Science and the normative

We have defined ideology as the political use of non-political knowledge and the fact is that the social sciences, and especially criminology, will always practically automatically be recuperated for ideological purposes. The very idea that someone decided to study crime implies that she is looking for some form of "solution": quite simply, no one likes or approves of crime. Why make a difference between ideology and knowledge, then? There are two different levels here: the subjective decision to study something does not make the study itself subjective. For example, if an oncologist chose her discipline because her mother died of cancer, she will still be held to the same scientific standards others have to respect. The added difficulty in the social sciences is that the object is not only already part of popular discourse and politics it actually exist only because it is, unlike cancer, which exists and kills whether it is studied of thought of or not.

So what does it mean to use expertise as basis for administrative action, to solve social problems?

a) The criminologist as problem-solver

Yaron Ezrahi (1976) looks at the idea (taken from French biologist Jacques Monod) of an "ethics of knowledge", the moral obligation of the scientist to better society with any new knowledge she may gain. The problem, typically, has been that such scientists have generally assumed that the political machine was an infinitely reorganizable tool separate from social reality and meant to be used for implementation of conclusions (p. 154). In other words, knowledge is assumed to "depoliticize" politics.

At this point we can unpack our "decision-making" category in the now well-worn conceptual triptych of agenda-setting, policy-making and evaluation/analysis. Solomon (1983) recommends that the political machine be studied in itself for two reasons: first, obviously, to improve the "rationality" of administrative operations (learn from past mistakes, progress, etc.), second, implicitly, to help scientists tailor their proposals to fit administrative necessities. The first part of the program, according to Solomon, works best in the "agenda-setting" phase; there, scientific knowledge can permeate government agencies and provide new ideas. This is also the conclusion of Weiss (1977), that research does not have to be "geared to the operating feasibilities of today, but [...] provides the intellectual background of concepts, orientations, and empirical generalizations that inform policy" (p. 80). This, based on asking a sample of officials questions such as "how likely is it that you would take [...] study results into account?" (p. 75)! Methodological objections aside, one would indeed expect, in the ideal (or discourse) of "administrative rationality" we have been discussing, that officials -- assuming they are being truthful -- would genuinely expect to "consider" any form of knowledge. It is rather like asking fishermen whether they might consider using a net. Now, reality, on the other hand, is quite different: Brodeur (1995) for instance recalls asking a top U.S. justice official (it was Phillip Heyman, then Assistant Attorney-General) what importance criminological knowledge had had in a recent reform of the U.S. penal system; the answer, not surprisingly: none whatsoever (see also Roger Hood, 1974: 380).

Observing that politicians and policy officials use criminological or other social sciences jargon is exactly as probative as noticing that ordinary people use expressions like "freudian slip": it does not make them psychoanalysts. In fact, we have just demonstrated that use of a language is the very worst example one could find, especially in the social sciences. The use of a word-concept is in no way an indication that the speaker understands or is using the concept "properly" (i.e. like the listener understands/uses it). This kind of improvement on "rationality" is largely mythical, and irremediably condemned to failure, and here is why: rationality, in the way it is conceived of here, simply does not fit government operations. Again, this is not a criticism of modern democratic administrations (not that we are saying it is perfect either): we are not simply deploring the capriciousness of politicians or the incompetence of bureaucrats. The concept of rationality does not fit, because it cannot; it is an artificial and essentially normative construction based on largely outdated ideas about human nature and the production of thought. While it works rather well in the natural sciences, engineering, etc., in policy making and other normative activities it is misleading. Maximizing rationality, approaching the "one best way", in these circumstances, amounts to exchanging political words for scientific ones, without ever touching the core concepts. In fact, it practically guarantees the core concepts' status quo, because it gives the impression of healthy progress, that the job was done. Mentalities must be articulated in language, but not specific words; this is not a semantic problem. There is very little evidence that expertise has indeed modified the way we do business in the crime business. The words have changed, the real estate has changed, but most of the transformations seen today come from much broader changes in the political discourse, towards a neo-liberal reorganization of knowledge.

b) Value-oriented solutions

Ezrahi (1976: 159) points out that sociological "reality" may not always be the best basis for the constitution of the state: "ideas that are valid in the context of scientific discourse may be considered dysfunctional for the purpose of maintaining 'conditions of peace' [according to an expression of Hobbes']" (p. 159). This is typical, according to him, of experts who see their role as "diffusers" of knowledge: they forget that, like a straight wooden stick dipped in water appears to be crooked, "straight" knowledge, once injected in the social body, will be reshaped by political forces. The obvious problem with that is that the "conditions of peace" are a sociological question in themselves, and thus there is no way to clearly differentiate knowledge from normative judgement, and at the same time there is no logical path from observation to action, no objective way to deduct ethical principles from empirical reality.

We have seen that both values and knowledge are constituted in political terms; furthermore, this opposition itself is also an entirely artificial/contingent element of political discourse. One can situate oneself as a defender/believer of tradition or of science on one particular issue, of course; but the point is that this position comes not from the individual, or from the definition of values or science, but on the particular definition of the issue itself. And there is absolutely nothing a scientist or a politician can do about that (or at the very least in the short/medium term). Abortion is a good example; once posed as a moral/religious question, it became a medical issue a few decades ago: how early is the fetus viable? Now, it seems to revolve around whether the fetus is a complete subject of rights or not, and perhaps at what point in time (note that the mother is conspicuously absent of all these debates, a patriarchal oversight that seems to run through all these discourses -- they all exclude the main actor). These points of view are not where the battle is waged; there is a complete battlefield in each of them. The changing discourse constantly offers new ways to be contentious, new ways to be pro-this or anti-that. When doctors perform abortions in a clinic surrounded by conservative christian demonstrators, you have a fight between two discourses, not between a scientific and a moral opinion or judgment that could be changed by better information. And no scientist will ever disperse the crowd outside, no matter how good a speaker she is, no matter how probative her knowledge appears. That is because the protesters know different (not to mention that they can find doctors who will support them). And if this isolation from fact is visible with contentious issues, one can imagine how little fundamental debate there can be over something like "crime".

One major difference between ethics and science is that the former tolerates no contradiction, while science, and especially the social sciences, tend to produce experts who are aware of at least a measure of uncertainty in their knowledge. This largely comes from the truly fantastic potential of science to be wrong, as is painfully obvious to any student of its history. Fortunately, we are no longer measuring skulls to determine dangerosity; but how do we know that what we are doing now is any better? Judging by the speed at which half-baked "genetic" explanations of crime splash in the media there is more than enough reason to be weary. All the while, there are quite a few now established facts in criminology which should be taken in consideration when forming policy, but almost never are.

On the other hand, values are never wrong. When they become inadequate, they are simply said to have been the product of different times, and adequate, if not necessary, then. Values can change without being wrong, a paradox that is impossible to apply to knowledge: if it changes, it has to have been wrong then, or now. And if it continually changes, when does one know it to be definitively right? In the exact sciences, like physics, at first sight sending a spacecraft in orbit seems like a definite proof that physics have evolved. Unfortunately, social sciences do not produce such startling achievements⁽⁷⁾. Why? The traditional answer -- and it is getting a little tired -- is that they are still in their "infancy" and not enough is known about the "social" yet. Paradoxically, it seems, as we underlined above, to already be moving away from the social, perhaps "giving up" on the whole idea of understanding and finding problems and solutions in it.

So how can the expert influence policy, and should he try to?

NOTE

7. And, we might point out, fewer Chernobyls as well. Social sciences seem to be essentially "low-intensity" affairs, both in their successes and failures.