The fallacy of attribution

I think, on some level, when we marvel at how the world has become more egalitarian, we think about King and Lincoln. But do we think about that changes relationship to, say, Louis Pasteur or Henry Ford? How much of our march toward humanism is really technological, a shift from a slavery of human, to a mastery of nature? What were the effects of, say, Koch’s work on tuberculosis on how we came to see ourselves in relation to each other? Was the retreat of death a boon for humanism? Was it a bane for religion? Laundry used to be a heavy-duty chore, mostly performed by women. When washing became automated, how did that effect women’s notions of self? How did it influence feminism? Is humanism a luxury, afforded by individual longevity? —TNC

As any good STS aficionado, my first instinct was to point out that humanism itself was partially responsible for the retreat of death. Human progress is far too often cleanly and directly attributed to the advance of science. Because of science and technology, human beings conquered abject deprivation. Because of science humans now live longer, fuller lives. But of course the story is not so simple. At least for the retreat of death, social reformers played a non-trivial role. There’s even some supporting research on this.

Also typical for STS aficionados (myself included) is the tendency to take this analysis too far. To downplay the the very real benefits science has wrought. To marginalize the importance of rationality and (as problematic as the term may be) scientific thinking. In a sharp follow-up comment abk1985 suggests I did so, and notes that “humanistic agendas [were not] pursued in a void.” The drive to improve sanitation in poor immigrant communities hinged on intelligent medical guessing and enlightened self-interest as well as an honest desire to help the less fortunate.

For better and for worse, facts about the world do affect our moral intuition. Though at times humanism did stand on its own terms, it’s often not possible to isolate it as the driving motivation. It makes no sense to say that early 20th century reformers were motivated 33% by altruism, 33% by self-interest, and 33% by science. For some, a latent desire to address poverty was surely buttressed by data that indicated epidemics in certain communities often spilled over to everyone. All of our actions are imbued with such motivational complexity.

And so while it may be interesting to argue over a beer, and while I’m sure many academics will study it, TNC’s “how much” questions are ultimately not helpful. Even attempting to determine how much of human progress can be attributed to science, or to humanism, or to the rule of law, obscures more than it illuminates. The arc of human history cannot, even in principle, be decomposed and deconstructed like this. Science and technology have always been there and have always played a role. At times they were very important and at others not so much. And that’s pretty much all we can say about that.


  1. Way to end with a cop-out. “And that’s pretty much all we can say about that.”

    The reminder that science and technology were always there and always involved is not as obvious or as well-understood as it might be to those of us working with it daily. It’s all too easy to fall into traps of thinking about science and technology as simply the high-tech and the cutting edge to forget that what we accept as common – as old tech – has profound impact. It’s just hidden by history and ubiquity. I see TNC as trying to articulate those things. Edgerton’s “Shock of the Old” and Ruth Cowan’s work (especially “More Work for Mother,” since it directly addresses the laundry question) are just two entrees into that kind of re-exposure of influence.

    Asking “how much” does not necessarily mean finding (or trying to find) a quantitative answer. (Most of the STS sub-fields don’t even try.) What I take from TNC’s questions is an effort to expand considerations of influence beyond what – at least for him and others not steeped in science and technology – are seen as the typical suspects. And while STS scholarship all too often tends to substitute one set of typical suspects for another, it is set up to broaden considerations of influence and networks of activity.

    I respect your effort to push back against the appearance of technological determinism here. But I think it was overshadowed by a resignation from trying to identify trends and influences in acknowledgment of the incredible complexity of the task. Some STS theory suffers from the same problem. Comparative measurements make lack the precision and universality of empirical data, but they are not without value. The acknowledgment of complexity and the challenge of capturing context need not require the throwing up of hands. Consider it an incompleteness theorem like Godel’s in mathematics – they didn’t throw out the logical systems for their incompleteness, they recognized their limitations and muddled through. We should do so well.

  2. Hey David. Thanks for the comment, and sorry for the slow response. As you noted, I instinctively do push back against the appearance of technological determinism. It’s probably a reaction to my experience in the hard sciences, where you do often hear in casual conversation people giving science all (or most) of the credit for everything.

    I like your phrasing “broaden considerations of influence and networks of activity.” I think you’re right that TNC was trying to do as such for his audience, and I probably over-interpreted his “how much” question.

    I agree that comparative measurements have much value, and I wasn’t trying to say that we should throw up our hands. And you’re right, I did have a weak ending and I should have elaborated more. What I was trying to say is that if you’ve sufficiently broadened your network to include science, humanism, civil society, etc., it is very hard to make general attribution claims. You probably can do comparative or interpretive analysis to study specific cases, and try to glean whether certain parts of the network were more or less important.

    Finally, I like your analogy to the incompleteness theorem!

  3. The devil of it is that a reasonable strategy to fight determinism (technological or not) is to try and do exactly what STS and TNC are struggling with – recognizing the multiple influences on any particular phenomenon. (It also makes for the best – and hardest – history) It’s really easy to go too far in one direction or the other.

    Drawing the line(s) between what to include and what not to include is critical, and could be much more transparent than it currently is. Some of that is unintentional – developing methods and tools for experiments leads to some assumptions and influences being hidden (the ‘black-boxing’ cited by many an STS devotee). But including everything leads to theories and observations collapsing under their collective weight.

    Yes, it will be challenging to make general attribution claims in situations where influences (call them variables if you like) are difficult to control. But several fields of study work on case methods with some level of success. Perhaps the ‘fuzziness’ and ‘softness’ of the social sciences comes from the need to go with less exact methods because the precision found in natural science methods often leaves too much out for their purposes.

    As long as we try and recognize our biases and limitations, we can go further in trying to chart the extent of our understanding and our ignorance. Sometimes our good faith efforts just need a swift kick.

Leave a Comment

Your email address will not be published. Required fields are marked *