week3-language-as-data-1

Up: 011_data-and-disc-lect Prev: week2-what-is-discourse-analysis Next: week4-language-as-data-2

Language as Data 1

Online Lecture Notes:

• Overview of “Data”

  • Etymology and Definition
    • Derived from the Latin word datum (something given).
    • Defined by the Oxford English Dictionary as collective information, often numerical, used for analysis, reference, or computation.
    • In computing: quantities, characters, or symbols processed collectively.
  • Public Perception of Data
    • Often linked to concepts like big data, AI, and digitization.
    • Popular images depict data as abstract, impersonal, and intimidating.
    • However, these representations often exaggerate the power of data systems.

• Algorithms and Their Relationship to Data

  • Definition:
    • A set of rules or procedures for calculation or problem-solving.
    • Algorithms are implemented in programming languages (e.g., Python, Java) and executed as software components.
  • Interaction with Data:
    • Algorithms process input data to generate output, forming the backbone of software systems.

• Contextualizing Data in Social and Technical Discourses

  • Two Approaches to Conceptualizing Data:
    1. Traditional View:
       • Data as “given” an objective, forming the basis for knowledge and decision-making.
       • Criticism: Data collection always involves selection and contextualization, making it a subjective process.
    2. Critical Perspective:
       • Data as “reinterpretation” or formalization of complex phenomena.
       • ISO/IEC defines data as a formalized representation suitable for communication and processing.
  • Implications:
    • Data is not neutral but shaped by social, cultural, and institutional contexts.
    • Collecting and interpreting data is a constructive process, influenced by perspectives and interests.

• Critical Data Studies

  • Key Points:
    • Data reflects and reinforces power structures (e.g., corporate control, algorithmic biases).
    • Researchers emphasize that data is “cooked,” not “raw,” shaped by selection and recontextualization.
    • Examples of biased data systems:
      • Gender and racial biases in facial recognition.
      • Misguided word associations in language models (e.g., Man is to Computer Programmer as Woman is to Homemaker)
      • Algorithms optimized for efficiency, often ignoring individual and societal impacts.

• Applications and Challenges

  • Data in Journalism and Politics
    • Data journalism analyzes structured information to uncover newsworthy stories (e.g., emergency services coverage, election analytics).
    • Data-driven campaigns, such as Barack Obama’s election strategy, illustrate the growing influence of big data in politics.
    • The misuse of data (e.g., disinformation) highlights its role as a form of power.
  • Ethical Concerns:
    • The need for iterative improvements and debugging of data systems.
    • Questions about fairness, transparency, and accountability in algorithmic decision-making.
    • Challenges in identifying and mitigating unintended harms caused by data systems.

• Key Quotes and Insights

  • “Data is always reinterpreted information”
  • Algorithms often prioritize efficiency and profitability over individual well-being and societal health.
  • “Data are never raw but always cooked” — emphasizes the constructed nature of data.

• Closing Thoughts

Lecture Notes

• Two paths of conceptualisation of “data” as a term:

  1. Data is something given that can be captured and further processed. Data does not have a meaning by itself.
     • Criticique: Data gets always selected, perspectivated and contextualised when collected.
  2. Data is something that can be obtained by reduction, selection or formalization from unites with a higher degree of complexity. Data is always cooked, and is never raw.
• IT perspective:
  • A reinterpretable representation of information in a formalized manner, suitable for communication, interpretation, or processing. Examples of data include a sequence of bits, a table of numbers, the characters on a page, the recording of sounds made by a person speaking, or a moon rock specimen.
• In Digital Discourse Analysis:
  • A datum is a phenomenon interpreted as a sign, which in the course of a research process is extracted from a given complexion and recontextualized.

• Critical Data Studies

  • “Data as a form of power”
    • Controlling data
    • Manipulating data
    • Gathering Data