The estimated reading time for this post is 16 minutes.
The recent discovery of gravitational waves has impressed many people and has caused considerable stir in the community of physicists. Surprisingly this commotion has not spread to the community of historians of science. This is surprising because I believe that the claim to have detected gravitational waves constitutes a serious blow to the stronger versions of social constructivism, which arguably has deeply influenced the profession of historiography of science in recent decades. The aim to find empirical confirmation of the existence of gravitational waves has occupied physicists from the 1960s onward. Sociologist of science Harry Collins (Cardiff University) has turned the activities of this group of ‘wave’ physicists into one of his central case studies. Collins has long been one of the most important proponents of the social approach to the study of knowledge formation. His programme of methodological relativism may be more ‘practical’ and less ‘philosophical’ than the perhaps better-known strong programme of the Edinburgh School, but in essence both approaches in the sociology of scientific knowledge (SSK) boil down to the same thing. They share the radical view on science as a social process, not in the sense that other factors such as ‘nature’ or ‘reason’ (whatever they might be) do not play a role in determining the course of scientific development, but in the sense that social factors are always ultimately decisive in determining such things as the acceptance (and rejection) of evidence, experimental methods and claims to knowledge. In tracing the search for gravitational waves, Collins’ aim has been to show how scientific data can be subject to interpretative flexibility, and how social or ‘non-scientific’ means are used to close scientific controversies. In what follows I will argue that the discovery of gravitational waves seriously undermines the SSK perspective on science because it cannot be fitted into SSK’s explanatory scheme. It follows that this discovery has rippling effects on the study of past science and I close with a brief reflection on the direction in which these ripples are heading.
One of Collins’ famous arguments comes from a reflection on expertise. Science ‘at the frontier’, he argues, is uncertain about the very empirical evidence it is gathering. How to gather empirical data and how to properly account for it depends on expert opinion, which involves theoretical explanations (and expectations), proper calibration of instruments, assessments of various forms of data interpretation, etc. But whether such theories are correct, and whether our instruments and modes of interpretation work well, can only be decided on the very same evidence we are trying to come to terms with. There is thus a regress: we already need to be experts to properly consider new empirical material, but we can only become experts through the accumulation of such material. This issue is particularly important in situations where there is no community consensus regarding the relative value of various competing theories, and where possible sources of experimental error are not well known. Collins argues that the regress cannot be broken by appealing to an ultimate super expert (these are not there) nor to canons of reasoning (these are not uncontested) and of course certainly not to something like a ‘crucial experiment’ because the valuation of empirical evidence is what is at stake in the first place. Yet, we do vest some individuals, or groups of people, with expertise in particular domains and we deny others this trustworthiness. This then can only be explained as the outcome of a social process, which is characterized by social strategies such as competition, negotiation, persuasion, sheer use of power, etc.
This mode of explanation can be applied to the whole range of social interactions that have occurred in the history of science, running from small divergences of opinion within a particular research group, to whole scale divergences between systems of belief, paradigms, patterns of thought, or whatever you would like to call them. The implication obviously is that, had the social process been played out differently, other beliefs, distinct from the ones we now hold to be true, could have ‘won’ and we would as a consequence now have a different system of beliefs. It follows that the ‘losers’ (or the ‘bad guys’ in Collins parlance) in the history of science should be treated with equal respect as the ‘winners’. Their ideas could have had a lasting impact on our world and should therefore be given a fair treatment in historiography.
Earlier claims of gravitational wave discovery
The double status of gravitational wave physics as fringe science and frontier science is the reason why it has been one of Collins’ main case studies throughout his career. One ‘bad guy’, Joel Weber, stood of the start of all the attempts to detect gravitational waves by experimental means. He was also the first person to claim to have detected the waves. Collins is inclined to defend Weber by showing that he occupied a genuine position in the debates surrounding his discovery claim. However, the majority of physicists rejected Weber’s claim. In my opinion Franklin (1998) demonstrates that this rejection was fully justified. Other research groups produced discordant results, they also carefully cross checked their results and tried multiple ways of interpretation, using both linear and non-linear algorithms. Next to this they carefully calibrated their experimental apparatus once more in order to check whether they functioned properly. Weber did none of these things. To make things worse, errors in the computer programs used by Weber were revealed. In spite of all this Collins has not accepted the conclusions drawn in Franklin’s paper. In Collins (2011) p.51 he writes: “It is now widely believed, though never decisively proved, that Weber managed to get the results he got by manipulating his statistics retrospectively. Weber had a tendency to ’tune to a signal’.” According to Collins only this ‘mild’ error of ‘tuning into signal’ can perhaps be attributed to Weber, besides that Weber had a technical answer to every point: he just failed to convince other scientists. From Collins (2004) we learn that Franklin’s mistake has been to consider published sources only, whereas Collins’ own investigation includes interviews with the participants. According to Collins this makes Joe Weber appear as a much less isolated figure than Franklin portrays him.
This ‘saving’ of Weber is rather unconvincing in my opinion because during the time of the 2004 book publication there had been six detection claims spread over a total of fifteen papers, including ten of Joe Weber during the 1960s and 1970s, one by the Frascati group in 1982 together with Joe Weber, then in 1996 Joe Weber alone again and in 2002 the Frascati group again. The scientific community has turned all of these into null-results after careful checking. So Weber may not have been a lone wolf, but it is clear that his research also did not gain much success.
The US government has nonetheless decided to invest staggering amounts of money (over 300 million dollars) in new experimental settings to find the gravitational waves. It makes one think of investments in SETI research (search for extraterrestrial intelligence), which has also produced nothing more than silence. Yet the difference with gravitational waves physics is that in theory (Einstein’s theory of general relativity) the waves are generated when stars or black holes explode or collide and a portion of their mass becomes energy that ripples out ‘like a disturbance on the surface of a serene pond’. Hence, although it is very difficult (see below) it must be possible to measure such a disturbance. The motto appears to be that we will not rest until we have tried everything to find them. SETI research may work on the same motto, but there is no theory involved that clearly predicts the existence of other intelligent beings outside the solar system. SETI just works on the assumption that if an extraterrestrial civilization exists it would emit electromagnetic signals, which because of certain regularities, would be distinguishable from natural electromagnetic radiation. So far no such signals have been found.
Do gravitational waves make SSK waver?
Collins continuing defense of Weber shows how hard it is to refute the SSK approach by argument. While this is true, we can nonetheless point to a number of undesirable consequences of the approach for the study of past science. In my view the SSK approach leads to an overly schematic treatment of contestants in scientific dispute in the sense that differences between opposing parties are given more weight than similarities between them, which are often fully neglected. Next to this I believe that the SSK approach asks for closure of scientific controversies too soon. As a consequence there is little room for uncertainty to linger on after a (provisional) choice has been made. My point is that, if we accept that disagreements and agreements between contestants in scientific dispute are complexly intertwined, and if we allow closure of scientific controversies to emerge only gradually, we arrive at picture of scientific development that is altogether different from SSK.
A welcome addition to this argument is given by episodes in the history of science that are hard to account for within the SSK framework. The discovery of gravitational waves at the end of 2015 represents such an episode. This important empirical result has been widely accepted by the scientific community without any controversy. In light of this consensus, and in the absence of any controversies, who needs to appeal to social factors to explain the broad acceptance of this result? This does not fit the SSK way of explaining things. As Collins wrote in his 2011 book: “From a sociological point of view it would be a pity if the interferometers had the stroke of scientific luck that would suddenly put an end to the debates and uncertainties that have lasted for nearly fifty years.” Well, this is now reality, and given the fact that gravitational wave physics plays such an important role in Collins work, and hence for SSK as a whole, it is natural to ask whether SSK scholars can produce a satisfactory response to the challenge that the discovery of the gravitational waves presents for them. I see four options:
- Underdogs feed the Mainstream
Siding with the ‘bad guys’, i.e. the small group of people who kept pursuing a research programme in spite of the string of null results, demonstrates that work done outside the mainstream can be promising pursuit. It shows that the tug of war, between those with faith in the endeavor and those without, has been a just fight. Should we earlier have written off the underdogs perhaps the detection breakthrough would have never been achieved. Perhaps such things have happened in the history of science more often, which shows why it is so vital to treat all past scientists (natural philosophers, etc.) with equal respect. This line of defense may look appealing but it is easy to see that it cannot be maintained by the SSK scholar because it is in fact a defense of the idea that discovery can occur in the periphery and lead to correction of the mainstream. This line of thought then puts discovery before consensus and that is in diametrical opposition to the central tenets of SSK.
- Relentless Skepticism
One can maintain a skeptical attitude towards the detection result. Checks for all kinds of possible noise in the signal are still being performed. Next to LIGO (Laser Interferometer Gravitational Wave Observatory) in the USA, where the result was claimed, there is the European VIRGO interferometer located in Pisa, which may possibly produce research outcomes that force us to accept a different interpretation of the LIGO result. Collins has always argued that empirical data allow for interpretive flexibility and that: “A result is not a result until the community gives it meaning.” And indeed, does only one detection result prove anything? In order to gain credibility results should be replicable. So who knows? Now, this line of defense is also not very promising when we consider that after gravitational wave physics switched to the interferometer technique no other competitive research strategy is in place anymore. There is thus consensus on the issue of how to check the experimental data in order to be absolutely sure that gravitational waves are the cause of the result. Collins keeps pursuing the search for controversy because the SSK approach to science almost demands it: “Social coherence on an international scale does not eliminate the possibility for disagreement; it means only that any disagreements will take place within one large group rather than between smaller groups. What may change however is the visibility of such disagreement to the outside world.” However, as far as I know, he has not succeeded in making such a dislocation of controversy visible.
- Consensus before the Facts
A more promising strategy, which is in line with the basic tenets of SSK, is to point out that consensus has been reached before discovery. The dominance of measuring with an interferometer, over alternatives such as the bar measurement technique, had to be won first after a series of battles. Interferometry means looking at changes in the way two light rays interfere. Around the year 2000 it gained prominence over the resonant masses technology. The interferotechnique is however much older, it was for example famously used in the Michelson-Morley experiment of 1880. Yet the apparatus requires enormous sensitivity in order to detect the weak signal of a gravitational wave, as the change in movement is less than one-thousandth the width of an atomic nucleus. Next to this incredibly small scale, it is an enormous challenge to free measurement from all kinds of possible noise signals.
Although he seems to regret it, because interferometry as the superior technology has led to a decrease in freedom of interpretation, this is in fact the outcome of the process Collins has so painstakingly documented for most of his own professional life! Why not accept it? Collins’ points of critique are diverse. He considers that interferometers work with lower thresholds for what could be (should be?) brought into the signal-noise analysis. He also points to increase of pressure to affirm or deny gravitational wave detection in publications because of the large invested sums of money. Finally he argues that the relative weight given to theory over empirical claims has been crucial to the transformation of gravitational wave physics to interferometrology, implying that this is a contingent matter. Perhaps indeed “…without LIGO and its funding battles of the 1980s the data from the bar groups would have been handled differently.” But why bother? The community of physicists unanimously selected a specific technique to detect the waves as superior. In contrast to Joe Weber’s experimental results, the measurement obtained with LIGO are therefore widely accepted. The key issue is not the ‘discovery’ of the waves but the fact that consensus has emerged about the method of their detection together with the funding obtained for building the instrumental set up. Thus closure in this case has been reached long before the actual detection of the waves. It was not ‘nature first’, hence SSK survives!
- Caving in
But still, something is rotten in the state of SSK when no real dispute occurs over obtained experimental data. Furthermore, while we may need the sociology of knowledge to understand how interferometry became the superior technology in the field, the very process has undeniably spurned technological advance beyond imagination. Gravitational wave physics went from small to big science, leading to a vast increase in precision of measurement and signal sensitivity. Investigators created optical instruments, kilometers long, and capable of sensing changes in the position of mirrors of less than one ten-thousandth the diameter of a proton. Thus, it is hard to escape the notion that in this case there is more to the application of technology than installing ‘purpose built antennae’. It involves such a high-tech and specialized science, that it is very difficult to come up with a challenging instrumental technique c.q. research programme. Are we then, for this reason, to conclude that the strong programme is too strong? Collins (2011) contains the surprising message that this is indeed the case. In this book there is, quite suddenly, no talk anymore of ‘methodological relativism’. Collins points out that the aim of his project is to understand science as the most reliable source of knowledge. The group of gravitational wave scientists is presented in the book as a model example how science ought to be practiced and even sets a behavioral example for the rest of society: “ … a high standard in the search for truth is better than academic Realpolitik, both in theory and in life…I try to elevate this point into a political philosophy arguing that science done with real integrity can provide a model for how we should live and how we should judge.”
These are pretty strong normative statements for a self-proclaimed relativist. Collins has come to the conclusion that the science studies require a 3rd wave (the strong programme within SSK was the 2nd, positivism the 1st), which grants the epistemic warrant of expertise a central place. Expertise is warranted, he argues when central values of ‘the institution’ of science are being honored. These include integrity, honesty, readiness to criticism (which is also why a simple theory is better than a complex theory), openness to falsification, allowing for dissenting views/lone voices, a high level of craft skills and replicability of results. Following these values: “Science has the potential to provide an object lesson in how to make good judgements in a society beset by technological dilemmas.” The essential values are normative guidelines of behavior and it is clear that Collins thinks that honesty and integrity are the most important of all.
This turn in Collins’ thinking could be witnessed in 2009, when I had the pleasure to sit next to him during a conference dinner in Bielefeld, Germany. Participants at the table started to make an inventory of colleagues in science studies based on an assessment of their honesty and integrity. No one present at the table was classified as dishonest or being of questionable integrity of course, but others were. At the time I was just getting started with my Ph.D. project and could not see the ‘turn’ in thinking that this discussion actually exemplified. Now that I have reached my own conclusions (see again my Ph.D. thesis) I can appreciate Collins’ concern, but also think he is still not willing to bite the crucial bullet. His list of essential values does not concern the epistemic content of our theories and methods of research but only the process of science and how scientists should conduct themselves. His argument essentially boils down to a defense of Mertonian values, which Collins wholeheartedly endorses as ‘elective modernism’. One could, for example, argue that honesty is not crucial to scientific development. Stealing (part of) an idea, meddling with data to fit them to a theoretical explanation and fencing of critical discourse for a period of time are dishonest activities, but they can lead to opening fruitful paths of investigation. What matters in the long run is not the dishonesty involved in such actions, but the epistemic content of the theory, i.e. its fruitfulness, predictive accuracy, consistency, problem-solving capacity, etc. Hence such values (perhaps it is better to speak of virtues) are in my view better candidates than the set of values Collins has suggested for what we may take to be the essential aspects of science.
To be sure, dishonesty is despicable. This is not a plea for dishonest behavior, but an argument, which aims to show that while this behavior can be involved in the core of science, this does not prevent the endeavor from moving forward, epistemically speaking, even though dishonesty may often have an opposite, hampering effect. Judgments of credit can be left to posterity, especially as we nowadays have so many skilled historians of science around. I agree with Collins when he writes that: “… the essential values of science are far more important than science’s products and findings.” But only if we realize that these essential values are nonetheless strongly attached to the process of acceptance and rejection of the content of these ‘products and findings’. Precisely this is lost in Collins argument.
Thus the fourth option actually involves moving away from the strong variant of SSK. Collins himself has already started to shift his position. While we still may not agree, the discussion about what the core values of science are, and how they can be related and ranked, is in itself already a discussion that takes us beyond the SSK programme. In conclusion it seems to me that we are in a boxing match, the round after the referee has called an eight count. So SSK may not be K.O. yet, but remember: for scoring purposes, a standing eight count is treated as a knockdown.
With thanks to Ivan Flis for valuable suggestions for improvement on an earlier draft.
 See Golinksi (2005).
 Next to a number of papers Collins devoted three books to the subject: Changing Order. Replication and Induction in Scientific Practice (1985), Gravity’s Shadow. The Search for Gravitational Waves (2004) and Gravity’s Ghost: Scientific Discovery in the 21st Century (2011).
 McAllister (2013) p.130.
 For an elaborate exposition of this issue, see my Ph.D. thesis Pluralism within Parameters, which can be accessed online: https://openaccess.leidenuniv.nl/handle/1887/36396.
 For a splendid example of gradual development towards consensus see Rudwick (1985) on the Devonian controversy in geology. Not surprisingly Collins has never accepted that the two mavericks at the end of the story, who decided not to join the consensus, come out as ageing buffoons instead of defenders of respectable positions. Note that Collins repeats this stance towards Rudwick in Collins (2004).
 Collins (2011) p.153.
 Collins (2004) p.784.
 In his column in De Volkskrant (one of the main Dutch newspapers) Bert Wagendorp cynically mentioned the incredible coincidence of the black hole collision billions of years ago, sending out waves signals, which reached the Earth just at the moment when our apparatus to detect them was finally ready. This is a nice piece of rhetoric but it is simply not true: LIGO has been in operation more than a decade and it is not the only event it should be able to detect. Moreover LIGO recently (June 2016) claimed a second measurement of gravitational waves due to another collision of black holes 1,4 billion years ago.
 Collins (2004) p.788-789.
 Collins (2004) p.785.
 Collins (2011) p.3.
 Collins (2011) p.155.
 cf. Collins (2011) p.157. See also Collins (2009).
 Collins (2011) p.157.
Collins, H. M. (1985) Changing Order: Replication and Induction in Scientific Practice. Beverley Hills and London: Sage.
Collins, H.M. (2004) Gravity’s Shadow: The Search for Gravitational Waves. Chicago: University of Chicago Press.
Collins H.M. (2009) ‘We Cannot Live by Skepticism Alone’. In: Nature 458, pp.30-31.
Collins H.M. (2011) Gravity’s Ghost: Scientific Discovery in the 21st Century. Chicago: University of Chicago Press.
Franklin, A. (1998) ‘Avoiding the Experimenter’s Regress’. In: N. Koertge ed., A House Built on Sand: Exposing Postmodernist Myths about Science, pp. 151-165. New York and Oxford: Oxford University Press.
Golinski, J., (2005) Making Natural Knowledge. Constructivism and the History of Science. With a New Preface. Chicago: University of Chicago Press.
Karstens B. (2015) Pluralism within Parameters. Towards a Mature Evaluative Historiography of Science. Leiden University (diss.)
McAllister J. W. (2013) ‘Empirical Evidence that the World Is Not a Computer’. In: M. Emmer ed., Imagine Math 2: Between Culture and Mathematics, pp. 127-135. Milan: Springer.
Rudwick, M.J.S. (1985) The Great Devonian Controversy: The Shaping of Scientific Knowledge among Gentlemanly Specialists. Chicago and London: University of Chicago Press.
Wagendorp, B. (2016) ‘Rillingen’. Column in De Volkskrant, 13-2-2016