Questions: Mary Poovey: History of Statistical Thinking

Question 1

A historian uses 19th-century British census data to analyze unemployment trends and notices that domestic servants, subsistence farmers, and informal traders are inconsistently categorized or omitted. What would Poovey's framework suggest she do?

• A. Exclude the inconsistent categories to maintain analytical cleanliness
• B. Abandon quantitative analysis — the inconsistencies make the data unusable
• C. Interrogate the categories themselves: who defined 'employment,' for what administrative purpose, and whose labor was thereby rendered invisible
• D. Supplement with narrative sources, which are inherently more reliable than statistics

Answer: (C)

Poovey's intervention is not to abandon quantitative history but to demand reflexivity about it. The inconsistencies are not errors to be corrected or obstacles to be ignored — they are evidence of interpretive choices embedded in the data at the moment of its production. When British census-takers used a category like 'employed,' they imposed an industrial conception of labor that excluded domestic work and informal exchange. Analyzing the dataset without asking who made these choices and why reproduces those choices uncritically. The inconsistency is itself the subject of analysis, not just a methodological problem.

Question 2

What is Mary Poovey's central argument about the 'modern fact'?

• A. Modern statistical methods are more accurate than pre-scientific narrative accounts of the same phenomena
• B. The idea of a fact as self-evident, impersonal, and detachable from context is itself a historical construction, produced through specific practices like double-entry bookkeeping and political arithmetic
• C. Facts are purely subjective constructions and should be replaced by interpretive narrative
• D. Statistics became genuinely objective in the twentieth century when mathematical probability theory was formalized

Answer: (B)

Poovey argues that 'the modern fact' — the notion of a discrete, context-free, self-evidently true unit of knowledge — is not a natural feature of the world but a historical invention. Before the seventeenth century, 'fact' was a legal term (from Latin factum, meaning a deed). The idea that facts could be impersonal and self-evidently true was constructed through practices like double-entry bookkeeping and natural philosophy that developed genres of writing claiming to separate observation from interpretation. Statistics emerged as a technology for making populations legible to states — but those numbers required prior choices about categories, and those choices were never neutral.

Question 3

Statistics are more objective than narrative sources because they record direct observations of social reality rather than interpreted perspectives.

• T. True
• F. False

Answer: False

This is precisely the view Poovey challenges. Statistical datasets require prior decisions about what to count, how to categorize, who gets counted, and which differences matter. Nineteenth-century British censuses embedded legal and moral definitions of poverty into the 'pauper' category; colonial censuses imposed European racial and occupational categories onto populations they did not fit. These decisions are interpretive, political, and context-dependent. Numbers present themselves as impersonal and objective — that presentational style is itself historically constructed — but they require the same critical scrutiny as narrative sources. The question is not whether to use numbers, but how to read them.

Question 4

According to Poovey, historians doing quantitative work should ask who produced a dataset, for what purpose, and using what categories — treating numbers with the same critical scrutiny as documentary sources.

• T. True
• F. False

Answer: True

Poovey's challenge to historians is methodological, not a rejection of quantitative practice. She calls for what can be described as 'critical quantitative history': use the numbers, but read them the way you read a document — with attention to authorship, purpose, and context. Before analyzing a historical dataset, ask: who compiled these numbers, for what administrative or political purpose, using what categories, and who was not counted? Slave manifests counted human beings as cargo; baptismal registers undercounted dissenting communities. Every dataset carries its producers' assumptions. Reflexivity about those assumptions is the professional obligation Poovey's work establishes.

Question 5

Poovey argues that 19th-century census categories like 'unemployed' or 'pauper' did not simply describe pre-existing social reality. What does she mean, and why does this matter for historians using these datasets?

Model answer: Poovey means that before a census counted 'paupers,' someone had to decide what made a person a pauper — a decision that embedded a legal and moral definition of poverty into what appeared to be a neutral numerical observation. The category 'unemployed' required prior adoption of an industrial conception of labor that defined regular wage work as the norm, making domestic labor, subsistence farming, and informal exchange invisible or marginal. The statistics did not describe a pre-existing social reality; they produced social categories and made those categories appear natural and quantifiable. For historians, this matters because analyzing trends in 'unemployment' or 'pauperism' without interrogating the category definitions inadvertently reproduces the assumptions of the administrators who created the data, treating historically contingent distinctions as if they were objective facts about human beings.

This question gets at the deepest level of Poovey's argument: that the act of counting is always also an act of categorizing, and categorizing is always a value-laden, historically situated activity. Quantitative historians who ignore this risk mistaking the census-taker's administrative framework for social reality itself.