In this article, we have described and demonstrated the benefits and also the logistical and ethical conundrums inherent in machine learning as a means to investigate the human remains trade. Moving forward ethically and creating a better app could open up new means for researchers, law enforcement, and concerned citizens to identify and even begin to categorise examples of human remains for sale that they might encounter 'in the field' (e.g. at a flea market, a gallery window, an online seller). Doing so would place the use of machine learning and neural networks for this purpose in the same category as the now numerous applications that attempt to identify unknown plants from photographs that a user takes (Goss 2018). However, as we have discussed, it is not as straightforward as we would have hoped and the ways in which classification algorithms break down or learn unintended signals implies that their use for any archaeological classification question does not represent a 'panacea.' How do we share our data or our code ethically while simultaneously using these methods 'at scale' to continue to understand the full extent of this traffic and identify the corners of the internet it might move to so that law enforcement can be better placed to prevent sales and shipment?
We do not yet know the answers to these questions, which is why we are posing them in this public space. Research in neural network approaches to data analysis (and to creative, generative work) is unfolding at breakneck speed, and archaeologists need to be having these conversations now. To reiterate, the issues we see in using neural networks, machine learning, and computer vision technologies in our particular corner of archaeology begin with:
We began with the story of Abraham Ulrikab, and our desire to restore something of the humanity to the other individuals out there, whose remains are bought and sold as macabre or exotic fetishes for a technologically mediated colonialist gaze. We might not be able to do this. But in trying, we begin to see the complex nexus of questions that such research must address.
Internet Archaeology is an open access journal based in the Department of Archaeology, University of York. Except where otherwise noted, content from this work may be used under the terms of the Creative Commons Attribution 3.0 (CC BY) Unported licence, which permits unrestricted use, distribution, and reproduction in any medium, provided that attribution to the author(s), the title of the work, the Internet Archaeology journal and the relevant URL/DOI are given.
Internet Archaeology content is preserved for the long term with the Archaeology Data Service. Help sustain and support open access publication by donating to our Open Access Archaeology Fund.