LL2-00 — Acknowledgements, Preface, Ch1 Introduction, and Part A introduction (Lessons from health hazards)#

Late lessons from early warnings: science, precaution, innovation, EEA Report No 1/2013. Report pp. 1–16 (PDF pp. 3–18).

What the section contains: title page and colophon (pp. 1–2), contents (pp. 3–4), acknowledgements (p. 5), preface (pp. 6–8), Chapter 1 Introduction (pp. 9–11, including the “Twelve late lessons” box on p. 11), a blank page (p. 12), the Part A divider (p. 13), the Part A contents (pp. 14–15) and the list of Part A panels (p. 16).

Reading and verification. I read the whole text extract in order, through the last marker (PDF p. 18 / report p. 16). Page images could not be rendered in this environment, so I used a layout-preserving text extraction of the PDF to check the pages most likely to be garbled. That check confirmed four things: - p. 12 is blank; - p. 13 has only the title “Part A — Lessons from health hazards” and a stock-photo credit, with no introductory prose; - the “Twelve late lessons” box (p. 11) and the panel list (p. 16) were extracted completely; - the two-column text on pp. 10–11 was read in the right order.

To characterise authors’ standpoints and check one number, I also looked at material outside this page range. It is marked wherever it is used: - Annex 1 author biographies (actual printed pp. 685–700); - the opening lines of four Part A commentaries that are missing from the panel list (pp. 121, 125, 130, 145; the audit read the full first page of each); - the Chapter 2 summary (p. 17); - the “In memory of” pages (pp. 683–684) and the back-matter page numbers.

Later-evidence notes are marked [later evidence], with sources. Points I add from general knowledge that I did not verify in this session are marked [general knowledge, unverified].

Authors and standpoint#

Who speaks in this section#

• Jacqueline McGlade (preface). Annex 1 says she became EEA Executive Director on 1 June 2003. She was on leave from a mathematics professorship at University College London, and her research covered spatial data analysis, informatics, climate change, scenarios and “the international politics of the environment and natural resources” (Annex 1, pp. 693–694). She is also a co-author of Ch 17, “Ecosystems and managing the dynamics of change” (p. 4). The preface is written as a personal and political essay, not a methods statement.
• David Gee is named as the “Originator of the Late lessons from early warnings project at the EEA” (p. 5). His Annex 1 entry (p. 689) says:
• he graduated in economics and politics;
• he worked in occupational and environmental health “since 1974, for trade unions and non‑governmental organisations”;
• he is “a former Director of Friends of the Earth, in the UK”;
• he has been at the EEA since 1995, and was “the catalyst, EEA editor and a chapters author” for both volumes.

He co-authors Ch 3 (lead), Ch 8 (vinyl chloride) and Ch 21 (mobile phones), and is sole author of Ch 27, “More or less precaution?” (pp. 3–4). The introduction is unsigned but reads as the editor’s framing.

Institutional position#

The report is published by an EU agency but written “with substantial contributions from many external authors” (p. 5). A footnote to Annex 1 says the views in the chapters “are those of the authors and do not necessarily reflect those of the institutions where they work” (p. 685). The front matter never says how far the chapters’ conclusions are EEA positions. The preface and introduction speak with the agency’s voice and adopt the chapters’ conclusions with only light hedging (“by and large”, “for the most part”, “carefully designed”, “can”, pp. 10–11).

Editorial team, advisers and intellectual lineage (p. 5)#

• Editorial team: David Gee, Philippe Grandjean, Steffen Foss Hansen, Sybille van den Hove, Malcolm MacGarvin, Jock Martin, Gitte Nielsen, David Quist and David Stanners (p. 5).
• Editors who are also authors. At least five editors also write chapters. Hansen writes the false-positives chapter (Ch 2) and co-authors the nanotechnology chapter (Ch 22). Grandjean writes Ch 26 and a Minamata commentary (p. 121). Gee writes Ch 27 and co-authors three case chapters. Van den Hove co-authors Ch 17. Quist leads Ch 19 (pp. 3–4).
• What Annex 1 adds. Van den Hove “serves as chairman of the scientific committee of the European Environment Agency” (Annex 1, p. 692). Quist is at GenØk, the Centre for Biosafety, and is “a founding member of the European Network of Scientists for Social and Environmental Responsibility (ENSSER)” (Annex 1, p. 696). Grandjean’s research on prenatal methylmercury neurotoxicity “helped inspire international efforts to control mercury pollution” (Annex 1, p. 690).
• Why it matters. The chapter that answers the strongest objection to precaution (Ch 2, false alarms, led by editor Hansen with Tickner, who is not on the editorial team) and the two signed concluding governance chapters (Ch 26, Ch 27) were written by editors. Ch 28, “In conclusion”, carries no author name in the contents (p. 4), so it is presumably editorial. The Part B closing chapter (Ch 17) is by the Executive Director and an editor (McGlade, van den Hove; p. 4).
• Overlap with the advisory board. Silvio Funtowicz (advisory board) co-authors Ch 19, and Owen McIntyre (advisory board) is also on the contributing-authors list (pp. 4–5).
• Advisory board (p. 5): Joan Martinez Alier, Philippe Bourdeau, Sylvaine Cordier, Sven Dammann, Michael Depledge, Frans Evers, Silvio Funtowicz, Thomas Jakl, Martin Krayer von Krauss, Owen McIntyre, Jerome Ravetz, Nicolas de Sadeleer, László Somlyódy, Sofia Vaz and Theo Vermeire.
• Funtowicz and Ravetz are the originators of “post-normal science” (Annex 1, p. 689, confirms Funtowicz’s post-normal science work with Ravetz).
• De Sadeleer is a leading legal scholar of the precautionary principle, and Martinez Alier a founder of ecological economics and environmental-justice scholarship [general knowledge, unverified].
• “Continuing sources of inspiration” (p. 5): Domingo Jiménez Beltrán, Philippe Bourdeau, Jane Keys, Malcolm MacGarvin, Andy Stirling and Brian Wynne.
• Stirling and Wynne are science and technology studies (STS) scholars known for work on ignorance and uncertainty, on the framing of risk appraisal, and on lay knowledge [general knowledge, unverified].
• The Vol 1 lessons reproduced on p. 11 (ignorance, lay knowledge, values, institutional learning) reflect that lineage directly.
• Remembered (p. 5): Poul Harremoës (spelled “Harremöes” on p. 5), who chaired the Vol 1 editorial team, and Masazumi Harada, co-author of Ch 5, “who worked throughout his career on mercury pollution” and died in June 2012 (“In memory of”, p. 683; Annex 1, p. 691).
• Reflexivity flag. The contributing-authors list includes “Andrew Maynard” (p. 5), a co-author of Ch 22 on nanotechnology (p. 4). The person commissioning this analysis is therefore a Vol 2 contributor. That matters most for the notes on Ch 22 and for the framing on p. 10, which applies historical lessons to nanotechnology.

How cases and authors were chosen (pp. 9–10)#

• Topic selection. Topics were “selected on the basis of advice from the editor, in collaboration with the editorial team and an advisory board, members of the EEA Scientific Committee and the Collegium Ramazzini” (p. 9).
• Footnote 1 (p. 9) describes the Collegium Ramazzini as “an independent, international academy founded in 1982 by Irving J. Selikoff, Cesare Maltoni and other eminent scientists”.
• Its stated mission is “to advance the study of occupational and environmental health issues and to be a bridge between the world of scientific discovery and the social and political centers, which must act on the discoveries of science to protect public health” (p. 9). A body whose stated purpose includes moving findings into protective action is a reasonable source of advice, but not a neutral one.
• The Collegium is one of five named sources of advice (the editor, the editorial team, the advisory board, members of the EEA Scientific Committee, the Collegium). The section does not say how much weight each carried.
• Links to Part A authors (Annex 1, outside this range): Morando Soffritti (Ch 8) is Scientific Director of the Ramazzini Institute (p. 698), and James Huff (panel 8.2) is a Fellow of the Collegium Ramazzini (p. 692).
• Author selection. Authors “to varying degrees, have had substantial involvement in the subject area”, and “would not have been approached if they had not already extensively studied the case” (pp. 9–10).
• Objectivity. Authors “were expected to be as objective as possible in answering the questions put to them by EEA”, in line with “their professional scientific training” (p. 10). The report names two supports for this: the seven structuring questions (“To support this, and to develop consistency between chapters”, p. 10) and peer review. It relies mainly on professional norms; there is no adversarial design such as co-authorship across opposing positions.
• Consistency. Authors were given “seven structuring questions” (p. 10). The questions are not reproduced in this section.
• Peer review. Chapters were “peer‑reviewed by recognised experts in the respective fields” working “within a set of editorial guidelines provided by the EEA” (p. 10). The acknowledgements say review came “from a broad range of perspectives” (p. 5). No reviewers are named and no review outcomes are reported, so the breadth claim cannot be checked from the report.

Part A chapter authors and their stated standpoints (authors from pp. 3, 14–15; standpoints from Annex 1)#

Pattern. Part A’s authors are senior domain experts, and many were also participants in the histories they write about: researchers whose findings triggered regulation, expert witnesses, regulators (Michaels serving at the time, and formerly at DOE; Bingham and Monforton formerly at OSHA or MSHA), and NGO or public-interest scientists. That gives them unusual knowledge of the evidence and archives, and a clear standpoint. On the authorship lists, no Part A chapter is written by an industry scientist, a company, or a regulator defending a past decision.

Panels and commentaries in Part A#

The listed panels (p. 16) are the following. Authors’ positions come from Annex 1.

Four commentaries missing from the list. The Part A contents (pp. 14–15) show four commentary pieces that are not in the p. 16 panel list. I confirmed their authors from their opening pages in the PDF: - “From Minamata to global health risk”, Philippe Grandjean (p. 121). - “Mercury science and policy since Minamata: four insights for policy”, Noelle E. Selin (MIT) (p. 125). - “Mercury in fish — the need for better information about contaminant exposures”, Argelia Castaño (Instituto de Salud Carlos III, Spain; also on Spain’s delegation to the UNEP mercury negotiations, Annex 1 p. 688) (p. 130). It opens on a genuine benefit–risk dilemma: fish are healthy, but methylmercury is harmful. Its first page also urges being “pragmatic in establishing the safe levels and not exaggerate the risks” (p. 130), a caution against over-warning that sits inside Part A. - “Corporate behaviour in the regulation of beryllium: could there have been a different outcome if the company had room to turn around?”, Tee L. Guidotti (p. 145). Annex 1 describes Guidotti as a consultant physician and vice president at a private firm, Medical Advisory Services (pp. 690–691). - A footnote says it is “an adaptation of a critical review prepared for the European Environmental Agency”. The piece calls itself “this panel”, so leaving it off the p. 16 panel list is an inconsistency. - Its opening argues that corporate leaders more often “live in a world of cognitive dissonance and denial rather than cupidity”, and that “The challenge is not to condemn the behaviour — that is easy. It is to understand it in order to control it” (p. 145). - It notes that Michaels and Monforton accuse the company of “cupidity and arrogance”, says “Doubtless, the company has a different narrative, but this panel is not about the company’s culpability”, and reframes the story as organisational behaviour and “pathway dependence” (p. 145). - It also says the company “was initially abetted by the US Department of Energy (DOE), which initially resisted an evidence-based precautionary protective standard” (p. 145). That is evidence, inside Part A, of a state actor sharing responsibility. - I have read only its first page. On that basis it is the closest Part A comes to a complicating voice on corporate motive, though it does not dispute the harm or the delay. It sits beside a chapter titled “Beryllium’s ‘public relations problem’”. It needs full treatment in the LL2-06 notes.

The p. 16 list is therefore an incomplete map of the commentary in Part A. Judging by the affiliations given in Annex 1, none of the 13 contributions (9 listed panels plus 4 commentaries) is written by a company whose conduct is at issue, a trade body, or a regulator defending its own record. Guidotti works for a private consultancy (pp. 690–691), and Castaño and Haigh have government or agency roles (pp. 688, 691), so the set is not drawn only from academia and NGOs.

Evident stance of this section#

The section is explicitly normative and pro-precaution. It treats corporate conduct and institutional inertia as the main causes of harm, while also naming causes that involve no one’s bad faith: “the novel and challenging nature of the issues themselves” and “poorly or inconsistently evaluated information” (pp. 10–11). It treats scientific knowledge as shaped by power. It aims to make “normally invisible” dynamics visible so that communities can take part in governing innovation (p. 7). It is open about its values: the preface opens with a moral and political diagnosis (p. 6). It is an argued, advocacy-inflected synthesis, not a neutral systematic review.

Section-by-section notes#

Title, colophon and contents (pp. 1–4)#

• Subtitle. Vol 2 is subtitled “science, precaution, innovation” (p. 1). Vol 1 was subtitled “the precautionary principle 1896–2000” (p. 9). My reading: the new subtitle marks a deliberate widening, from the precautionary principle as such to how science is produced and how innovation is steered.
• Colophon. ISBN 978-92-9213-356-6; doi:10.2800/73322; “Reproduction is authorised, provided the source is acknowledged” (p. 2).
• Structure (pp. 3–4). There are 28 chapters in five parts:
• Part A, Lessons from health hazards: Ch 2–11.
• Part B, Emerging lessons from ecosystems: Ch 12–17.
• Part C, Emerging issues: Ch 18–22 (Chernobyl and Fukushima, GM crops and agroecology, invasive alien species, mobile phones, nanotechnology).
• Part D, Costs, justice and innovation: Ch 23–25.
• Part E, Implications for science and governance: Ch 26–28.
• Annex 2 gives an overview of Vol 1, and Annex 3 updates some Vol 1 cases.
• Pagination error. The contents page gives the back matter as “In memory of” 683, Annex 1 715, Annex 2 731, Annex 3 747 and Index 768 (p. 4). “In memory of” is correct: pp. 683–684 carry memorials to Masazumi Harada (1935–2012) and Poul Harremoës (1934–2003). Everything after it is out by 30 pages. The PDF shows the actual printed pages are Annex 1 pp. 685–700, Annex 2 pp. 701–716, Annex 3 pp. 717–737 and Index pp. 738–760, and the colophon on the last page says “760 pp.”. Anyone citing the annexes should use the actual printed pages.

Acknowledgements (p. 5)#

• “This report has been prepared by the European Environment Agency (EEA) with substantial contributions from many external authors” (p. 5).
• Editorial team, advisory board, production support and 81 contributing authors are listed (p. 5). The EEA is “particularly grateful to those authors who wrote the case study chapters” (p. 5). Standpoint implications are covered above.
• “Case study chapters have been peer reviewed from a broad range of perspectives” (p. 5).
• Inspiration and remembrance are covered above.

Preface (pp. 6–8), Jacqueline McGlade#

1. Epigraph (p. 6): Benjamin Franklin, “An investment in knowledge pays the best interest” (The Way to Wealth, 1758). It frames knowledge (monitoring, research) as a prudent investment, which foreshadows Vol 1 lesson 2 (p. 11) and the concern with how knowledge is “financed” (p. 10).
2. Moral and political diagnosis (p. 6). - It opens: “There is something profoundly wrong with the way we are living today.” - It names “corrosive pathologies of inequality” in access to a safe environment, healthcare, education and clean water, reinforced by “short‑term political actions and a socially divisive language based on the adulation of wealth” (p. 6). - A progressive response needs both more knowledge and “an awareness that many aspects will remain unknown”, plus “a more ethical form of public decision‑making based on a language in which our moral instincts and concerns can be better expressed” (p. 6). - “These are the overall aims of Volume 2” (p. 6). The report’s stated aims are ethical and political, not only technical.
3. The world when Vol 1 appeared, around 2001 (p. 6): an economic slowdown; China joining the WTO; an EU of 15; a third decline in global grain production in four years; meat recalls, foot and mouth disease and BSE; rising temperatures and falling bird populations; US rejection of Kyoto; the first human genome sequence; chemicals managed through the Basel, OSPAR/HELCOM and Montreal instruments; and the World Trade Center attack, which “had just happened” (p. 6).
4. “A period of extraordinary hubris” (p. 6). - The financial crises of 2007–2009 are attributed to “indebtedness, mispriced products, lax monetary policies and mis‑engineered protection against risks and uncertainty”. - “The world experienced more not less volatility.” - Political systems “became silted up by vested interests and a determination by citizens to protect assets accumulated in easier times”. The blame here falls on citizens as well as elites. “Beneath it all lay a deeper environmental crisis epitomised by climate change and biodiversity loss” (p. 6).
5. Collapse of trust (p. 6). - Trust collapsed in financial institutions and big companies “as they abandoned staff, pensions and health care schemes”. - “Recent evidence from social psychology” is said to show that people trust only those they know; no reference is given. - Stephen Green’s Good value (2009) is quoted on “a massive breakdown of trust” across finance, business, politics, media and globalisation (p. 6).
6. Trust in science, and a causal claim (p. 6). “The scientific elites have also been slowly losing public support”. The preface attributes this “in part” to “the growing number of instances of misplaced certainty about the absence of harm, which has delayed preventive actions to reduce risks to human health, despite evidence to the contrary” (p. 6). - This is a significant hypothesis: expert overconfidence in safety delays prevention, and the resulting harm erodes trust in science. - No evidence is cited for either the decline or its cause. See [later evidence] below.
7. Wicked problems (pp. 6–7). - Problems “have grown into what Charles W. Churchman in 1967 termed wicked problems”. They are difficult to solve “because of incomplete, contradictory and changing requirements”, hard to recognise, “resistant to resolution because of the complexity of their interdependencies”, and need “many forms of social power” (p. 6). - “Solving them requires a new combination of hierarchical power, solidarity and individualism” (pp. 6–7). - Attribution: Churchman’s 1967 guest editorial in Management Science credited the term to Horst Rittel. The preface’s wording resembles common later paraphrases rather than Churchman’s text [general knowledge, unverified]. - The triad of hierarchy, solidarity and individualism echoes cultural theory’s “clumsy solutions” (my interpretation; the preface does not cite it).
8. Chemicals governance (p. 7). - It cites “the 100 thousand chemicals currently in commercial use” (p. 7). - “More conventions and treaties in place than a decade ago”: the 2004 Rotterdam Convention on Prior Informed Consent (PIC), covering “24 pesticides, four severely hazardous pesticide formulations and 11 industrial chemicals”, and the 2004 Stockholm Convention on persistent organic pollutants (POPs), covering DDT, PCBs, industrial chemicals and by-products such as dioxin (p. 7). The 2004 dates are when the conventions entered into force; the preface gives only the year [general knowledge, unverified]. - “But these conventions only address the top‑down hierarchical approach to power” (p. 7).
9. EU legislation “to achieve a global regulatory influence” (p. 7). - The instruments listed: the Cosmetics Directive’s ban on chemicals known or strongly suspected to be carcinogens, reproductive toxins or mutagens; RoHS, covering lead, mercury, cadmium, hexavalent chromium, PBBs and PBDEs, and encouraging “substitution to safe/or safer alternatives”; “the closely linked 2006 EU Waste Electrical and Electronic Equipment Directive”; the 2006 SAICM; and the 2007 REACH regulation, “to assign greater responsibility to industry to manage the risks from chemicals and to provide safety information on substances” (p. 7). - Verdict: these tools “once again point to the main economic actors rather than communities or individuals” (p. 7). - Date note: the original WEEE Directive is 2002/96/EC, so “2006” may refer to implementation dates [general knowledge, unverified].
10. Persistence, bioaccumulation and vulnerable windows (p. 7).
    • “One thing that has become clearer over the past decade” is that some chemicals are “highly stable in nature” with “long‑lasting and wide ranging effects”.
    • They bioaccumulate in fatty tissues, and predators (“polar bears, fish and seals”) biomagnify them “with devastating consequences for both humans and ecosystems”.
    • “So exposure to toxic chemicals and certain foodstuffs are at risk of causing harm”, especially to vulnerable groups: foetuses and children “when the endocrine system is being actively built”.
    • “Even with small dose exposures, the consequences can in some instances be devastating”, including cancer, developmental effects, chronic disease and learning disabilities.
    • “Here the power to act could be more properly set by well‑informed individuals and communities” (p. 7).
11. Knowledge and power as the core theme (p. 7).
    • “The relationship between knowledge and power lies at the heart of Volume 2.”
    • Many chapters expose “implicit links between the sources of scientific knowledge” and “strong vested interests, both economic and paradigmatic”.
    • Authors explore “the short‑sightedness of regulatory science”.
    • The aim is to illuminate “these normally invisible aspects” so that “communities and people” become “more effective stakeholders and participants in the governance of innovation and economic activities” (p. 7).
    • The phrase “vested interests… paradigmatic” is analytically important. It places vested interest inside science (commitments to prevailing theories and methods) as well as in commerce.
12. Extending lessons to finance, and history as insight (p. 7).
    • Lessons from past environmental and health mistakes are “directly applicable to the better regulation and governance of global institutions and financial and economic risks” (p. 7).
    • R. G. Collingwood (Autobiography, 1939) is quoted: history offers “insight” rather than “[scientific] rules”, and “we study history in order to see more clearly into the situation in which we are called upon to act” (p. 7).
    • This is a methodological self-description. The cases are meant to yield insight for judgement, not predictive laws. It matters for how far any “lesson” can be pushed.
13. Moving into contested territory (pp. 7–8).
    • Beyond “such widely accepted tragedies as leaded petrol, mercury poisoning in Japan’s Minamata Bay and older pesticides which sterilised many men”, the volume has “ventured into the uncertainties of potential yet contested harm”: GM products, nanotechnologies, “chemicals such as Bisphenol A”, new pesticides and mobile phones (pp. 7–8).
    • There is also “an examination of the 80 or so potential ‘false positives’”, cases where there were indications of harm “but where it was subsequently claimed that there were in fact no risks to prevent” (p. 8).
    • Number check: the Ch 2 summary says “In total 88 cases were identified to be alleged false positives” and that analysis “revealed four regulatory false positives” (p. 17). The preface’s “80 or so potential” cases therefore refers to the alleged cases, loosely rounded down, not to confirmed false positives.
    • The preface calls BPA an issue of “potential yet contested harm” (p. 8). The introduction lists it among nine “false negatives” (p. 10). The report’s own classification of BPA is inconsistent.
14. Framing (p. 8).
    • “A major part of effective decision‑making lies in the way issues are framed.”
    • On climate change, “US Vice President Al Gore chose to make the question a matter of choice between believers and sceptics”. The public was then asked “to make a scientific decision when too few people had the qualifications to make any kind of reasoned judgement”, which was “a false choice” (p. 8).
    • Instead, “the question should have been framed around which areas should people and governments make decisions and which should be delegated to experts” (p. 8).
    • The claim about Gore is unreferenced. The proposed reframing is a division of labour between democratic and expert decision, which sits in some tension with the preface’s own call to change “the power structures of knowledge” (p. 8).
15. Ostrom (p. 8).
    • “There are few certain and enduring truths” in the ecological, biological and social sciences.
    • One exception comes from “Elinor Ostrom, a late and widely missed colleague”: “complex problems can be solved if communication is transparent and open, visions are shared, trust is high and communities are activated to work from the bottom up as well as from the top down” (p. 8).
16. Closing (p. 8).
    • In “the Anthropocene”, more people must be involved in solving wicked problems, “through gathering local information or becoming more aware of the many uncertainties”.
    • “The power structures of knowledge will need to change.”
    • To respond “more responsibly to the early warning signals of change”, governance must be redesigned to reflect “a future defined by the local and specific rather than only the global and the average” (p. 8).
    • My reading, not stated in the text: “the average” may be a quiet critique of governance and risk assessment built on population averages, which would link back to the vulnerable groups on p. 7. The preface does not spell this out.

Chapter 1 Introduction (pp. 9–11)#

“Why further late lessons from early warnings?” (p. 9)

• Vol 1 (2001) “looked at the history of a selection of occupational, public health and environmental hazards and asked whether we could have been better at taking action early enough to prevent harm” (p. 9).
• The twelve lessons came “from cases where public policy was formulated against a background of scientific uncertainty and ‘surprises’ — and where clear evidence of hazards to people and the environment was often ignored” (p. 9).
• Analytical note: this sentence runs together two different failures. One is acting under genuine uncertainty. The other is ignoring clear evidence. The second is not a precaution problem in the strict sense. It is a failure to act on what is known.
• “The 14 case studies and 12 key lessons from the 2001 report remain highly pertinent today” (p. 9).
• Four reasons for Vol 2 (p. 9): 1. Expand coverage. Add “long‑known, important additional issues with broad societal implications” (lead in petrol, mercury, environmental tobacco smoke, DDT) and issues with more recent lessons (the contraceptive pill and fish feminisation; insecticides and honeybees). 2. Fill an acknowledged gap. Analyse false positives, “where government regulation was undertaken based on precaution but later turned out to be unnecessary”. Most cases are “‘false negatives’ — instances where early warnings existed but no preventive actions were taken”. 3. New society-wide challenges. Mobile phone radiation, GM products, nanotechnologies and invasive alien species: “if, how and where precautionary actions can play a role”. 4. Systemic challenges. How precautionary approaches can help manage “fast‑changing, multiple, systemic challenges”, and underpin “opportunities for sustainable innovations and, supported by information technologies, greater public participation in their selection”.

Overall approach (pp. 9–10). External case studies sit alongside chapters by editorial-team members. The selection advice, the five-part grouping, authors’ prior involvement, the expectation of objectivity, the seven structuring questions and peer review are covered under standpoint above.

Scope: five aims (p. 10). The report is designed, “inter alia” (so the list is not exhaustive), to help “politicians, policymakers and the public” to: - (i) “understand better the ways in which scientific knowledge is financed, created, evaluated, ignored, used and misused in taking timely and precautionary decisions about how to reduce harms, whilst stimulating benign innovations and generating useful employment”; - (ii) “learn from some very expensive ‘mistakes’ in the past”, especially “with some of the relatively new, largely unknown, yet already widespread technologies like nanotechnology and mobile phones”; - (iii) be aware of “the skewed ways in which the costs of actions and inactions for hazardous technologies have been estimated, and the role that some businesses have played in ignoring early warnings and in manufacturing doubt about the science supporting such warnings”; - (iv) consider how “the law, or administrative arrangements, could be better used to deliver justice, to those people (and ecosystems) that have been, or could be, harmed by poorly designed, or badly deployed, innovations”; - (v) “explore how best to engage the public in helping to make strategic choices over innovations, and their technological and social pathways”, in ecosystem management, and “in long term monitoring through citizen science”.

Notes on the scope: - My reading: the scare quotes around “mistakes” in (ii) suggest that not all past harms were innocent errors. The same sentence then uses “mistakes” without quotes (“make fewer mistakes now”), so the signal is modest. - The phrase “largely unknown, yet already widespread” in (ii) names a core structural problem: deployment has run ahead of knowledge. - Aim (i) presents precaution as compatible with “benign innovations” and “useful employment”. That pre-empts the objection that precaution costs jobs and innovation.

Part summaries (pp. 10–11)

• Part A (p. 10).
• It “commences with an analysis of ‘false positives’ showing that these are few and far between as compared to false negatives and that carefully designed precautionary actions can stimulate innovation, even if the risk turns out not to be real or as serious as initially feared”. This sentence repeats the Ch 2 summary almost word for word (p. 17).
• The other nine chapters are false negatives: lead in petrol, perchloroethylene (PCE) in mains water, Minamata disease, occupational beryllium disease, environmental tobacco smoke, vinyl chloride, dibromochloropropane (DBCP), bisphenol A and DDT.
• Three common themes are drawn from them: “there was more than sufficient evidence for much earlier action; slow and sometimes obstructive behaviour by businesses whose products endangered workers, the public and the environment; and the value of independent scientific research and risk assessments” (p. 10).
• Part B (p. 10). Degradation of natural systems: booster biocides, the pill and fish feminisation, climate change, floods, insecticides and honeybees, and ecosystem resilience. It considers “scientific evidence as the basis for action/inaction”, “the multiple, often complex factors and feedback loops in play, many of which are not fully understood”, and science–policy–society interfaces, “in the context of heightened systemic risks and substantial unknowns” (p. 10).
• Part C (pp. 10–11).
• Covers “newly emerging and large‑scale products, technologies and trends”, which “potentially offer many benefits but also potentially much harm to people and ecosystems and thereby ultimately economic development” (p. 10).
• The cases are Chernobyl and Fukushima, GM crops and agroecology, invasive alien species, mobile phones and brain tumours, and nanotechnologies (p. 10).
• “There is often little science, and very little direct hindsight, to assist in the management of these emerging technologies but the lessons from the historical case studies need to be applied if hazards are to be avoided” (p. 10).
• “The evidence from the chapters in Part C is that, by and large, societies are not making the most use of the costly lessons that can be gleaned from their histories” (p. 10).
• Reasons case studies identify for delayed action (pp. 10–11):
  • “the novel and challenging nature of the issues themselves”;
  • “poorly or inconsistently evaluated information”;
  • “strong opposition by the corporate and scientific establishments of the day”;
  • “the tendency by the decision‑making institutions, practices and cultures to favour the status quo and the short term perspective”.
• Part C also “illustrates the value of bottom‑up as well as top‑down approaches to innovations” for “the directions of technological pathways, the equitable distributions of benefits, costs and knowledge ownership, and the diversity of locally sensitive technological options” in the food, energy and ecosystem crises (p. 11).
• Part D (p. 11).
• “The historical chapters illustrate numerous harms which for the most part have been caused by irresponsible corporations.”
• This, “coupled with shortcomings in how decisions are made by governments on when to act on early warnings, and in the law when it comes to compensating victims of harm”, is analysed in three chapters.
• They offer insights on “how cost calculation methods can be improved”, on “how insurance schemes could be used to compensate future victims of harm”, and on “the reasons why businesses frequently ignore early warnings” (p. 11).
• Part E (p. 11).
• Governance implications for science, public policy and public engagement, aiming “to maximise the benefits of innovations while minimising harms”.
• Main insights: “science could be more relevant for precautionary decision‑making”; “the wider use of the precautionary principle can avert harm and stimulate innovation”; and late lessons and precautionary approaches are “highly pertinent to today’s multiple and inter‑connected crises — such as those arising from finance, economics, the use of ecosystems, climate change, and the use and supply of energy and food” (p. 11).
• “Finally, many of the historical and recent case studies illustrate the value of engaging the public in broadening the knowledge base and stimulating robust innovations” (p. 11).

Box: “Twelve late lessons” (p. 11), “Based on the case studies of Volume 1”. The EEA authorises reproduction with acknowledgement (p. 2). The lessons below follow the box’s wording. 1. Acknowledge and respond to ignorance, as well as uncertainty and risk, in technology appraisal and public policymaking. 2. Provide adequate long‑term environmental and health monitoring and research into early warnings. 3. Identify and work to reduce “blind spots” and gaps in scientific knowledge. 4. Identify and reduce interdisciplinary obstacles to learning. 5. Ensure that real world conditions are adequately accounted for in regulatory appraisal. 6. Systematically scrutinise the claimed justifications and benefits alongside the potential risks. 7. Evaluate a range of alternative options for meeting needs alongside the option under appraisal, and promote more robust, diverse and adaptable technologies so as to minimise the costs of surprises and maximise the benefits of innovation. 8. Ensure use of “lay” and local knowledge, as well as relevant specialist expertise in the appraisal. 9. Take full account of the assumptions and values of different social groups. 10. Maintain the regulatory independence of interested parties while retaining an inclusive approach to information and opinion gathering. 11. Identify and reduce institutional obstacles to learning and action. 12. Avoid “paralysis by analysis” by acting to reduce potential harm when there are reasonable grounds for concern.

Two notes on the box: - The source line misdates Vol 1 as “1986–2000” instead of 1896–2000 (p. 11), a typo. - The front matter reprints Vol 1’s lessons without updating them. No new consolidated list of Vol 2 lessons appears here.

Part A introduction (pp. 13–16)#

• p. 13. Divider page: “Part A — Lessons from health hazards” and a photo credit (© iStockphoto/AVTG). There is no introductory prose. Part A’s substantive framing is the Ch 1 paragraph on p. 10.
• Part A contents (pp. 14–15). The section headings signal each chapter’s argument. They are signals only: I have not read the chapters, which belong to other notes.
• Ch 2. Sections cover “‘False alarms’, ‘regulatory abuse’ and ‘regulatory false positives’” (2.1), “Mistaken false positives” (2.3), “Identified false positives” (2.4), “Swine flu” (2.5) and “Food irradiation and consumer health” (2.6) (p. 14).
• Ch 3. “‘Progress’ or precaution?” (3.4), “Lead contamination is ‘normal and safe’” (3.5), “1966: US Congress asks awkward questions” (3.6), “Lead in petrol poisons catalytic converters — so it’s got to go” (3.7) and “Public funding to study lead poisoning in children” (3.8) (p. 14). Heading 3.7 suggests the phase-out was driven partly by a technological incompatibility, not by health evidence alone.
• Ch 4. “Foreseeable harm?” (4.2) and “The view from 2013” (4.6) (p. 14).
• Ch 5. “Chaos implementing the Minamata disease accreditation system (1968 to present)” (5.4) (p. 14). This points to a long tail of disputes over recognising victims.
• Ch 6. “The science and its use” (6.3) and “Lessons for policymakers” (6.4) (p. 14).
• Ch 7. “Tobacco industry strategies to subvert scientific knowledge” (7.3) (p. 15).
• Ch 8. “A veil of secrecy” (8.1), “1930–1961: early warnings from animals and humans meet industry indifference” (8.3) and “1974: governments swift response” (8.4) (p. 15). The 8.4 heading suggests that once the hazard was acknowledged, the response was fast.
• Ch 9. “The discovery 1977: ‘our union members are sterile’” (9.1), “Pesticide registration and inadequate ‘hazard control’ 1961–1977” (9.3), “DBCP exports from 1969 to the 1980s: spreading sterility?” (9.7) and “Banana workers bring compensation cases: 1990–2010” (9.8) (p. 15). These point to workers as the ones who detected the harm, to hazards exported after domestic action, and to compensation through the courts.
• Ch 10. “Identifying the risk was an accident, not the result of a regulatory process” (10.3), “Bisphenol beyond Paracelsus” (10.4), “The time makes the poison” (10.5), “Concern or no concern: that is the question” (10.6) and “Spheres of influence” (10.10) (p. 15). These signal a challenge to the dose-makes-the-poison paradigm and a focus on the timing of exposure.
• Ch 11. “Current uses” (11.3) and “The legacy of Rachel Carson” (11.7) (p. 15). DDT is still in use, so it is not a simple closed “false negative”.
• Panel list (p. 16). Tabulated above. Four further commentaries (pp. 121, 125, 130, 145) are in the contents but not the panel list.

Case timeline (not a case study: project and context chronology from this section)#

This section is front matter, not a case. The chronology below covers the dates the section itself uses to set context. The next table records how the section classifies the Part A cases.

How the introduction classifies the Part A cases (p. 10; chapter authors from p. 3):

The lag between warning and action cannot be assessed from this section. The introduction presents “there was more than sufficient evidence for much earlier action” as one of three “common themes” that “emerge” from the nine false-negative chapters (p. 10). It does not say explicitly that the theme holds equally for every case. The case notes (LL2-02 to LL2-11) need to test that claim against each chronology.

The authors’ own lessons and conclusions#

A. Findings presented as derived from the case evidence#

1. False positives are rare. They are “few and far between as compared to false negatives”, and “carefully designed precautionary actions can stimulate innovation, even if the risk turns out not to be real or as serious as initially feared” (p. 10, drawn from Ch 2).
2. Three themes run through the nine health-hazard false negatives (p. 10): - there was “more than sufficient evidence for much earlier action”; - businesses behaved slowly and sometimes obstructively; - independent research and assessment proved valuable.
3. Societies “by and large” are not using the costly lessons of history in managing emerging technologies (p. 10).
4. Four reasons for delayed action (pp. 10–11): - the issues were novel and challenging; - information was poorly or inconsistently evaluated; - corporate and scientific establishments opposed action; - institutions favoured the status quo and the short term.
5. Harms “for the most part” were caused by “irresponsible corporations”, compounded by governments’ decision practices and the law’s failure to compensate (p. 11).
6. Persistence, bioaccumulation and vulnerable windows. Stable chemicals bioaccumulate and biomagnify, and small doses during development can be devastating, as “has become clearer over the past decade” (p. 7).
7. “Misplaced certainty about the absence of harm” has delayed prevention and is part of why scientific elites are losing support (p. 6). The preface asserts this; it does not derive it from evidence.
8. The 14 Vol 1 cases and 12 lessons “remain highly pertinent” (p. 9).

B. Recommendations and advocacy#

• A “more ethical form of public decision‑making” (p. 6).
• A “new combination of hierarchical power, solidarity and individualism” for wicked problems (pp. 6–7).
• Power to act on chemical exposures “more properly set by well‑informed individuals and communities” (p. 7).
• Making the knowledge–power links visible, so communities can take part in governing innovation (p. 7).
• Framing questions around which decisions belong to people and governments and which to experts (p. 8).
• Following Ostrom’s conditions: transparency, shared visions, trust, and action from both the bottom up and the top down (p. 8).
• Changing “the power structures of knowledge”, and designing governance for “the local and specific rather than only the global and the average” (p. 8).
• Applying historical lessons to emerging technologies “if hazards are to be avoided” (p. 10).
• Improving methods for calculating costs, and using insurance schemes to compensate future victims (p. 11).
• Wider use of the precautionary principle (p. 11).
• Public engagement in strategic choices and citizen-science monitoring (pp. 10–11).
• Applying these lessons to finance and the economy (pp. 7, 11).
• The twelve Vol 1 lessons (p. 11), which are themselves prescriptions.

Distinguishing A from B: in this section even the “findings” are summaries of chapters, and the evidence behind them is not presented here. Items A1–A5 are claims whose strength depends on Ch 2–25. Item A7 is the preface’s own hypothesis and has no supporting evidence at all.

Mechanisms and dynamics#

Only mechanisms the section itself names or clearly implies are included.

1. Power shapes knowledge production. The organising idea of Vol 2 is that “the relationship between knowledge and power” determines what becomes known and acted on (p. 7). - Scientific knowledge is “financed, created, evaluated, ignored, used and misused” (p. 10). Each stage is a point where interests can act. - Funding comes first in that list. My inference, not stated in the section: who pays for research shapes which questions are asked (compare Vol 1 lessons 2–3 on monitoring and blind spots, p. 11).
2. Paradigmatic vested interests. Vested interests are “both economic and paradigmatic” (p. 7). Opposition to early warnings comes from “the corporate and scientific establishments of the day” (p. 11). - Resistance is not only commercial. Scientific communities invested in prevailing assumptions and methods can also discount anomalies. - The Part A contents hint at examples, such as the challenge to “dose makes the poison” in Ch 10, “Bisphenol beyond Paracelsus” (p. 15).
3. Expert overconfidence: “misplaced certainty about the absence of harm” (p. 6). The preface describes certainty that there is no harm, held “despite evidence to the contrary”, which “has delayed preventive actions” (p. 6). So the failure it names is holding on to safety claims against contrary evidence, not only over-reading thin evidence. My gloss: confidence in safety out of proportion to what is known. Vol 1 lesson 1, distinguishing ignorance from uncertainty and risk, is the prescribed corrective (p. 11).
4. Manufactured doubt and business obstruction. Some businesses have played a role “in ignoring early warnings and in manufacturing doubt about the science supporting such warnings” (p. 10). Part A shows “slow and sometimes obstructive behaviour by businesses” (p. 10). Panel 7.1 on the tobacco industry’s role in shaping risk-assessment rules (p. 16) points to influence over the rules of evidence themselves, not just the evidence.
5. Institutional inertia and short-termism. Decision-making institutions tend “to favour the status quo and the short term perspective” (p. 11). The preface speaks of “the short‑sightedness of regulatory science” (p. 7) and political systems “silted up by vested interests” (p. 6). Vol 1 lesson 11, on institutional obstacles to learning, is the corrective (p. 11).
6. “Paralysis by analysis” and the standard for action. Vol 1 lesson 12 sets the trigger for action at “reasonable grounds for concern” (p. 11), below proof of harm. The implicit dynamic is that demanding ever more evidence before acting becomes a way of not acting. The section never defines what makes grounds “reasonable”.
7. Deployment runs ahead of knowledge. Technologies can be “relatively new, largely unknown, yet already widespread” (p. 10). For emerging technologies “there is often little science, and very little direct hindsight” (p. 10). Widespread exposure accumulates before the evidence base exists.
8. Persistence, accumulation and irreversibility. Stable compounds persist and bioaccumulate “at concentrations many times higher than in the surrounding environment”, and biomagnify up food chains (p. 7). Harm can be locked in long after emissions stop, which raises the cost of waiting.
9. Vulnerable windows and the tyranny of the average. Foetuses and children are singled out as vulnerable during development (p. 7). Governance must move beyond “the global and the average” (p. 8). Protections calibrated to the average can miss the groups most at risk.
10. Who bears costs and who gains.
    • The preface opens with “corrosive pathologies of inequality” in access to a safe environment (p. 6).
    • Costs of action and inaction have been estimated in “skewed ways” (p. 10).
    • Harmed people and ecosystems often lack legal redress (pp. 10–11).
    • Innovation raises questions of “equitable distributions of benefits, costs and knowledge ownership” (p. 11).
    • The DBCP headings (p. 15) point to hazards borne by workers and by export markets.
11. The form of governance limits what it can see.
    • International conventions “only address the top‑down hierarchical approach to power” (p. 7).
    • EU instruments, REACH included, “point to the main economic actors rather than communities or individuals” (p. 7). REACH places responsibility for risk information on industry (p. 7).
    • The implicit argument: regimes that work through large producers and states are blind to local, specific and community-level knowledge and harm.
    • The alternatives offered are the triad of hierarchy, solidarity and individualism (pp. 6–7), Ostrom’s bottom-up plus top-down model (p. 8), and lay and local knowledge (lesson 8, p. 11).
12. Framing decides outcomes. “A major part of effective decision‑making lies in the way issues are framed” (p. 8). The example is a binary of belief versus scepticism on climate change, which pushes a technical judgement onto the public and produces “a false choice” (p. 8). The report’s own vocabulary is also a framing choice:
    • “false positives” and “false negatives”;
    • “mistakes” in scare quotes (p. 10);
    • “surprises” in scare quotes (p. 9);
    • “benign innovations” (p. 10);
    • “irresponsible corporations” (p. 11);
    • “hubris” (p. 6).
13. Trust dynamics. The preface hypothesises a feedback loop: overconfident safety claims lead to delayed prevention, then harm, then loss of trust in science (p. 6). It places this within a wider collapse of trust in finance, business, politics and media (p. 6).
14. Precaution and innovation.
    • The claim: “carefully designed” precaution “can stimulate innovation” even when the risk proves unfounded (p. 10), and wider precaution “can avert harm and stimulate innovation” (p. 11).
    • The implied mechanisms: substitution pressure (RoHS “encourages the substitution to safe/or safer alternatives”, p. 7), examining alternatives (lesson 7, p. 11), and innovation steered by diverse, locally sensitive options (p. 11).
    • Two panel titles point to alternatives that already existed but were neglected or later adopted: the “alcohol alternative to lead in 1925” (panel 3.1) and wet-cleaning instead of PCE (panel 4.2) (p. 16).
    • The mechanism is asserted here and is not tested in this section.
15. Independence of science and assessment. “The value of independent scientific research and risk assessments” is one of the three Part A themes (p. 10). Vol 1 lesson 10 asks for “regulatory independence of interested parties” while keeping inclusive information-gathering (p. 11). Panel 4.1, on different risk assessments drawn “from the same basic data” (p. 16), implies that judgement and framing, not data alone, drive conclusions.
16. Complexity and systemic risk.
    • “Wicked problems” with interdependent, changing requirements (p. 6).
    • Ecosystem “feedback loops… not fully understood” (p. 10).
    • “Multiple and inter‑connected crises” across finance, climate, energy and food (p. 11).
    • The preface lists “mis‑engineered protection against risks and uncertainty” among the causes of the 2007–2009 financial crises (p. 6). It later says lessons from environmental and health histories are “directly applicable” to governing “financial and economic risks” (p. 7). The analogy is implied by putting the two together, not drawn explicitly.
17. History as a source of insight rather than rules. Collingwood (p. 7) supplies the method: historical cases sharpen judgement about present situations; they do not supply algorithms.

Transferable insights (technology-neutral)#

Strength ratings apply to the support in this section. Most claims here summarise chapters whose evidence sits elsewhere, so ratings are provisional until those chapters are read.

1. Separate two failure modes: deciding under genuine uncertainty, and failing to act on strong evidence. - Evidence: Vol 1 cases involved both “scientific uncertainty and ‘surprises’” and cases where “clear evidence… was often ignored” (p. 9). “More than sufficient evidence for much earlier action” is one of three common themes the introduction draws from the nine Part A false negatives (p. 10). - Implication: many “late lessons” concern the institutional capacity to act on what is already known, not precaution in the strict sense. - Strength: moderate. The distinction follows directly from the report’s own framing; whether each case fits is for the chapters to show.
2. How knowledge is financed, produced, evaluated and ignored is a site of power. Economic and “paradigmatic” interests shape it. - Evidence: pp. 7, 10 (scope i); p. 11 (“corporate and scientific establishments”). - Strength: asserted here. It is the organising thesis, and its evidence is in the Part A chapters, especially Ch 3, 6, 7 and 8.
3. Confident claims that there is no harm, held “despite evidence to the contrary” (p. 6) or while evidence is incomplete, are a recurrent failure. They may also erode trust in expertise. - Evidence: p. 6. - Strength: asserted for both parts. The failure mode recurs in the report’s cases; the trust mechanism is unevidenced, and later survey data do not show a broad collapse (see bias check).
4. When deployment outpaces knowledge, exposure accumulates before the evidence arrives. Learning then happens on an exposed population. - Evidence: p. 10, “relatively new, largely unknown, yet already widespread”; “little science, and very little direct hindsight”. - Strength: asserted here. It is plausible and is the premise of Part C.
5. Persistence, accumulation and exposure in vulnerable windows turn delay into long-lived or irreversible harm. Protections set for the average population can miss the groups most at risk. - Evidence: p. 7 (bioaccumulation, biomagnification, foetal and childhood windows, small-dose effects); p. 8 (“the local and specific rather than only the global and the average”). - Strength: moderate. Bioaccumulation and developmental vulnerability are well-established science [general knowledge]; the section itself cites no evidence. “Irreversible” is my inference from “long‑lasting” effects (p. 7). The low-dose claim is hedged (“in some instances”) and was strengthened for BPA by later assessment (see bias check).
6. Recurring causes of delay: - issues are novel; - information is poorly or inconsistently evaluated; - incumbent corporate and scientific establishments oppose action; - institutions are biased towards the status quo and the short term. - Evidence: pp. 10–11. - Strength: moderate. It is presented as a cross-case synthesis consistent with Vol 1 lessons 4 and 11 (p. 11), but the cases were selected as failures.
7. Producers of a hazard can delay or obstruct, including by manufacturing doubt about the evidence. Influence can extend to the rules of risk assessment itself. The section’s own wording is hedged: “slow and sometimes obstructive behaviour” and “the role that some businesses have played” (p. 10). - Evidence: p. 10 (Part A theme 2; scope iii); panel 7.1 title (p. 16). - Strength: asserted here. Case-level strength has to come from the chapter notes (Ch 3, 6, 7, 8 especially). The generalisation on p. 11, that harms were “for the most part” caused by “irresponsible corporations”, is broader than anything shown in this section. It may underplay the role of states and public bodies. Guidotti’s commentary, for example, says the US DOE “initially abetted” the beryllium producer (p. 145). This is to be tested in the case notes.
8. Independent research and assessment, and regulators free of interested parties, matter. - Evidence: p. 10 (Part A theme 3); lesson 10 (p. 11); panel 4.1 title, “same basic data”, different conclusions (p. 16). - Strength: moderate. It recurs across Vol 1 and Vol 2 syntheses; the evidence is in the chapters.
9. The false-alarm objection to precaution is weaker than often claimed. - Evidence: p. 10; p. 8 (“80 or so”); Ch 2 summary: 88 alleged cases, 4 confirmed (p. 17). - Strength: suggestive. It rests on one chapter’s classification, led by an editorial-team member (Hansen) with Tickner. False negatives have no comparable sampling frame, so “few and far between as compared to” has no common denominator. It needs scrutiny in LL2-02.
10. Precaution can stimulate innovation, especially through substitution and a wider search for alternatives.
    • Evidence: p. 10; p. 11; p. 7 (RoHS substitution); lesson 7 (p. 11); panel titles on existing alternatives (p. 16).
    • Strength: suggestive. Illustrative examples exist; the general claim is asserted and not tested in this section.
11. Cost accounting is itself contested. How the costs of acting and of not acting are estimated can be “skewed”.
    • Evidence: p. 10 (scope iii); p. 11 (Part D).
    • Strength: asserted here; Ch 23 carries the evidence.
12. Those harmed are often left without remedy, and legal and administrative arrangements lag behind. Mechanisms such as insurance for future victims are proposed.
    • Evidence: p. 10 (scope iv); p. 11; the Minamata “accreditation system” heading, “1968 to present” (p. 14); DBCP compensation cases, 1990–2010 (p. 15).
    • Strength: asserted here; Ch 5, 9 and 24 carry the evidence.
13. Regimes that work through large producers and states are structurally blind to local, specific and community knowledge and harm. Combining top-down and bottom-up approaches, and using lay and local knowledge, is proposed.
    • Evidence: p. 7; p. 8 (Ostrom); p. 11 (lessons 8–9).
    • Strength: suggestive. Ostrom’s evidence is strong for local common-pool resources; carrying it over to hazards from widely distributed technologies is asserted, not shown.
14. How a problem is framed determines who is asked to decide what. A binary of belief versus disbelief can push technical judgements onto publics and value judgements onto experts.
    • Evidence: p. 8.
    • Strength: asserted. The single example is unreferenced, and the preface’s proposed fix (delegating to experts) sits uneasily with its own call for participation.
15. Treat history as a source of insight, not rules. Lessons inform judgement about a present case; they do not predict it.
    • Evidence: p. 7 (Collingwood).
    • Strength: asserted as the report’s methodological stance. It limits how mechanically any lesson should be applied.
16. Twelve standing procedural safeguards (lessons 1–12, p. 11):
    • distinguish ignorance from uncertainty and risk;
    • monitor long term;
    • look for blind spots;
    • bridge disciplines;
    • test in real-world conditions;
    • scrutinise claimed benefits as hard as risks;
    • compare alternatives and favour diverse, adaptable options;
    • use lay and local knowledge;
    • take account of the values of different social groups;
    • keep regulators independent;
    • remove institutional obstacles to learning;
    • act on reasonable grounds for concern.
    • Strength: moderate. They derive from 14 Vol 1 cases (assessed in the LL1 notes), are mostly procedural and modest, and are not re-tested here.

Limitations, contestation and bias check#

Where the section is advocacy rather than analysis#

• Explicitly normative aims. The preface opens with a moral and political diagnosis (“profoundly wrong”, “adulation of wealth”, “extraordinary hubris”, p. 6) and declares ethical aims for the volume (p. 6). That is transparent, but readers should treat the synthesis as argument.
• Sweeping generalisations:
• harms “for the most part” caused by “irresponsible corporations” (p. 11);
• late lessons “directly applicable” to financial governance (p. 7);
• wider precaution “can avert harm and stimulate innovation” (p. 11).

None is supported with evidence in this section. - Unreferenced empirical claims: - declining public support for “scientific elites” and its cause (p. 6); - “recent evidence from social psychology” on trust (p. 6); - the characterisation of Al Gore’s framing (p. 8).

Case selection and design#

• Sampling on failure. The only stated criterion for the new Part A cases is that they are “long‑known, important additional issues with broad societal implications” (p. 9). Advice came from the editor, editorial team, advisory board, EEA Scientific Committee and the Collegium Ramazzini, of which only the last is described with a protective mission (p. 9). The cases are widely regarded as failures, and the finding that “there was more than sufficient evidence for much earlier action” (p. 10) is likely to be partly built into how they were chosen. The false-positives chapter is the report’s attempt to correct for this, which is to its credit.
• The false-positive comparison. “Few and far between as compared to false negatives” (p. 10) compares 4 confirmed false positives out of 88 alleged (Ch 2, p. 17) with a set of false negatives that is neither enumerated nor sampled in the same way. There is no common denominator. The classification was done by an editor (Hansen) and Tickner, an alternatives-assessment specialist at UMass Lowell’s Lowell Center for Sustainable Production (Annex 1, p. 699) who is widely known for work advocating the precautionary principle [general knowledge, unverified]. Ch 2 sorts the non-confirmed cases into “real risks”, “‘the jury is still out’”, “unregulated alarms” and “risk‑risk trade‑offs” (p. 17). How cases were assigned to these categories determines the headline ratio. This needs detailed testing in LL2-02.
• Authors as protagonists. Authors were chosen for prior involvement (pp. 9–10), and many were researchers, expert witnesses or regulators in these very histories (Annex 1). The benefit is deep expertise and access to archives. The risk is that a partisan account goes unchallenged. Judging by author affiliations, industry and regulators defending past decisions have no voice in Part A’s panels. Guidotti’s commentary (p. 145) appears, from its first page, to be the main complicating voice on motive, and Castaño’s warns against exaggerating mercury risk (p. 130). Both are absent from the panel list.
• Opaque review. Reviewers are unnamed; review was conducted “within a set of editorial guidelines provided by the EEA” (p. 10). The “seven structuring questions” (p. 10) are not given, so readers cannot see how the questions may have steered chapters towards “lessons”.

Hindsight#

• Judging that evidence was “more than sufficient” (p. 10) is done after the fact. The section does not address the prospective problem of telling which of many early signals deserve action.
• Lesson 12’s “reasonable grounds for concern” (p. 11) gestures at a threshold but never defines one. A fair reader would want operational criteria for what makes a warning credible.

Internal inconsistencies and small factual slips (minor, but they show looseness in the front matter)#

• BPA is “potential yet contested harm” in the preface (p. 8) but a “false negative” in the introduction (p. 10).
• The preface’s “80 or so” potential false positives (p. 8) loosely rounds down Ch 2’s 88 alleged cases (p. 17). This is imprecision, not a contradiction.
• The twelve-lessons box dates Vol 1 “1986–2000” (p. 11).
• The contents page numbers for the back matter are wrong by about 30 pages (p. 4).
• The WEEE Directive is dated 2006 (p. 7) [general knowledge, unverified: original 2002/96/EC].
• “Wicked problems” is credited to Churchman; he credited Rittel [general knowledge, unverified].
• Expertise versus participation. The preface’s framing fix (decide which questions to “delegate to experts”, p. 8) pulls against its call to change “the power structures of knowledge” and to empower communities (pp. 7–8). The report does not resolve this.

Counter-arguments acknowledged or ignored#

• Acknowledged:
• the false-positive objection, via Ch 2 (pp. 8–10);
• the thinness of evidence for emerging technologies: “little science, and very little direct hindsight” (p. 10);
• the contested status of some harms (p. 8);
• risk–risk trade-offs, implicitly, in the Castaño fish commentary (p. 130) and DDT’s “Current uses” (p. 15);
• the risk of over-warning, in Castaño’s call not to “exaggerate the risks” (p. 130), outside the front matter proper.
• Not addressed in the front matter:
• the opportunity costs of precaution, such as benefits forgone or delayed;
• risk–risk trade-offs as a general problem, as opposed to a category in Ch 2;
• regulatory capture of precaution itself, for protectionism;
• the case that some harms were driven by public programmes or consumer demand rather than corporations;
• evidence that regulation can hinder as well as stimulate innovation.

Fairness in the other direction#

• The report is open about its values. It includes a false-positives analysis that it did not have to run. It also printed Guidotti’s commentary, adapted from “a critical review prepared for the European Environmental Agency” (p. 145), which reads corporate motive less harshly than the chapter beside it. “Critical review” here may simply mean a scholarly review, not a dissent.
• The Vol 1 lessons it reprints (p. 11) are mostly procedural and modest: monitoring, scrutinising benefits, comparing alternatives, independence, lay knowledge. The trigger for action, “reasonable grounds for concern”, is a measured standard, not a demand for proof of safety.
• Most of the Part A false negatives (lead, second-hand smoke, vinyl chloride, beryllium, Minamata, DBCP) are widely accepted as regulatory failures by mainstream public-health history [general knowledge, unverified]. The thrust of Part A is not fringe.
• The preface does not only blame corporations. It also points to “a determination by citizens to protect assets accumulated in easier times” (p. 6), and the introduction lists causes of delay that involve no bad faith (novelty; poorly evaluated information; pp. 10–11).
• The Collingwood framing (p. 7) is itself a caution against over-generalising.

Later evidence on checkable claims in this section#

• Trust in science (p. 6).
• Source: Cologna, Mede et al., “Trust in scientists and their role in society across 68 countries”, Nature Human Behaviour, 2025 (doi:10.1038/s41562-024-02090-5; abstract read via Europe PMC).
• Method: a preregistered survey of 71,922 respondents.
• Finding: “in most countries, most people trust scientists”, and “there is no widespread lack of trust in scientists” (abstract re-checked during the audit). It also notes concern about distrust among a small minority.
• Implication: this does not support a general collapse of public support for science. The survey is a single cross-section, however, so it cannot test the preface’s claims about a decline over time or its cause.
• Number of chemicals (p. 7).
• Source: Wang, Walker, Muir and Nagatani-Yoshida, “Toward a Global Understanding of Chemical Pollution”, Environmental Science & Technology, 2020 (doi:10.1021/acs.est.9b06379; abstract via Europe PMC).
• Finding: “Over 350 000 chemicals and mixtures of chemicals have been registered for production and use, up to three times as many as previously estimated”. Over 50,000 identities are claimed as confidential, and up to 70,000 are ambiguously described.
• Implication: the preface’s “100 thousand” understates the scale, although “in commercial use” and “registered” are different measures. The confidentiality finding reinforces the preface’s point about knowledge and power.
• Rotterdam Convention coverage (p. 7). The Convention’s website (pic.int, Annex III page; re-checked during the audit) now lists 57 Annex III chemicals: 38 pesticides (including 4 severely hazardous pesticide formulations), 18 industrial chemicals and 1 in both categories. The latest additions (carbosulfan; ultra-low-volume fenthion formulations) entered into force on 22 October 2025. The preface’s figure was 39 (24 + 4 + 11). The coverage has grown, but slowly.
• Low-dose and vulnerable-group effects (p. 7), and BPA’s status (pp. 8, 10).
• Source: EFSA’s 2023 re-evaluation of BPA (EFSA Journal 2023; abstract via Europe PMC).
• Finding: EFSA set a tolerable daily intake (TDI) of 0.2 ng/kg bw per day, about 20,000 times lower than the 2015 temporary TDI of 4 µg/kg bw per day. The immune system was identified as the most sensitive endpoint, and dietary exposures “exceeded the TDI by two to three orders of magnitude”.
• EU follow-up: the EU then restricted BPA in food-contact materials (Commission Regulation (EU) 2024/3190) [general knowledge, unverified].
• Continuing disagreement: other assessors, notably Germany’s Federal Institute for Risk Assessment (BfR), disputed the magnitude of EFSA’s TDI [general knowledge, unverified].
• Implication: this supports the preface’s concern about low doses, and suggests the introduction’s “false negative” label for BPA was closer to where the EU ended up than the preface’s “contested”. The science was still contested in 2023.
• Precaution and innovation (pp. 10–11). The economics literature on the “Porter hypothesis”, for example the review by Ambec et al. (2013), finds reasonable support for the claim that regulation induces innovation. Support is weaker and mixed for the claim that it improves competitiveness [general knowledge, unverified]. The report’s claim is closer to the weaker, better-supported version.

Notable quotes#

1. “There is something profoundly wrong with the way we are living today.” (p. 6)
2. “The scientific elites have also been slowly losing public support. This is in part because of the growing number of instances of misplaced certainty about the absence of harm” (p. 6)
3. “The relationship between knowledge and power lies at the heart of Volume 2.” (p. 7)
4. “strong vested interests, both economic and paradigmatic, are exposed” (p. 7)
5. “History can offer something altogether different from [scientific] rules, namely insight.” (Collingwood, quoted p. 7)
6. “A major part of effective decision‑making lies in the way issues are framed.” (p. 8)
7. “a future defined by the local and specific rather than only the global and the average” (p. 8)
8. “relatively new, largely unknown, yet already widespread technologies” (p. 10)
9. “there was more than sufficient evidence for much earlier action; slow and sometimes obstructive behaviour by businesses whose products endangered workers, the public and the environment; and the value of independent scientific research and risk assessments” (p. 10)
10. “the tendency by the decision‑making institutions, practices and cultures to favour the status quo and the short term perspective” (p. 11)

Open questions#

1. The “seven structuring questions” (p. 10). What were they? They shaped every chapter and may have built lesson-drawing into the design. They may be in EEA project documents or the author guidance. Worth asking whether Andrew can retrieve them.
2. Peer review. Who reviewed which chapters, from what “broad range of perspectives” (p. 5), and were any industry or regulatory reviewers involved? Were any criticisms rejected?
3. Case selection. Which candidate cases were considered and rejected, and on what criteria? What role did the Collegium Ramazzini play in practice (p. 9)?
4. Operational criteria. How does the report define an “early warning”, “sufficient evidence” and “reasonable grounds for concern” (p. 11)? Is a prospective test offered anywhere, perhaps in Ch 26–28?
5. The false-positive comparison. Is there any defensible common denominator for comparing false positives and false negatives (p. 10)? How robust is Ch 2’s classification of its 88 cases? Have later studies revisited it?
6. The trust hypothesis (p. 6). Is there evidence that trust in scientific institutions falls specifically after exposures of “misplaced certainty” about safety? Longitudinal trust data (for example Eurobarometer series) could test this; Cologna et al. (2025) cannot.
7. Does precaution stimulate innovation? Which Vol 2 chapters actually document this, beyond assertion (pp. 10–11)? The candidates are Ch 2, panel 4.2 on wet-cleaning, panel 3.1 on the alcohol alternative, and Ch 25.
8. Who caused the harm? How does the claim of “irresponsible corporations” (p. 11) hold up against cases where governments, utilities or public-health programmes were central (PCE in water mains, Minamata’s government response, DDT for malaria, swine flu, and the US DOE’s early resistance on beryllium noted by Guidotti, p. 145)?
9. Did the report make a difference after 2013? Did it influence EU policy, for example on REACH implementation, the EU Chemicals Strategy for Sustainability (2020) or the handling of endocrine disruptors? Did the EEA revisit the project?
10. Resolving the tension between the preface’s expert-delegation framing (p. 8) and its call to redistribute epistemic power (pp. 7–8). Do the Part E chapters settle it?

Audit log#

Independent audit against the text extract (all of PDF pp. 3–18) and the PDF (Ch 2 summary p. 17; first pages of the commentaries on pp. 121, 125, 130, 145; “In memory of” pp. 683–684; Annex 1 pp. 685–700; back-matter pagination). Verified without change: all Annex 1 page citations and quoted bios except those noted below; the Ch 2 figures (88 alleged, 4 false positives, named cases, “some unintended deaths”); the back-matter page numbers; the twelve-lessons box; the quotes in “Notable quotes”; the Cologna et al. (2025) abstract and the Rotterdam Annex III count (both re-checked online).

• Institutional position: “adopt the chapters’ conclusions without qualification” softened to “with only light hedging”, with the hedges quoted (pp. 10–11).
• Editors as authors: noted Tickner is not on the editorial team; added that Ch 28 is unsigned and Ch 17 is by McGlade and van den Hove; added advisory-board overlap (Funtowicz, Ch 19; McIntyre).
• Collegium Ramazzini: quoted its full stated mission; replaced “exists to turn findings into protective action” with a description of its stated purpose; noted it is one of five named sources of advice; added the Soffritti and Huff links (Annex 1 pp. 698, 692).
• Objectivity: corrected the claim that no design feature supports objectivity. The report names the seven structuring questions and peer review as supports (p. 10); the point now is that there is no adversarial design.
• Part A authors table: Hansen’s editorial-team role re-cited to p. 5 (not in his Annex 1 entry); added Michaels’s 1998–2001 DOE Assistant Secretary role (Annex 1 pp. 694–695), a significant omitted standpoint fact given DOE’s role in the beryllium story; panel 7.1 page cite now includes Fooks (p. 689).
• “Pattern” paragraph: specified which authors were regulators (Michaels, Bingham, Monforton).
• Commentaries: added Guidotti’s Annex 1 affiliation (private consultancy, pp. 690–691); added that the piece calls itself a “panel”; extended its quotes and noted its statement that the US DOE “initially abetted” the company (p. 145); added Castaño’s caution against exaggerating risk (p. 130) and her treaty-delegation role (p. 688); qualified the “13 contributions” sentence with the affiliations of Guidotti, Castaño and Haigh.
• Evident stance: added the introduction’s causes of delay that involve no bad faith (novelty; poorly evaluated information).
• Reading and verification: listed the additional outside-range pages consulted.
• Pagination: “In memory of” (p. 683) is correctly paginated and holds the Harada and Harremoës memorials; the 30-page error starts at Annex 1.
• Acknowledgements: “about 90 contributing authors” corrected to 81 (counted from p. 5).
• Remembered: added the source spelling “Harremöes”, Harada’s “worked throughout his career on mercury pollution”, and p. 683 as a source for his death.
• Preface point 4: restored the omitted clause blaming “a determination by citizens to protect assets accumulated in easier times” (p. 6).
• Preface point 8: marked “entered into force” for the 2004 conventions as general knowledge.
• Preface point 10: restored “certain foodstuffs” from the source sentence (p. 7).
• Preface point 13: clarified that “80 or so” refers to alleged cases, loosely rounded from 88.
• Preface point 16: the “critique of population averages” reading is now marked as interpretation.
• Scope: added “inter alia” (the aims list is not exhaustive); the “mistakes” scare-quote reading is marked as interpretation and qualified.
• Twelve-lessons box: lessons 5 and 8 restored to the box’s exact wording.
• Part A contents, Ch 8: “response was fast” is now attributed to the heading.
• Timeline: 2004 “in force” and Ostrom’s 2012 death marked as general knowledge; Harada death re-cited to p. 683.
• Classification table: the beryllium row adds DOE’s role (p. 145); the DDT malaria point is sourced to Annex 1 (p. 687), not the front matter.
• Case-timeline note and Transferable insight 1: “in all nine cases” corrected. The source presents “more than sufficient evidence” as a common theme emerging from the nine.
• Authors’ lessons A2: “sufficient” restored to the source’s “more than sufficient”.
• Mechanism 1: “who pays shapes the questions” marked as my inference.
• Mechanism 3 and Transferable insight 3 (notes and digest): reworded to match the source. The failure named is certainty about safety held “despite evidence to the contrary”, not only over-reading thin evidence.
• Mechanism 16: “direct analogy” to finance corrected. The preface lists mis-engineered risk protection as a cause of the financial crisis (p. 6) and separately says the lessons are applicable to finance (p. 7).
• Transferable insight 5: the bioaccumulation science marked as general knowledge; “irreversible” flagged as inference.
• Transferable insight 7: dropped the unjustified “strong” rating; added the source’s own hedges (“sometimes”, “some businesses”); the claim that the generalisation underplays state actors is softened to a possibility, with Guidotti’s DOE statement as a pointer.
• Transferable insight 9: “editorial-team authors” corrected (only Hansen is an editor).
• Sampling on failure: replaced “selected… because they are known failures” and “bodies with protective missions” with the stated criterion and the actual list of advisers (p. 9).
• False-positive comparison: Tickner’s “pro-precaution research centre” label replaced with his Annex 1 description plus a general-knowledge note; listed all four Ch 2 categories (p. 17).
• Authors as protagonists: added Castaño alongside Guidotti as complicating voices left off the panel list.
• Inconsistencies: “80 or so” versus 88 recast as imprecise rounding, not a contradiction.
• Counter-arguments: added Castaño’s caution about over-warning.
• Fairness: “commissioned at least one critical review” reworded (the term may mean a scholarly review); “widely accepted failures” marked as general knowledge; added the preface’s blame of citizens and the causes involving no bad faith.
• Later evidence: noted that the Cologna abstract and the pic.int Annex III figures were re-checked during the audit.
• Open question 8: added the DOE beryllium example.
• Digest: matching fixes: “slow and sometimes obstructive”; “some producers”; “long-lasting” instead of “irreversible”; the Hansen/Tickner attribution; the “skewed” cost-estimate wording; the full list of advisers and the Ramazzini mission; Michaels’s DOE role; the unsigned Ch 28; the four omitted commentaries (Guidotti’s DOE point and Castaño’s caution); “not given in this section” for the structuring questions; the specific pagination error; a reworded expert-delegation tension; causes involving no bad faith; the citizens clause; and the Rotterdam figures.
• Contemporary-technology scan: no references to specific contemporary technologies or companies outside the source were found; nothing removed.