L'Initiative de démocratisation des données (IDD) a eu ce résultat extraordinaire de rendre les données canadiennes aisément accessibles aux chercheurs en sciences sociales. Mais il ne s'agit là que de la pointe de l'iceberg: au delà de la satisfaction de la demande existante, l'IDD devrait, dans sa deuxième étape, tenter d'impliquer le plus grand nombre possible de gens dans la production de la connaissance sociale. En effet, la complexité des enjeux auxquels notre société est confrontée exigera des citoyens et des travailleurs de plus en plus compétents. L'aptitude à lire, à compter et, plus récemment, à utiliser un ordinateur sont devenues des prérequis d'une participation pleine et entière à la vie sociale contemporaine; de la même façon, un minimum d'aptitude à interpréter des données sociales est en train de devenir une composante essentielle d'une vie de travail productive et d'une participation à la vie politique. Ceci ne signifie en rien que la manipulation de données peut remplacer la lecture et la réflexion sur les idées; bien au contraire, ces idées prennent une signification nouvelle et deviennent plus claires quand nous nous demandons comment on peut les évaluer au moyen d'une démarche empirique.

The Data Liberation Initiative (DLI) has achieved extraordinary success in pursuing its first objective: making Canadian data easily accessible to social science researchers. Even in these times of extreme fiscal restraint, government agencies have scraped together part of the necessary funding, and most institutions of higher learning in Canada have proved eager to participate, contributing financial resources, personnel and equipment. The administrative structure for the program has been put in place and Statistics Canada, the key source for the data, is participating with enthusiasm.

But this is only the tip of the iceberg. As important as it may be to satisfy existing demand for data, our long term objective should go much beyond. We want to increase the amount of socially relevant knowledge available to Canadians in public debates; we also want to involve as many people as possible in creating such knowledge and in transforming it so as to fit various purposes at hand.

Extending the data culture entails not only producing more knowledge, but also making more people knowledgeable. It means involving ever larger circles of people in appropriating and understanding knowledge, in criticizing it, and even in producing it themselves. This process is indispensable for reasons that have to do with both the nature of social scientific knowledge and the challenge of a democratic modern society.

The validity of scientific knowledge, social and otherwise, rests entirely on its total openness to criticism; this in turn presupposes that there are enough appropriately trained producers and critics of this knowledge. The difficulty is particularly acute for a relatively small society like Canada; our problems are not the least bit simpler than in larger ones, and yet our resources are much less abundant. We need to involve as many people as possible in the task of understanding the dynamics of our society, and of helping it through its current crisis. Moreover, social science is no magic elixir, potent with ready made cures: its application to issues requires the involvement of all interested parties, and acting otherwise usually brings about misery and perverse effects.

Indeed, the complexity of issues facing our society will require an increasingly sophisticated citizenry and work force. Democratic debates around questions such as the coexistence of different cultures brought together by modern communications and migrations; ecological problems; the dialectics between the values of equity and freedom; new forms of concerted socio-economic action, etc., require complex information and relatively sophisticated participants. Economic prosperity will also increasingly rest on intellectual, managerial and cultural inventiveness rather than on the sheer availability of resources and physical capital; much of that inventiveness will depend on the widespread ability to decipher the signals coming out of social processes themselves.

Literacy, then numeracy, and, even more recently, some computer literacy have become preconditions for a full participation in contemporary social life; similarly, a minimal capacity to interpret social data has become key to leading a productive life, and to being an informed citizen. In this sense, we cannot be too ambitious in setting the goals for the second phase of the Data Liberation Initiative: our objective should be to extend the data culture, to reach various publics with appropriate amounts of knowledge about how to interpret social data, to generalize recourse to such data in social, political and economic debates. Of course, quantified data are not, by far, the only pertinent source of social information; indeed, we are increasingly awash in a world of archives, textual, verbal, and visual, and we should also learn how to exploit such data. In this sense, the current phase is just a way station in the generalization of social sciences education; we focus on social statistics because the DLI currently provides us with an invaluable opportunity, but we should situate our enterprise in a broader context.

Wishing for a generalization of social science education may seem far-fetched. A few examples demonstrate, however, that elements of social science knowledge have indeed already become part of current discourse. Opinion surveys, for instance, have become familiar to most voters, who probably even know a few things about response rates, intervals of confidence, various methods of redistribution of non-responses, etc. Many key economic concepts, like inflation rate, unemployment rate (even in their de-seasonalized version), trade balance, Gross Domestic Product, income polarization and the like are used in debates and in the media, and sections of the public are even beginning to consider criticisms of them (such as the contrast between the Gross Domestic Product and a recently proposed Genuine Progress Indicator). Similarly, some amount of knowledge about socio-professional categories, immigrant groups, the changing roles of women, new family forms, and the like regularly shows up in discourse. Extending the data culture simply means strengthening this trend, systematizing references to precise definitions and data, and increasing the capacity to criticize these definitions and measurements.

One should also consider the marvellous opportunities for dynamic training processes opened up by the availability of data. It has become increasingly necessary to learn not only contents, but also the methods through which knowledge can be transformed, extended, and created as circumstances change. There is no better way to promote such autonomy in thinking than to have people conceptualize for themselves, and try these methods on data in a systematic way. This of course does not mean that data manipulation can replace reading and thinking about ideas; on the contrary, ideas assume new meanings and come in sharper focus when one asks oneself how they can be empirically assessed .

Figure 1 identifies the different publics this new phase of the DLI should reach--at least in principle. Within each sector indicated by the figure, social science is produced and used in different ways. At the centre are the professional producers of social science knowledge; at the periphery, the public is at once the object, the beneficiary, and a shaper of this knowledge; in between are different categories of users and creators of such knowledge, identified according to their needs and to their level of sophistication in understanding and performing research. The sector corresponding to private enterprises has been blanked out, not because it does not need and use social science knowledge (quite the contrary), but because extending the data culture outside the public sphere partly implies a different logic, that of profitability; for this reason, the private sector has not been included in Phase 1 of the DLI, and it continues to obtain its data in its own way. Nevertheless, cooperation between the public and private sector is not only conceivable, but might turn out to be key to some aspects of Phase 2 of the DLI.

Under each of the publics identified in figure 1, numbers indicate similar training needs with respect to the use of social data. Figure 2 cross-tabulates these publics--regrouped as indicated by these numbers--and the various training needs; the latter correspond to a number of technical areas of expertise (from elementary to advanced), but also, and even more importantly, to the art of asking fruitful questions, and of asking them in appropriate ways. An "x" in any cell of figure 2 indicates that a given public, or set of publics, should be provided with training in a particular area.

This is, obviously, much too much for Phase 2 of DLI; Phase 1 has demonstrated that we should not shy away from ambitious objectives, but also that we should aim at achieving a few successes within a reasonable time. As a consequence, these figures should be used to select our priorities. We should launch the discussion of Phase 2 with questions such as the following.

• What has to be done on a short term basis, as a precondition for DLI as it stands to produce results? Training of research librarians in ways of identifying, acquiring, storing and making available pertinent data would seem to belong in this category, for instance.
• Which publics can be regrouped for the purposes of common training activities, depending on their interests, their needs, their institutional affiliation (or lack thereof), the compatibility of their timetables, etc.?
• To what extent should these activities be organized around substantive subject areas, such as health, welfare, ecology, the economy, culture, etc.?
• Which of these needs should be served through existing structures and activities, and which require new organizations? How and with whom should the Phase 2 Task Force undertake discussions about priorities and division of labour?

Notes

Bernard, Paul, "Phase 2 of the Data Liberation Initiative: Extending the Data Culture," Government Information in Canada/Information gouvernementale au Canada 3, no. 1 (1996).
<http://www.usask.ca/library/gic/v3n1/bernard/bernard.html>
Back to text.

[2]