Review: Nonsense on Stilts: How to Tell Science from Bunk

Nonsense on Stilts: How to Tell Science from Bunk by Massimo Pigliucci

My rating: 3 of 5 stars



(unedited and lightly proofed)

I found this book well worth reading, esp. the last 30 pages or so. I withheld two stars from my rating mostly for what didn't happen in Chapter 6. If you're short on time, here's a shortcut approach to getting the book's chief take-away:

Start with Chapter 11, § "Science as a Bayesian Algorithm" (page 275 in the edition I read from), and read from there through Chapter 12 and the Conclusion. That'll be about 31 pages. The essence of the book and its take-away for readers is right there. Glance at the footnotes for that portion too. After this, if you're wanting more depth, rewind to the beginning and start. Otherwise, you'll have the most important bit and will be primed to think about Goldman's Five Criteria when evaluating the judgements of experts you may be exposed to, no matter what the topic of interest. That will serve you quite well. This last part of the book is really the only part that to me affirmatively addressed the statement in the subtitle, "How to tell science from bunk." That there wasn't more to flesh out the premise of the subtitle perhaps was also a reason I withheld some stars.

As for the rest of the book, I felt there was somewhat less than I wanted on the philosophy of science and science-vs.-pseudoscience. The types of logical fallacy are important, and didn't get much more than an incidental treatment here. I need bolstering in this area myself, and didn't get it here. Types of logical fallacy seem to show up very often in arguments portrayed in media, so having a toolkit for this would help in making quick evaluation of what some "news" program's analyst or magazine article's author is trying to sell you.

The history of the development of science is interesting backstory, and worth including, but I think if other parts of this book were edited down to make room for it, I would have appreciated a different cut.



That's about it for the hard review. In the remaining space, I'm going to talk about a silent parade of elephants that sit quietly among you as you read Chapter 6, which is good, but could really be a whole book in itself if you wanted to address said elephants. As it is, the author pretty much confines himself to an orthodox narrative and leaves the elephants unaddressed (perhaps he wasn't even aware of them).

Chapter 6 addresses the relationship of science and politics by focusing on the global warming issue, which is too bad in my view, because this issue has too much baggage attached to it for many lay readers, myself included, and as a result it makes too narrow a survey of the relationships between them.

There is much more to say about the relationship of politics and science, and I won't touch on any of it except to suggest that this alliance of politics and science (where it gets the majority of its funding now, a relatively recent innovation for good or ill) is a key reason behind the anti-intellectualism addressed in the book. "Scientific American" magazine, for example, today could be regarded more as a public-policy advocacy magazine which derives its authority from the science it portrays, rather than a magazine devoted to the promotion of science, and dissemination of its discoveries. The politically focused parts of the magazine have become dominant over the pure science articles. The pure science features have also been cut down in length and scope to make room for the political, with more reliance on freelancing science writers over the researchers themselves, who used to write up their own features more often.

When the name of science is used to justify political ends, and those ends run counter to basic human liberties, or even simple expectations about the role of government in a polity, the resentment for the politic bleeds over into a mistrust of science too. Science then gets tossed out wholesale, at worst, as just additional government propaganda. The solution may be greater separation. Science should retain its typical conservative (in the non-political sense) approach and disclaim public policy action when the science upon which the policy is relying for its authority, doesn't support that policy. We have good science supporting bad policy, and when the policy fails or backfires, the public will (and has) link the failure with the science, and this may be unjustified if the science didn't support the policy action in the first place...so scientists should say so...beforehand!

I have serious reservations about the conventional presentation of current climate change science. I have to do several gut-checks each time I confront it to try and evaluate if my own biases can be justified. Perhaps I'm wrong, but I think they can, yet I will listen. But I am also tired of being labelled a "denialist" just for having my doubts, and being able to conceive of a mechanism whereupon the science process could be hijacked for a political result. This chapter doesn't soothe this ache any, but instead left me with an impression along the lines of: "You shouldn't take my word for it, but if in fact you don't, you're a denialist because Al Gore got it mostly right. Just sayin'..." Oh man.

Not addressed in Chapter 6 were issues like the co-opting of science for political ends. And climate science could be the poster child for this. We could say that the IPCC's pronouncements that current global warming is human caused are true. Our national and supernational governments immediately leap to a conclusion that they now have license to take action, and that they also know the appropriate action to take and the proper way to assess and balance the consequences of any such action. We've been presented truckloads of new proposals or in some instances been saddled with new law and regulation about various forms of anti-carbon or carbon-trading tax schemes. Evaluations of the likely impact of such schemes all so far as I am aware, all suggest that these will ultimately have insignificant effect, but do significant economic damage in the meanwhile, and simultaneously bolster and centralize government authority over economic affairs. Whoops! Was that what we wanted? The science may be sound, and careful analysis should answer any skepticism in time, but what we really ought to be skeptical of are governments and their plans to fix things. The track record is terrible, and the incentives almost always line up in ways that are destructive to long term freedom, prosperity, and enjoyment of life.

Few realize or take the time to contemplate the implications that the IPCC is primarily a government-consensus body, not a union of concern scientists. Further, it's divided into two groups, Working Group 1, which is charged with public relations, advocacy, and public policy, and whose work is usually what is reported by media. The other is Working Group 2, which in my opinion is the group to actually focus your attention, for this group is more interested in doing the research and integrating the scientific picture.

The results of present science and their interpretation between the two working groups, to my present understanding, has a wide gulf. While WG1 tells you we're sure what's going on here, it's going to be terribly catastrophic for all life on earth, especially human, and the fast majority (>90%) of scientists agree, WG2 is far more conservative, restricting itself to a more reserved opinion that the climate is changing, and that humans are influencing that change, but how bad it could be, and how it might go anyway if humans had no role is not clear. WG2 is more open about its confidence in current understanding of the mechanisms behind the climate system, and shows that the agreement among scientists is not so uniform as WG1 would like you to think; that the basis and confidence for selecting any particular sort of mitigation strategy is very premature at present.

WG2 shows that this science is largely one of measuring and modelling, making ever more complex data models to fit the observed historical data, then extrapolating these models to attempt to forecast the future. But why would such a model work? Answering this key question requires actually understanding the climate system, and unfortunately the focus has been too much on the data modelling (it's easier and gets funding). Here's a lesson on what that can get you:

Ptolemy had presumed the orbits of the planets were circular, and made predictions for their motion that soon strayed from observation. Not wishing to abandon the idea of circular orbits, he was able to introduce increasingly complex modelling terms he called "epicycles" that permitted this original concept to approach any desired level of accuracy and permit the observations to match his model. Great! At least you'll be able to predict where in the sky to look for a planet at any arbitrary time, but holding onto the Ptolemaic model, no matter how precise, would deprive you of actually understanding what was really going on: the planets' orbits are elliptical, not circular, and the reason is gravity. If you don't know this, you cannot progress in understanding the solar system or motion itself. Ptolemy's epicycles would've sufficed to predict the map of the solar system on the sky, but if we had held to them, we would have become stuck in his era, and our modern world would likely look more like his did then, and less like ours does today.

Our climate science is very possibly "Ptolemaic" at this point. And disagreeing with or remaining justifiably skeptical of the IPCC's orthodoxy gets you shunned as a "denialist." I think using such labels at this point is more a way to foreclose legitimate dissent, give license to policy ends for political reasons, and perhaps safeguard government funding streams to certain corners of the science establishment.

Okay, I shall end this massive digression and this review with a nod to Chapter 5, which is very useful if you are a "denialist" because you'll need to consider the sources that are feeding your "denialism". A process of re-normalization and filtering may need to be applied if there is any hope at screening out bias which may be present in opposition arguments. Ultimately, this could require you to become expert in the subject yourself. The end of the book, cited at the top of this writeup, is most helpful in giving you a strategy to proceed short of making that great leap to becoming expert.



View all my reviews

Fukushima Radioactivity a Complete Non-issue

I write this in reaction to a discussion group posting I saw recently, citing this article from The Register

I admit to having a starry-eyed youth regarding nuclear power. Growing up in the 80s, I viewed it as a boogeyman that made all the adults wary, but which just made the industry interesting to me. I saw all the old 50s "Our Friend the Atom"-type films. It seemed like _the_ miracle that would solve all our energy needs. We could be fabulously wealthy with essentially limitless electrical power available for just fractions of a penny/kWh.

Then, in April 1986, I lived under a mild pandemonium as Chernobyl became a household word, and civic officials from my town began patrolling sidewalks with civil-defense issued Geiger counters from the 50s, waiting for the arrival of the fallout clouds (despite living in Northern Minnesota; the fallout never came...of course). Terrible, frightening, and captivating for the curiosity of a young boy. Gradually, the bloom came off the rose. I witnessed recountings of the Three-Mile Island accident. I was scared by the spectre of films like "Silkwood" and "The China Syndrome". I Began to view it like other magic-bullet cure-alls of past and present. Hypothetical promise unable to live up to testable reality.

Before the 2011 Tohoku earthquake/tsunami, I was already doubtful about the future of nuclear power. I thought that, much like wind/solar/etc., it was a power technology that was just-competitive only because of massive subsidy and investment by government, making its real cost far greater than it's advertised price.

I've since learned that there is too much conflation in the minds of the public of the two kinds of nuclear tech: weapons and power generation. Weapons tech is completely incompatible with civilian power generation, and the two technologies overlap very thinly at one end of their Venn diagrams. The inputs into civilian use simply cannot be repurposed for weapons use. Power reactors of any design simply cannot become nuclear explosives (though one design, Chernobyl's RBMK, can explode (by steam) if operated incompetently). I've also learned that next to the modern-era renewables subsidy (most especially wind power), the subsidies to the nuclear industry are just about non-existent, and only very slightly more than that handed out to fossil fuel sources (yeah, we subsidize it all).

I'm opposed to subsidy on general principle, and I think that the cost disadvantage of nuclear power today can be explained by the largely unnecessary and over-reactive nature of government regulation of the industry (safety standards too stringent vs. real risks, and protection and waste mitigation policies administered in inefficient and idiosyncratic ways, and a culture of secrecy bourne mostly out of US Navy pressure (read: ADM Rickover) to ensure that Soviet nuclear tech remains behind the US), and also due to available economies of scale going unexploited (each plant is a unique and custom facility, insufficient total reactor buildout, etc.). With private market development (which definitely would require public acceptance, presently unlikely), I believe the already small price differential compared with coal would diminish to negligibility, and definitely be superior to oil and nat-gas.

In the wake of the Tohoku quake/tsunami, I became further pessimistic of nuclear power, esp. a completely private industry without government involvement, namely: how could the industry secure for itself the necessary liability insurance to cover any disaster, like  Fukushima, absent a government backstop?

Over the past several years, I've gradually come back around on the potential of nuclear power. I still think, for the present, it's unlikely we'll see much growth in the industry worldwide, but I now think this will mostly be for reasons of public attitudes, not of economic or technological impracticability. If you're one of the many who are convinced by the present climate science, to believe that we are responsible for a carbon-fueled global warming trend that can only be mitigated by forced de-carbonizing (I am not)...then you especially must re-examine nuclear power, because it is the only economically viable alternative whose technology is mature, and that can be produced affordably and at scale, compared with other renewable sources.

Since 2011, I've been studying radiation. It's /the/ great bugaboo in the public mind when they think of nuclear power. People are too afraid of it. If you want to find a nuclear power station, with government regulations what they are today, your best bet is to map and search for the conspicuous hole in the natural background levels! Our fear is caused in part by an obsolete model (the Linear/No Threshold model) of biological response to radiation doseages. Especially at the small end of the curve described by this model, most of our data is interpolated or simply guessed at, the signals of damage have been insufficient to reliably measure. Yet this old model implies that there is "no safe level" of radiation. That conclusion is highly dubious, borne mostly out of an abundance of caution given significant gaps in our scientific understanding. New research (too preliminary as yet) suggests that the may be a hormetic effect produced by low-levels of radiation above the average background level of most places. But even as considered under current practice, the LNT model /is/ saying that the lower the dose, the lower the risk...so the question which must be asked is, what /is/ the specific risk? Perhaps in the totality of your living situation, and relative to other risks, living in a previously contaminated disaster zone would, for you, represent an insignificant added risk to enjoying a healthy, happy, long and satisfying life!

Even with the handful of nuclear power disasters, including Japan's, an insufficient amount of radioactivity has been released to do much harm to anybody. Most people will be totally unaffected, medically. In fact, it can already be said that more people were harmed or killed as a result of mandating evacuations in Japan, than would have come to harm in the future (due to elevated cancer rates) from the added radiation exposures. Both in the Chernobyl exclusion zone and in Fukushima's, the real threat from contaminated soil and water is too small to warrant having those exclusion zones persist. Your skin is adequate shielding from the radioactivity of the released substances. In most situations, the volume of air or structure nearby will stop any emitted radiation long before it even reaches you. The real threat is from ingesting radioactive isotopes which would then have an extended opportunity to do damage to your cells while your body worked to remove them. But even here, the quantity of isotopes available for ingesting is simply too small to hardly warrant mention. The fear factor would preclude it being legal, but it's likely you could drink milk from a cow that was eating hay grown on contaminated soil, and eat fish modestly contaminated from swimming in the contaminated waters near the reactor sites, and live a full lifetime free from any side effects.

Folks served electricity from a coal-fired plant are generally exposed to more radioactivity from the coal itself, and the air emissions and ground waste after its combustion, than they would be had they lived right next door to a nuclear power station and been powered by it, instead.

All that to say that, even when nuclear power goes horrifically and catastrophically wrong, compared to other sources of power, it's orders of magnitude safer for human populations. How many coal miners died last year, and every year since coal became a primary fuel source? How many oil workers? How many refinery plant accident deaths? How much death, disability and destruction from flyash retention dams which have given way?

And Fukushima taught the industry lessons, such that in the future (if the public will accept it) reactors will be better designed to withstand such cataclysm without releasing anything dangerous. Most the operating reactors in the US are pressurized-water types. The designs evolved from similar designs used for the nuclear-powered carriers and submarines in the US Navy. Now you talk about robust: the Navy nukes are put through more strenuous regimes of operation than the earthquake/tsunami gave Fukushima Daiichi: extremes of load, extremes of temperature, extremes of vibration and angles of tilt. They've all operated without fail. There hasn't yet been a nuclear-plant related disaster aboard a US Navy ship. Simply building to the standard of these Navy reactors eliminates most risks.

This is already way TL;DR, but the worries about nuclear waste are similarly vastly overblown. I'll simply indicate that what is regarded as waste today, the retired and casked spent-fuel assemblies currently stored on-site for lack of a geologic repository a-la Yucca Mountain, Nevada, are not especially dangerous even if they were left un-casked. They represent a harvest of only 3-4% of the obtainable nuclear energy within, and so are actually a valuable future fuel resource (to alternative reactor designs than could extract it). Even if unpursued as a future fuel source, the operational lifetime of a plant and the relative harmlessness of the material itself is such that the most prudent course of action would be to just secure the sealed casks permanently on-site after decommissioning. The earth will long be a cinder from the expansion of the sun as a red giant, before we could burn up enough nuclear fuel to make the spent-fuel waste require any serious land area to store.

So...now I'm circumspect. I once again believe in the untapped potential of nuclear power. I think that future research and education may show that the post-catastrophe exclusion zones of Chernobyl and Fukushima may not be necessary to protect human life in the aftermath of what are the sufficiently rare worst-case scenarios for nuclear-power gone wrong. With adequate demonstration of contingencies to minimize the rate of such incidents (these would become actuarial and public-image considerations), I think a key critical stumbling block for nuclear power, a privately generated and maintainable market for accident liability insurance to bulwark the remaining risk, becomes realizable. With that, there is no great economic barrier to entry for nuclear power tech to achieve dominant adoption rates. /If/ the public will accept it in their backyard, and if government will permit it, plants could be built and come online w/o any coersive economic incentives that we currently use for renewables. The true cost of nuclear would be measurable, and I think it's likely to be low enough to once again have a shot at being the /magic-bullet/ promised us in the 1950s.

There will never be the magnitude of obtainable power from all the various renewables, that there is right now from nuclear. Whether we actually tap into and exploit it will depend on education and public attitudes, which are currently very poor. If we must de-carbonize, unless we shift to nuclear we will be forced into accepting a very austere energy future, as renewables cannot reliably come up to replace what would be lost from fossil-fuel, even if we could afford it (which we couldn't...a close look at subsidy levels to renewables will make that clear).

Most humans still find it quite difficult to accurately measure and quantify risk. Since 2011, I've become convinced that overall negative risks from non-nuclear power technologies are at least equal to or greater than those from nuclear power. Greater use of nuclear energy would allow us to reduce and perhaps eliminate reliance on coal, the dirtiest of the fossil-fuel sources and the backbone of electrical generation now; saving our oil-based sources for transportation that cannot be converted efficiently to utilize renewables (e.g. jet travel, diesel shipping). We could also perhaps feed more people, as this hopelessly misguided push for biofuel grown from land that might otherwise be growing food would be priced rightly back out of economic viability.

Not only is the Fukushima radiactivity a complete non-issue, it may just be the story that points the way to a new future course. It may be a phoenix rising.

More resources:

An important independent source of factual nuclear energy information.
A podcast on the impact of Fukushima radioactivity.
Take a look at Rod Adams' Atomic Insights, and follow @atomicrod on twitter.