Day Two: Epistemologies of Big Data
SDS 237: Data Ethnography
Lindsay Poirier
Pull out a piece of paper, and write three things you know to be true.
Turn to your neighbor and discuss:
- What does it mean that you know this to be true? What counts as knowing?
- How/through what means do you know this to be true?
- What are the limits of your knowledge on this?
Epistemology
- Greek words
- Episteme”: knowledge; understanding
- “Logos”: reason; argument
- Philosophical study of the nature and limits of knowledge
- What conditions must be met for us to say that we “know” something to be true?
- How do a group of people come to acquire knowledge?
What counts as knowledge?
Historically knowledge defined as “justified true belief”
Does knowledge exist independently of a knowing mind?
- Positivists claim yes, there is objective truth independent of a knower
- Interpretivists and constructivists claim no, truths are subjective or tied to a knowing mind
How do we acquire knowledge?
Empiricists claim that knowledge emerges from direct observation
Rationalists claim that knowledge emerges from logic and reason
…and then there’s testimony, memory, intuition, feeling
Interrogating Epistemic Beliefs about Data
- How does a community understand the relationship between numbers and truth?
- How does a community characterize data origins or originary forms? Where do they suggest that data comes from?
- How does a community understand the means by which knowledge and meaning emerges from patterns in data?
- What analogies or metaphors does this community engage to characterize what data is or what data is like?
- How does this community frame the role of culture, human judgment, and bias when evaluating data?
Group Exercise
- There are large sheets of paper and markers on each table.
- In one section of the paper, write down phrases that capture the hype/imaginaries/mythologies of Big Data/AI.
- In one section of the paper, write down phrases that capture the actual work of machine learning.
- Select one question from the previous slide and discuss how epistemic beliefs about Big Data/AI differ from each perspective.
What social forces are at play in separating hype/imaginaries/mythologies of Big Data/AI from the actual work building and interpreting models?
Feminist Epistemologies
- All knowledge is embodied
- Contrast with disembodied knowledge - i.e. not tied to a specific body
- Bodies are situated in certain social positions and have a finite point of view (Haraway 1991)
- Critique of the “unmarked body,” the “God trick,” or the “view from nowhere”
- Knowledge is tied to particular standpoints
- Our experiences, what we’ve read, our education, our social positions, and what our bodies enable us to do, see, hear, taste, touch, and smell
- Factors are innumerable and unique to every person
marsroverdriver, CC BY-SA 2.0 https://creativecommons.org/licenses/by-sa/2.0, via Wikimedia Commons