Is Slow Dangerous?
Scholarship should be open, inclusive, and slow, but is slow dangerous to some?
In a recent post, Emily Bender laid out a thoughtful argument focused on establishing better values for scholarship. Bender believes that scholarship should be:
– Open, with open amounting not only to accessibility but also for accountability, as in honest.
– Inclusive, because the more approaches there are to a problem, the more fully we realize the nature of the problem.
– Slow addresses the problem with the publish or perish model which has not only continued in the American academy (and elsewhere) but has probably like everything else only sped up.
Like Bender, for me open and diverse are intertwined. The more scholarship is held tightly behind paywalls, the more it costs to access it, the fewer people can engage it, leading to fewer voices in the conversation. That’s a starting principle without any need to remind us of the historical realities of who can afford to pay for things and who, in fact, is charging for these things.
Most of the big names in science and scholarship are European or American publishers, many of whom are increasingly under a smaller and smaller number of roofs. (Like so much else, the defense of lack of competition is scale: one has to be bigger to survive everyone else who is getting bigger.) At the same time, most of the scientists and scholars who can afford to access all the increasingly limited, and also increasingly more expensive, publishing resources come from a small number of European and American institutions. Limited options. Limited resources. Limited access. Limited audiences. Limited producers. Limited insights.
Limited views mean limited review.
And it means limited research. Bender is quite right here. And at no time has there been more at stake in having diverse vantage points than in the moment when so much of the world’s computational resources are increasingly concentrated in the hands of a few very powerful large language models (LLMs) and their associated algorithmic enterprises (AIs with chat-like interfaces).
We have already seen that models trained on biased data have real world effects, though typically those effects are on people who do not occupy the offices of elite universities and large corporations. We have witnessed the frustrations of women and non-whites when trying to make a case for more diverse perspectives. (Those interested should check out Brandeis Marshall’s Data Conscience: Algorithmic Siege of Our Humanity as well as Catherine D’Ignazio and Lauren F. Klein’s Data Feminism.)
Bender’s argument for slow scholarship is as solid as they come, and much of her argument, and by no means am I demeaning it by saying this, is really an attempt to remind people of what good, normal science is supposed to look like. Science moves slow because it allows for lots of input through collaboration, through review, and finally through being published in places where it can be seen by as many people as possible.
Perhaps it is a function of the erosion of support for public higher education that has led to universities and/or individual investigators seeking more ways to monetize their work, but such monetization almost always means competing in the private sector where first-to-market or being able to scale quickly are more highly prized than, and in fact are contrary to, something like openness.
It is certainly true that science has regularly profited by highly competitive environments. Science rushes forward during times of conflict like wars, be they actual or cold. But at what cost? And are we willing to say that the only way humans got to the moon was because there was a cold war space race? If anything, the highly collaborative, and open, nature of much of science in American universities during the second half of the twentieth century might have been “peak science.”
So, I am completely behind Bender’s argument for slowing things down, for returning to normal, as it were, but like so many things, I don’t think this call is as simple as it sounds. One thing speed has led to has been more people making more things available. Not unlike the internet itslf which let loose a torrent of voices, of people speaking into the great digital void in hopes of being heard, and, in some cases actually being heard, found. (And, sigh, thus the rise first of the blogger and then, sigh once more, the influencer — listen, I never said democracy wasn’t messy!)
My fear for slowing things down is that it also amounts to a return to the usual suspects, the people who can afford to slow things down. For all the postdocs, upon whom too many institutions depend upon for cheap labor, there can be no slowing down. For all those at lesser institutions funded less and less by increasingly hostile state legislatures and administered increasingly by counters of beans, there can be no slowing down. For all those hopeful graduate students seeking a toehold in a very tight job market, there can be no slowing down.
My worry is that slowing down profits those already in pretty good, if not, to use a word that will make some wince, privileged positions. I don’t think it has to be that way, but that doesn’t stop me from worrying.
Yours in the struggle.
