uncategorized

What is a "Scientific" category?

Friday, June 2nd, 2017

Dalton Conley Speaks Out on Race in an interview at Nautilus (which I saw via 3QD). He says a lot that is sensible but his denial that "race" can be a "scientific category" is based more on wishful thinking than on reason. Most scientists think of science as "whatever works" and so would entertain any category that can be used to make consistently better than chance predictions as "scientific". And even the crudest out-dated concepts of "race" do have some predictive power. Much of what was claimed as such was probably false but some is undoubtedly real.

Conley's reference to "the chopstick problem" is a case in point.

If I desperately needed to quickly find ten people who were highly dextrous with chopsticks, and whom I was prepared to reward highly for that, and had a room of a thousand eager applicants to consider, then it might make sense to test just those with epithelial folds in their eyelids and ignore the rest. Yes, I might miss the best one in the room, but depending on the skill level needed and the urgency of that need speed might be more important than optimization. Many employers, educators and law enforcers *think* they are in an analogous situation. Often they are wrong but sometimes they may be right. The question of how to either avoid or somehow compensate for such unfair exclusions is not resolved by declaring that there is no "scientific" connection between epithelial folds and chopstick skill.

It may be morally wrong to make use of (or even talk about) some of these things, but that moral position is not advanced by misidentifying moral error as scientific error. Indeed such misidentification both undermines public scientific literacy and discredits the moral position that it was intended to advance.

Should we outsource our moral beliefs?

Thursday, June 1st, 2017

In this article at 3quarksdaily, Grace Boey almost changes my mind. I have long felt that the delegation of moral authority that religion almost always leads to is harmful, but the way Boey puts things makes me feel I need to be clearer about what I mean by that.

Mythical Myths #27 - The Galileo Trial

Monday, May 29th, 2017

In 3quarksdaily: The Galileo Trial: Faux News from the 17th Century, Leanne Ogasawara refers to several misconceptions about Galileo but her main point is suspect. She refers to a number of resources on the issue, some of which I have not yet read, but the most interesting to me is the reaction of (ex)Vatican Observatory Director George Coyne to his experience on the church's Galileo Commission in the 1980s. Coyne was clearly frustrated by the abuse of process by which the Galileo Commission (of which he was a member) was purported to have identified the failings of the Church (and esp of the popes of the time) as a "myth", so unless some of the other links convince me otherwise I will have to classify Ogasawara's "Faux News" as yet another Mythical Myth.

Many Paths to Equity

Monday, May 29th, 2017

Prompted by the impact of increased labour costs on American Airlines' stock valuation, Matt Breunig writes in Jacobin (coming to my attention via 3QD):

If Naidu and others are right, Piketty’s theory of how wealth and income inequality develop may be exactly backwards. And his prescriptions for reversing skyrocketing inequality may suffer accordingly.

 

But I don't see any "getting it backwards" here. Piketty's observation is just that when, for whatever reason, the rate of return on capital exceeds the rate of growth of the economy, ownership will concentrate in fewer and fewer long term holders of capital. Perhaps the reverse is also true - ie that preventing the concentration of capital ownership brings down the rate of return, but none of the critics seem to be making this claim. The difference is rather about how to prevent the concentration - by taxation and/or redistribution of assets or by raising the cost of labour (which then brings down the rate of return and/or causes continued demand for a high rate of return to lower the face value of the assets). But that is a false dichotomy. We can do both!

Mythical Myths #74 - Tacoma Narrows Resonance

Friday, May 26th, 2017

According to a story by astrophysicist Ethan Siegel in this week's Forbes magazine: Science Busts The Biggest Myth Ever About Why Bridges Collapse . Wow! That headline reads as if "science" has just made a new myth-busting discovery and the article tells us that the previous understanding of what happened was based on the kind of resonance by which a singer can break a wineglass and is all wrong. But the "myth" is pretty much itself a myth here as the Federal Works Admin report way back in 1940 stated that "It is very improbable that the resonance with alternating vortices plays an important role in the oscillations of suspension bridges. First, it was found that there is no sharp correlation between wind velocity and oscillation frequency such as is required in case of resonance with vortices whose frequency depends on the wind velocity." And with regard to whether or not the actual nonlinear effect should be described as a resonance I agree with the comments by Gregg Collins two years ago and Jeff Hester (also an astrophysicist by the way) one year ago in response to munutephysics' video on this from 2011. Both of these commenters point out that despite not being a case of simple response to a periodic applied force "it is far ~more~ incorrect to say it was not a resonance at all". What really happened is that the bridge had an inadequately damped torsional vibration mode which was excited not by a periodic change in the environment but by a nonlinear interaction with steady wind which pushed harder on the twist when the twist itself was greater (much as a child on a swing pumps herself up by reacting to her position rather than moving at a predetermined frequency).

Misrepresentation of Genetic Science?

Monday, May 22nd, 2017

Razib Khan at Gene Expression: The misrepresentation of genetic science in the Vox piece on race and IQ  objects to this Vox piece about Charles Murray.

I have a number of concerns about the Vox piece but curiously the main claim that Razib objects to is not one of them. I am not close enough to the field to know whether or not it is *now* true that "no self-respecting statistical geneticist would undertake a study based *only* on self-identified racial category as a proxy for genetic ancestry measured from DNA", but what Razib presents is not a counterexample. Rather it is a reference and quotes from Neil Risch in the early 2000's which looked (to me) more like an analysis of how well racial/ethnic self-categorization *might* be (or might have been) used as a proxy for genetic structure [based on microsatellite markers] (back when the cost of the latter was prohibitive) rather than an actual use case. And, as Razib himself admits, "This isn’t 2005".  So, for what little it's worth, I do find it hard to imagine how a "statistical geneticist" (as opposed to a sociologist for example) might need to use only self-identification of subjects rather than actual DNA analysis as input into a study at this time.

Back in the 1970's (20 years *before* 'The Bell Curve'), I joined demonstrations against seminar speakers promoting the work of Shockley and Herrnstein because they treated as gospel (and advocated social policy supposedly based on) the conclusions of twin studies whose data had substantial unexplained patterns of what looked like fraudulent manipulation.

I did not necessarily believe that the conclusions were false though, and had some academic interest in whether or not they might be - but felt that that investigation would be premature without some prior thought about how to deal with the impact of whatever turned out to be the case. My general stance in favour of freedom to investigate does accept limits based on the possibility of knowledge causing social harm  (so I don't really want to know whether or not some "races" are intellectually inferior to others) - as well as on some rights of personal privacy (so that I also hold in check any curiosity I may have about my neighbours' sex lives for example).

Ever since then I have been disappointed by the left's strategy of complete denial rather than addressing the hypothetical question of how best to deal with the possibility of eventually being faced with an "inconvenient truth".

What would be the best social policies for dealing with a situation in which some readily observable characteristic turned out to be correlated with something real and valuable but less easily measurable? (Say if people with red hair had a measurably higher average rate of dishonesty for example)

Perhaps it's not the mean but the variance!

Saturday, April 8th, 2017

Why are only two of the world’s top 100 chess players women?
There may be many factors at play here - including social expectations etc, as well as possible gender differences in average measures of either ability or interest. But it's also possible that a large part of the explanation is just a matter of variance (ie of spread of the distribution of ability or attitude).
As Larry Summers was criticized for pointing out, having one group dominate the top tier is not necessarily evidence that either they are better "on average" or there must be some kind of unfair selection process. The group with higher average of some measure may be under-represented at both extremes if it is more tightly concentrated. And the fact that variations on the y-chromosome may be less likely to be balanced by a partner could be a source of greater variability of some characteristics in males.
Of course, even if this is true, the apparent excess of males at the top may be more visible than the corresponding excess at the bottom, and that may both give mediocre males false confidence and discourage some females from even trying. So it is always worthwhile to counter these effects.

Source: Why are only two of the world’s top 100 chess players women? | Aeon Essays

PostModernism

Saturday, April 8th, 2017

Stephen Downes comments on critique of post-modernism with reference to whether or not it is a "fact" that tennis balls don't fit into wine bottles.

Despite the failure of a particular bit of language to unambiguously cover all aspects of a situation, It seems likely to me that there are nonetheless real facts about tennis balls and wine bottles. The "proof" of this is in the fact that we would be at least as surprised (though perhaps less entertained) to see a tennis ball in a bottle as a model ship - and so would immediately look for the illusion or trick that needs to be excluded in order to "save" the statement. (This process is a large part of what Imre Lakatos is on about in his book 'Proofs and Refutations' which deals with the same phenomenon in Mathematics.) Yes, language is an inadequate tool for expressing the full content of human understanding, and human understanding may be an inadequate tool for capturing the true nature of reality, but I think it is a mistake to infer from those sad facts that no such reality exists.

Why DT tweets nonsense in the middle of the night

Monday, April 3rd, 2017

I like to think of Donald Trump as just obsessively glued to his computer in the middle of the night much like the fat boy he speculated was responsible for the DNC hacking (though more devoted to taking offense and picking fights than to picking digital locks). But that's not entirely right.

The method in his madness (as described in Bradley Eversley's answer (on Quora) to Why do you think Trump always sends disturbing or controversial tweets in the middle of the night? ) is to always find something to grab attention at the beginning of the news day (and the more outrageous the better for this purpose)  as a distraction from whatever more really serious allegations are being made against him or damage he is about to do.

 
Update: And here's a related quote from an article by Stephanie Hayes in the Atlantic:

As a Time interviewer aptly summarized during a recent chat with the president: “Whatever the reality of what you are describing, the fact that [the facts] are disputed makes them a more effective message, that you are able to spread the message further, that more people get excited about it, that it gets on TV.”

Why Were We All Surprised?

Sunday, April 2nd, 2017

Source: Statistician Nate Silver says conventional wisdom, not data, killed 2016 election forecasts

Actually, I suspect that many people were confusing chances of winning with expected share of the vote. Perhaps not consciously, but not surprising as we are so often presented with percentages as representing the latter. So seeing the number 70 associated with Hillary naturally created a false sense of security in her supporters.

Flying "Cars" were silly - until now

Thursday, February 9th, 2017

The reason we don't yet have the flying cars imagined by early 20th century "futurists" is not because of mechanical infeasibility (though the energy cost may be substantial) so much as because of the problem of traffic control. Accident avoidance in crowded airspace is beyond the capacity of either expert pilots or centralized traffic controllers, and any vision of replacing cars with flying vehicles had to fail if it depended on human controllers. And it is only now that automated control based on full 3d "vision" in all directions  is becoming feasible. But I don't agree with Uber's new hire that "There will be an evolution from professional human pilots to autonomy over time" because I don't think that human piloted flying cars should ever be allowed off the ground. Source: Uber brings in NASA engineer to build flying cars

Two Challenges

Saturday, February 4th, 2017

Two very different recent events raise the same issue. How to confront the temptation to abandon one's principles when one's opponents do so without penalty. And in both cases I think The Atlantic has presented a view worth noting.

One is the attack on Al Quaida in Yemen which ended with the deaths of several civilians and a US soldier. Many on the left are crowing that this was authorized by Trump without proper analysis, and I agree that the sources of his advice were unbalanced in that they failed to combine military analysis with any evident input about the long term radicalization effect of a possible disaster (such as what actually happened).

But in Trump's defense, former Obama defense advisor Andrew Exum writes in  Don't Politicize the Failed Yemen Raid:

For the recommendation to have gone forward to the president, the senior leadership of the Department of Defense would have signed off on this operation. And for that to have happened, special operations and regional U.S. commanders would have had to have blessed the planning that went into the operation itself.

The left cannot on the one hand claim Donald Trump is ignorant of military and security affairs, and then on the other hand expect him to second-guess the professional recommendations of his uniformed and civilian military leadership.

I don't agree with Exum that non-military input is properly left out of decisions like this - though perhaps not to the extent of items as trivial as "debating whether or not it made sense to move three helicopters" (and the role of Benghazi nonsense in prompting that level of overdue diligence is worth noting). The temptation to exploit this event in the same dishonest way that the right did with Benghazi is clearly irresistible to some on the left - especially since that sleazy tactic was so spectacularly successful. But it really should be avoided. It's hard to take the high road when the low one is seen to work, but in the end winning without integrity is just another way of losing.

 

The other item is Peter Beinart's piece on how Milo Yiannopoulos Tested Progressives—and They Failed .

...when Trump’s opponents use the danger he and his supporters pose to restrict basic freedoms, there’s a problem.

Which is what happened earlier this week at the University of California, Berkeley, when a violent protest prevented Milo Yiannopoulos, a Breitbart News writer who has made his name by viciously mocking women, trans people, and African Americans, from speaking on campus.

With or without violence, the de-platforming was probably misguided although the protest itself was not. Let us rather all hear the trash with which the young Republicans want to be identified - and not provide any excuses for them to play the martyr or for those who wish to de-platform other views that we may share but which they declare offensive.

Update(Feb3):But . like Robert Reich, I wouldn't "bet against" a suggestion that the violent protesters who closed down the Yiannopoulos event were right wing provocateurs rather than real left wing opponents.

Giant Pterosaurs 

Saturday, January 28th, 2017

 

How did they fly? (or did they always have strong winds or cliffs to jump off?)

And why do they have beak-like snouts like birds? (while bats do not)Source: Fossils of giant pterosaurs found in Transylvania

The Science is Clear?

Saturday, January 28th, 2017

Anyone distressed by the amount of distrust of real science in the world today need only look at the overstated (and actually false) headline in this bit of "science" journalism to find an explanation.

Of course torture doesn't always work. And it very likely would only rarely be the most effective way of getting information. But I can tell you for sure that it can sometimes work because I know that there are circumstances in which it would work on me.

What's in the article is mostly not false

Enhanced interrogation techniques often lead to unreliable information; instead, we might want to follow the lead of a noted "nice-guy" Nazi.

but it does not support the headline. And regardless of what the authors intended, that headline stands as the position of a self-styled "science" magazine.

Source: The Science is Clear, Torture Doesn't Work - D-brief

Robert Reich (Trump’s Plan to Neuter the White House Press Corps)

Sunday, January 15th, 2017

And he repeatedly lied. But the media in attendance weren’t allowed to follow up or to question him on his lies.For example, Trump wrongly stated that “the Democratic National Committee was totally open to be hacked. They did a very poor job. … And they tried to hack the Republican National Committee, and they were unable to break through.” Baloney.FBI Director James B. Comey said there was evidence that Republican National Committee computers were also targeted. The critical difference, according to Comey, was that none of the information obtained from the RNC was leaked. Also, according to Comey, the Russians “got far deeper and wider into the [DNC] than the RNC,” adding that “similar techniques were used in both cases.”

Source: Robert Reich (Trump’s Plan to Neuter the White House Press Corps)

Unfortunately this is a poor example. The DNC was hooked by a very simple phishing operation and whoever was responsible for advising Podesta did indeed do a "very poor job" (and so by implication did whoever hired him right up to Podesta and the DNC leadership themselves). Almost cetainly the RNC was "also targeted" by "similar techniques" but as Comey said, the Russians  “got far deeper and wider into the [DNC] than the RNC.” This is totally consistent with Trump's claim that "they tried to hack the Republican National Committee, and they were unable to break through.” That claim may be "Baloney", in that other techniques did get in to the Republicans and the discovered embarassing information was not released, but there is nothing in Reich's rebuttal to justify that counter-claim.

Does Networking Make Us Smart or Stupid?

Saturday, December 10th, 2016

Stephen Downes links to Amy Burvall on her daughter's reliance on her friend Ellie as a source of information.

I think the question in her antipenultimate paragraph: "Do we crave more temporal, less formal interactions, even if it means the information we receive is at risk for being less accurate?" should be answered with a resounding "Yes!" - which means that it is quite possible that the way we most naturally use networked knowledge is making us stupider rather than smarter. Of course other possibilities exist and the network also allows us to access central sources which are subject to corrective review from a wide variety of directions, so contra Nicholas Carr I am not worried about Google Making Us Stupid but neither do I share David Weinberger's confidence that Networked Knowledge Makes Us Smarter.

In any case, it's the long-range structure of the network that matters, and the evidence from Nov 8 and June 24 is that the network is disconnected to such an extent that those of us in the "smart" bubble were unaware of the extent of the "dumb" one.

Source: #rawthought: Your Ellie – On the Primacy of Networked Knowledge | AmusED via #downes

Should Schools Teach General Critical-Thinking Skills? 

Monday, December 5th, 2016

This article claims that "critical thinking" should not be taught as an independent discipline - largely because success in any particular area is allegedly more dependent on detailed knowledge of that area. But one of the main reasons for teaching "critical thinking" is to train people to evaluate the arguments and claimed expertise of others in areas where they do NOT have loads of experience or knowledge themselves. As such, it is perhaps THE most important thing a person can learn .... if indeed it can be taught.

But an important point made or implied in the article is that there is precious little positive evidence that those who claim to be teaching critical thinking are having any measurable success. This is certainly an appropriate application of critical thinking to the claims of those who claim to teach it. But it does not address the question of whether or not those skills should be taught if possible. A proper application of critical thinking to the article itself requires us to investigate whether or not the graduates of a "critical thinking" class are  or are not better able to identify a charlatan than those who took the "placebo" class. Being too lazy to do so myself I will leave this as an exercise for the reader.

Some Math Ed Stuff 

Monday, December 5th, 2016

Inverse Functions are certainly a minefield for students, and the situation is not helped by teachers' use of sloppy language to describe the concept and the prescription of a mindless ritual for answering assigned questions.

This article points in the right direction although it's not quite perfect in my opinion1. But what got my friend Bruce to comment was one of the authors taking the objection to explanation by procedural prescription into another area where it might be less apt - namely the concept of average value.

Note 1:
Expressions like "the inverse of y=f(x)" are problematic because the relation defined by y=f(x) is the same as that defined by x=f^-1(y) and does have inverse relation defined by y=f^-1(x). So, contrary to the article, it is in some sense correct to say that the "the inverse of y=f(x) is y=f^-1(x)", and the formal definition of functions as sets of ordered pairs does justify "switching x and y" if this is interpreted and explained properly.

Prepare to be inspired (fromBBC via Butterflies and Wheels)

Monday, December 5th, 2016

Connie, this is for you:

The BBC has an annual series called 100 Women. BBC 100 Women names 100 influential and inspirational women around the world every year. We create documentaries, features and interviews about their lives, giving more space for stories that put women at the centre.

Source: Prepare to be inspired - Butterflies and Wheels

Liberal Elite Give Thanks to Trump Voters 

Saturday, November 26th, 2016

By Rob CoxNEW YORK (Reuters Breakingviews) - This is the week when all Americans give thanks for the bounties bestowed on them by the Blessed Creator...

 

Source: Cox: Liberal elite owe gratitude to Trump voters | Reuters