Bruno Latour—Steve Woolgar, Laboratory Life. The Social Construction of Scientific Facts. Sage Library of Social Research, Volume 80, Sage Publications, London 1979, 272 S.

The authors have undoubtedly made a considerable contribution to the sociology of scientific research; in this book, however, they are also writing as philosophers, particularly in the chapters "The Construction of a Fact: the Case of TRF(H)" (pp. 105‑150) and "The Microprocessing of Facts" (pp. 151‑186).

On p. 187 they write: 'Similarly in the last chapter we attempted to examine the operation of microprocesses without committing ourselves to either a realist or relativist position.'

On p. 178 they explicitly reject the realist philosophy that physical reality exists independently of our act of knowing. They do indeed try to avoid relativism, but they cannot avoid choosing between philosophical realism and idealism.

A serious omission in the bibliography of Latour and Woolgar, to my mind, is Lenin's 'Materialism and Empiriocriticism'. Had they read this work, they would have been aware of the principal arguments against positions as those they adopt and could have prepared to refute them. It is now only possible to guess at what their answers to Lenin's objections would be.

We need a consistent philosophy which can serve as a foundation of science. If we neglect various forms of objective idealism, we are left with only one consistent form of idealism: solipsism. But since the essential pursuit of science is socially accepted knowledge, and a solipsist denies or considers unproved the existence of any consciousness except his own, solipsism is useless as a foundation for science.

Latour and Woolgar are not solipsists. On pp. 168‑171 they even state that, in a laboratory the development of scientific ideas is so much the result of continual interaction between scientists, that it is often difficult to tell who originated an idea. As the title of the book indicates, they are telling us that scientific facts are socially constructed, and that scientific knowledge exists by virtue of the consensus of scientists. What this amounts to is that an epistemology that can serve as a foundation for science is not required to guarantee the existence of a physical reality independent of our observation, but only the existence of other observers.

The fact that nobody is either such an ubiquitous observer, that he can procure all the data on his own or a thinker of such genius, that he alone can construct all the theories needed to account for the data, so that science can only develop by the collaboration of several scientists, is not at issue. What is at issue is whether, if one denies or considers unproved the physical reality of individual observations, one can base a philosophy of science on the correspondence between the observations of several observers.

The first and most important objection to a non‑realistic philosophy of science is the impossibility of consistently escaping solipsism. This is evident, as soon as we ask ourselves, how it is that we know that there are other people and what they think and observe. The answer is: by observation. I know that there are other people because I see or hear them. I know what they think or observe, because I hear what they say or read what they write.

So if I consider it unproved that all observations reflect a reality outside the observer, I must also consider the physical reality of the observations by which I see my fellowmen and hear what they say or read what they write as unproved. Then there is no means by which I can know whether there are other people, and if there are, what they think or observe; and so I have arrived at solipsism.

Latour and Woolgar think that they have proved what they call constructivistic epistemology by analysing the scientific research process down to its smallest constituents. But these smallest constituents are readings of instruments and discussions of scientists. If Latour and Woolgar were consistent in their constructivistic epistemology, they would also apply it to their own discussions with scientists. Then a sentence spoken by one of the scientists would have to be socially constructed too, i. e. in order to ascertain that a scientist had actually said what he had said Latour (who did the field work for the research) would have had to ascertain that others had heard the same thing as he and would have had to ask other persons present what they heard, etc., and so, if such a way of working could be maintained and had not exhausted the patience of his interlocutors, he would have become involved in an infinite regress. Actually, of course, like any normal person carrying on a conversation, he assumed that, because he had good ears and his interlocutor spoke intelligibly, his interlocutor had actually said what he heard, in other words he considered the words of his interlocutors as a philosophical realist. That is just the pitfall for the opponents of epistemological realism: epistemological realism is so natural for most people, that even the philosophers opposing it theoretically apply it often nonetheless in practice thinking and so become involved in internal contradictions.

A second difficulty for a non‑realistic epistemology is the question of the existence of the world before the appearance of mankind. According to the precepts of science, the earth, the sun, the stars existed long before mankind. But according to the constructivistic epistemology of Latour and Woolgar, scientific truth is socially constructed by men and women. There have been several attempts to reconcile the existence of the world before the human race with an idealistic or constructivistic epistemology, but they have all turned out to be unsatisfactory. (See Lenin, pp. 59‑70.)

A third difficulty for a non‑realistic epistemology is the difference between changes in the world and changes in our knowledge of the world. "Knowledge may change without objects and objects change without knowledge." (Bhaskar, p. 250).

On p. 178 Latour and Woolgar argue that they cannot see in the distinction between physical reality and our representation of it much more than a useless doubling of reality. An omniscient being such as a god could indeed equate his thoughts with physical reality. But we human beings know nature only imperfectly and so we are forced to distinguish between the world and our knowledge of it. The extent of Latour and Woolgar's difficulties in the philosophy of science is not yet fully apparent in a science like neuro‑endocrinology-‑the area of research of the laboratory they studied--engaged in studying the timeless laws of nature. The formula "TRF=Pyro‑Glu‑His-Pro‑NH2" has been valid according to the realistic epistemology anyhow since the time immemorial when in the evolution of the animals TRF developed in their brains (the determination of that moment would be a completely separate and probably far more difficult research program, which in any case was not an area of concern for them). In the case of such timeless truths only one time parameter is important: the time of their discovery; and the absurdities due to identifying the moment of their discovery with their first moment of being can only be clouded over for a short time. But even here there are difficulties as soon as we consider individual events that have occurred before the discovery of particular natural laws. "Ancient parchment rolls were eroded by oxygen in the air" is also said, if those parchment rolls were eroded before Lavoisier discovered oxygen in 1777‑1783 (the discovery of oxygen is epistemologically completely analogous to the discovery of TRF discussed by Latour and Woolgar). From this it is apparent that scientists and the man in the street assume that physical things also exist before their discovery.

The difficulties of the constructivistic epistemology become still greater in cases in which the described part of nature changes across time. For example the statement 'Before 1846 it was not known that the planet Neptune existed' can definitely not be paraphrased as 'Before 1846 the planet Neptune did not exist' and if someone calculates the orbit of Neptune in the past according to Newton's laws (which have to be replaced by Einstein's relativity theory according to the present state of science), (s)he does not stop at the moment that Neptune was discovered.

In order to explain such phenomena Latour and Woolgar have devised 'splitting' and 'inversion'. 'At the same time, the past becomes inverted. TRF has been there all along, just waiting to be revealed for all to see' (p. 177). It is almost like Orwell's '1984', in which the state repeatedly changes the representation of the past in order to bring it into agreement with the requirements of present politics. Such are the absurdities of an epistemology that confuses our representation of the things with the things themselves!

A fourth reason for distinguishing the world from our representations of it is that especially in natural science the sequential order of the underlying bases in nature is almost systematically opposite to the order in which we are aware of their existence and also usually opposite to the order in which they were discovered. See Bhaskar, p. 169.

Against this Latour and Woolgar adduce the difficulties which relativity theory and quantum mechanics present for a realistic epistemology. 'In addition, he (Bhaskar in 'A Realist Theory of Science') wisely confines his discussion to physics, and to pre-Newtonian physics at that. Perhaps the 'independence' of 'intransitive objects of scientific knowledge' would seem less unproblematic in relation to more recently constructed phenomena such as chromosomes or non‑Newtonian physics' (p. 178). But they describe Bhaskar's argumentation completely incorrectly. Actually Bhaskar does treat relativity theory (p. 34, 155, "There is a similar break involved in the transition from Newtonian to Einsteinian dynamics') and quantum mechanics (p. 109: 'Quantum mechanics strengthens the anti‑determinist's hand') and also treats the quite specific methodological problems of sciences in which experiments are not performable at all (sociology) or only with a far less complete control of the variables than in the natural sciences (psychology, pp. 244‑246).

The theory of relativity produces, if it is well understood, no difficulties for a realistic epistemology. The term “theory of relativity" is a somewhat misleading term, but it has been so generally accepted that it is probably not possible to replace it. Still in order to describe the character of the theory of relativity some people have proposed the name "invariant theory", the doctrine of the relations invariant to the location and the acceleration of the observer. It is, by no means, a consequence of "invariant theory" that the meaning of all physical statements is relative to the observer. But it does require a drastic revision of the physical conceptual systems of classical physics, so that among other changes separate space and time have to be replaced by a four‑dimensional space-time continuum, but that in this revised conceptual system the relations among events can be formulated without reference to a definite observer, that the so‑called interval between two events is invariant to the observer's location and acceleration.

Einstein himself never explained his philosophical ideas in great detail, but he did describe his epistemological realism on several occasions, most nuancedly on p. 684: "He (the scientist) therefore must appear to the systematic epistemologist as a type of unscrupulous opportunist: he appears as a realist insofar as he seeks to describe a world independent of the acts of perception; as an idealist insofar as he looks upon the concepts and theories as the free inventions of the human spirit (not logically derivable from what is empirically given); as a positivist insofar as he considers his concepts and theories justified only to the extent to which they furnish a logical representation of relations among sensory experiences. He may even appear as a Platonist or Pythagorean insofar as he considers the viewpoint of logical simplicity as an indispensable and effective tool of research."

For a realistic epistemology quantum mechanics presents more difficulties. The deepest cause of the counter‑intuitive character of many phenomena in quantum mechanics is, that no observation that does not disturb the object observed is possible, and so we do not observe the micro‑object as it was before the observation but as it is modified by the observation. Much has been written about the philosophy of quantum mechanics and a more detailed treatment would lead us too far astray. Yet I have to point out, that the general philosophical objections to an idealist epistemology also hold for quantum mechanics (consistently pursued it renders all communication among scientists impossible; the transition of electrons from one orbit to another in atoms (typically one of the phenomena in the domain of quantum mechanics) must also have occurred innumerable times without someone's observing them, and the adoption of an idealistic epistemology in all natural sciences except quantum mechanics only raises more problems than it solves. It may be a solution to combine a realistic epistemology for the macroworld that includes interactions between people and a philosophy of bipolar interconnectedness between the observer and the object for quantum mechanics.

A second argument of Latour and Woolgar against realistic epistemology is the great and prolonged effort often needed to reveal the truth about a certain topic, and the many contrary opinions often pronounced before the definitive solution of a problem. They give the example of the determination of the formula of TRF, which they had been able to observe in their laboratory, and which eventually appeared to be Pyro‑Glu‑His‑Pro‑NH2. But this only confirms Bhaskar's observation on p. 169, that the layers of material reality most difficult to find, which are found last, are usually the most fundamental in the causal order of the dependency of being.

On p. 186 the authors offer a third argument for their constructivistic epistemology: 'Many of the substances (and their analogues) mentioned in earlier chapters are patented. Substances 'discovered' in the laboratory are described in the texts of patents as having been 'invented'. This shows that the ontological status is rarely likely to be finally settled: depending on the prevailing interests of the parties concerned, the 'same' substance can be given a new status'. But the fact that a laboratory was granted a patent for, a chemical substance by no means proves that the distinction between discovering something and making something is doubtful, but only that a minor conceptual clarification is necessary. If one Means by TRF 'chemically pure TRF', then of course it did not exist before it was made in the laboratory. If one means by TRF TRF molecules, mixed with other molecules or not', then of course TRF has existed for a long time in the brains of innumerable organisms before it was ever isolated in the laboratory. It does not clash with a realistic epistemology that in some cases scientists actually make things and not merely discover them. For example an epistemological realist also says that Kamerlingh Onnes made fluid helium and did not discover it, and that MacMillan and Seaborg made plutonium and did not discover it.

Conclusion: in spite of the merits of this book as a sociological study of a laboratory its plea for the replacement of a realistic by a constructivistic epistemology is not convincing*.

* I am grateful to Mrs. Jean Winner for correcting my English.


Bhaskar, R. A Realist Theory of Science, Leeds Book Ltd., Leeds, 1975, 258 pp.

Einstein, A. Philosopher‑Scientist, Tudor Publishing Company, New York, 1951, 781 pp.  

Latour, B., Woolgar, S. Laboratory Life. The Social Construction of Scientific Facts, Sage Publications, London, 1979, 272 pp.

Lenin, V. I. Materialism i empiriokriticizm. Kriticheskie zametkie ob odnoy reakcionnoy filosofii. (Materialism and Empiriocriticism. Critical Remarks about a Reactionary Philosophy), Gosudarstvennoe izdateljstvo politicheskoy literaturi, Moscow, 352 pp.

Orwell, G. Nineteen Eighty Four, A Novel, Secker & Warburg, London, 1949.


Zeitschrift für allgemeine Wissenschaftstheorie X111/12 (1982)
© Franz Steiner Verlag GmbH, D‑6200 Wiesbaden

Themaat, [Willem A.] Verloren van. 'Review: Bruno Latour-Steve Woolgar, Laboratory Life: The Social Construction of Scientific Facts', Zeitschrift für allgemeine Wissenschaftstheorie XIII/12 (1982), pp. 166-170.

Esperanto Study Guide / Esperanto-Gvidilo

Home Page | Site Map | What's New | Coming Attractions | Book News
Bibliography | Mini-Bibliographies | Study Guides
My Writings
| Other Authors' Texts | Philosophical Quotations
Images & Sounds | External Links

CONTACT Ralph Dumain

Sign Registry View Registry

Uploaded 20 May 2006

Site ©2006 Ralph Dumain