3 ms·

Humanities Data: A Necessary Contradiction

28 points by diodorus 11y ago

acbart 11y ago
I help teach a "Computational Thinking" undergraduate course for non-CS majors. The topic at heart here is essentially Abstraction. Specifically "Data Abstraction", since there's more than one kind of abstraction in a bigger picture, but we don't try to tackle that nuance with our students. We say that Abstracting an entity (the most vague word we've found, referring to people, places, things, events, and even other abstractions) causes you to lose details. That abstraction is contextualized to a stakeholder, since you have to have some criteria for deciding what details you keep and lose. An abstraction for a person would be different if it was for Facebook ("Name", "Age", "List of Friends") or for a doctor ("Name", "Age", "Sickness"). Abstraction is powerful because it lets us concretize real, vague things into something that a computer can understand and process ("Because computers are very stupid!"). I often discuss how an abstraction loses a lot of valuable data. One student was working with NFL games, representing them a set of properties like ("away team name", "home team name", "away team wins", "home team wins"). That representation ignores how many interceptions were made, the weather that day, the feelings that the crowd had, how much effort the players put into training - a lot of "humanist" data. This article makes me wonder what goes through their minds. No one's complained about making these abstractions - they've never suggested we're ruining a game, a historical event, or whatever we're abstracting. But I think maybe we should encourage the students to think about what happens when you abstract something and what it means for the entity and not just the stakeholder.
- miriamkp 11y ago
  That's really interesting and helpful — thanks. I've heard CS people talk about abstraction a lot, but hadn't realized they were using the term in the way you describe.
  - acbart 11y ago
    I don't know how common my definition is among the CT community. Frustratingly, few researchers operationally define the term (cardinal sin in Education research!) and establish graded metrics based on it. I haven't met any other researchers who have resisted our definition, but there are other definitions. Repenning has one that is particularly... distinctive from my own: http://www.cs.colorado.edu/~ralex/papers/PDF/SIGCSE10-repenning.pdf http://www.cs.colorado.edu/~ralex/papers/PDF/SIGCSE10-repenn...
mcguire 11y ago
I would like to recommend Matthew Jockers' book, Macroanalysis if you saw something interesting in this article. Jockers is, among other things, a scholar of Irish and Irish-American writers in the 19th and 20th centuries (IIRC). Apparently, according to previous work, Irish-American authors had "gone quiet" in the first decades of the 20th century, a fact that needed explanation. One of Jockers' early results, however, was that the quiet period was an artefact of the limited set of authors studied---the "close reading" model that produces statements like, "we’ve immersed ourselves so deeply in our source material that we’re attuned to its nuances" also means that the majority of source material hasn't been examined at all. In this case, a certain subgroup of authors, male and on the east coast of the US, may have stopped producing, but that output was more than replaced by other authors.

28 points by diodorus 11y ago

I help teach a "Computational Thinking" undergraduate course for non-CS majors. The topic at heart here is essentially Abstraction. Specifically "Data Abstraction", since there's more than one kind of abstraction in a bigger picture, but we don't try to tackle that nuance with our students. We say that Abstracting an entity (the most vague word we've found, referring to people, places, things, events, and even other abstractions) causes you to lose details. That abstraction is contextualized to a stakeholder, since you have to have some criteria for deciding what details you keep and lose. An abstraction for a person would be different if it was for Facebook ("Name", "Age", "List of Friends") or for a doctor ("Name", "Age", "Sickness"). Abstraction is powerful because it lets us concretize real, vague things into something that a computer can understand and process ("Because computers are very stupid!"). I often discuss how an abstraction loses a lot of valuable data. One student was working with NFL games, representing them a set of properties like ("away team name", "home team name", "away team wins", "home team wins"). That representation ignores how many interceptions were made, the weather that day, the feelings that the crowd had, how much effort the players put into training - a lot of "humanist" data. This article makes me wonder what goes through their minds. No one's complained about making these abstractions - they've never suggested we're ruining a game, a historical event, or whatever we're abstracting. But I think maybe we should encourage the students to think about what happens when you abstract something and what it means for the entity and not just the stakeholder.

miriamkp 11y ago
That's really interesting and helpful — thanks. I've heard CS people talk about abstraction a lot, but hadn't realized they were using the term in the way you describe.
- acbart 11y ago
  I don't know how common my definition is among the CT community. Frustratingly, few researchers operationally define the term (cardinal sin in Education research!) and establish graded metrics based on it. I haven't met any other researchers who have resisted our definition, but there are other definitions. Repenning has one that is particularly... distinctive from my own: http://www.cs.colorado.edu/~ralex/papers/PDF/SIGCSE10-repenning.pdf http://www.cs.colorado.edu/~ralex/papers/PDF/SIGCSE10-repenn...

mcguire 11y ago

I would like to recommend Matthew Jockers' book, Macroanalysis if you saw something interesting in this article. Jockers is, among other things, a scholar of Irish and Irish-American writers in the 19th and 20th centuries (IIRC). Apparently, according to previous work, Irish-American authors had "gone quiet" in the first decades of the 20th century, a fact that needed explanation. One of Jockers' early results, however, was that the quiet period was an artefact of the limited set of authors studied---the "close reading" model that produces statements like, "we’ve immersed ourselves so deeply in our source material that we’re attuned to its nuances" also means that the majority of source material hasn't been examined at all. In this case, a certain subgroup of authors, male and on the east coast of the US, may have stopped producing, but that output was more than replaced by other authors.