Black boxing is a process that makes the productions of actors and artefacts opaque (Latour 1999, 183), and the degree of invisibility of that process can be seen as a measure of its success. It is only when one of our digital devices stops working or behaves unexpectedly that we become aware of it as constituting one or more black boxes, especially when we realise we are unable to fix it or make it work again. Such success – and hence invisibility – is perhaps why it is all too easy to overlook the need to understand the structures, operation, and outcomes of a technology by opening the box. Consequently, difficult or not, arriving at a critical appreciation of our computational cognitive artefacts is key to understanding their effects on our own cognition.
Four distinct approaches may be characterised to this end: a critical reading of the program code/algorithm itself, a critical appreciation of the software package, a critical engagement with the creative process of the design of software, algorithms, and other structuring aspects of the computational environment, and a critical understanding of the use of the hardware/software applied within an archaeological context.
It is generally appreciated that programming languages affect the way that program designers and developers think. In 1975 Dijkstra defined this as one of the truths of computing science:
'The tools we use have a profound (and devious!) influence on our thinking habits, and, therefore, on our thinking abilities.' (Dijkstra 1982, 129).
'Software developers as a species tend to be convinced that programming languages have a grip on the mind strong enough to change the way you approach problems – even to change which problems you think to solve.' (Somers 2015, 81).
As a result, Berry argues for a need to understand program code and the systems so constructed since:
'A close reading of code can … draw attention to the way in which code may encode particular values and norms … or drive particular political policies or practices.' (2011, 9).
Furthermore, Berry demonstrates the ways in which code can be designed to disguise or mislead, whether deliberately or inadvertently (Berry 2011, 75ff). Consequently, access to code is key to the concept of reproducible research (for example, Marwick 2016). As Ducke argues, 'the ability to track data through complex processing chains is key to successful collaborative research' (2015, 93) and the only way to understand the mathematical models and formal reasoning behind a computer-based archaeological study is through examining the source code (2015, 99).
This implies a degree of digital literacy that is rare among archaeologists. However, that aside, the transparency that ought to be implied by reading the source code is not straightforward. Understanding the breakdown of computational tasks and the formalisation of practices and methods underpinning the computational analysis through making the code and algorithms explicit will frequently be highly challenging. Access to the code itself may not be feasible (hence Ducke's argument for open source software in archaeology) but, if available, it may not be easily interpretable – either through obfuscation or the sheer complexity of the code, especially problematic if that code is itself created by the software rather than the human agent. For example:
'… certain techniques imported from the computer sciences may never be understood in the same way we understand statistical concepts like variance or regression because there no longer is a 'manual' equivalent of the automated approach.' (Rieder and Röhle 2012, 76).
So for commercial reasons and/or lack of expertise, the actions of code may score very low in terms of Heersmink's dimensions of access (see section 4), procedural and informational transparency, although high in transformative terms and, perhaps paradoxically, very high in terms of trust.
Internet Archaeology is an open access journal. 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.