Skip to main content


A 40-year-old tome’s prescient observations about scientific fact-making resonate today

Laboratory Life: The Social Construction of Scientific Facts

Bruno Latour and Steve Woolgar
Sage Publications
271 pp.
Purchase this item now

In the 1970s, sociologist Bruno Latour wanted to observe how scientific facts emerged in real time. Preferring not to rely on researchers’ own accounts of how discoveries were made, he was keen to witness scientists work in situ. To achieve this, he served for 2 years as a resident anthropologist in Roger Guillemin’s laboratory at the Salk Institute for Biological Studies.

The account of his stay, coauthored with Steve Woolgar and published 40 years ago under the title Laboratory Life: The Social Construction of Scientific Facts, raised contentious questions about the nature of scientific truths. Challenging the idea that facts are things discovered by scientists that reflect an objective reality, the authors argued that facts are instead tools constructed by scientists themselves. The book’s central claim divided scientists, while sociologists hailed it as a watershed moment in the study of science.

Since the 1970s, the public’s faith in science as an arbiter of what is and what is not fact has waned. As a seminal work that casts fact-generating as a sort of culture, Laboratory Life can help readers make sense of­ science’s ambiguous role in today’s society.

Since the Enlightenment, philosophers have debated whether anyone can claim to know anything about reality. Spurred by rapid technological change and the rise of postmodernism, this question experienced a resurgence in the latter part of the 20th century. Laboratory Life was part of this revival. Its authors were particularly interested in claims associated with specialized knowledge—the sort of facts that require teams of scientists and sophisticated equipment to reveal.

During his time in Guillemin’s lab, Latour noticed the defining role that language played in bringing well-hidden facts to light. He argued that it was not the experiment per se but the scientist’s interpretation of an experiment’s result that made phenomena real.

When an experiment was conducted for the first time, any discussion about its findings was often hedged with qualifying statements (e.g., “it seems like” and “this appears to be”). However, as a finding became more robust—if, for example, it was validated using a different method—the number of possible interpretations scientists drew upon grew smaller. Finally, when a finding became widely reproducible or attained a certain degree of precision, it began to be regarded as having obtained a sort of stability. At this point, the existence of the fact no longer hinged on a scientist’s interpretation; it was considered a truth unto itself. But, as Latour and Woolgar noted, the consensus point where an observation was deemed to reach this stability was, by nature, arbitrary.

Sociologists found this troubling. If language is the only way to signal what is a fact and what is not, then one could argue that facts themselves reside in language, not in an external reality. After all, if reality is essentially limitless, then language, which is finite, can never faithfully describe reality. Scientists may dismiss this caveat as academic, but thinking about scientific facts as tools constructed by language helps clarify why current anti-scientific rhetoric is so effective.

Because language allows for a gray area between unsubstantiated claims and concrete facts, the role of the scientific community is to carefully shepherd findings along this continuum. In 1979, when Laboratory Life was published, this journey was fairly insulated. At that time, discerning between what was fact and what was artifact was left mostly to scientists themselves.

Today, it is a different story. The internet is a wide web of contradiction, where information can move rapidly within social networks. This renders the complicated phenomena that scientists study vulnerable to the turbulence that rhetoric can create. Findings that can upset the status quo—such as the biological mechanisms governing sexual orientation, or the safety of genetically modified food—can be diluted through rhetoric, regardless of whether the findings in question reflect an objective reality. If facts are tools made by language, they can be undone by language too.

Forty years after its publication, Laboratory Life remains prescient in its ability to encourage scientists to see that descriptions of reality and reality itself are not the same thing. The gap that separates the two can, however, be made smaller by good science. While we may never touch reality, we can certainly get very close.

About the author

The reviewer is at the Salk Institute for Biological Studies, La Jolla, CA 92037, USA.