Drag Queen Data Science

Comments

• quinnanya

  February 7

  I love this idea. To riff on it further: one of the things that frequently bugs me about data science projects is the tendency to reach for easily-available data, without asking all the questions that you should be about where it comes from, who put it together, for what purpose, what is it missing and why, etc. Data as something you just buy off the rack at Target.

  The weeks before the annual DragFest on campus are always some of my favorites at the Textile Makerspace I run, because the contestants show up with the most wonderful, whimsical plans, and even when they're cobbled together with just the scraps we have lying around, they're so thoughtful and deliberate in their work. That's the kind of relationship with data I'd like to see more often. Because, oof, sometimes those off-the-rack data sets are just three awful problems in a trenchcoat.
• cleogirl2525

  February 8 edited February 8

  Hi! I am reading this and catching on to a few things ~ I'll list some keywords here. Not sure exactly where to locate the central direction of this exchange just yet but for me, I'm reading something about "intentionality" and about "organization," and maybe even "success" or "authentication." Which are things I know to important when thinking through large datasets.

  I'm wondering if the beauty of drag-data or drag-as-data sets is that it actually can kind of be "three awful problems in a trench coat" (or maybe always are, lol - I've never heard this phrase but it's profanity is kind of extremely perfect for this...?) and it can still be sensational.

  For instance, can we argue/begin to question, like, isn't part of transgression of queer performativity (not just drag but any stagework/performing of identity markers) in fact being precisely that messy-trouble data set (messy/regular/basic because there's no world in which intentional construction exists because the mainstream rules cannot encompass the queer individual and also don't even recognize the individual as an individual or their identity construction as a kind of data) and also as a result, making a big and deviant departure from a kind of intentionality that follows logics that conform to rules-making bodies?

  Is performing something "too well" or "too intentionally" or even "successfully" or even "towards success" (as to satisfy a rule-making body's perimeters ~ say an application someone has written if we apply this logic to real data) something that creates the kind of performativity queerness and also drag is attempting to begin breaking away from?

  On the other side of that - in the Target model - the replication of foundational models feels extremely important in drag (we would still love a Target queen!!) What could this mean for our conversation around data sets and more importantly, how could this give us a lens into what unites drag performers as individuals just as data object markers or locations can mean something to both the collective of a dataset and to its self as a member of the collective within the larger body of other data? What could this say about how we work through, tend to, or build space for these problem-sets?

  If we move away maybe from the live stage and move into campy queer drag moments in cinema, I'm beginning to think alongside Susan Sontag's Notes on Camp here. Maybe we can look at definitions of "success" and "intentionality" and 'foundational modeling" as compared to performances of identity or self-fashioning that are sensational and fun or inspiring or even related to queer joy and ecstasy.

  @jordanfreitas your slides and especially your "Make it make sense!" moment feels very in line with this line of questioning.

  
• jeremydouglass

  February 20

  @jordanfreitas -

  Thank you for sharing the slides from this lightning lecture. Pulling out your "what if" questions from the middle of the deck:

  What if data science were more like drag?

  • if quantification were more apparently only a part of a more nuanced reality,
  • and combined with other ways of knowing.
  • if statistical analysis were more often used as a tool for challenging the status quo and bringing visibility to societal problems

  • instead of pressure to conform or not be counted
    What if data science were more like drag?

  • allowed for contradictions instead of reinforcing binary modes of thinking

  • avoided imposing categories and inventing boundaries where the reality is more fluid

  ...I was struck by the way that you found in drag answers that then raised questions about data science, and how the form of those question-answers (drag queen data science) are both specific to the metaphor and yet also seem (to me) to share a common language with a number of related inquiries into alternatives to computational knowledge/power. I'm thinking of words like "quantification" "binary" and "imposed categories" here.

  When I looked back over recent meetings of the CCS Working Group in our Archive and thought about related past conversations in the working group on getting beyond "imposed categories," I recognized them for example in part of Jon Corbett, Outi Laiti, Jason Edward Lewis, Daniel Temkin writing in Indigenous Programming about different ways of knowing, in particular discussion of paradigms in Jon Corbett's Cree#. Trying to think beyond categories can also generate violent pushback, as discussed in e.g. Elizabeth Losh, Judy Malloy, and Jacqueline Wernimont writing in Gender and Programming Culture:

  When Ari Schlesinger asked if there could be a feminist programming language on the HASTAC blog in 2013 (this later became a central thread of CCSWG14), based on experiences with the critical code community, Reddit commenters at The Red Pill attacked her as an opponent of logic and mocked her with satirical guides to pseudocode.

  I suppose where this chain of thought is leading me is how, when engaging with power at a technical level, many people may find themselves walking shared roads, even if their frames of reference (e.g. drag queen culture or Cree childhood education) are very different. In 2018 Safiya Noble, Jessica Marie Johnson, and Mark Anthony Neal writing in Race and Black Codes also critiqued the separating the discourse by fields (e.g. gender studies here, race there, queer studies over there) in this way:

  This week is a reminder that when dealing with race and code, there is no race without gender, sexuality, and related identities or the structures that enforce/maintain them. As Johnson and MAN noted in "Wild Seed in the Machine,” Black Code Studies IS "queer, femme, fugitive, and radical." Which is to say, there is no discussing Blackness outside of or beyond a discussion of gender, sexuality, ability, ethnicity, power, and precarity.
  Which is also to say--Black Code Studies will be feminist, queer affirming, trans* defending and invested in social justice or it will be bullshit.
• markcmarino

  March 2 edited March 2

  I am loving this idea of applying Drag to data science, and wanted to offer a few ideas...

  In their introduction to a special issue on "algorithmic antagonisms", Heemsbeegen, Treré, and Pereira offer two examples of 'queering data,' which is I think adjacent to data science drag, and keeping with our Queering Code week. Drawing heavily on Cox and Soon, they offer the examples of

  Searchatlas.org (Ochigame and Ye, 2021), which questions the normativity of search results by making the ‘information borders’ that search engines produce visible, where they otherwise would not be. This tactical data practice remediates the bordering of Google’s mission to organize the world's information by showing the hidden normalisation of Google’s model across regions and languages. The example given by the researchers is how searching for God

  produces thematic tapestry of images that mirrors geographies

  Their second example is admittedly more on the nose from Jake Elwes'

  The Zizi Show (2020) creates deepfake drag performances that agitate normativities around sexuality and the objectivity of video through a novel algorithmic perfomativity. Zizi queers drag as much as it queers the algorithms in question by offering interactive Deepfake versions of songs from multiple and blended performers at zizi.ai/. As Elwes puts it “As an artist, I’m so interested in when these [computational] systems break down…in seeing these algorithms not do what they were intended to do” (Elwes, 2021 qtd. in Heemsbergen et al.).

  Do these examples fit what you were thinking of @jordanfreitas? Let me offer my own example which comes from our ELIZA and DOCTOR code critiques. I often ask our team (esp @anthony_hay and @aschwarz ) whether the DOCTOR script is data or code. I believe their answer is data. But I am tempted here to call it data in code drag or vice versa. I don't mean to appropriate a cultural practice for my metaphor but am trying to follow your example of seeing drag as a performance that plays with boundaries and helps people see the construction of those boundaries by transgressing them with flare.

  I think the difference in the previous examples is in what Heemsbergen et al. call "the tactical nature of the algorithmic practices." In other words, perhaps drag requires an intention to play performatively with boundaries, rather than something we find latently there.

  But what I am getting most from your slides is that data science as drag is not a practice of looking for drag in the data but instead developing a scientific practice or mindset or -- orientation toward data inspired by the disruptive performative play of drag. Is that getting close?

  Heemsbergen, L., Treré, E. and Pereira, G., 2022. Introduction to algorithmic
  antagonisms: Resistance, reconfiguration, and renaissance for computational life. Media
  International Australia, https://doi.org/10.1177/1329878X221086042
• anthony_hay

  March 2 edited March 2

  Hi @markcmarino. Code or data. Is this analogy useful? When I play a record on my gramophone it is the geometry of the grooves on the disk, combined with the design of my gramophone, that determines the sound it emits. (The gramophone is the ELIZA code, the wiggle in the grooves is the ELIZA script.) Is the wiggle code or data? We might say the sound is encoded in the groove. Do the grooves "program" the gramophone to output Bauhaus rather than Beethoven? Code and data are both the same thing and different things, for some definition of "same" and "different." (See Data And Code Are The Same Thing) Maybe this is similar to a person being both the same and not the same when in and out of drag?

  Hi @jordanfreitas. I would love to have seen your lightning talk though I may not have understood it because I'm not a data scientist. My impression was that data scientists tried to be honest and embrace messiness. E.g. I think there are 100 different climate models in CMIP6. Source

  

  avoided imposing categories and inventing boundaries where the reality is more fluid

  I don't know exactly what you are referring to here. Isn't the point of science to be a tool to help us understand (model) reality? We could keep talking shop forever (and will do) but don't we need to reach tentative conclusions (and keep them under review in the face of new evidence) if we are to stop babies dying from measles?