Review: Representation in Scientific Practice Revisited

Representation in Scientific Practice Revisited
Reviewed by Amy Ione

Representation in Scientific Practice Revisited offers an explicit sequel to the discussion featured in the 1990 book Representation in Scientific Practice [1]. I use the word sequel because this more recent volume is not an update so much as an effort to show that the questions surrounding representation inhabit a quite different theoretical and conceptual landscape 25 years later.

The 1990 book grew out of a workshop on “Visualization and Cognition” held in Paris in 1983 [2]. Although a compilation of already published articles, the book is now remembered as a contribution that helped to coalesce the late 20th century discourse on scientific visualization among historians, philosophers, and sociologists of science on visualization and representation. In some ways it was also representative of how Kuhnian paradigms had changed thinking. Thomas Kuhn introduced paradigmatic thinking in his 1962 book The Structure of Scientific Revolutions [3]. His thesis about distinctive ways of thinking in historical eras, in turn, laid the foundation for a focus on scientific context and a more nuanced approach to ideas and practices. With the first Representations volume it was clear that the discussion had shifted accordingly and included enhanced sensitivity to how humanists and social scientists perceived and modeled reality. Within this framework, epistemological thinking and practices were elevated.

The second volume demonstrates that this sea change brought about a focus on ethnographic studies within Science and Technology Studies (STS). The systematic study of scientists working and the environments in which they practice is so predominant in the articles of the second volume that an unacknowledged subtheme of the book is the degree to which practices within environments are now representative of what Kuhn might call a “normal” approach in historical, humanistic, and sociological investigation. Indeed, as author after author explained the design of his or her ethnographic study it is hard to miss how standardized the approach is. No doubt this is why some of the authors ask if the time is ripe for a shift from an epistemological to an ontological treatment of the representations concept.

Representation in Scientific Practice Revisited itself is comprised of 14 lengthy papers primarily by younger scholars and seven brief, reflective pieces by established academics.

Five authors contributed to both books: Bruno Latour, John Law, Michael Lynch, Lucy Suchman, and Steve Woolgar. Three elements stand out within this array. First, the majority of the articles highlighted contemporary modalities of “representation” and “visualization,” in effect stressing that a number of tools related to imaging and digital technologies were developed after the first volume was published. Second, the narrowly-based ethnographic studies of most articles tended to mute the historical context that generally adds bite to work within the STS field. Third, the failure to include a comprehensive essay charting the last 25 years gave the impression that the book was intended for the specialized STS community more than a broader audience. It seemed the short essays were intended to serve this overview purpose, but because they were general outlines rather than detailed critical analyses, they did not. Thus, like many anthologies, the diverse and thought-provoking articles in this collection do not provide a cohesive, in-depth window from which to view representations in scientific practice. Instead, each author aims to identify particular contemporary concerns, constituents, mechanisms, and animating features within the parameters of his or her research topic. Despite the lack of a comprehensive assessment, the volume still succeeds in holding one’s attention due to the high quality contributions. The upshot is that the volume shows what some people working in this area are thinking and invites the reader to seek out more studies.

The reason that the short reflective contributions failed to bring cohesion to the study is that there was no organizing overview to integrate of the critical concerns of the authors. Bruno Latour, who contributed to both Representations volumes, began to argue for the term “visualization” in the 1980s. His view is that visualization more accurately connotes how practices make things visible than terms like perception or observation. This view is based on the idea that often what we perceive is invisible to the eye directly. His essay in this book asserts that scientific imagery allows us to move the gap between what we know and what was previously known, and argues for more manipulations. “Only once the mimetic and scopic obsession of an image as a copy has been put aside will it be possible to study scientific imagery” (p. 350). Martin Kemp, by contrast, does not focus on a gap we cross so much as a need to reconcile the degree to which we can trust what images say, particularly when they are manipulated. It wasn’t just that the Latour and Kemp essays were too short to adequately show the tension between their views; they also reminded this author that science does not take place in a vacuum. Indeed, Kemp’s art history perspective seemed like an outlier in the book. I find this strange because so much of a science/social science owes a large to debt to Kemp as well as historians like Martin Rudwick and art historians such as Samuel Edgerton and Svetlana Alpers.

Cyrus C. M. Moody’s engagement with Thomas Kuhn’s work in his essay “Essential Tensions and Representational Strategies” shows that Kuhnian ideas still retain a place within STS discussions. Moody’s title refers to what Thomas Kuhn called the “essential tension of science”. He argues that Kuhn incorrectly treated it as a tension between innovation and tradition that demonstrated the degree to which conventions help scientists frame and communicate about their work. Using probe microscopy, Moody makes his case by looking at the complex relationship between representational conventions and (un)conventionality in this research, concluding that contemporary instruments aid in seeing why the “essential tension” is a misnomer because innovation and tradition are not in conflict when scientists go about their work.

Moody’s argument, like most of the papers in the book, used an ethnographic component to state its case. Janet Vertesi’s ethnographic study explores how digital tools allow us to both construct knowledge of an alien planet (Mars) and inscribe this same knowledge into an image. She calls this framework “drawing as. Catelijne Coopmans writes about visual analytics, the use of data visualizations to share and comment on inaccessible information. Her case study of software demonstrations by the visual analytics vendor Tableau Software demonstrates how web seminars work in helping people see and use data. Rachel Prentice’s examination of surgical practices, like Morana Alac’s effort to see how practitioners communicate in cognitive neuroscience laboratories, likewise underscored that systematic studies of practices and communication strategies are now the core driver of STS work. There were also excellent essays by Michael J. Barany and Donald MacKenzie, Sarah de Rijcke and Anne Beaulieu, and Emma K. Frow.

The most compelling essay is Joseph Dumit’s discussion on the use of brain scans in legal cases. Its appeal comes from a sociological thrust that connects the subject to broader themes. In this paper the author looks at the arguments presented in the Supreme Court case, Roper v. Simmons 543 U.S. 551, 2005, where both the defense and the prosecution used representations, brain images, to argue their legal positions. The defense asserted that their 17-year old client was ineligible for capital punishment due to his age using a “bright line” argument. A bright law is a judicial rule that helps resolve ambiguous issues and the law has established a “bright line” between maturity and adulthood. The gist of the defense argument was that their client was legally defined as an adolescent, and adolescents are not as mature as adults. Given his age he should not be held culpable for murder. Their expert witnesses used brain scans to explain that the brain develops over time. The prosecutor came back with a different brain narrative, also supported by brain images. He argued that this case was not about developmental immaturity. Rather, the youth was truly dangerous because his actions revealed he was born with a “criminal brain.” Because Dumit positions the two arguments within the long history of behavioral labeling, the author makes it is clear that the tension between the two views in this courtroom is not merely representative of contemporary legalisms, but also are a part of historical debates on how we become who we are. Dumit also points out that neither argument in the Roper case explains why many mature adults do untenable things, just as many teenagers are careful, cautious, and appear quite “mature.” More importantly, Dumit correctly shows that the polemical use of neuroscience analysis is tempting but a comprehensive look at the range of viewpoints makes it difficult to separate the wheat from the chaff.

I found the papers that tried to define the problems within the field particularly fascinating because the majority of papers seemed to be cut from the same cloth. A few authors tried to expand the ethnographic methodology beyond its epistemological emphasis. For example, in “Toward a New Ontology of Scientific Vision” by Annamaria Carusi and Aud Sissel Hoel tried to offer a framework for dismantling the qualitative-quantitative distinction by weaving a case study with an ontological approach. They analyzed the practices and instrumentation of computational biology and tried to critique it using Merleau-Ponty’s later work, which was based on the biological thought of Jakob von Uexküll. Their argument seemed like a step backwards. The views of particular philosophers, who operated within their own context, seem to tie things up more than they open up avenues for new ways of thinking.

Another effort was Lorraine Daston’s comment in “Beyond Representation.” Daston points out that much of the literature on scientific visualization in the 1990s was focused on the process of transforming (her italics) the object. This included using context, images and detailed relationships to demonstrate that technical products are in effect tools of the work/learning process. Now, she asserts, it is time to think about images beyond representation because representation is “an intrinsically epistemological notion” and “has a Kantian ring.” Whereas she is right that “[t]orepresent conjures up associations with spectacle” (p. 320), her brief comments did not leave the impression that she had developed a well thought out program in terms of how ontological notions would encompass how both mutability and scale so as to inform the complex issues surrounding representation.

Indeed, as I read it seemed that mutability and scale were the missing kernels in the book, perhaps due to the book’s largely contemporary vantage point. I kept thinking of this omission in terms of both the kind of representation offered in The Powers of Ten by Charles and Ray Eames [4] and the histories that have accompanied our perceiving our various frames of reference. For those unfamiliar with the 1977 Eames film, it uses logarthnmic magnification to demonstrate orders of magnitude. Beginning with two figures having a picnic on earth, the documentary expands out from the Earth using a logarithmic scale based on a factor of ten until the entire universe is surveyed. The original scene quickly fades from prominence, followed by the earth itself, the solar system galaxies and so on. Once we are lost in the vastness of space they incrementally return the viewer down to the picnickers and then they reverse the equation. Going inward they reduce the scene so that a single atom and its constituent quarks are observed. Needless to say, the film conveyed the macro and micro levels as known in 1977. If made today it would show an expanded story. The essays in the book do not capture this multi-dimensional complexity, however.

In summary, whereas I highly recommend all of the contributions on their own terms, the book as a whole does not off a robust commentary. It reads more like a list, strangely detached from the complexity of the world. As such it raises the question of what exactly do academics do in the large and small sense. In this case, all of the discrete pieces have value and yet the composite brings the Metterling Lists to mind [5]. Woody Allen’s story on Metterling’s lists mocks the sociological trivia that academics collect, explaining that the data points used to penetrate Metterling’s philosophical thinking even included the scholar’s changing preferences in sock colors as noted on his laundry lists. While laundry lists are omitted in this sequel, it does raise the question of how our now standard sociological orientations to representational practice (and other modalities) will move from the currently normalized procedures for study into new and pathbreaking insights.


1. Lynch, Michael, and Steve Woolgar. 1990. Representation in Scientific Practice. Cambridge, MA: MIT Press).

2. The workshop on “Visualization and Cognition” was held at the Centre de Sociologie de l’Innovation at the Ecole Nationale SupÉrieure des Mine de Paris on 12-15 December 1983.

3. Kuhn, Thomas S. 1962. The Structure of Scientific Revolutions. Chicago: University of Chicago Press.

4. The Powers of TenTM (1977) is available on YouTube, see

5. Allen, Woody. 2007. “The Metterling Lists.” In The Insanity Defense: The Complete Prose, (New York: Random House, 3-10).

Reviewed for Leonardo Reviews:

Representation in Scientific Practice Revisited
by Catelijne Coopmans, Janet Vertesi, Michael Lynch, and Steve Woolgar, Editors
The MIT Press, Cambridge, MA, 2014
384 pp., illus. 27 col., 37 b/w. Paper: $38.00
ISBN: 9780262525381.