Peter Wills' Testimony To GM Commission
GE Commission , First Precautionary Evidence
GE Royal Commission Peter Wills Testimony Date: given on an (Auckland)formal hearing on 13/11/OO at the District court.
Testimony to: Royal Commission on Genetic Modification
by Peter R Wills
Associate Professor, Theoretical Biologist
Department of Physics, University of Auckland, Private Bag 92019, Auckland
On behalf of
Physicians and Scientists for Responsible Genetics
Greenpeace New Zealand Inc
Friends of the Earth (NZ)
Sustainable Futures Trust
Pacific Institute of Resource Management
(09) 373 7599 ext 8889
fax: 373 7445
Descriptions, analyses and reviewuations of genetic change and its consequences are, of necessity, expressed within a special scientific conceptual framework, but they depend on a number of questionable assumptions. The limitations and uncertainties inherent in our current knowledge of molecular biology, ecology and evolution severely constrain our ability to draw valid conclusions about the outcomes of genetic modification. This is so to the extent that we must regulate with the utmost caution any current human enterprise in this field.
Regulation of genetic engineering and its products in New Zealand, especially by ERMA and ANZFA, gives overriding weight to a scientific perspective without recognising its fundamental limitations. While formal opportunities for public participation have expanded, no fair balance has been achieved between competing interests in the questions at stake. Many people feel powerless to have any real influence over decisions that have the potential for enormous effect in their lives.
Scientific and the Maori worldviews present incommensurate approaches to questions concerning genetic engineering. The disjunction shows up as a confused competition between differing senses of the value of life, conflicting interests in control over domains of culture and Nature, and incompatible visions for the future. Resolution of problems concerning genetic engineering will require a rapprochement between disparate constructions of the questions involved. Failure to take real notice of the incompleteness of science and the integrity of Maori is hindering progress from both sides of the divide.
The ambition of the biotechnological governmental-industrial-academic complex to bring the processes of genetic change under global human control is driven and supported by the misrepresentation of scientists themselves as a morally neutral egalitarian community of scholars who work for the common good. Scant consideration has been given by the broader scientific community to the overall effects of the enterprise of genetic engineering and its capability of transforming biological reality beyond historical recognition.
1. Perspectives from Biology
1.1 The character of genes
1 It is impossible to discuss the question of genetic engineering without first coming to some understanding of what a gene is. The classical definition of a gene is as a unit of inheritance, specifying some characteristic of an organism.
2 Since the discovery of the structure of DNA and the elucidation of the processes molecular biological translation (protein synthesis), a gene has been taken to be either
(i) part of a DNA molecule, or
(ii) a sequence of nucleotides (A, G, C or T),
and it has been presumed that everything known previously about the processes of inheritance can now be understood in these terms.
3 This modern view is highly problematic, not only because of the ambiguity in the definition of the gene as matter (DNA) or as information (a sequence), but also because genes are not autonomous determinants of identifiable characteristics of organisms.
1.1.1 Material objects
4 If we take the view that a gene is a piece of DNA, then we will be concerned with its material origin and we may ask after the identity of the original, individual organism from whom the material used in any experiment was first acquired for sequencing or copying. There are marked cultural differences in the importance attached to the question of material origin.
5 The question of particular material identity, connection and continuity is of special significance to Maori, especially in relation to establishing the whakapapa of human trans-genes that have used in experiments in Aotearoa/New Zealand. By way of contrast, scientists conduct calculations of the simple stepwise dilution of the matter comprising any original sample. Material dilution occurs through succeeding generations of genetic replication whether in vivo or in vitro.
1.1.2 Informational sequences
6 If we take the view that a gene is a sequence of nucleotides, then it is information that can be transmitted and stored symbolically, by using electronic computers for example, and we are likely to be concerned with issues of intellectual property.
7 The extant connection between DNA sequences and the characteristics of organisms that sustain their stable existence within complex, interacting eco-systems has been established historically through four billion years of geographically dispersed evolution.
8 Claims that genetically engineered organisms are "inventions" amount to human arrogation of Nature's "prior art" of how genetic information can be used to record biological characteristics.
1.1.3 Operational view
9 The term "gene" is often used by molecular biologist to refer to a DNA sequence that encodes a protein product, taking into account that it may be fragmented in the actual genome of the organism (due to introns), that it may have an obligatory association with a promoter sequence and that it may contain unusual signals (e.g., for ribosomal frameshifting1). There is no simple universal set of rules relating genetic sequences to protein production in cells.
10 Molecular biology offers no convincingly detailed account of the connection between genetic information and organismic characteristics. The connection is treated as a quite arbitrary outcome of innumerable, unstructured historical events that have occurred and whose consequences have become entrenched during evolution. However, it is assumed universally that a firm, orderly connection exists.
11 What has been discovered by molecular biological investigation is not the causal connection between genes and characteristics but rather an appreciation of the manner in which minor genetic variation gives rise to the variation in organismic characteristics all else being equal.
12 The qualification "all else being equal" is constitutive of the definition of genes as information. The association between variation in a particular gene and variation in a corresponding trait, and thus the supposition that "gene X encodes characteristic y", depends on innumerable other contingencies which are themselves subject to arbitrary variation capable of annihilating the perceived association.2
1.2 The genotype-phenotype relationship
1.2.1 Complexities of the relationship
13 There is no simple one-to-one relationship between genes and biological traits.3
14 Many genes within the population of a species display polymorphisms, meaning there are different versions of the gene in different individuals (or different alleles of a single individual). Sometimes the effect of a polymorphism is clearly recognisable. In other cases it can appear to be "neutral".
15 Genes can display pleiotropy, meaning that a single gene may be related to multiple, seemingly independent characteristics of the organism.
16 Polygeny refers to the common observation that the statement of (or alterations in) more than one gene contributes to a single characteristic.
17 Many genes appear to be redundant. There can be multiple encodings, not necessarily identical, of a single protein product. However, in many cases the statement of redundant genes appears to confer some advantage on the organism, not just in respect of being able to lose a functional protein as a result of mutation and still survive.
1.2.2 Character of the relationship
18 The relationship between genetic information and its statement in terms of biological traits is exceedingly complex. It is the view of many theoretical biologists (including myself) that this relationship is irreducibly complex.4
19 An organism cannot be constructed from knowledge of its inherited (genetic) information alone. It seems likely that the amount of information needed about the process of construction (what you might call the "algorithm") is of the same order as the amount of information in the genes.
20 When thinking about manipulating the relationship between genetic information and its statement in biological characteristics, we must consider the effects our actions have on determinants of events that are not encoded in the genes that are being subjected to alteration.
1.2.3 Relevant research
21 Work I have done in two fields of theoretical biology,
(i) the replication of prions, and
(ii) the origin of genetic coding,
illustrates that assessments of the consequences of genetic change based on the accepted ideas of molecular biology and evolution (the "modern synthesis" of Darwinian natural selection and Crick's Central Dogma) face profound difficulties. The validity of such assessments are subject to very restrictive qualifications.
22 Work in progress in a further field,
(iii) ecological population dynamics,
casts doubt on the validity of some conclusions, such as those reached by genetic engineers and regulatory bodies like ERMA and ANZFA, starting from accepted principles.
1.3 Replication of Prions
23 Starting in the early 1980s I was a defender of Prusiner's idea, then held to ridicule.5 but now vindicated, that prions containing no nucleic acid, neither DNA nor RNA, are the aetiological agents responsible for the transmission of spongiform encephalopathies like ovine scrapie, "mad cow disease" and Creutzfeldt-Jakob Disease.
24 I defended Prusiner because it was clear from my own theoretical work, contrary to the thinking of the vast majority of molecular biologists, that proteins, like genes, could possibly act as semi-autonomous determinants of inherited biological characteristics
25 This possibility flew in the face of the Central Dogma of Molecular Biology which states that the flow of biological information is generally one-way, from DNA and RNA to protein, but more particularly, that "once information has got into protein it cannot get out again".
1.3.1 Proteins equal genes equal organisms?
26 The existence of forms of proteins (also dubbed "prions") that determine certain stable phenotypes of yeast and fungi by transmitting information from mother-cell to daughter-cell in proteinaceous form is now an established fact. In this context prions are considered to be genes.
27 Under one instrument of international law, the Biological Weapons Convention, prions are categorised as "organisms" following New Zealand's raising of the matter at my instigation.6 Prions are not considered to be organisms in terms of the HSNO Act.7
28 More generally, prions demonstrate the complete inadequacy, with the potential for devastating consequences8, of analyses of biological causation in terms of a stable and fixed "genotype-phenotype" relationship. Even the universal genetic code, which is taken as providing an unshakeable foundation in virtually all analyses of the consequences of genetic engineering, gives no guaranteed way of understanding what may follow from genetic change.
1.4 Origin of coding
29 Certain assumptions are inescapable in technical analyses of the consequences of genetic engineering. One of those assumptions is that there is a well-defined relationship between alterations to the genotype of an organism (the sequence of nucleotides in its DNA) and the manifestation of that alteration in the phenotype (identifiable characteristics of the organism).
30 For about 20 years I have concerned myself with the question of the molecular evolutionary origin and stability of this genotype-phenotype relationship.9
1.4.1 Determinants of the genotype-phenotype relationship
31 The main finding of work in the field of "the origin of genetic coding" is that the processes of biological self-organisation that guarantee the regularities and patterns that we observe in the genotype-phenotype relationship (thereby making genetic engineering per se possible),
(i) are not encoded in genetic information,
(ii) are fundamental to the origin and sustenance of all life, and
(iii) have a continuous evolutionary development longer than that of
32 In two recent papers,10 I have discussed some of the semiotic (as opposed to physical) conditions that have to be fulfilled in a system before it is able display functions typical of biological processes. Molecular studies of genes and the characteristics that they encode are inherently incapable of shedding any light on such matters.
1.4.2 Lessons for GE regulation
33 Our lack of knowledge and understanding of the diffuse, intertwined processes of self-organisation on which all of life depends creates an over-riding uncertainty in analyses of the consequences of genetic manipulation and undercuts any confidence expressed in the relative value of the technology.
34 According to the standards of judgment that society generally requires be applied in making assessments which affect its members, the confidence expressed by genetic engineers in their inventions is groundless. They have achieved a perception to the contrary by hiding molecular biology's fundamental ignorance of the intrinsic interconnectedness and integrity characterising the many modes through which genetic information influences the living world.11
1.5 Genes, ecology and evolution
35 Biological causation is such that dramatic effects can be very remote from the original change to which they can be attributed. A good example, the elucidation of which has been described by Ricard Solé,12 is the extinction of the large blue butterfly (Maculina arion) as a result of the introduction of myxoma virus to English rabbit warrens. Competition between different types of grass selectively eaten by rabbits and consequent changes in the habitat of ants, symbionts of the butterflies, is concluded to have mediated the extinction.
36 All of the régimes that have been invented for the regulation of genetic engineering and the assessment of its environmental consequences rely on being able to forecast events of this sort in complex, natural and domestic ecosystems. The example gives only a hint of the true complexity of ecosystems and their modes of response to perturbation.
37 In many submissions to the Environmental Risk Management Authority I have stated that we scientists have a profound lack of detailed knowledge of ecological interactions and the manner in which those interactions depend on genetic factors. Any guess concerning the scale of biological change that seemingly innocuous genetic manipulation may ultimately cause could be wrong by as many orders of magnitude as the difference the KT event of 65 million years ago (loss of 70% of all species, including the dinosaurs) and the advent of melanism in moths (an apparently minor adaptation generally attributed to industrial smog).
1.5.1 Dynamic character of ecosystems
38 While we have a tendency to think of ecological systems in terms of "stability", the idea is actually inappropriate to the way in which they function and change. All ecological systems are dynamic not only in terms of their constitution (perpetual variation in relative population numbers of different species) but also in terms of their adaptation to the ever-changing environment (both physical and biological).
39 On a very long evolutionary timescale, the only predictable feature of ecosystems is the constant pattern of adaptation and change. There are some simple regularities in the relative scale of different events. These regularities of scale characterise mechanisms of ecological and evolutionary change (speciation, extinction, displacement, etc ) that are mediated by genetic variation and statement.
1.5.2 Scaling in self-organised systems
40 It has become recognised that an enormous range of phenomena13 (from stockmarket fluctuations, through the geographical distribution of human population to major extinctions during evolution) that depend on the combination of numerous individual small-scale causal events displays the same sort of scaling:
events with consequences of relatively large magnitude occur with relatively low frequency but on a long timescale events of all magnitude occur inevitably with characteristic unpredictability and irregularity.
41 In the last few years we have begun to glimpse how the regularities of scale observed in ecological and evolutionary change are related to genetic variation.14 It is suspected that the maintenance of stable patterns of ecological and evolutionary change depends on the maintenance of stable patterns of genetic change. This is currently being investigated by theoretical biologists like Ricard Solé15 and myself.
1.6 Consequences of genetic changes
42 Of central importance in assessing the consequences of genetic engineering are considerations of:
(i) the character of the desired genetic alteration,
(ii) the process causing the alteration,
(iii) collateral genetic alterations
(vi) genetic network effects
(v) potential non-genetic ("environmental") changes.
1.6.1 Changes sought be genetic engineers
43 Genetic engineering can have many purposes. It is possible to engineer organisms so that they amplify genes, produce proteins or have new characteristics. The new characteristic sought may be defined in purely genetic terms (as in "gene knock-out" studies) or it may be some phenotypic characteristic.
44 Here I concern myself only with the consequences that ensue from genetically engineered organisms being released or escaping into the environment, ignoring for the moment problems of containment or ethics involved in the regulation of genetic engineering in the laboratory.
45 In the case of crops, for example, the change sought by the engineer may be herbicide resistance, vitamin synthesis, altered metabolism or some other characteristic. Usually, a new gene is inserted into cells of the organism in the expectation that a new protein will be produced in its cells (or synthesis of a normal protein suppressed) conferring the desired characteristic.
46 Intracellular statement of a protein with a particular function can often be associated quite unequivocally with some organismic characteristic, as in the case of the enzyme EPSPS16 and resistance to the herbicide glyphosate (Monsanto's Roundup).
47 It is impossible to know in advance all of the functions a protein may have when it is expressed in a new biochemical milieu. For that matter it is not possible to know all of its functions in its natural milieu. That is why extensive testing of genetically engineered organisms is needed to characterise their properties once they have been created.
1.6.2 The engineering process
48 Genetic engineering is exact in only one limited sense, that of the extent of our knowledge of the change in DNA sequence that is effected. This can be determined very accurately. In every other respect genetic engineering is a very haphazard process.
49 In almost all cases it has so far proved impossible to achieve any precision in the process of inserting new genes into the genomes of host species. Incorporation of transgenes usually occurs more or less at random, as far as we can determine.17
50 Even if perfect control over transgene insertion were achieved it would still be necessary to investigate the effects of the engineering process by characterising thoroughly the new organism.
1.6.3 Effects of genetic transposition
51 Transgene insertion can disrupt virtually every kind of genetic function in a cell: transcription, translation, promotion or suppression of statement, replication, recombination, etc. Furthermore, the effects of any disruption may not be particularly evident until the organism is placed in a particular environment.
52 In addition to potential disruption of normal genetic statement due to the presence of a transgene in the genome, statement of the transgene can alter existing cellular functions or give rise to new ones, quite apart from the desired change.
53 Extensive testing and selection of genetically engineered organisms is necessary - to find those on which the transgene has conferred optimally the desired characteristic without other deleterious effects.
1.6.4 Genetic network effects
54 Genes continually switch one another off in complex dynamic patterns that have so far not been characterised to any great extent and are not very well understood. There has been some theorising but there is precious little data against which to test theoretical predictions.
55 No gene's function can be defined in isolation from those of other genes that are simultaneously expressed. Ultimately the way in which a cell with a particular genetic makeup behaves under any circumstances depends on the complex response of its genetic network.
56 The insertion of a transgene is inherently capable of altering the characteristic dynamic states of a cell. These changes could be quite subtle and could have consequences that are difficult to detect but nevertheless of some importance to the internal functions of the organisms cells or the interaction of the whole organism with its environment.
1.6.5 Interactions with surroundings
57 Any change to the phenotype of an organism is likely to alter the manner in which it interacts with its surroundings, both its response to the physical environment and its ecological relationship to other organisms. The change can affect its rate of reproduction ("fitness") as well as that of other species.
58 Small changes in fitness can be of enormous evolutionary significance and may be mediated through some seemingly minor rearrangement within a habitat.
1.7 Effects of genetic engineering
59 There is no basis in either theory or observation for the assertion that patterns of genetic change effected by genetic engineering are of the same character, in respect of their effect on ecosystem dynamics and adaptation, as past evolutionary patterns of genetic change.
60 Genetic engineering changes the characteristics of organisms. That is the only reason for doing genetic engineering, even if the goal is art18 rather than medicine, pharmacy, agriculture, environmental modification or warfare. Assessing the ecological consequences of introducing an organism with new characteristics into the environment requires consideration of every function of the organism and the way in which each functionality is affected by the new characteristics that the organism has acquired.
61 The task is essentially impossible. First, not every functional effect of a genetic change can be detected and assessed. There are simply too many minor effects to be considered. Second, novel functional effects, not evident in the original organism, can emerge from cooperation between the change made and other functionalities already present. This is because there is not a simple one-to-one relationship between genes and traits.
1.7.1 Justification for genetic engineering
62 The main assumption of genetic engineering is that genetic changes made artificially are no different in character from those that occur naturally.
63 According to the genetic engineer's way of thinking, all of the restrictions and limitations imposed on genetic transposition by natural reproductive processes are purely arbitrary. Any gene could, at least in principle, end up in any organism. If the functional genes were all scrambled between species we would have a different set of organisms inhabiting the biosphere, but we would still have the same sort of well-adjusted world (after an appropriate settling-down period).
64 Genetic engineering produces new organisms with constellations of genes in arrangements that have never occurred before. In one sense, natural reproduction and evolution produce the same result. However, the results of natural reproduction and evolution are restricted by the limitations of the processes of genetic change that can take place as a result of mitosis, meiosis, mutation, selection, horizontal gene transfer and other natural mechanisms.
65 Genetic engineers argue that the natural limitations and restrictions on gene transfer are of no significance, that genes can be transposed with impunity, that only immediately recognisable functional effects of a genetic transposition are of significance. They argue that what is possible in Nature is restricted and limited only by extraneous incidental effects, what organisms can mate with what other organisms and so on.
1.7.2 Analogy to species transposition in ecology
66 In assessing the possible validity of this assumption as a basis for reviewuating the consequences genetic engineering it is reasonable to draw an analogy to the transposition of species across geographical boundaries that humans have effected, especially during the last few centuries.
67 The ecological damage caused by mammalian pests in New Zealand provides a prime example of the unrecognised hazards entailed in human actions. Stoats and other predators would not be any threat to our native birds if they had other, more readily available, desirable food and their numbers were adequately restricted by other ecological factors. However, within the functional context of our native bush habitat they have, over the course of a century or so, caused a disaster.
68 Adjustment to perturbations of the scale inflicted by human activity has amounted to a complete transformation of our natural ecosystems where they have not been completely destroyed. The evolution of new species comparable with those driven to extinction, if it can occur, will require times many orders of magnitude longer than the duration of the perturbation that caused the extinctions.
1.7.3 Comparison with selective breeding
69 It has often been argued that genetic engineering is no different from selective breeding (of crops and animals) that has been practised by human societies for millennia. It has even been argued that genetic engineering is safer than selective breeding because of the greater control which engineering offers over what genes actually end up in the new organism.
70 It is false to say that genetic engineering is a way of achieving more quickly what can be achieved by selective breeding. Evolutionary processes have led to results that are absolutely unique, on any cosmic scale of reckoning.19 and the means of genetic transposition through selective breeding and horizontal gene transfer are extremely limited.
71 One can only maintain that genetic engineering and selective breeding are the same on the a priori basis that the uniqueness of evolutionary precedents is of no cellular, organismic, ecological or evolutionary significance. In my judgment there is every reason to believe that the relationship between evolutionary precedents is what maintains order and structure at all levels of biology and is therefore of the utmost significance.
72 The argument concerning the safety of genetic engineering in comparison with selective breeding is similarly spurious. Any new hazards (whether to consumers or within the ecology of the organism in question) associated with the new constellations of genes that can arise as a result of selective breeding are limited to those that can propagate by means of the similarities between organisms that allow them freely to exchange genetic material - sexual reproduction essentially.
73 Genetic engineering places no restrictions on the new constellations of genes that can be created in an organism and then all of the normal means of propagation are available to disperse any new hazard throughout populations of species with which the engineered organism exchanges genes.
2. Issues of science and regulation
2.1 Regulation of genetic engineering in New Zealand
74 We have invented permissive regulatory charades that provide a semblance of rigour by concentrating on the obvious and immediately observable effects of the designed genetic change. Thus committees, authorities and commissions are easily able to persuade themselves that they are much wiser than people were a century ago and now they would never permit an action anything like the small-scale introduction of possums into New Zealand for commercial purposes. This is believed, in spite of our complete lack of experience of the longterm consequences of artificially causing radical changes to the genetic constitution of species.
75 The régime set up under the HSNO Act to regulate genetic engineering is very comprehensive, by international standards. It provides for extensive public participation. The implementation of these provisions of the Act by Environmental Risk Management Authority (ERMA) have been exemplary, including the establishment of a separate Maori advisory body, Nga Kaihautu Tikanga Taiao. On the other hand, the public, including Maori20, have had little influence on decisions of the Authority. Consultation with the public has been a matter of form rather than content.
76 The dominance of expert scientific opinion has been even more marked in the functioning of the Australia New Zealand Food Authority (ANZFA). The doctrine of "substantial equivalence" on which all of its decision-making is based precludes any consideration of broader implications of genetic engineering and alienates non-expert members of the public from the decision-making process. New Zealanders have been disenfranchised as a result of an international agreement that was entered into without Parliamentary approval.
2.2 Laboratory microbes
77 In New Zealand, as elsewhere, there is much more genetic engineering of microbes, by a factor of a hundred or perhaps a thousand, than of higher organisms. This activity has been conducted increasingly over the last two decades to the extent that it is now absolutely routine in molecular biology laboratories, schools even, all over the world.
78 Approval for the conduct of much of this genetic engineering is delegated by ERMA to Institutional Biosafety Committees. There are rules concerning the species of microbes that can be used for approved experiments and the manner in which material must be handled and disposed of.
2.2.1 Potential hazards
79 It is a matter for some concern that humans are introducing an extremely wide range of genes from diverse taxa into laboratory microbes. These microbes are "crippled" so that their chances of living in the wild should be virtually non-existent and live material is not supposed to be released into the environment. However, it is not at all inconceivable that genes introduced into laboratory microbes, or parts of them, could be transfected into wild microbes as a result of laboratory disposal. Possible dangers arising from such events were discussed in the mid 1970s by authors such as Chargaff21 and Sinsheimer.22
80 It is difficult to determine what impact the last two decades of human genetic engineering have had on the evolution of microbial flora. It would be foolish to think that there has been no impact. Human activity has made functional DNA sequences from other taxa available for transfection into wild microbes through pathways that have never been available before.
81 The scale of the potential problem bears relation to the frequency with which microbes are genetically engineered and the diversity of taxa from which transgenes are derived. The pattern of possible horizontal gene transfer into microbes has been changed, perhaps markedly.
82 I would advise that rules for the disposal of DNA from genetically engineered laboratory microbes be revised with a view to minimizing further the possibility of transfection into wild microbial flora. There are lessons to be learned from the evolution that microbes underwent in response to the widespread use of anti-biotics by humans.
83 Such a suggestion is not incompatible with the desire of laboratory scientists to be freed from unnecessary bureaucracy in respect of gaining permission for experiments deemed to be of low risk.
84 I would advise that if the procedures governing the granting of approvals by Institutional Biosafety Committees (IBSCs) for low-risk laboratory experiments are to be freed up, then the IBSCs should be constituted so that those making decisions and granting approvals should be much more independent of the scientists wanting to do the experiments.
2.3 Concerning ERMA
85 The HSNO Act requires ERMA to make decisions based on technical risk analysis. Sections 5 and 6 of the Act require that various ecological and social factors be considered, including the relationship of Maori to their taonga, and Section 7 requires that a 'precautionary approach' be adopted. Section 9 requires ERMA to develop a methodology for dealing with risks, weighing costs and benefits.
2.3.1 Treatment of "general concerns"
86 In practice, ERMA has been increasingly flooded with submissions opposing applications for field trials of genetically engineered organisms.23 Most of the reasons given for opposition to applications are described by ERMA as "other" because they do not fit into any of the categories that are afforded weight under their interpretation of the HSNO Act.24
87 I have argued in submissions that the weight of general concerns would justify the rejection of applications for field trials, but all such concerns have been relegated to the irrelevant periphery of the decision-making process.
88 In its decision25 on AgResearch's application for field trials of transgenic cattle the Authority ruled that the matters raised by submitters under Sections 5(b) and 7 of the HSNO Act were not particularly relevant to the conduct of research in containment. In its latest decision26 on AgResearch's application grow a flock of transgenic sheep, no consideration is given to general concerns except to note in relation to Section 5(b) of the HSNO Act, that the risks to New Zealand's 'clean green image', export relationships and organic farming "are negligible for current and future generations alike".
2.3.2 ERMA's permissiveness
89 Having satisfied itself that there are no hazards described in scientific, technical terms that cannot be mitigated by appropriate measures, ERMA has approved (with controls) every application that it has considered for field trials of genetically engineered organisms.
90 So permissive has ERMA been that it has given approval for field trials of genetically engineered organisms not even known to exist. Applications GMF98007/8 described hypothetical transgenic potatoes27 that had not been created, but the Authority considered them well enough defined in terms of their intended genetic constitution to approve field trials to be conducted, should they be successfully created.
91 ERMA's latest decision acknowledges that the applicant changed the emphasis of the purpose of the application (from agricultural to medical research benefits) after submissions were received and before the hearing took place. The Authority argues that it had to consider this change of emphasis as irrelevant, lest it "encourage applicant's (sic) to misrepresent potential benefits".28
92 The decision records no consideration of the possibility that the change in emphasis was disingenuously constructed to undermine the thrust of submissions opposed to the application. I have received anonymously information to indicate that in a previous case the same applicant (AgResearch) was less than frank with ERMA about the origins of and motivation for a similar project to create transgenic farm animals.
2.3.3 Gap in Jurisdiction
93 In relation to a legal argument concerning animal welfare that I raised in submissions, ERMA has acknowledged that there is a jurisdictional gap that needs to be addressed by government29. I had already raised this matter with the Prime Minister in relation to the Warrant of the Commission and the associated moratorium on applications for field trials.30
94 The point31 is that the creation of transgenic sheep is a matter that ERMA delegates to the AgResearch Biological Safety Committee, who are answerable to the Ruakura Animal Ethics Committee in relation to matters covered by the Animal Welfare Act 1999. However, that Act does not apply to foetuses during the first half of a term of pregnancy.
95 Thus, a genetic engineer has no responsibility for any suffering which the act of creating the animals causes. Such suffering, even if reasonably envisaged, concealed or ignored, is a fait accompli that confronts the relevant Animal Welfare Committee in the second half of gestation, or ERMA when an application comes forward for field trials.
96 I do not accept that the stop-gap measure,32 whereby ERMA assures the public that their concerns have been met by the relevant Animal Ethics Committee, is adequate.
2.4 Inadequacies of quantitative risk analysis
97 The methods of quantitative risk analysis that are recommended so highly for assessing potential problems with genetically engineered organisms are fraught with conceptual and practical difficulties. This is especially true when they are applied to events which are rare, poorly defined, or catastrophic on a large scale (worst case scenarios).
98 Genetic engineers and regulatory authorities tend to dismiss worst case scenarios as scare-mongering and ascribe them no credibility.33
2.4.1 Risk and hazard
99 Practitioners of risk analysis usually fail to make a proper distinction between "risk", which is the probability of an event occurring, and "hazard", which is the scope for harm entailed in conducting some activity. The best known applications of risk analysis to worst case scenarios are to nuclear catastrophes.
100 Comparisons are usually made by multiplying the risk by some assessment of the hazard (like number of deaths). According to such analyses a steady death rate of one person every decade from radiologically induced cancer within a given population is considered equivalent to a hundred thousand deaths from a large-scale disaster which has a probability of one in a million per year and, when it occurs, affects a large sector of the population living at that time.
101 Risk analysis is unable to give satisfactory measures of the absolute probabilities of different harmful events. It is now widely acknowledged that the multiplication of different probability factors gives estimates of low risks which are useful only for comparing relative probabilities in closely similar situations. This is relevant to the use of the Brenner scale to assess the risks associated with the production of transgenic micro-organisms.
102 A claim of the form "The actual chance of causing physiological damage to any individual as a result of creating this transgenic organism is smaller than one in a billion" must be regarded as meaningless for practical purposes and should certainly not be used as the basis for judging the wisdom of taking the risk, especially if it entails the potential creation of a novel pathogen.
2.4.2 Interdependence of risk factors
103 Another source of difficulty with risk analysis is that probabilities can be assigned only to events described in purely mechanical terms. The multiplication of the probabilities is then valid only if the different failure modes are truly independent.
104 However, engineering catastrophes tend to occur when human actions put the system under consideration into a mode such that the probabilities of different failures are drastically altered and the prior analysis no longer gives any worthwhile indication of the real risks associated with various hazards. Factors which are considered to be independent from an engineering point of view turn out to have unforeseen dependencies imposed by human actions.
105 In biology the problem is much worse because human intervention is not necessary to produce "quirkish" interdependence between particular members of different classes of events. New phenomena can appear that are so novel that their character cannot even be guessed at in advance.
2.4.3 Risks peculiar to biological systems
106 No matter how we classify biological events and entities, we will have no guarantee that rare members of apparently independent categories will in fact interact to produce a new self-sustaining phenomenon. Prions are entities that defy normal categorization,34 but they have caused a catastrophe on British farms and BSE has now been transmitted to humans. This could probably have been avoided if more stringent measures were imposed in about 1990.
107 The risks associated with the creation of novel biological situations cannot be measured. The integrity of the defined categories of events and entities which underpin risk analysis cannot be guaranteed to the same extent in biology as in physical and chemical engineering. To make matters worse, many biological events are threshold-regulated. No risk analysis could have assigned a probability to the possibility of the BSE epidemic prior to its occurrence
108 We should regard the conspiracy of events to produce unusual and unpredictable outcomes as a characteristic of biological systems and be extremely wary of analyses based on the sort of reasonable common sense with which committees and Commissions function, especially when dealing with the novel creations of genetic engineering.
2.5 Concerning ANZFA
109 There is no mechanism within the whole process of ANZFA's function that allows any consideration to be given to what might be called the "intangible" aspects of matters within its field of jurisdiction. The principle on which it judges food comprised of, containing or derived from genetically engineered organisms is that of substantial equivalence. The matters of substance in terms of which equivalence between GE food and traditional food is judged all fall into areas that are framed by scientific, technical enquiry.
2.5.1 Substantial equivalence
110 Three dictionary definitions35 of "substantial" are of relevance:
(i) "having substance, actually existing, not illusory",
(ii) "of real importance or value", and
(iii) "deserving the name in essentials, virtual, practical".
111 The first meaning is not what is intended. ANZFA has declared foods with different chemical compositions to be substantially equivalent. Monsanto's genetically engineered Round-up Ready Soy (RRS) has been found to be substantially equivalent to their parental lines36 even though RRS contain a protein ingredient novel to soy, the enzyme EPSPS.
112 Neither is the second meaning what is intended. For reasons of real importance and value to a very large number of people, RRS is not equivalent to ordinary soy. This perceived non-equivalence of genetically engineered food to ordinary food has nothing to do with scientific analysis directly. It is a matter of personal, perhaps ethical, choice. If ANZFA were to take this definition of "substantial" then they would simply be dictating that people cannot expect to exercise personal or ethical choices in respect of the food they eat unless the choice is provided by the Authority. In effect that is the attitude ANZFA has taken.
113 It is the third definition on which ANZFA actually relies and the Authority has taken control of what deserves to be called "essential", or what is "practical" in terms of differences between foods. The only questions of considered relevant by ANZFA are those of safety (including allergenicity), nutritional quality (wholesomeness), composition, and end use.
2.5.2 Problems of "substantial equivalence"
114 People who wish to have nothing to do with food derived from genetically engineered organisms are not opposed to the ANZFA's regulation of these important factors, but they can rightly complain that ANZFA is telling them what to think and denying them the opportunity to exercise freedom of statement when they are told that two foods are "substantially equivalent" when one is genetically modified and the other is not.
115 The 1989 poisoning of hundreds of people with Showa Denko's preparation of tryptophan from genetically engineered microbes is an illustration of how the principle of "substantial equivalence", even in ANZFA's interpretation, can fail.37
2.5.3 ANZFA bias
116 ANZFA has shown open bias in favour of industry interests. This bias is demonstrated, by way of example, in its Assessment38 of the use of RRS in food.
117 The Assessment contains two tables. One shows absolutely no benefit, to government, industry or consumers, but potentially high costs to all, associated with the option of banning the sale of RRS food. The other shows universal benefit and tolerable costs associated with the option of permitting the sale.
118 However, the categories of costs and benefits used to compare the two options are not at all equivalent. For example, the benefit to consumers from permitting sale of RRS food is said to be that they "can be assured that [RRS] have been through a premarket assessment and found to be as safe for human consumption as conventional soybeans", but there is no corresponding assurance (that ANZFA has protected the consumer) registered as a benefit against the option of not permitting sale of RRS food. With this rather blatant stacking of the evidence, ANZFA's approval of RRS was a foregone conclusion.
2.5.4 Inadequacies of industry testing
119 Of equal significance is the manner in which ANZFA bases its assessments of food safety on studies that come from almost exclusively from the applicant seeking approval for the sale of a novel food. In relation to the assessment of RRS, ANZFA reports39:
"A full data package for [RRS] was submitted by the applicant for assessment. Quality Assurance certification was provided that the studies were done in accordance with Good Laboratory Practice and that the information presented in the application accurately reflects the raw data generated during the studies."
120 I do not believe that any serious scientist would give very much weight to data which was presented in such a manner. In the case of testing for drugs there are three phases of carefully designed clinical trials that involve the (often blind) judgments of independent physicians. Even so, there is often residual suspicion that large pharmaceutical corporations are able to wield undue influence at various stages of the regulatory process.
121 In the case of ANZFA's assessments, safety considerations are finally weighed against financial concerns40 "Good Laboratory Practice" allows researchers enormous leeway in determining what experimental results are accepted as raw data.
122 Inconvenient results are routinely cast aside when investigation detects some irregularity in experimental protocol. Convenient results do not demand the same investigation. Really "clean" results, that would be obtained exactly if the experiment were carefully repeated by independent researchers, cannot usually be obtained in studies looking for marginal biological effects without the honing of experimental conditions over a considerable period of time and many repetitions of the same protocol.
3. Issues of scientific and Maori epistemology
3.1 Disparity of worldviews
123 The regulation and control of genetic engineering's role in our national life has been dominated by technical, scientific considerations. However most of the public discussion has relied on a context in which political, ethical or cultural values are of greatest importance. This has been particularly so in relation to the contribution that has come specifically from Maori.
124 New Zealand society faces the unresolved generic problem of deciding how fairly to give the proper weight in decision-making to the cultural perspective of the Crown's treaty partner - Maori. In the case of genetic engineering the problem is exacerbated because the terms used in the discussion are defined from a perspective that is foreign to Maori. This disparity of perspective, coupled with the claim of science to deal in universal truths, has marginalised the contribution of Maori. Maori are seen as expressing concern for what is in the realm of the "intangible".
125 Non-Maori with non-scientific reservations about genetic engineering have experienced similar treatment of their concerns. However, some success has been achieved by Maori through effective political action, but it has been impossible to draw official discourse and decision-making into the context of what might be called a "Maori worldview" or "Maori epistemology".
126 The discussion of genetic engineering from Maori perspectives can do much to illuminate hidden assumptions, especially in analyses that seem purely scientific and technical.
3.2 The universal versus the particular
127 Science seeks to explain phenomena in terms of order and structure that is permanent and fixed, not contingent on anything local or historical. Traditional Maori express a sense of order and structure that is intrinsically local and historical, contingent on events and relationships established by precedent, not given unalterably.
128 Scientific analysis relies on the prior establishment of universally applicable categories that can be used to describe things and events. These categories may specify things like "electron" or "gene" or events like "chemical reaction" or "translation of genetic information".
129 Scientific categories are abstract constructs that have been built up and refined through a process of observation and experimentation. The definition of the categories and their relationships is always, at least formally, open to question. However, in their day to day work scientists treat basic categories of description as if they gave a true representation the one and only physical reality. That reality is taken ultimately to be "given" by unalterable laws of Nature and to have universal properties.
130 In Maori tradition knowledge of something is concerned with achieving a proper perception of its location in time and space. Knowledge of things and events is concerned with the particularities of whakapapa - layers of genealogy and lines of descent, their patterns and linkages.
131 For Maori, everything is ultimately related to everything else, but the true character of something belongs to the particular thing itself and its historical origin. The character of things is not described as a set of properties derived from an abstract world beyond what is here and now.
132 For Maori, everything is rooted, not only to its origin in time, but also to its origin in space - the place and tradition of the tangata whenua to which it belongs. This relationship with the earth and its local geography, something amounting to an umbilical connection41, is of particular poignancy in the contrast between scientific and Maori explanations of the causes of things.
3.3 Mechanism versus agency
133 The most fundamental character of reality in Maori cosmogony entails a conception of agency within Nature that has been systematically exorcised from intellectual discourse within the Western scientific tradition.
134 In science, the final explanation of things, events and possibilities is expressed in terms of what "happens" and its mechanism. Everything we observe derives from the properties of a single, unchanging material substance (which physicists, since Einstein, have identified as energy rather than atomic matter).
135 In the original conception of the Ancient Greeks, this material Nature, physis, was not distinguished from the divine power that was thought to pervade it. Later Aristotle expressed the idea that everything in Nature had an internal goal-directed drive, telos, to find its rightful place.
136 Only in the seventeenth century did Galileo and Newton come up with a purely formal, mechanistic description of motion. There was then no need to think of Nature as being alive with any of the attributes we now associate with subjectivity or conscious intent.
137 Darwin dealt the final blow to any scientific idea of élan vital by describing the entire history of life in terms of the mechanistic principle of natural selection.
138 In Maori tradition, things, events and possibilities cannot be reduced to the properties of a material substance and mechanistic laws. Marsden and Henare42 identify Tua Uri43 as a representation of the 'fabric of the universe' in which whakapapa begins with mauri, divine power or agency.
139 Mauri precedes hihiri, pure energy, in the cosmological genealogy and hihiri is refined to give rise to Mauri-ora, the life principle, and thence Hau-ora, the spiritual breath of animate life. These precede shape, form, space and time.
3.4 The secular versus the sacred
140 The defining political event marking the advent of modern science was the trial of Galileo (now the subject of a New Zealand opera of bicultural origins44). Through his refusal, on the basis of scientific judgment, to capitulate to ecclesiastical power Galileo emancipated 'natural philosophy' from arbitrary strictures imposed by parties for whom the truth was predetermined. Science then established for itself an intellectual authority that transcended the foundation on which the Church had relied.
141 Theology itself eventually underwent a revolution45 of 'secularization' in which Christian belief was fully accommodated to scientific methodology. Although scientific findings are regarded as being independent of cultural or ideological bias, scientific research is still subject to general ethical and regulatory controls.
142 There is nothing internal to science that associates a value, according to any scale whatsoever, with any thing, event or possibility.
143 Everything in the Maori world is imbued with a natural sanctity or tapu. The tapu ascribed to things is derived from divine association and establishes a prima facie untouchability that humans are bound ritualistically to propitiate in all of their actions. There is no exemption within the sphere of Maori influence.
144 The contrast with the perspective of science could not be starker.
For Maori: everything is in some sense intrinsically sacred and a demand for respect arises from the very nature of things.
For science: nothing at all is sacred; no attribute could be more foreign to physical reality.
3.5 Politics of rapprochement
145 The crux of the problem in relation to genetic engineering and Maori is to find a way of giving force to considerations of whakapapa, mauri, tapu and other precepts of tradition without their remaining a sideline to the recognised discussion based on scientific principles and analysis. What is done in New Zealand to solve this problem will have implications far beyond our shores.
146 It is not just a matter of whether Maori are given some measure of political power in decision-making. Much more is at stake in what humans decide to do with genetic engineering and we should act according to the best principles that we can conceive from our uniquely bicultural constitution.
147 Up until now contributions to debate about genetic engineering from alternative scientific and Maori perspectives have amounted to competition for control over domains of culture and Nature.
3.5.1 Dialogue between scientists and Maori
148 Maori have been consulted in various fora (like ERMA's Nga Kaihautu committee and the Patenting of Life Forms Focus Group of the Ministry of Commerce) but there has been little attempt to forge agreement, common understanding and joint action based an appreciation of the real differences in worldview. This cannot be done quickly simply to facilitate business interests, as has been attempted in processes of "consultation" by parties applying to ERMA.
149 Ammunson and Cairns46 recommend bringing together for dialogue those separately well-versed in biotechnology and tikanga. While such dialogue between scientific experts and tohunga would be important, it would be limited in two very important respects. First, it would not engender the kind of criticism that is needed to get to the fundamental assumptions that separate the parties. Second, the approach is gratuitously elitist and would exclude virtually all opponents of both Ammunson and Cairns47 and their