To illustrate the fact that they will alienate all comers at some point, their views on animal experiments are rather equivocal – risking ire from both sides. This is their position:

"There is always some uncertainty about applying results from experiments on mice to men and women, even if the small size of a mouse is taken into account. Different species of mice show different sensitivities to radiation, and mice may be more or less sensitive than humans, or worse, may be simultaneously more and less sensitive, depending on the type of cancer of the specific mutation. As a result, animal experiments are used to supplement, rather than replace, what we can learn from observations on human beings exposed to radiation."

After alienating half the science community, who, as a brief perusal of the pages of the UK magazine New Scientist will confirm, are quite vociferous in defending the scientific use of animal experiments, the passage effectively loses most environmentalists by then defending what had just been effectively rubbished. This kind of nonsensical fence–sitting is surely detrimental to their argument, rather than being proof of their purity as "objective" commentators.

Hardcover jacket, OUP, US

Of course, everyone will find something missing, or treated too fleetingly. Areas considered include radiation, breast cancer [with regard primarily to pesticides], landfill and air pollution. However, though the examples chosen are fairly wide–ranging, there is too much space given to radioactive pollutants, from A–bombs, through Radon and Nuclear Power plants, to power lines. Ironically, the current fears surrounding mobile telephones as a source of microwave radiation are not mentioned at all.

In the past both sides in most pollution debates display a conflict of interests, such that objective facts are effectively not allowed. The science is questioned by those asking questions, in part because industry uses science in its defense. Those who respond to the assault from "experts" are pushed over to a Luddite position, that also represents particular [activist] interests. Meanwhile the pro contingent will brook no comment from amateurs – they are "not scientists", and are dismissed as mere "greenies". The truth may often be drowned in the middle, rather than emerging phoenix–like from the conflagration.

The authors set out to achieve their goal of informing "concerned readers" – where 'informing' equates with objectifying the readers perception of how environmental hazard issues are resolved, with reference to logical tools and concepts in scientific investigation. They also especially promote the embryonic field of molecular biology as our best hope of resolving some of the misunderstandings on the part of both science and citizen. This is an area of modern science that apparently offers great promise in pinning down the facts about particular environmental pollutants effects in the body. This is because it reveals more accurate ways of measuring chemicals and their effects. It is obvious that the authors feel such clarity would show up many past positions, held by either or both sides, to be untenable, and thus lead to less polarized debate in the future.

However, as the human genome patent issue hots up, it is hard to see how this bright new hope can be realized. The TNCs / industrialists are well on the way to side–lining it already – for if only an elite techno–clergy can be allowed permission to divine truth from DNA, we are back to square one, regardless of what could be achieved by sequencing and molecular biology.

This highlights one of the failures of the book. Despite wishing to force an acknowledgement that science should be seen as the activist's friend, the authors fail to indulge much controversy, and steer well clear of outright criticism of past "science", or of the corporate nature of science in an industrialist laissez–faire world economy. Yet this criticism is needed, in order to convince any of those who are concerned with environmental justice that the science they dismiss has a point.

And, of course, the science does have a point. What the authors offer, teased out from the cases they study, are a number of basic principles for environmental forensics. They cover causation, obtaining evidence, and using evidence found, describing the book as part of the "mission" that honest scientists are on, claiming:

"Science is the only antidote to ignorance and fear, and the best answer to the distortions and omissions of those engaged in advocacy for a cause."

This statement is applied to the authorities, industry and other scientists, not just to "greens". For example, in a discussion of Nuclear Test fall–out in Nevada in 1951 it is made plain that the AEC refused to acknowledge responsibility, instead denying that the tests could cause harm whilst suppressing and falsifying data. They then run through various [meteorological, geologic…] reasons why measuring exposures in order to find "proof" is not often fruitful with radiation – even though later in the book they describe the peaks in leukemia that follow weapons–test radiation fallout peaks like clockwork. They merely mildly castigate the AEC for not being honest, and opine that if they had done more, then "there would have been less suspicion of government, less anger, and, just possibly, less cancer" and refer to the events in Nevada as a "sad story".

This just is not a strong enough statement – the "ifs" and "buts" here are at the heart of the matter, but are left on the periphery of the argument throughout this book. However, it does yield the point that "the better we know exposures the more confidence we can have in the conclusions we draw." This leads to the first of several forays into describing how dose–response relations, and the accompanying comparison studies, using "exposed and unexposed human populations", work – and how they fail.

Logical Principles Misused

Several of the case studies bring us up against this point – that on the one hand we need to have reliable measurements, and on the other that these must also be meaningful measurements. Industry consistently use scientists who have little respect for science–as–truth when faced with cuts in their research grants, to claim that links cannot be found due to, say, other sources of the chemicals involved; or the problem of dissipating concentrations over time; or lack of measurements from before a problem is revealed; or data merely shows statistical or probable links – not proof; or the parameters that their reports take to be relevant reveal no problem…

All these arguments can be meaningful, but all too often they are used disingenuously. It will be an uphill battle for the Goldstein's to play this – valid – card against such a background. When there is meaning in such arguments, there can often be counter–intuitive ideas at play.

For example, with many environmental pollutants, including radiation and many hormone–mimic chemicals, there has been a sensible feeling that the lower the dose the less the danger. Yet in these cases real arguments have been proffered to suggest that some materials have a lessening effect per volume in higher concentrations, or, conversely, that vanishingly small amounts are sufficient to deal a lethal blow. Some forms of radiation are now seen as being in the former camp, allowing background radiation to be seen as the real problem, with diminishing effects attributable to higher doses. And some estrogen–imitators are seen as being in the latter camp, such that the US EPA and the WHO are having to try damn hard to avoid saying that there is NO safe minimum threshold [since such a position is politically unacceptable to the backers and funders of these institutions], despite considerable evidence that this is so.

Other Sources of Exposure

That both sides can make errors of judgment is highlighted by the case of a Pennsylvanian nuclear power plant worker, who set off detectors at the new plant before proper operations had even begun. It transpired that the worker's house had natural radon levels "a thousand times greater than federal standards permit in mines." It is clear how such a resolution may quite reasonably be distrusted by anti–nuclear activists, as they have set beliefs that a – rare – case such as that above will not be readily encompassed by. The case highlights the whole question of variable background levels. Is there something suspicious about plants like Sellafield being built in some of the most radon permeated territory in Britain, an area rich in granite and other uranium rich rock? Or that the fall–out from Chernobyl just happened to be highest over the very fells that are closest to Sellafield? As long as science and the uses of science are conflated in the activists' imagination, an easy resolution to this natural suspicion will not be had.

Risk Perception is Colored by Politics

As a natural hazard, as well, radon has not attracted the fervent lobbying that pollution with a human cause engenders. This means that a very real hazard, that can be remediated relatively easily, has not been on the political agenda and therefore goes largely unaddressed. That most lobbyists are honest and well–meaning makes an even greater irony of the fact that:

"…radon in homes presents a real health hazard, not so bad as smoking cigarettes but much worse that any yet–demonstrated effect of pesticides, toxic waste dumps, incinerators, or living near nuclear facilities."

Although some "green" groups try to avoid shitting in their own back yard, it is also the case that the source of much exposure to toxins – the home – is the cause of a good deal less voluble campaigning than it should be. Yet "cigarettes, chemical products like deodorants, household cleaners, paints, sprays, even the furniture itself and the fabrics that cover it", and clothes are the source of 76% of human exposure to chemicals like, for example, benzene and tetrachlorethane. This is obvious to asthma sufferers, who can often have more problems indoors than out. Industry is not off the hook here, of course, for who produces these products that end up in the home? However, an awareness of the issues on the part of the ordinary person is obviously necessary if we are to truly address the problems.

The well–meaning activist is also often ideologized, such that greedheads and impersonal agencies are more likely to be noticed and demonized than is either Nature or the behavior of the common man.

Association is not necessarily indicative of Causation I

The fact that individual exposures reflect the pattern of local geology rather than averaged measurements highlights another problematic area. Some average measures of a substance can appear to overlay or map onto particular disease, but it is not possible to simply assume that the apparent correlation is meaningful – though it is natural and logical to do so. An example given is a perceived correlation between "total countrywide consumption of fats as an estimate of the consumption of fats by each individual in that country" and colon and breast cancers. Later studies based on actual individual consumption, rather than group averages, failed to confirm the correlation. It appears that a logical relationship masked the reality.

A similar problem may occur when limiting studies to heavily exposed groups. Often these will consist of workers, who are by definition usually not representative of the population as a whole, so that statistical aberrations may show up simply as a result of not comparing like with like. A conscious effort must be made to take into account things like better than average health in a studied group.

The assumption on the part of environmentalists over increased leukemia rates at the British Sellafield Nuclear Plant has been that radiation is responsible. However, it is often contended that in fact there may be a viral factor in the cancer, and that the number of workers entering and leaving the area is artificially overexposing a vulnerable population. The latter would be effectively masked by assumptions about the former. This appears rather odd, especially given that surely places like London have far more exposure related to population flux than Cumbria does. It is easy to say that the argument is just another ploy from industry to move the goalposts and avert blame. I suspect this myself, especially given the known lies about levels of radiation release during the 1957 fire, etc, but what if there was some truth in it, and the core local population in this previously isolated community do have a predilection for leukemia, possibly exacerbated by increased exposure to triggers? Is this trigger for leukemia radiation or some previously unsuspected viral agent? Or, as yet others have argued, is the Sellafield cluster just a statistical aberration? Place your bets now…

The statistical aberration argument is not necessarily always wholly unreasonable. However, as the authors put it:

"The assumption of randomness is a confession of ignorance. When we make it we are not saying that there are no causes for a disease but rather that whatever the unknown causes are, the chance of any one person being exposed to them is, as far as we in our ignorance can tell, the same as that of any other person."

In the case of Sellafield there seems to be ample reason to question whether the chances really are the same for other people, with the known high background radiation, plus the artificial [and due to mismanagement and lies higher than we know] leaked radiation.

"Chance is lumpy"

Still, why the particular individuals were affected and not others seems to require us to invoke randomness again… The overall pattern of flipping coins will be 50–50, but it would be odd if a pattern that averages out at 50–50 was exactly even on all scales. It may go Tails Heads Heads Heads Tails Heads Tails Tails Heads Tails – five of each, but a cluster of three heads near the start… – but the chance of getting heads again in the fifth position, making four heads in a row was still 50%. "Getting exactly the average is less common than getting either more or less than the average."

This means that diseases that are genuinely random will still appear to cluster. And some smart–alex will happen along and draw a line round the cluster and see it not as part of a larger pattern but as some weird event. He could be right – as the example comparing the individual and group–aggregate above shows, the cluster could be real.

How can you tell? Here is one suggested solution – at Sellafield the cluster could be, and was, watched. If it had been merely a chance cluster it would have disappeared over time, and if a local cause remained it would not. The Sellafield cluster remained for a decade. This undermined the argument that the cluster was just an aberration. Other cases may have different logical elements that need to be analyzed in order to show whether the seeming pattern is "real" or an illusion caused by clumpy chance. However, the principle is the same – it is the duty of the scientist to attempt reasonable levels of falsification before taking a particular interpretation of the "facts" to the public. This is an area where many scientists fall short – whether because of blind spots and incompetence, or because of a debt to sponsors [whether industrial or environmental].

Use & Abuse of Risk Factors

Another problem is that many of the diseases associated with industry are also associated with bad diet and low social class.

How do you disentangle these lifestyle factors from industrial sources of harm, when the physical location of people in a lower social class coincides with the physical location of a known pollution source? As they invariably do, of course. The risk factors complicate the diagnosis, a fact that you can rely on companies to milk for all its worth – no, their pollution didn't cause it, look at the other factors.

These other factors, of course, exist. Genetic factors are particularly polarizing. The fact that a group may have a genetic proclivity for a given disease will be played on heavily in order to exonerate a particular industry / pollutant. However, it will be a rare industry / pollutant that can reasonably be argued to be ultimately blameless whether this or that group is more susceptible than average or not. It is not reasonable to blame asthmatics, for instance, for their susceptibility to attacks in the presence of a given pollutant. If they are sensitive in such a overt way, what other effects will the pollutant be having to everyone?

But just as the company will exaggerate the effect of other factors in order to mask the effect of their contaminants, those opposing the company will be unwilling to accept that anything is a factor but the contaminant. A campaign to demonize some particular pollutant, possibly emanating from a local factory, that is thought to be the cause of increased asthma in a community may be overlooking something as simple as the fact that the modern trend of using double glazing in the home cuts down the circulation of air, which "[leads] to an increase in home humidity, which favor[s] the growth of molds and fungi known to exacerbate asthma, and [provide] more congenial conditions for the house–dust mite".

Once again any honest observer, whether a scientist or not, will be shot by both sides.

The chapters on breast cancer point up many of the same category errors and ideological conflicts. Hormone mimic chemicals, such as pesticides, are almost universally accepted by the public as having a role in breast cancer, and many scientists agree. Many do not – are they all liars? One theory is simply that the modern lifestyle that women have is at odds with their "natural" lifestyle.

This is often greeted with derision, especially by feminists, who have ideological reasons to refuse such an explanation. This does not mean they are wrong, but it is an argument that, in the past, women – acting as, in effect, baby machines – spent a great deal more time lactating and a good deal less time menstruating [they also died in childbirth more…]. That they do not anymore means that most women in the West could be said to have by definition an unnatural lifetime hormone balance. It has been argued that this is the cause of increased breast cancers. Such an argument is hard to resolve. But one thing is sure, getting into it will obliterate any sensible discussion for a while, and just when pesticides and plasticizers are beginning to be taken seriously as a major hazard.

Comparing Like With Like?

And what factors are relevant, anyway? Are we comparing like with like? For example, are the average ages of two populations studied the same? Are the racial profiles the same? Does the group who show purportedly higher breast cancer rates self–examine more than the other? This is often linked to education and so to social class – is this a factor in other ways, such as diet? Are the "case–control" groups representative? When may these things be relevant to the study, and when not?

In the words of Rachel Carson, perceived as one of the founders of the modern environmentalist movement, who herself died of breast cancer:

"When one is concerned with the mysterious and wonderful functioning of the human body, cause and effect are seldom simple and easily demonstrated relationships. They may be widely separated in space and time. To discover the agent of disease and death depends on a patient piecing together of many seemingly distinct and unrelated facts developed through a vast amount of research in widely separated fields."

These kinds of problems and errors immediately encourage scientists to feel that they must be sacrosanct as expert interpreters of the data, and cannot allow mere citizens to be allowed to influence science politics. As an argument against citizen science this stinks, as, of course, scientists often make the errors, with later research by other scientists correcting them. The role of activist lobbying here is not clearly negative. The science tends to ossify a mistaken result as fact for longer than may be the case if there is political pressure for testing the results. And citizens, too, can handle logic.

Outdated Ideology Locked–In?

This does not get away from the fact that inertia can operate on both sides, with lobbyists being slow to accept new evidence and becoming stuck in an out of date protest long after the game has moved on. This is a particular problem for the larger and more "corporate" environmental activist groups. In fact corporatism, hierarchy and economies of scale are one of the environmentally costliest things imaginable, with both pro and anti groups becoming locked–in to ridiculous battles like dinosaurs with truth having no look–in on either side.

The authors look at several campaign areas, including breast cancer activists. They point out that the sense of urgency these primarily middle–class women have engendered, while having achieved increases in funding and research in the area, also create untested public perceptions which tie activists in to a sometimes blinkered viewpoint. There may well be other relevant factors in breast cancer, that are not being properly researched due to pressure from those who have become convinced that hormone mimics, and less direct pesticide effects on hormone metabolism, is the be all and end all. Other lifestyle factors may be underestimated in the shadow of the emerging ideology of the activists. They may also limit intensive study to well–off urban communities, where concern is high [or, more to the point, heard], rather than pesticide spraying zones, where poorer women are much more exposed.

The coverage of the power–lines / magnetic field issue looks at the extreme positions in order to highlight how much contention can exist even between scientists. My overall impression here, though, is that real issues are being buried between complacent industry scientists, and the 'nuttier' pundits who claim the most extreme negative effects. Manufacturers of electric blankets did come to take seriously the possible leukemia risks of high fields [now banned] around their products, though this may really have been them taking consumer concern [=money] seriously.

Association is not necessarily indicative of Causation II

Once again, the probable proper interpretation of the evidence is generally held to be that links between leukemia and appliance use [including humidifiers, hair dryers, TV's...] actually mask the real causes. The lack of a clear dose–response relationship is part of the reasoning here. There is also a lack of convincing explanations for how magnetic and/or electric fields around appliances might cause the effects being studied. This has dominated the argument – perhaps understandably – but this does not negate the correlation, so complacency is to be avoided. It may still be the case that the appliances the source of leukemias, and that the magnetic fields the appliances create are the cause.

But in the absence of a known mechanism for such a vector, it is not illogical to argue that there may be a chance association here, and the real cause of the problem is elsewhere. It would be a shame if a proper explanation was overlooked because of the politics of those at loggerheads in the debate.

Here is a classic example of a false assumption, as described in the book, the acceptance of which may well have cost many lives before an alternative was powerfully enough represented to overturn the established position:

The London fogs in the nineteenth century seemed to correlate with cholera epidemics, as low–lying areas suffered the most from both. This association was shown not to be causation once it was understood that sewerage in drinking water was the real reason. The low–lying areas received more highly polluted drinking water, as the cities sewage flowed down to the river.

This is the essence of the argument some scientists use against many environmental health activists – the association the activists point to are not really causes, just coincidences. However, the logical frameworks that the scientist may use to prove causality are not always as transparent as they should be, and must be analyzed to see if they in any way "rig" the outcome.

An example of this is "expecting" a strict dose–response relationship in time between a release to environment of a chemical and noticeable effects in a population. Many industry scientists have argued that such a relationship is necessary to "prove" the link between dioxin release and disease in a community, and this will be accepted as a reasonable logical explanation – unless you know that dioxins bioaccumulate in fatty tissues, which release the chemical to the organs of the body at different times in different people depending on metabolism, lifestyle, etc. In which case you will see that the expectation of such a link is naive at best and a downright lie at worst. When industry is the sponsor, the latter interpretation is understandable.

Sometimes "3+3=9"

One area that environmentalists are often quite bullish about is the effect of synergy. It is often the case that an anti–pollution group will use straight–line extrapolation, where we all know that in the real world these often cannot really continue upward forever, as other factors usually intervene [of which lobbyists having a positive effect may be one]. And at least as often we will see graphs showing exponential increases. These too may often not come to pass.

However, it is also the case that synergy does happen. Industry pundits often present straight–line extrapolations that seem worrying but not overly so, where they blankly refuse to present data that incorporates the interaction of their problem with other factors. Both parties can be wrong in these things, but if there is one thing certain it is that sometimes "3+3=9". This is an even more complex set of interactions. You are more likely to be wrong. The synergy could even work to dampen effects.

But the costs of positive synergy can be astronomical. An example in the book looks at lung cancer. Those exposed to asbestos and those who smoke each have a ten–fold risk of developing lung cancer, yet those in both camps experience "not a twentyfold risk, but rather an almost one–hundred–fold risk".

Where "Ideology" Matters I – the Precautionary Principle

I have made several references to "ideological" positions on both sides in environmental health hazard disputes. Often I may have sounded as if some Third Way exists that makes both groups look foolish. This is sometimes true, whether those implicated like it or not.

However, there are some points about which neither side can agree where I personally feel the need to agree with the activist position more strongly.

The scientific method tends to hold great stock in the concept of "statistical significance". This is undoubtedly an important concept. However, in environmental disputes the usual position that the establishment holds is increasingly seen as untenable. To the question "How large an increase in leukemia would have to occur, before we rule out chance as responsible?", the scientific community's position is expressed by the authors, who evidently embrace it, like this:

"We first calculate what the probability is that the difference observed could be attributed to chance alone. This probability will lie between 0%, meaning that it is 'impossible' that it is due to chance, and 100%, meaning that it is 'absolutely certain' that it is due to chance. […] The practice the scientific community has chosen to follow in studies of this type is to draw the line at 5%: we call the difference 'statistically significant' if the probability that it is due to chance alone is 5% or less. This is equivalent to saying that the probability that the difference is not due to chance in 95% or more…

"[…] When we use a strict standard, like 5%, we will make the false negative mistake more often (attributing real increases to chance), and the false positive mistake less often (accepting a chance increase as real), than if we used a more lenient standard.

"If both mistakes were equally harmful, there would be no reason to be so strict. Scientists, however, do not regard the two mistakes as equally harmful…

"[…] guilt must be proved 'beyond reasonable doubt.' Our practice favors the defendant. […] We have chosen these safeguards because the thought of an innocent person found guilty and going to prison of being executed (a false positive) is more offensive to our moral sense than the thought of a guilty one going fee (a false negative)."

It is clear that a strict standard for statistical significance may make sense scientifically, but that is little consolation for those who have just lost a loved–one to a leukemia that they deem "unnatural". Of course, if you are close to someone who dies of a relatively rare disease that the media suggests may sometimes be caused by chemicals produced by industry, it is natural to ask questions. This does not make you right. Yet you maybe right, and science, uncovering possible links, may never agree that the statistical evidence is sufficient to blame the industry concerned. This is because a degree of occurrence outside the average occurs naturally.

It is also the case that even when there is statistical significance in a rise in disease, that in itself does not point at any particular source. The "test of statistical significance… [is used] properly when we start with a hypothesis about the cause of some disease and then test it by looking for higher rates in an exposed population. We use it improperly when we start by noticing higher rates of disease in some group and then look for a hypothesis about what exposure might have caused them."

A strict standard for variation may make sense, as a low level of deaths caused by industry can be drowned in the space between statistically insignificant and statistically significant. And in many countries, courts will not accept such statistical evidence even when it is deemed significant. But, your friend is still dead, and dead from pollution. In cases involving environmental problems few of the culprits risk jail, and the only people "executed" are those killed by pollution. The eminently sensible insistence on "Guilty until proven innocent" in murder trials and other criminal cases is actually detrimental to stopping ecological destruction and punishing dirty industry. A more environmentally cautious approach would favor a less strict standard.

There is, in fact, a definite move towards accepting a less stringent standard than that embraced by the science community and the courts. I agree that a less strict standard for ascertaining "statistical significance" is necessary in order that we follow, instead, the Precautionary Principle in environmental matters. After all, it is better to have descendants to apologize to for having been over–cautious, than none due to lack of caution. [Some suggest that this principle may not go far enough – see Individualist Greens]

Where "Ideology" Matters II – Acceptable Risk

"Standards are set using the concept of 'acceptable risk'. The word 'acceptable' is a matter of values, of weighing health consequences against other costs."

Certainly, the delay in publication of the 1993 US EPA report on dioxins would appear to be a result of different interpretations of "acceptable risk". Suggesting a considerably more stringent safe threshold for organo–chlorines, it has had an uphill struggle to be accepted by the rest of the political and business community. It is obvious that when the "other costs" are primarily financial the decision–making establishment has a skewed set of values regarding human well–being. Lives, especially those of the poor and represented, are cheap.

Nonetheless, costs are not always transparent, and further analysis may make a clear course of action more problematic.

"A report on electromagnetic fields and human health by the Harvard Center for Risk Analysis stated as follows:

The potential economic costs of mitigating the possible health effects associated with [electromagnetic fields] are substantial. […] The cost to the average consumer of changes such as these may be significant, not only in the size of the citizens' electric bills, but also in the potential availability of electricity for heat, air conditioning, and other household necessities.