Terms for Midterm

Datafication

transformation of human activities, relationships, and aspects of life into machine-readable, quantifiable records to create value or make decisions

Four Pillars of Ethical Reasoning

• Utilitarism/Consequentalism

• Deontology

• Virtue Ethics

• Care and Justice

Utilitarism/Consequentalism

focuses on the outcomes of actions and emphasizes benefit over harm

Deontology

focuses on universal rules, rights, and regulations

Virtue ethics

focuses on the character, intention and integrity of a creator or institutions

Care and Justice

focuses on relationship retention and acknowledging vulnerabilities to distribute benefits, burdens, and opportunities evenly

Ethnography

researcher immerses into a specific community to observe behavior, culture, and social interaction in its natural context, turns researcher into “human sensor”

Different contexts of ethnography

• Cultural Context

• Historical Context/Relativizing

• Situational Context

Cultural Context

evaluating moral rules, taboos, and traditions within the framework of the society practicing them

Cultural Context

demonstrates that identical metrics or behaviors carry different meanings depending on where they occur

Situational Context

acknowledges that an individual’s outlook or declared truth is shaped by their unique experiences

Situational Context

cuts through absolute thinking and shows that data cannot be evaluated correctly without knowing the environment in which it was produced

Historical Context/Relativizing

analyze figures and events within the social, scientific, and ethical knowledge available during that time

Historical Context/Relativizing

stops researchers from projecting modern, Western biases on foreign and ancient cultures and rejects ethnocentrism

Rejecting Ethnocentrism

using relativizing to offset the human habit of assuming our immediate environment is “correct” or “natural”

Ethnography

where research immerses themself into a community to observe culture, behavior, and social interaction in its natural context

• allows researcher to understand the social systems from people who actually live it

Thick descriptions

ethnography captures context, nuance, and human accounts for data scientists, providing the “why” to datasets

Participant observation

• taking field notes while actively participating in the daily lives and routines of a community

• valid knowledge requires engagement

Adminstrative recoding

selecting a single feature out of a complex social practice and reclassifying it into a standard metric to place value/pass judgement

Mechanical objectivity

scientific ideal that says instruments and photographs are ideal observers because they are not affected by human biases and imagination that will “pollute” raw data

Truth-to-nature

illustrators drew perfected versions of specimens rather than raw versions

Blind sight

Daston and Galison term for elimination of expert interpretation, replaced by a objective machine

Depersonalization of knowledge

knowledge has to be communially agreed upon and requires checks and balances for public numbers (common metrics) to keep us from lying to each other

The Great Romantic Rupture

• imagination demonized and artists viewed as vital but dangerous and needed to be kept to a rational mind

• separating the artist from the scientist

• subjective vs. objective

Objectivity as Communicability

defining objective knowledge as information that can be shared globally without any information loss which produced national languages in school

Teology

Aristotle’s pre-modern principle that all objects have an internal purpose they are trying to carry out

• 17th century revolution smashed this worldview as common sense was valued over abstractivity

Primary vs. Secondary Qualities

• Galileo and Locke divide primary geometric properties (size, shape, motion) from secondary human sensations (taste, color, warmth)

• Reading the human observer completely out of legitimate knowledge

Newtonian Synthesis

• Issac Newton unifies terrestrial and celestial mechanics under a single equation

• Refuses to hypothesize and makes perfect nature science —> equation for gravity

Value-Neutral Illusion

statistical data, collection method, and math analysis should be free of politics, morals, and bias —→ hides political choices behind objective figures

Legibility

• Modern states create simplified, standardize models of complex and dense societies

• makes it easier to tax, govern, and control societies

State simplifications

Map, census, and standard measures introduced to reduce infinite detail and shift power away from local figures

Esprit geometrique

Geometric layout of communities to organize space to clear at illegible alleys, pushes the poor to the perimeter

Standardization of surnames

• Assign permanent inherited and alphabetized names to people

• Governor Clavaria 1849 decree in Philipennes

• For taxation in military deployment

Linguistic centralization

• official national language over regional dialect and vernaculars

• Devalues local knowledge and make sure everyone speaks a language discernible to the state

High modernism

• Idea that geometrically ordered environment will cultivate more orderly, moral citizens which produces well-designed disasters

• State planning breaks human networks needed for organic social survival and the paradox of state ethics

Métis

• represents practical, local, and deeply-contextual knowledge born from direct experience

• Cannot be translated into bureaucratic data metrics

The disinterested observer

• Ethics of justice uses legal agents (juries and judges) to determine right or wrong with required emotional detachment

• erases local tribunals, blood feuds, and informal mediation

The paradox of state ethics

State treat everyone equal by atomizing them and removing context, they end up treating them inhumanely

Recursive Co-evolution

• Administrative tools for state governance and ethical framework produced and justified one another

• ethics shift from contextual and relational to the detached third person

First and second person ethics

• ancient/medieval systems relied on relational, virtues, honor, community, accountability, kinship and loyalty

• ethics are Co-sustained

The third person shift

Enlightenment philosophy, built “the view from nowhere” and ethics are independent of personal identity or history

The God trick

• Donna Haraway critiques a third person shift and says ethical and scientific truth are always contextual and therefore cannot be state-leveled flattened

Situated knowledge

• Haraway says the ethical and scientific truths are always partial and contextual

• True moral accountability cannot be had when these truths are flattened

The self-defense affordance

Webb Keane says humans can move between first second and third person perspectives

• how universal rule enforcement becomes possible and highlights clash when these rules clash with relationships

The interaction order of ethics

• Jack Sidnell says true ordinary ethics are embedded in real time

• interaction is always moral balancing, people are making sure they’re being listened to and respected

Sequential accountability

• moral accusations happen linearly and in real time

Service mode

Götz Hoeppe recognizes the shift from “visitor mode” where researchers gather their own data —> data production line style we researchers are disconnected from the creators of the data

Control room co-Presence

Re-entry of data consumers in the workplace of data producers which allows for mutual second-person, social accountabilities

Disadvantageous inequity aversion

Frans de Waal term for an emotional outbreak when someone receives unfair treatment compared to their neighbor

Commensuration

Transforming fundamentally different quantitative categories into data through a common metric

• suppresses notable differences for comparison

Ontological transformation through number

Quantification with Shifts human mindset from “what kind” of entity to “how many”

Incommensurables

categories of life that cannot be quantified (sacred lands, children, and friendships)

Ontologies of census taking

Staffan Müller-Wille statistics derive their power not from the numbers themselves, but from the pre-structured content

Reification

once a person or object is assigned a numerical index, it turns into a stable fact and masks human choices that went into creating a metric

The chain of traceability

• A record showing how a measurement connects back to an official standard

• shows measurements are trusted because they are consistently used

Abstract Labor

• Marx critique of how industrial capitalism completely erases the specific, unique traits of human work

• shows how businesses can reduce human work to numbers to calculate costs and profits

Metrological Reformation

using physical objects to set measurement standards to using scientific experiments

The Circulations of Particulars

• measurements become accepted everywhere because people share tools, objects, and practices between different places

• universal standards are created through human effort

In-Plant Quality Tracking

written records to track equipment and its accuracy instead of regularly checking the equipment in person

Uncertainty Absorption

• messy or incomplete information gets simplified and passed to managers

• results in clean, confident numbers without realizing how much information was missing or uncertain

Black-Boxing (Latour)

Hiding the complicated steps and decisions behind a result so people trust numbers without questioning how it was created

Captation

Guiding people through information in a way that makes them accept a particular conclusion as numbers make an argument seem correct

Funny Numbers

• Changing how numbers are calculated or presented to make results look better

• Institutions can make themselves appear more successful to receive money or support

• E.g. hospital only counts certain patients when calculating success

Circular Discharges

same patient is repeatedly counted as successfully treated after being discharged even if they return later

Thin Prescription

Judging a complicated person or institution using only a few numbers people may focus on improving their scores instead of solving the real problems

Neoliberal Centralization via Decentralization

Letting local organizations make decisions about their work while forcing them to meet strict targets set by leaders

• organizations look independent but the numbers control what they do

Curricular Re-Engineering

Changing what schools teach to help students get better test scores focusing on passing tests more

The Statistical Capacity Lens

Believing social problems can be solved by collecting and analyzing numbers overlooking deeper causes, such as unfair treatment, culture, and unequal power

Continuum Segmentation / Arbitrary Breaks

• Dividing human trauma into artificial categories to make it easier to compare

• may dismiss victims’ fear as too subjective, causing psychological abuse to be classified as only “moderate” rather than recognizing its full severity

Data Double

• Organizations often make decisions about people based on these simplified version of a person represented by their personal information, records, and numbers

• borrowed from Haggerty and Ericson’s “survelliance assemblage”

• represents the quantified version of individuals as they look toward the government

Advocacy acting as Objectivity

• Using data to make a personal belief or prejudice look like an objective scientific fact

• real measurements can support unfair ideas if the research is designed around bias

Dynamic Nominalism (Ian Hacking)

naming and classifying a type of person can change how people understand and behave as that type of person

Aggregates vs. Data Doubles (1930s Clash)

disagreement over whether to collect broad information about groups and the economy or detailed information about individual people

The 1935 Turning Point (Contributory Pensions)

change to Social Security that connected a worker's benefits to their recorded earnings over time

Data Gravity

Old records can be used for new purposes, increasing the amount of information collected about people

The National Data Center Backlash (1965–67)

Public opposition to a proposed government system that would connect many types of personal records in one place

• attention to the need for laws that protect peoples private information

Discourse of Dispossession

(1951) Henrietta Lacks's cells extraction post-death without her permission, economic exploitation and biological control

Folder Thicker Dynamics

Black students had thicker school records than other students because more information and notes were collected about them

Ownership Talk Critique (Sarah Igo)

treating personal data as property that people can own or sell may not solve the deeper problems with collecting it

• Instead of only asking who owns the data or who gets paid, we should also ask whether the data should be collected at all