efforts pale in comparison. Even the dose-response curve which DAV alludes to, is weak in comparison.
K.Kilty: I didn’t say the liberals were stupid–I said it was some form of mental illness to deride oil and gas then take paychecks from their taxes. You may interpret that as you wish.
Agreed that Wyoming is at least 10 years behind the times. I now refer to Wyoming as “the place technology goes to die”. I’ve been here through two other bust cycles, so belt tightening is all too familiar.
I had never heard the ideas referred to as Koch’s postulates, but I looked it up. That pretty much is what I’m saying.
Only about 15% of smokers get lung cancer—85% of more do not.
Hard to find, yes. It’s closer to 1:10000 while the probability of not getting lung cancer regardless of your smoking history is around 98.98%. This info used to be on the CDC site but has disappeared. I have only personally known only one person with lung cancer and even then it was only suspected. It’s a relatively rare disease. I suppose though if I were in a family prone to these things then I would likely know more than one.
An interesting quote regarding Koch’s ideas: So eventually, Koch’s requirements had to be discarded in some avenues of medicine such as carcinogenesis; the bar was simply too high, and the need for more cancer science far too great. Under the title a scientific definition of causation.
So since Koch’s criteria were too hard, cancer “science” picks an easier definition for declaring causation. Maybe that’s why the title is in lower case — to underline the irony.
Typo in a previous post. Should have been: the probability of not getting lung cancer regardless of your smoking history is around 99.98%.
DAV: Very interesting quote indeed. If the bar is too high, just disregard the bar. The idea is similar to global warming believers who complain the bar is just too high for the usual degree of prediction, so we should disregard the bar. What we really are doing is creating possibly false hopes and, bare minimum, very poorly done science with limited value. It falls under the “it’s too important to wait for real scientific results”.
I have always considered much of science to be playing fast and free with truth. This leads to many erroneous treatments and causes being announced. The usual example of this is ulcers, but there was the estrogen treatment of menopausal in women that was supposed prevent heart attacks. It was later linked to cancers increasing. (Maybe we shouldn’t treat naturally occurring aspects of aging?)
My dad had lung cancer and died. He did smoke, (though not in the house) and spent a lot of time around weed sprays. His half-sister died of breast cancer. I did know one other person who had a lung transplant but I was not sure if that was due to lung cancer. It’s kind of a matter of chance. I also knew people with pancreatic cancer that survived for years. Cancer is so multifaceted claiming “causes” is very similar to claiming that global warming is absolutely caused by humans. No one really knows.
I’m not entirely sure how your words fit together. However, I must, out of intellectual curiosity, engage in the social science of what-did-Will-mean and ask what was the cause behind the effect of your words.
I suppose that it does make sense that after abandoning formal and final cause, one would be lead to give up on efficient cause. Out of sincere curiosity (though it is not a cause of what I am to write), what do you imagine the future of science to be that does not destroy science or sneak causality in.
Please re-read what I wrote. Obviously you failed to understand it.
I’m happy to answer straightforward question asked in good faith, but have no time or patience for the not very clever ramblings of a man desperately trying to appear clever. Rambling questions suggest a rambling mind. A clear and precise question suggests a clear and precise mind.
“First, if Briggs is on to something new, Wyoming is about ten years behind and will not take the risk of inviting him here and finding itself on a frontier. ”
Unless Dr Brigg’s provides evidence to the contrary I expect he is onto something very old, not something very new. The project here, in roundabout ways, may be to remind academics about uncertainty. Some of the things Dr Brigg’s laments were hammered into me on my very first day of my university science lectures. So when he writes that academics believe this or claim that, well that was not my experience at all. I was warned about these sorts of mistakes from the beginning. But that was many decades ago, and most scientists are not competent epistemologists, so what he is doing strikes me as important and valuable; but not new. (I also read incredibly dumb claims especially by those in the global warming movement who are academics and I wonder if they skipped all those classes, perhaps, or they are no longer taught.) Academics seem to go through fads where they forget and remember what certainty means. Marxists once claimed they were scientists, psychoanalysis also, and eugenics, and many other subjects we now know to be ‘completely uncertain’ after all.
Please re-read what I wrote. Obviously you failed to understand it.
I’m happy to answer straightforward question asked in good faith, but have no time or patience for the not very clever ramblings of a man desperately trying to appear clever. Rambling questions suggest a rambling mind. A clear and precise question suggests a clear and precise mind.
Your answer to Semiotic Animal serves as my answer to you. Substitute answer for question.
As for U of Wy, they are completely satisfied to teach that 1800’s wind mills are appropriate in 21st century life, adhere to superstitious teachings about global warming reminiscent of the early witch doctors, so if you are correct, they should love Briggs going back in time. (Those of us who went to college before the 80’s may have been taught that uncertainty existed–that’s why it’s often older people writing skeptic blogs. They were taught to think, not just grind out meaning repetitive information.)
Pardon me for taking your seriously. You are quite require right that I lack the necessary wit to even appear clever and I am a rambling man.
Let me ask the question as precisely as I can: if cause and effect are only contingent concepts for intelligibility, what would such a science devoid of cause and effect look like? If possible, please make no reference to any sort of causation in your response. If not possible, why should I suppose that such is even possible generally for science? It reminds me of the man (I forget his name) who insisted that intentionality was illusory, but could not and refused to provide an account of his position that did not presuppose intentionality.
Given that my statement about the sun makes reference to the need for causation, in what sense do you say “Correct” then go on to say that you dislike cause and effect? Were the sun not to rise tomorrow, the question the scientist and indeed anyone would be “Why did that happen?” or more explicitly “What caused that?”
Sciences that don’t incorporate cause and effect are all our most advanced sciences, such as relativity, QM, standard model, etc. Even Newton’s Mechanics dispensed with Cause & Effect centuries ago. In our best theories the laws of nature are described in terms of mathematical relations, not in terms of X (time interval), then Y.
Since for example the WHO estimates 50% of those who contract Ebola die from the disease and you insist that something that causes something must always cause something, you clearly don’t understand what I wrote. But it’s not my task to try to defend you when write nonsense or drivel, so let’s leave it that you don’t believe Ebola is dangerous and that everyone else thinks it is.
You will need to elaborate a bit more. What you say about physics is manifestly false and requires more than a list of putative examples to defend and an off hand comment about using mathematics to represent causal relations.
Your example, to put it bluntly, is childish, somewhat irrational, and not helping your case at all. I said there is a virus that causes Ebola in people and no one gets Ebola that did not contract the virus. That is the causality. The causality of death in Ebola cases is complex. Not all events have a single cause. I have never claimed they do. The causes of death would more than likely be a combination which would cause death in 100% of the cases in which it occurs. Since there is probably a genetic component, overall heath, diet, medical care, etc, involved mapping the causal factors would be incredibly complex. If we had the ability to track that many factors, it might be possible to ascertain the cause in each case. Then, if we have correctly identified the cause, it will occur in each individual where all of the factors are present. Considering the incredible amount of combinations of factors, it’s unlikely we will be able to identify all of the factors any time soon. Researchers study questions like this all the time, seeking what combination of factors cause a certain outcome. Said researchers often decide it’s too hard to identify all factors so they just spew out probabilities and pretend they know cause.
Your comment that I somehow believe that Ebola is not dangerous is frankly incrediblity idiotic. It’s a definate attempt to avoid the actual subject and very closely resembles a straw man, since I never made any such claim. However, if you’re out of ways of refuting my claims, that probably seems like a great cover to hide behind.
A reminder of why I called what you wrote “drivel” –
“Causality requires 100% correlation. If it’s not 100%, then it’s not causality. (Only about 15% of smokers get lung cancer—85% of more do not. This was the highest percentage I could find for how many smokers get lung cancer. This is not causality.)”
Hence you’ve argued that smoking is not a cause of cancer. Which is not the case, which is why your statements were ignorant. Now if you want to qualify yourself to such an extent that you’ve essentially retracted all of the above, that’s fine, and smart. But I’d suggest you can’t complain and name call when what you originally wrote was such utter nonsense.
Not really my role to become your private tutor in terms of bringing you up to speed on 500 years of the study of the history and philosophy of science. (Probably the grandiose pomposity isn’t called for then, if you can’t understand the point I made.) As good a place as any for someone in your situation would be to start here:
It’s a little out of date but because Bertrand Russell is such a clear and wonderful writer, we can look past that. And anyway, it’s not as if this strand of philosophy is even something you were aware of until now.
You proposed to answer a question sincerely asked. You responded by giving a list of supposed examples rather than answer the question. You proposed that by writing equations we somehow escape causation without explanation. You then point me to an article (I will point to your criticism of Briggs’ doing the same).
What you have not done is answer the question, namely, what you envision a science without causation to be. You have provided no argument or explanation, but only question begging examples. You have provided no argument why these are examples as you envision but rather throw out a red herring as if my not being knowledgeable of your philosophy is reason for you not to answer or provide an argument. Indeed, that is exactly why I asked about in the first place. What you say is manifestly false, but let an argument actually be proposed if it is true.
I give up. This is like arguing with Sylvain, who doesn’t read what’s written and never admits any error. I’m putting you in the “do not address category” with your twin. Irrational arguments should not be addressed, as you yourself said about Sylvain. (Oh, the happy monkey dance to celebrate the fact that you “won” is fine. I’m sure Sylvain does one every time people give up trying to interact with someone who cannot or will not listen or learn.)
I fully admit that at the point I hit name calling, I’ve ran all out of my patience and willingness to understand that which does not seem to be rational or reasoned. When I pull out name calling, it’s an indication there’s no point to trying to be nice or informative. While Bob just ignores people, I tend to make an attempt to try and understand. I tried that here and name calling is pretty much where we are—I cannot argue with someone whose logic and arguments make no sense at all. I am sure you put me in the same catagory, which IS logical, because if I think you are impossible, it follows that you may believe the same. There, you hit one logical conclusion in all of this.
So you are so lazy that you cannot read the short article I pointed you to which succinctly addresses the entire issue you asked about? You would prefer me to re-word what is clearly written and explained there? And you’re complaining that I won’t this for your personal benefit? Let’s add egomania to your list of unpleasant personality traits.
You gave up because you wrote drivel and can’t defend it. Hence the ad hominem attack. Here is a simple thought experiment for you to consider. Imagine I am evil dictator (possibly somewhere in Syria) and I inject people with various dangerous diseases for my amusement. I give you a choice of three injects. Disease (A) has a 90% probability of death. Disease (B) a 40% probability of death and Disease (C) a 10% probability of death. My statistics are based on the last 1000 people I injected.
Here is my question: I am prepared to give you this information if you ask for it. Will you ask for it? That is to say, do you believe what you wrote that probability does not demonstrate cause, and it’s impossible to determine in the case of an individual anyway. Since you believe in what wrote, you would not be interested in my data, correct?
You and Sheri are talking past each other. Your definition of cause is a pragmatic one and one I often use, X sometimes causes Y and don’t really care why it doesn’t at time. Sheri’s point is that X should cause Y all the time except when inhibited as in, firing bullets at a person causes injury unless you miss. The cause chain would be G & (not Miss) causes Injury.
For smoking I think it quite apparent that Smoking & Something Else causes Lung Cancer. Since the incident rate is low, Something Else must be missing a lot. If the presence of both is necessary it’s quite acceptable to say smoking is a contributor and only so if that Something Else is present. This does not preclude some additional, mysterious factor X that is the cause of lung cancer in those who don’t smoke. The causal chain would then be (S & SE) → (C) ← (X)
You are of course perfectly correct. The definition I use is the one commonly used, especially by scientists. It may have various philosophical meanings but it doesn’t have a single agreed one. When Sheri declares Cause is this and not something else, she has to argue that position. Which is a very difficult thing to do, as I agree with Bertrand Russell that its such an amorphous term its gone past its use by date in philosophy.
While I would still like to know the cause of Will becoming Sylvain, DAV is correct. A phenomena can be caused by 1 factor, 2 factors, or 20 factors, not all of which are equal. Which is precisely what I did argue. That seems to be a defense if DAV types it but not if I do–wonder what causes that?
Will: As for Bertrand Russel and his claim that cause is an amorphous term, the RICO people love him because they got a huge pile of cash for the Treasury by claiming cigarettes CAUSE cancer. Now, the American Cancer Society clearly states that cigarettes are one known risk factor. ANYBODY can get lung cancer, and that the majority who smoke never do get lung cancer. Not having an agreed upon cause but rather a “pragmatic cause” is wonderful for suing rich corporations and shaming people into desired behaviors (why is that no one says people caused their own AIDS infections? That’s straightline, yet I don’t hear that one). It also lets the EPA outlaw thousands of useful chemicals because they CAUSE cancer the same way as cigarettes do. Yet this is what is considered “pragmatic” for “cause”? As noted, your government and tacky scientists love you. Really, really love you. What you are advocating is actually bad science. Russel’s philosophy really does not belong in the study of illness/cause. Diseases are not philosophical. I also wasn’t aware that philosophy trumped evidence and methodical experimentation.
Also, CO2 does cause global warming using this definition.
You wrote, “…As for U of Wy, they are completely satisfied to teach that 1800’s wind mills are appropriate in 21st century life…” This is not so. There are a few people here who think we should eventually be content with what windmills can provide, but they are a distinct minority in at least two colleges of the university. We are not so far behind the times, or ahead, that we deny thermodynamics and economics, for instance.
By the way, being behind the times provides some advantages. If you have ever read Roger’s book on the diffusion of innovation, you will see much evidence for the cutting edge innovators to go broke and extinct. Laggards in the second wave of an innovation tend to do much better.
K.Kilty: I stand corrected. I was not aware that is was a minority that supported wind turbines. The media makes it appear U of W is all in on the wind and solar. My apologies and I thank you for letting me know.
I agree there are advantages to being on the second wave of innovation. Perhaps my objection is more that Wyoming seems to lack a great deal of things such as high-speed internet and for a long time there was very limited cell phone coverage. There were two television stations here in 1982 covering three networks. (Obviously, this has improved.) Such services are taken for granted elsewhere. Tourists remark about the absence of services and the desolate drives through the state. This is in part because of low population numbers and I do like the sparse population. Just wish my internet wasn’t so slow and spotty and my cell phone worked where I recreate.