THE COMPLEMENTARITY OF KNOWLEDGE AND IGNORANCE IN SCIENCE

Not many years ago, an eminent scientist with philosophical leanings described science as « the art of the soluble ». This elegant formulation reveals much about the implicit limits of scientific enquiry and its picture of the world. For what is not soluble is not scientific ; it does not count, it does not exist. This attitude is well known to those concerned with foundational problems ; they are accustomed to their work being vie­ wed by research scientists as irrelevant or worse, to the real tasks of science. The exten­ sion of this presupposition to technology has produced a host of unanticipated policy issues in risks and the environment. For the “downs­ tream” phases of production cycles rarely pre­ sent neatly “soluble” engineering problems; as in the example of the wastes of nuclear power, they typically suffer neglect until they sud­ denly present crises in all dimensions.


THE COMPLEMENTARITY OF KNOWLEDGE AND IGNORANCE I N SCIENCE
J EROME R. RAVETZ N ot many years ago, an eminent scientist with philosophical leanings described science as « the art of the soluble ». This elegant formulation reveals much about the implicit limits of scientific enquiry and its picture of the world. For what is not soluble is not scientific ; it does not count, it does not exist. This attitude is well known to those concern-ed with foundational problems ; they are accustomed to their work being vie wed by research scientists as irrelevant or worse, to the real tasks of science. The exten sion of this presupposition to technology has produced a host of unanticipated policy issues in risks and the environment. For the "downs tream" phases of production cycles rarely pre sent neatly "soluble" engineering problems; as in the example of the wastes of nuclear power, they typically suffer neglect until they sud denly present crises in all dimensions.
This restricted view of science as the « art of the soluble » may well have enhanced its power in the past, to the point where it now presents perils for the future. It also has other, even deeper effects on our vision of knowledge and the world. For it entails a total exclusion of ignorance from our view. Uncertainty exists only as far as it can be managed interestingly, in the form of the soluble research problems at the margin of our scientific knowledge. Ignorance is not soluble by means of ordi nary research ; therefore it does not exist. A scientist who explores problems which cannot be solved (such as Einstein in his later years) is generally considered by his colleagues to be wasting his time. In the classic image of science, purveyed by philosophers and publi cists, and imbibed by generations of teachers and their pupils, science is about certainty. Uncertainty is there to be banished, and igno rance is to be rolled back beyond the horizon.
For it to remain plausible, this sim plistic vision of the triumphal advance of science needed to maintain ignorance of one very practical problem in scientific activity : priorities and choice. For whenever a propo sed research project is given a low priority (in whatever sort of decision mechanism), it is not undertaken. As a result, the chance of gaining new knowledge is lost; and in that res pect we remain in ignorance. There is an asymmetry here : a project that is undertaken might fail, thereby not increasing our stock of knowledge in spite of our intentions; but if a project is not undertaken at all, then we are intentionally kept in ignorance. Such consi derations became salient in the era of "big science", when there were more interesting problems (and promotors) than resources for all of them. Priorities then became the name of the game, and the preservation of ignorance in the less favoured fields of science became established as an unavoidable aspect of science policy.
The asymmetry goes deeper, in the cognitive and social structures of science. It has generally been assumed that only trained, certified scientists could advance knowledge through research ; and also that in some important fashion their research is "valuefree". Thus the knowledge that has emerged is the product of a highly specialized, indeed esoteric, activity ; and when taught or popu larized it has the appearance of inevitability and incontrovertibility. By constrast, our igno rance is obviously very strongly conditioned by values. Thus, if our society is relatively less interested in (say) occupational health and alternative energy supplies than hi-tech medicine and nuclear power, we remain in ignorance about the former. Since ignorance has no positive content, it is very difficult for anyone to point to it, invoking authorita tive expertise to analyse it. So the focus is on the science that we have, with this appa rent inevitability and incontrovertibility. We are never made aware that it exists only in relation to the value-loaded decisions to obtain some particular sort of knowledge a the expense of remaining in ignorance of other possible, equally valid and worthwhile, sorts. In this way, we can speak of ignorance being "socially constructed", indeed to some extent politically constructed.
By focussing on ignorance rather than on knowledge we can escape some of the rela tivistic, sceptical implications of a contex tual theory of science and its products. It is possible to imagine ignorance as being more conditioned by values and power, while know ledge (or what is accepted as such at any time), being rather less conditioned. The dis tinction is not absolute ; although philoso phers of science have paid relatively little attention to this topic, the role of values in shaping our facts is clearly seen in the way that confidence-limits (and other parameters of statistical tests) embody prior decisions about the relative error-costs of the different sorts of erroneous inferences from statistical argu ments. Indeed, the methodologies for treat-ment of outliers among data could serve as an excellent case study for the shaping by values of both our ignorance and our knowledge.
A yet deeper consequence of the triumphalist, positivist image of science is that all of the understanding of ignorance has been systematically suppressed among stu dents of science. Who has ever seen an exam ination in a traditional science where there are questions involving ignorance, uncertainty, choice or values ? Examination questions with the instruction "Critically evaluate", com mon in all other disciplines, are rare in the extreme in natural science. Those who emerge from a prolonged training in scientific research, perhaps even with the title of Doctor of Philosophy, have been conditioned to remain ignorant of their ignorance. This is the basis for Kuhn's profound, eloquent and deeply trou bling image of the puzzle-solving scientist, with a training more restrictive than any but orthodox theology, and whose disciplinary histories resemble those of 1984. With this perspective we can better understand why Popper considered "normal science" (as so ambivalently named by Khun) to be « an enemy of science and of our civilisation ».
Worse yet, this reductionist concep tion of the human intellect has generally been identified as the guenuine rationality of our intellectual culture. Traditional spo kesmen of science have derogated other forms of knowing, involving the recognition of ambiguity, contradiction, uncertainty and ignorance, as inferior or unreal. In this way, the lived experience of individuals, and the historically developed culture of peoples, are all either crammed into the scientistic mould of simplistic certainty, or discarded. While this presupposition is so plausible as to be almost invisible in the case of the physical sciences and their applications, the imitation of physics in the sciences concerning huma nity shows clearly how ludicrous and damaging can be this restriction of knowledge, and this culturally imposed ignoring of ignorance. It is easy to detect "ignorance-squared" in supposedly scientific statements : just verify that numerical precision goes up as measure ment accuracy goes down. Those extra pseudo-precise digits conceal ignorance from all, usually including the author as well.
Such an "ignorance-squared" can almost be called wilful within our culture. Ignorance of ignorance is a very recent phe nomenon in European intellectual history. Continuously from the time of Plato, up to that of Descartes, the ignorance of ignorance was a recognized category among philoso phers. Indeed, all those who knew about Socrates knew that his quest was for aware ness, for self-knowledge, for discovering and coming to terms with his own ignorance of ignorance. And his method of elucidation was designed to help this interlocutors to cure their own ignorance of ignorance, on some issue that was crucial for right living. The Renaissance humanist writers, including Montaigne, gave prominence to ignorance of ignorance. The big break came with Galileo and Descartes, who imagined human know ledge as limitless in its scope and perfectibility. For them, ignorance was a void to be filled, as quickly as possible ; and each had a Method whereby this could be accomplished.
In connection with Descartes, it might seem that the method of Doubt involved a recognition of irremediable ignorance and hence of ignorance-squared. But this Doubt was of a very "philosophical" sort, which was there as a challenge to be overcome at the very beginning of his endeavour. Once removed by his heroic conceptual endeavour, it never returned. Thereafter, the Method would gua rantee a steady increase in knowledge and in certainty. Once Doubt had been conquered by Descartes, it hardly ever reappeared in the philosophy of science, until our own troubles times. Thus the philosophical condition of modem science, technically described as "bar baric" by E.A. Burtt, has its roots in a radical philosophical move that soon resonated with the demands of its cultural matrix. Religious dogmatism, responsible for the confusions and horrors of the long Reformation period, was discredited and in its place came scientific optimism, sometimes tentative but also some times as dogmatic as its rejected predecessor.
In subsequent centuries, it was left to isolated critics of science to voice concerns about ignorance. Each had a very personal basis in experience, as Bishop Berkeley in theology, and Ernst Mach in foundational studies ; and so their critique could never achieve coherence as a tendency, or compre hension by a broad reading public. Basic cri ticisms of science of any sort could be dis missed, for they did not contribute to the onward march of successful scientific resarch and technological development. Khun's "puzzle-solving within paradigms" expresses all that very well. Of course, the situation is now changing. The loss of mystique of science is reflected in the dominance of the sceptical, debunking scholarly analyses of scientific practice. And the new problems of risks and the global environment have put uncertainty and ignorance at the centre of public debate.
The need to take very large scale deci sions about industry, technology and life styles, in the face of the recognized global environmental threats, has shown clearly that science can no longer be assumed capable of providing certainty for our policy problems. Debates on risks and the environment some times focus less on the supposed facts than on the management of their uncertainty and quality. The crucial decisions are at the metho dological level : on the assignment of the appropriate degree of "precaution" ; on the estimation of "error costs" of any proposed policy; and ultimately on the placing of bur den of proof on one a another participant or stakeholder. None of these decisions can be reduced to a quantitative formalism ; they all involve personal judgements and the clash of values. In this way, such major policy issues in risks and the environment are a cases of what my colleague Silvio Funtowicz and I have called "post-normal science".
Similarly, in relation to small-scale local issues, the limitations of certified scien tific expertise have also become vividly appa rent. In connection with health and envi ronmental effects of toxicants, either in industrial processes or in the form of waste, communities have sometimes found them selves in debate against not only those who created the hazard, but sometimes also against those who are supposed to recognize and then regulate it, and even those who are supposed to provide objective knowledge for its control. Khun's image of the puzzle-solving normal scientists, constitutionally kept in ignorance of anything outside the walls of their disci plinary paradigm, has been discovered to be applicable not only in the laboratories and lec ture halls, but also in the spheres of the courts and regulation.
In such situations, ignorance of igno rance can be very damaging indeed. So long as it persists, whole communities can he lulled in a false sense of security. But when it is bro ken, those who have been implicated in its maintenance lose both credibility and legiti macy. Their expertise is revealed as applying to a small and not very relevant part of know ledge : that which concerns normal-scientific practice, of the reduced, artificial and control led environment of the research laboratory. The awareness of ignorance occurs, for these unfortunate persons, not in the playful inter rogation of a Socratic dialogue, hut in the harsh glare of a judicial or adversarial process.
Thus there is a guenuine politics of the ignorance of ignorance. In earlier times Science was invoked in the demolition of the ignorance-squared of a traditionalist, theo cratic culture ; now Ecology is invoked in the challenge to the modern technocratic version of ignorance-squared, in which esta blished science is now implicated. This struggle will have the varieties of conflict, including extremists on both sides, diversio nary tactics, and ultimately (we may hope) a reasoned debate where that is possible.
Among the diversionary tactics, there is a long series of attempts to tame uncer tainty (and thereby to suppress ignorance) by mathematical means. These have mainly appeared in connection with probability and statistics; their more recent versions involve sophisticated psychological theories on the hand, and gigantic computer programs on the other. In addition, analogues of genui nely uncertain situations in the natural world are sought in particular mathematical sys tems (e.g. game theory, decision support sys tems, catastrophe theory, fuzzy sets, chaos theory). It is perennially hoped that theo rems in such mathematical systems will pro vide beacons of certainty for rational decision making.
We can no longer afford to maintain ignorance of ignorance at the centre of our scientific culture. Only through recognising the complementarity of knowledge with igno rance, realized through open dialogue and debate, can science meet the challenge of the new environmental policy issues of our day. ■