@ Desrosières, A. (1998). Classifying and Encoding. The politics of large numbers: A history of statistical reasoning. Harvard University Press.
Chapter 8 Classifying and Encoding
- 01 Introduction
- follow the establishment of the formalisms and institutions that have made massive use of these objects socially and technically possible
- recall taxonomical questions deriving from intellectual and theoretical traditions
- examples on classification concerning natural species, branches of industry, poverty, unemployment, social categories, and causes of death
- 02 Statistics and Classification
- encoding: the decision to attribute a case to a particular class
- statistics is one tool for the passage from singularity to generality and the construction of consistent classes of equivalence
- double dimension
- nominalism-constituting and defining categories and on encoding singular cases
- realistic-the methods of classification resulting from data analysis
- 03 The Taxonomy of Living Beings
- Foucault’s archeology of taxonomy
- Linné: “all of nature could enter into a taxonomy”
- Buffon: “too varied and rich to be fitted into so rigid a framework”
- encoding as a sacrifice of inessential perceptions
- compensation for sense loss
- Linné: revealing the order of nature
- Buffon: constructing the order of nature
- compensation for sense loss
- the choice of pertinent variables
- variables
- measurable: numbers & sizes
- unmeasurable: forms & dispositions
- historical change
- 16c: the unique distinguishing feature
- 18c: networks of differences observed in botany
- 19c: the theories of animal organisms
- variables
- how to construct classes of equivalence
- Linné(system): chose certain characteristics, and created his classification on the basis of those criteria, excluding the other traits
- Buffon(method): compare, regroup categories, gradually defining the table of kinships
- the historicity of discontinuities
- Linné: based on the combination of a small number of criteria, defines theoretical places in a potential space. These places may be filled to varying degrees, but one does not a priori know why.
- Buffon: Categories exist only in our imagination.
- Foucault’s archeology of taxonomy
- 04 The Durkheimian Tradition: Socio-logical Classifications
- Durkheim and Mauss(1903)
- establish scrupulous correspondences between the social and symbolic classifications used by primitive societies -> inherited by (Bourdieu, 1979)
- exclude “technological classifications” and “distinctions closely involved in practice”(classifications were not linked to actions)
- Durkheim(1897): created an objective sociology based on statistical regularities, and a sociology of “collective representations” that did not apply to the tools of knowledge and to the representations underlying statistical objectification.
- This construction is thereby reduced to an operation of “measurement,” inferior to the process of conceptualization.
- Thus the theoretical and practical examination of the definitions, nomenclatures, and encodings used by the social sciences has long been rendered almost unimaginable by this division of labor, and by the distinction between an ideal of scientific knowledge aiming at truth, and daily action dependent on categories deemed impure and approximative, if not biased.
- Durkheim and Mauss(1903)
- 05 The Circularity of Knowledge and of Action
- encoding stage in three domains
- Judges apply the law -> Cases accumulate, become law
- Doctors describe diseases -> 1830s, description rises argument
- Schools test students -> Students are divided into categories
- A circle: action-data-information
- Data produced by statistical institutions v.s. Data produced from activities of management
- The dynamic periods in the life of these bureaus correspond precisely to the times when they manage to link their investigations closely to crucial current questions, simultaneously producing the categories of action and the means to evaluate them
- the key of data-action – connection of knowledge-action
- encoding stage in three domains
- 06 Industrial Activity: Unstable Associations
- Knowledge results from descriptions of actions, which are necessary for stabilizing these actions
- The field of productive activities was more favorable to reflective excursions than that of the population
- 1788, Tolosan, The major categories referred to the origin of the raw materials employed: mineral products, vegetable products, animal products
- 1841, rather than statistics, the survey presented a series of monographs on various establishments, aimed less at counting than at locating and describing the as yet small number of industrial establishments
- 1861, a fresh industrial survey presented activities as classified in terms of their destinations
- 1870s, Information on industry and other economic sectors was produced in a circuitous way, dealt with the trades of the people employed, prompted a rearrangement of activities and establishments
- The criterion for rearrangement was now the technique of production
- 07 From Poverty to Unemployment: The Birth of a Variable
- During the 1890s, the criterion for classifying activities shifted from a market-driven logic to an industrial and professional logic, constructing unemployment, which would gradually replace the idea of poverty
- Great Britain
- before 1890s
- based on the activities of relief institutions that had existed since 1835
- through the organization of special surveys
- The Poor Law of 1835 was the source of a dual system of relief
- before 1890s
- Great Britain
- During the 1890s, the criterion for classifying activities shifted from a market-driven logic to an industrial and professional logic, constructing unemployment, which would gradually replace the idea of poverty
- 08 A continuous, One-dimensional, Hierarchical Social Space
- 09 From Crafts to Skilled Jobs
- 10 Four Traces Left by the French Revolution
- 11 One or Several Urns: Taxonomy and Probability
- 12 Making a Story Hold