Here comes a section of my dissertation chapter "1999 - How to Study the Social Sciences". It is work-in-progress, and perhaps a bit hard to get into in the beginning. But here I try to show how a Latourian terminology can be applied when studying a concrete assemblage such as a social scientific research institute. The picture above is the "object" that orients analysis, the book "Det nya Samhället" ("the New Society"), a report written by 21 authors, all social scientists on a tour de force of quantifying society. Comments are appreciated (in Swedish or English, or why not German too). The LaTeX-to-HTML conversion didn't do footnotes very well, but a heavy pdf can be fetched here. Also, Levi Bryant's post on a related topic is worth reading to get further into what social sciences really are.


In this chapter I will further define the SOM-institute as my study object. The entry point for this mapping is the 1999 SOM survey. By disassembling this report, while introducing the theoretical approaches needed, it is possible to select further paths, that lead to a richer picture of how society is quantified by way of the social sciences.

In the Introduction chapter I approached social scientific knowledge as it was circulating in the news media. Facts were simple, and to a certain extent given in the sense that methodological problems, diverging interpretations and the complexities of scientific research remained unspoken of. But these brief flashes of facts barely revealed anything about how science was done, except for the short radio report about declining response rates.

To further develop my theoretical approach, and to refine how I shall proceed methodologically, I will step back roughly a decade in history, to the 1999 SOM-survey (Holmberg & Weibull, eds., 2000). Also, this will serve as a closer introduction to the SOM-institute, as taking apart its pieces also reveals what role the institute has in a broader context of social scientific research. But before proceeding, it is necessary to introduce the concept of blackboxing.

Concept I - blackboxing

The notion of black boxes, or blackboxing, is introduced in Bruno Latour's 1987 book Science in Action (Latour 1987), and has since then played central role in his works for describing not only how scientific knowledge is made durable, but also how the very fabric of societies hold together (See Latour 1999: 183-185). As mentioned in the introduction chapter Latour borrows the concept from cybernetics. However, it both has a richer meaning, and much wider consequences when used to study scientific activities as they unfold. It is time to leave the cybernetic meanings of static systems, feedback loops and self-regulation behind, and approach the concept anew, in pursuit of a powerful tool for studying scientific work empirically.

Science in Action begins with the example of the role of the Eagle microcomputer in discovering the double-helix structure of DNA. Usually, the story of this discovery is narrated through the involvement of Nobel Price winning scientists and their struggles in competing to become the first to prove the existence of a natural phenomena. Latour, however, goes back in time, to the moment when the scientific community still remained uncertain to what the structure of DNA really looked like. Then, not only are facts, models and theories uncertain, but also the equipment, methods and previous statements. As science is made, scientists need to get rid of fuzzy complexities, defunct hardware and contradicting theories.

When scientific instruments work, they are treated as unproblematic black boxes, that generate outputs from inputs. However, they first have to be made to work, and this is the moment when the black box still has not been sealed:

But it was not a good machine before it worked. Thus while it is being made it cannot convince anyone because of its good working order. It is only after endless little bugs have been taken out, each bug being revealed by a new trial imposed by a new interested group, that the machine will eventually and progressively be made to work. (Latour 1987: 11, italics in original)

This 'debugging' of black boxes occurs either before they work, or when they break down. But when they work, they remain almost invisible. According to Latour, this is not exclusive to complex scientific instruments such as telescopes, bubble chambers or lasers, but it also holds for everyday equipment that we use. Mobile phones, microwave ovens or credit cards are devices that we take for granted, even though most of us do not have much of a clue to how they function in every minute detail. To move along in our lives, and get things done, we are unable to hesitate and examine every component in our tools. Hence, a key element in practice is to ignore complexities, and surrender to a kind of technoscientific ignorance. Of course we may however at any time dissent and ask why there is no cell phone reception in a certain neighbourhood, why the fees are high on credit card transactions or whether microwave ovens emit dangerous radiation or not. The cost of such dissent will interrupt our lives when it comes to our everyday tools. We would have to seek evidence, gather information and form alliances with other concerned people, even run the risk of being questioned for questioning such mundane things that most people consider unproblematic. And questioning scientific instruments, in particular, would be an even more daring enterprise. To challenge, let us say, the accuracy of the Hubble Space Telescope or the Large Hadron Collider, we must learn and aquire the terminology, skills, and perhaps even the social status of a scientist to be taken seriously. We may even have to build a better telescope or particle accellerator in order to prove that the existing equipment contains errors. Simply walking down the street, claiming that the Large Hadron Collider will be the end of the world because it will create a black hole that implodes the universe, will probably do nothing more than labelling us 'conspiracy theorists' or 'lunatics'.

Philosopher of metaphysics Graham Harman thoroughly analyzes the status of black boxes in Latour, in his Prince of Networks (Harman 2009), a whole book dedicated to understanding Latour's philosophy anew, and also using it as one (of several) building blocks in Harman's "object oriented ontology":

For Latour, the black box replaces traditional substance. The world is not made of natural units or integers that endure through all surface fluctuation. Instead, each actant is the result of numerous prior forces that were lovingly or violently assembled. While traditional substances are one, black boxes are many we simply treat them as one, as long as they remain solid in our midst. Like Heidegger's tools, a black box allows us to forget the massive network of alliances of which it is composed, as long as it functions smoothly. Actants are born amidst strife and controversy, yet they eventually congeal into a stable configuration. But simply reawaken the controversy, reopen the black box, and you will see once more that the actant has no sleek unified essence. Call it legion, for it is many. (Harman 2009: 34)

As Harman points out, black boxes never come alone. And moreover, even though Harman is not primarily interested in methodology, he suggests (just like Latour does) that reawakening controversies and opening black-boxes is a way of rendering visible the vast networks of alliances that make a black box black. You may hit speed dial on your mobile phone and in matters of seconds talk to someone on the other side of the globe. But it is not the device in your hand making the phone call. It is a whole vast network of chipset manufacturers, telecom carriers, antennae, satellites and crews of administrators. The device in your hand will, and must do so to be usable at all, reduce this complexity into a few keystrokes and an invoice appearing in your letter box every month. When it fails, however, the alliances may start to unfold. Maybe your carrier has failed to make a deal with a carrier in the country you try calling to, perhaps did an antenna in the base station fall to the ground in a recent storm, or there could have been an unlikely failure in an undersea fibre cable connecting the continents. As a consumer you barely ever have the opportunity to inspect the business deals between carriers, the base stations or the undersea cables. Instead another black box enters the scene. You call customer support, and they might tell you that there is a "technical error".

As a consequence, following the processes of blackboxing can be a strenuous task, as one encounters even more black boxes along the way. A core activity of (techno)scientific work is to make sure the black boxes work, which means that when we encounter facts that are already there when the machines seemed to work almost flawlessly, the other components appear to be almost invisible. To proceed, Latour suggests that we ought to "arrive before the facts and machines are blackboxed or we follow the controversies that reopen them". (Latour 1987: 258).

In the introductory chapter the following was stated based on reading the tabloid Aftonbladet:

After the elections many columnists seek answers to why the right-wing populist party Sverigedemokraterna was voted into parliament in 2010. According to one article, this is partly because 45 per cent of the population want to lower migration quotas, a statistical fact which is repeated in several other news articles. These are numbers from the SOM-institute as well. The attitudes towards immigration have been measured since the 1990s by the institute (Aftonbladet, 2010-10-03).

The statement "45 per cent of the population want to lower migration quotas" is a typical example of ready-made science. It is backed up by a scientific 'author', the "SOM-institute", which in turn have been measuring attitudes towards immigration "since the 1990s". We learn nothing more about how the black box "SOM-institute" works, no more than we learn about antennas, cables and carriers when we dial a number on our mobile phone. We only get the output, and most readers of Aftonbladet are probably satisfied with that. A black box remains black when it stands uncontested, or when it does not break down. To learn more, the SOM-institute has to be followed to the point where we are able to see some uncertainty, controversy or failure. The next move must thus be to get closer to the black box, and a suitable way of doing that is to look at a SOM-report that was written a decade earlier.

Analyzing the 1999 SOM report

Each year since 1986 the SOM-institute has published a volume where social scientists present the latest results and trends found the yearly postal survey. The 2000 publication is called Det nya samhället (The New Society). It is a 400 page volume that also includes the questionnaire that was sent out in 1999, and a methodological appendix describing the procedure of the survey field work.

Already by reading the preface we are able to go from the lonely figures and graphs found in newspapers, to an initial disassembly of what this type of knowledge is about. It is written by Sören Holmberg and Lennart Weibull (2000), who state that the report is based on a survey of two randomized populations, each with 2800 respondents, where the response rates have been fairly equal between them. The fieldwork took place between October and December 1999, and the private contractor Kinnmark DM AB was hired to collect the data. A reference is given back to the methodological appendix, as an invitation to the interested reader. It is also mentioned that the survey has been running every year since 1986. This introduces a brief description of what I will call epistemic assemblage, and even though it is full of unopened black boxes (populations, field work, data, response rates), the experienced social scientist can still make a judgement about its qualities, and recognize it as a typically scientific survey in the way it presents its methodology already in the first page of the book. Compared to the average media reports, where there was usually nothing more than a simple output of percentages produced by the SOM-institute, the initial paragraphs of the preface contains more black boxes; what is a "randomized population"? How many is two times 2800? Who is "Kinnmark DM AB"? What is "fieldwork"?

Turning to the methodological appendix in the back of the report, there are answers to these questions. It is stated that this is the fourteenth survey, and that each of them are performed "under almost identical circumstances" (Lithner 2000: 395). Then a detailed description about the survey follows. Going from "randomized population" in the preface, this now elaborates to an "simple random sample". The simple random sample is not further explained (another black box), instead the population is explicated, from "two times 2800" to being of the ages between 15-80 years, including foreign citizens.

Then the two questionnires are introduced, and they are also included in yet another appendix. It is said that the questionnaires are developed in co-operation between then SOM-institute and affiliated research projects, which in turn are listed in another appendix. There are 21 of them, mostly research projects at Gothenburg University, but also including participants from Uppsala University, Institute of Public Affairs in Warzaw and Stiftelsen Institutet för mediestudier. The purpose of having two questionnaires is to be able to ask more questions without making the questionnaire too long, but also having identical questions on certain topics, to grow a large enough population to be able to break it down in smaller groups without compromising statistical verification.

The most interesting part of the methodological appendix is however the description of the field work during the late fall of 1999. Again the private contractor Kinnmark DM och Distribution is mentioned as having been commissioned to perform the actual data collection. Then a whole list of 'equipment' is presented; questionnaires, post-cards, telephone reminders, follow-up letters; each of them dated. Taking a closer look at this chain of events, it is possible to re-construct what happened the last couple of months of the previous millennium:

29 Sept. 1999 Dispatch of advance letter
4 Oct. Dispatch of questionnaire, follow-up letter, brochure and return envelope.
11 Oct. Dispatch of greeting and reminding letter20 Oct. Dispatch of reminding letter.
20 Nov. - 5 Oct. Dispatch of extra greeting- and reminding letter for Riks-SOM I.
5-16 Nov. Telephone reminder.
17 Nov. Dispatch of postal reminder.
23-29 Nov. Telephone reminder.
2 Dec. Dispatch of postal reminder.
7-13 Dec. Telephone reminder.
14 Dec. Dispatch of postal reminder.
23 Dec. Dispatch of mini survey.
18 Jan. 2000 Field work is terminated.

This assemblage contains the components of the black box referred to as "survey", which is practiced as "field work". Each postcard, telephone call and dispatch of questionnaires are dated precisely. Most readers of the report are probably satisfied by now, as to the degree of scientific rigour and openness in presenting how the survey was done. This is the Hubble Telescope or the Large Hadron Collider of the SOM-institute, and even though it is not as costly or famous, I will argue that it should be put on an equivalent level of scientific laboratories in the natural sciences. This means that more black boxes have to be opened.

In Give Me a Laboratory and I will Raise the World Latour (1999b) discusses the role of Pasteur's laboratory in Paris in the 1880s, and its relations to an "outside society". Latour's point is, however, that the notions of inside/outside and micro/macro need to be abandoned, or at least drastically revised, if we are to understand what is at stake when Pasteur discovered the anthrax vaccine. What had to be done to create the vaccine was for Pasteur to convince a number of actors through series of negotiations; farmers had to be convinced that a vaccine would save their cows (and income) and let Pasteur take samples from them and bring back to the laboratory. Once inside the laboratory, the anthrax bacilli are grown stronger and are progressively turned into a vaccine. Then, it could eventually be brought out back to the cows, fulfilling the promise Pasteur made to the farmers. (Latour 1999b, pp. 208-274)

Just like the anthrax bacilli needed to get in and out of Pasteur's laboratory, as shown in the figure above, so is the case of questionnaires that the respondents need to fill out and return to the SOM-institute. The respondents are the microbes of society, and not until they are captured and counted, they can be transformed into 'macrobes' that speak in the name of society as a whole. Whereas Pasteur had to travel to the countryside to get his samples, the SOM-institute is utilizing mediating actors; the questionnaires are sent via the postal system, they are gathered by the Kinnmark contractor and turned into computer readable data, which is handed over to the institute and sent further to the researchers in the involved projects. If everything goes as planned, articles and reports can be written, news media can print articles covering the findings and circulate them back out to anyone who has an interest in a society described by the social sciences. To put it in Latourian vocabulary, to say something about macro-society, the SOM-institute needs to amplify a sample of micro-society through trials of strength.

The questionnaires are flowing back into the SOM-institute safely during the fall of 1999. The response rate is according to the methodological appendix 67 per cent, a number that is considered high by the author. In a brief discussion of response rates, there is an external reference to "Ohlsson 1986" who states that "the response rate of social scientific postal surveys are generally between 60 and 70 per cent, depending on geographical area and types of respondents" (Lithner 2000: 398). How much is enough? And who decides when enough is enough? At what moment can a social scientist know when the survey works? What has to be examined, to further open up this black box is just as with the Eagle Microcomputer, to examine the 'de-bugging', tinkering, patching and analyzing of these progressive steps that had the researchers decide to assemble letters, postcards and telephone calls i specific configurations. This is the first limit of the 1999 SOM-survey. To find the next link in the chains of black boxes I will have to go back to 1986, perhaps even much further.

Picture above: Going from a black box, opening it up, then dislocating its components, which are still somewhat assembled, but lack the borders the box once gave them.

Skriven av admin den 19 juni 2011