regulated. Regulations are rarely introduced at the first evidential hint of a problem.
While much of the relevant evidence is scientific, judgements about how much
evidence, and of which types, are variously necessary or sufficient grounds for
restricting or prohibiting industrial practices are normative policy judgements, not
scientific ones. Some have argued that only causal proof of harm should be deemed
sufficient grounds to ‘disrupt’ market transactions, while many others have argued
from a precautionary perspective that industrial practices and products are almost
invariably regulated in conditions of scientific uncertainty. Judgements about how
much evidence should be required and how much uncertainty should be tolerated
have been the focus of fierce disputes. In practice, regulatory regimes have typically
been reactive rather than anticipatory.
Regulatory science: a contested domain
One common source of scientific uncertainties arises as a consequence of the fact
that many putative risks are not studied directly, but explored indirectly through
the use of models. Thankfully, scientists are sometimes reluctant knowingly to release
a potentially harmful product in order to study directly the nature and scale of
the harm it can cause. Rather than testing radioactivity or chemicals on human
subjects, experiments are typically conducted using laboratory animals or bacterial
and cell cultures as models of the effects on people, as well as on flora and fauna.
Rather than deliberately changing concentrations of greenhouse gases, the risks of
climate change are most commonly examined using computer-based models
coupled to available (but incomplete) empirical data. Policy-making is complicated
by the fact that we are often very uncertain about the relevance and reliability of
the models that the experts develop and deploy to the conditions that they purport
to model. It is not that too few models are available; rather, the problem is often
to make judgements about the relative reliability of competing models, their
predictions and forecasts.
Many protagonists in debates about (un)sustainability assert that they are uniquely
in possession of the best science; their assertions can come in two main forms. The
first alleges full scientific certainty for their perspective, while the second laments
the limitations of prevailing knowledge and the extent and/or severity of scientific
uncertainty. The former tactic often requires understating, or maybe even con -
ceal ing, uncertainties. The latter typically involves overstating, or perhaps
exaggerating, them. Both tactics have been used by protagonists on all sides, by
those in governments, the corporate sector and civil society. That does not mean
that everybody is always dissembling, but it does mean that ostensibly scientific
assertions concerning the risks and/or benefits of industrial products and processes
such as ‘fracking’, ‘nanotechnology’ or ‘climate engineering’ should not be taken
at face value because their apparent scientificity is invariably misleading. On the
contrary, questions should routinely be asked about how they were constructed
from mixtures of both factual and normative considerations. The fact that they
have all been constructed, using forms of hybridization, does not, however, entail
44 Erik Millstone
While much of the relevant evidence is scientific, judgements about how much
evidence, and of which types, are variously necessary or sufficient grounds for
restricting or prohibiting industrial practices are normative policy judgements, not
scientific ones. Some have argued that only causal proof of harm should be deemed
sufficient grounds to ‘disrupt’ market transactions, while many others have argued
from a precautionary perspective that industrial practices and products are almost
invariably regulated in conditions of scientific uncertainty. Judgements about how
much evidence should be required and how much uncertainty should be tolerated
have been the focus of fierce disputes. In practice, regulatory regimes have typically
been reactive rather than anticipatory.
Regulatory science: a contested domain
One common source of scientific uncertainties arises as a consequence of the fact
that many putative risks are not studied directly, but explored indirectly through
the use of models. Thankfully, scientists are sometimes reluctant knowingly to release
a potentially harmful product in order to study directly the nature and scale of
the harm it can cause. Rather than testing radioactivity or chemicals on human
subjects, experiments are typically conducted using laboratory animals or bacterial
and cell cultures as models of the effects on people, as well as on flora and fauna.
Rather than deliberately changing concentrations of greenhouse gases, the risks of
climate change are most commonly examined using computer-based models
coupled to available (but incomplete) empirical data. Policy-making is complicated
by the fact that we are often very uncertain about the relevance and reliability of
the models that the experts develop and deploy to the conditions that they purport
to model. It is not that too few models are available; rather, the problem is often
to make judgements about the relative reliability of competing models, their
predictions and forecasts.
Many protagonists in debates about (un)sustainability assert that they are uniquely
in possession of the best science; their assertions can come in two main forms. The
first alleges full scientific certainty for their perspective, while the second laments
the limitations of prevailing knowledge and the extent and/or severity of scientific
uncertainty. The former tactic often requires understating, or maybe even con -
ceal ing, uncertainties. The latter typically involves overstating, or perhaps
exaggerating, them. Both tactics have been used by protagonists on all sides, by
those in governments, the corporate sector and civil society. That does not mean
that everybody is always dissembling, but it does mean that ostensibly scientific
assertions concerning the risks and/or benefits of industrial products and processes
such as ‘fracking’, ‘nanotechnology’ or ‘climate engineering’ should not be taken
at face value because their apparent scientificity is invariably misleading. On the
contrary, questions should routinely be asked about how they were constructed
from mixtures of both factual and normative considerations. The fact that they
have all been constructed, using forms of hybridization, does not, however, entail
44 Erik Millstone