Today was an very good day in my dissertation writing. I managed to scribble down 10k characters and I decided to instantly put them here on interweb. This draft concerns what I will call “withdrawn hardened functions” in epistemic objects. It is partly influenced by object oriented ontology which I am trying to enrich the standard Science and Technology Studies terminology with. It’s quite a heavy read, so enjoy or surf along!
So far I have only dealt with the positive domains of scientific knowledge; blackboxing, interfaces and assemblages as productive elements. But a core problem in theory of science is what is unknown, what hides in the unconscious or hidden domains of imperceptibility, what is included and what is left behind. This means that in order to deal with this complex issue I need to adapt also a terminology for talking about what evades a concrete epistemic assemblage.
To these questions, at first glance, pure actualism gives little room to navigate. I will propose that the black boxes I encounter have withdrawn hardened functions, which makes them combinable and plastic. This feature has been called “immutable mobiles” by Latour (1999: 306-307), and it falls close to what Star & Griesemer (1989) call a “boundary object”. Since as mentioned already before, blackboxing actually is a process of forgetting, embedding and ‘hard coding’ tasks and processes that are needed to produce on a surface level something that is positive knowledge. For example, a pre-compiled dataset of statistical information gathered from a survey makes computerized statistical calculation possible, and quite user friendly compared to doing it manually, precisely because it in a given moment gives us the opportunity to forget thousands of questionnaires and how they were collected and assembled, to instead paying attention to creating bars and diagrams for a scientific report.
This is a wholly different approach than what is the outcome of a Kuhnian thinking. Whereas the process of forgetting is historically very dramatic in Kuhn, I shall argue that it is shallow compared to the sort of object oriented aspect of blackboxing. Lets take a look at a central passage in The Structure of Scientific Revolutions:
In short, they [textbooks] have to be rewritten in the aftermath of each scientific revolution, and, once rewritten, they inevitably disguise not only the role but the very existence of the revolutions that produced them. Unless he has personally experienced a revolution in his own lifetime, the historical sense either of the working scientist or of the lay reader of textbook literature extends only to the outcome of the most recent revolutions in the field. (Kuhn 1996: 137)
This leads Kuhn to thinking about different historical paradigms as incommensurable, and that the disguising of past revolutions leads to a view of scientific progress as being linear and cumulative. But, while these two points are valid and refreshing to the history of science, they are indeed very clumsy for the more close studies of scientific activities. A paradigm, would then appear as a monstrously large black box, where a whole generation of scientists are only able to think ‘within the box’, while the actual workings of the machinery is veiled. Only, according to Kuhn, when enough anomalies appear the scientists start to doubt that the whole paradigm might be wrong.
Two problems arise here. The ‘monstrous’ aspect of Kuhnian historicity leads to a sort of empirical over-determination. In the reports of the SOM-institute we find for example a terminology resembling the sociology of Durkheim, Parsons, Merton etc. The methods of surveys and quantification are also ‘borrowed’ from the intensified usage of these methods in sociology towards the end of the 19th century. Even though this is true on one level, I argue that it adds very little to our understanding what is done, and what that practice means. The abstractness of paradigms, rather ironically, makes the co-production of scientific objects and other objects invisible. To make a crude example (which is unfair to attribute to Kuhn himself); if I read in the local newspaper “The researchers talk about a Gothenburg effect and a slow norm shift” (as already quoted in the prelude section of this chapter), and then conclude that this is knowledge within a Durkheimian paradigm since it talks about norms and norm shifts, I would instantly remove myself from a process that has significant value for translating the research practice of the SOM-institute into a circulation of facts. The concept of norms is indeed built into theoretical tools used (which in turn may be blackboxed), but if we ignore the relevance of that another actor, the Göteborgs Posten local newspaper, made use of and valued highly enough the much debated question of corruption scandals, the role of science and its interfacing with other societal assemblages is abruptly veiled in darkness, and analysis would stop on what I consider to be a shallow level.
Another more serious flaw in Kuhnian-inspired theories of science is their human-centered character. For science to change, either the scientists need to change their beliefs, theories and everyday practices, or they have to be replaced by a new generation of scientists (1). This is not true for technology, and with technoscience, it is not valid either. Let me give two examples, one simple and one advanced:
Example 1 – The hammer
A carpenter uses hammers (2) as a routine piece of equipment when building houses. It is connected to other objects such as nails, human users, and wooden planks. Hammers are constructed objects, and in one respect they reconfigure the human user too, which has to learn how to use it. One could even say that hammers are paradigmatic technologies of house building, since they imply methods, can be calculated with by architects, etc. Now, the hammer may also be used for committing a brutal murder. Then it becomes a piece of evidence in a murder investigation, is placed in a plastic bag, checked for fingerprints and may even be a technical evidence putting the murderer away for prison for several years. A skilled carpenter knows the difference between a good and a bad hammer, but in the moments of driving nails into wood his or her attention lies elsewhere than with the technological advances, means of production, and the price of the hammer. It is precisely because it is blackboxed, that it may withdraw from full inspection and reflection, that is is a powerful tool. As the house is completed, and populated with new people, they in turn do not need to know anything about hammers, even though they may be ‘implicated’ in the house, and need to be brought forth once again as the house is repaired. The hammer is thus more than its use together with nails and planks, more than the carpenter’s skills, and more than evidence in a courtroom. The hammer survives the house.
Example 2 – Experiments in relativity
Even though I consider the Sociology of Scientific Knowledge to be unsuitable to my theoretical needs, Harry Collins and Trevor Pinch (1993) have produced a textbook example of how scientific experiments may reinforce each other throughout historical paradigms. In their chapter Two Experiments that ‘Proved’ the Theory of Relativity, Collins & Pinch set out to understand how the 1919 solar eclipse experiment led by physicist Arthur Eddington was accepted very swiftly, even though the results of the actual experiments were quite poor and inconclusive due to harsh conditions of photographing light as it was supposed be displaced by the large gravity field of the sun (and thus proving the theory of relativity). The experiment was very difficult to perform at the time, cameras had to be mounted on remote islands to be in time for the solar eclipse, and they were sensitive to temperature and vibrations due to the long exposures needed to make the photographs.
A contributing factor to the quick acceptance of the inconclusive results of the Eddington experience, was according to Collins & Pinch, that beginning in 1881 Albert Michelson (later in collaboration with Edward Morley) had performed series of experiments of a wholly different purpose. They wanted to measure the ‘aether drift’ that was thought to occur as earth moved across space. It was believed that light traveled through the medium ‘aether’, and thus the movement of the earth would produce slightly different speeds of light in different directions. These experiments, that over time took place for half a century, however failed to account for any significant variations, and thus many considered the speed of light to instead be constant.
Now, it may seem that the Eddington experiment and the Michelson-Morley experiments are disconnected. But Collins and & Pinch connect them despite being about two different things:
The way the 1919 observations fit with the Michelson-Morley experiment should be clear. They were mutually reinforcing. Relativity gained ground by explaining the Michelson-Morley anomaly. Because relativity was strong, it seemed the natural template through which to interpret the 1919 observations. (Collins & Pinch 1993: 52)
As the Michelson-Morley experiments kept failing, they unintentionally reinforced the Einsteinian relativity theory, because it presupposes the constant speed of light. The results of Michelson-Morley, even though they were a ‘failure’, could be a component part in strengthening the Eddington experiments, even though Eddington had a wholly different theoretical purpose. What I am getting at here is a somewhat dramatic comparison: Just like the hammer can be used both for carpentry and murder, scientific results, methods and machinery can be used for very different purposes, in different setups and epistemic practices. Even though carpentry and relativity physics are radically different activities, the point is that parts and components can be taken out of their contexts, since they are rendered mobile by way of blackboxing. Assemblages, architectural or scientific, mobilize and assemble their equipment, where most of them are already there. But assembling and selecting what components to choose is not only about actively knowing where to go. It is equally important to forget. Be it about the theoretical functioning about the hammer or the ‘aether wind’, exclusion is as important as exclusion.
This is, I will argue also the case for the social sciences, and especially concerning its use of quantification, which will be the topic for the next section.
(1) Of course paradigms may extend over centuries, but it can still be said that Kuhn also classifies the durability of scientific beliefs around scientists and communities of researchers.
(2) Selecting this example is a tribute to Heidegger’s tool analysis in §15 in Sein und Zeit (1972) where the hammer is used as an example on how a piece of equipment is always in relation to other objects, and that equipment has to withdraw from consideration to be used for something.