7 min read.
This is a “billet d’humeur.” An opinion piece, if you want, but this expression lacks the nuance of mood and humour that the French version suggests. I want to share my feelings. Because I am a little angry.
This morning, I was trying to compromise between two things. Write a passage on Donna Haraway’s notion of situated knowledges for my PhD dissertation (I’m late), or follow an online workshop, in a project I participate to, where Danish social science and humanities scholars are offered help in digital methods and data science. I decided to follow the workshop. It offered a promising talk by a Danish physicist who had studied social interactions by analyzing GPS data with network science techniques. It did not make me so happy.
I am not going to say more about that guy, because he is a great guy and my rant is not personal. And I must say, his work was really nice. But I have a problem with his narrative. He made a number of points that I did not expect from someone standing on the bridge between the natural and the social sciences. His words caught me by surprise. Here is a selection of his clichés, commonplace ideas from a natural scientist selling the big data rhetoric. I presume he was not realizing it. Note: I do not have a record, so all of my quotes are from memory. I would do better if I could, it would be worth it.
Self-evidence. This physicist shows a video of moving dots over the map of a building. We clearly see different rooms, and dots assembling in them over time. He says: “Those are real people, with feelings and goals, and we see them moving around, living their everyday life.” By saying so, he conflates the visualization with the phenomenon. Yet those dots are not actual people: those are, in this case, students in a rich Western country, and they do not represent all human beings. That kind of bias is a major concern in contemporary social science. Their phones are also not always on, and not always close to their body. Yes, we do see relevant patterns. But the picture is far from self-evident.
End of theory. He says: “The amazing thing with this technique is that you do not need any assumption, any theory. You get knowledge directly by analyzing the data, it does not come with the flaws of other social science methods.” This is so 2008. We have seen that this work, of course, comes with assumptions. More importantly, developing a statistical processing and a model, as he showcases, piles up even more assumptions. People do not equate their body. To meet is not to be close, and to be close is not to meet. etc. For instance, I was more with that physicist, listening to him live on Zoom, than with the co-worker in front of me, doing her own thing. And each of these piled-up assumptions comes with its own flaws. At the end of the day, the findings are narrow and their reach very limited. And if you want to generalize to “human beings,” then you have to face the same need for assumptions as anyone else (i.e. social scientists). That is not quite the end of theory.
Superiority of the natural sciences. He says: “I can help social scientists with more modern methods.” At this point I was starting to get angry. I don’t think I need to explain the problem, but my emotional reaction deserves a comment. I know such misconceptions are commonplace, and that it is a cultural issue. If all you have ever been doing is maths and physics, then assuming the existence of universal laws is a given, since it has always been unreasonably rewarding to you and your peers. There is genuinely no reason, for you, to suppose it works differently for the social world. Yet it does. You have a personal bias, and by definition you ignore it until you face a different reality. But I was disappointed to see these points from someone in close contact with the social sciences. I believe that if you just read seriously a few sociology papers, your figure it out by yourself. But there might be an explanation.
That physicist had not one, but two narratives. The first narrative is the pitch, the story intended to sell his work. The clichés come within that story. The second narrative is the work itself, e.g. the paper. On that level, there is no need for grand concepts such as the “human being.” It is about data points, statistical analysis, and modeling. The validity conditions are clear, and the claims narrow.
I suggest that these two narratives work like two legs. When one gets weakened, you move your weight onto the other one. This way you can argue both sides of the academic border. When criticized for the limited relevance of your findings, you can draw from the legendary reach of physics, and its quantitatively proven universal laws. You could not do that in a paper, because your results would not be universal at all. But you do not have to, as long as you can sell the dream informally, elsewhere. When criticized for not acknowledging social science’s norms (preexisting knowledge and specific epistemological commitments), you can let the quantitative nature of your papers speak for itself. Your method is legit and your findings robust, even if your paper’s way of warranting knowledge is different from a large body of sociological work, with which it is poorly compatible. The informal gospel provides the reach, while the paper’s narrative provides the scientific robustness. You cannot argue both at the same time, but defending them separately works well enough.
I observed something similar with Barabási, a famous scientist and pioneer of network science. What he writes in his general public books (e.g. Linked, The Formula) is different from his papers. He is criticized for his grand claims or (alleged) beliefs in the existence of universal laws of the social. He is explicit about this vision in his books. Yet such disputed claims are absent from his papers, and his work is much more solid than what his critiques seem to think. His papers have narrow findings with clear validity conditions. He might suggest interpretations, that is all. The inflated interpretation as findings of universal reach is made separately, outside of peer review, in other spaces. And to be fair, his vision has the benefit of giving a meaning to the otherwise dry findings of network science. His vision has to be challenged on a scientific ground, but regardless of the outcome, it has the benefit of existing.
Casually hiding methodological commitments
As researchers, beyond our cultural differences, we have a responsibility not to spread problematic ideas such as the self-evidence of visualization, or the absence of methodological commitments to data science. Visualization is never self-evident, because there is always a method. We must not hide it, because we must account for its flaws. Data is never raw, it is always obtained, and it comes with its own biases. We know that. We know that we know that. Yet we let ourselves go down this slippery slope in informal spaces. We must stop selling this rhetoric, especially when teaching. It is toxic, and unnecessary. It does not empower anyone worth empowering.
The funny and sad conclusion is that I went through this instead of writing on Donna Haraway’s situated knowledges. She was precisely making this point. At the time she wrote her famous paper, I was only 8 years old. I wish we had since learned her lesson. She calls this enactment of self-evidence and absence of epistemological commitments “the god trick of seeing everything from nowhere.” I leave you with her, a challenging but good company. Enjoy the verve.
The instruments of visualization in multinationalist, postmodernist culture have compounded these meanings of disembodiment. The visualizing technologies are without apparent limit. The eye of any ordinary primate like us can be endlessly enhanced by sonography systems, magnetic reasonance imaging, artificial intelligence-linked graphic manipulation systems, scanning electron microscopes, computed tomography scanners, color-enhancement techniques, satellite surveillance systems, home and office video display terminals, cameras for every purpose from filming the mucous membrane lining the gut cavity of a marine worm living in the vent gases on a fault between continental plates to mapping a planetary hemisphere elsewhere in the solar system. Vision in this technological feast becomes unregulated gluttony; all seems not just mythically about the god trick of seeing everything from nowhere, but to have put the myth into ordinary practice. And like the god trick, this eye fucks the world to make techno-monsters.