Tacit Knowledge and Tactile History: Otto Sibum and “Gestural Knowledge”

This post is the first in a short series on what I call “tactile history”: the practice of historical research that extends beyond examining documents to examining the objects of science and the locations they inhabited, and to the actual reenactment of historical scientific research.  The objective of tactile history is to recover aspects of historical work that would not have survived in the form of a written report.  In this vein, tactile history could be seen as a step beyond “notebook studies” — say, Gerald Holton on Robert Millikan’s oil drop experiments,* or Gerald Geison’s The Private Science of Louis Pasteur (1995) — which look beyond scientific publication to recover the messier day-to-day practices of scientific life.

Where laboratory notebooks merely recover otherwise hidden practices, tactile history attempts to recover something that was never expressed in any form, and is often referred to as “tacit knowledge”.  This could be an inexpressible Fingerspitzengefühl (a fine-tuned hands-on knowledge), a lack of understanding of why an experiment works, pattern recognition, or an unreasoned premonition about what new scientific knowledge will look like.  In the 1980s, tacit knowledge became a crucial part of the “controversy studies” literature, because it was understood to be elemental in successfully replicating an experiment.  By studying controversies surrounding replication, one could uncover the many tacit preconditions underlying successful replication.

On this note, here is sociologist Harry Collins from his seminal paper, “The Seven Sexes: A Study in the Sociology of a Phenomenon, or The Replication of Experiments in Physics,” Sociology 9 (1975): 205-224, on p. 207:

…the problem becomes one of explaining the successful copies of experiments rather than the failures.  The model which seems most appropriate is one which involves the transmission of a culture which legitimates and limits the parameters requiring control in the experimental situation, without necessarily formulating, enumerating or understanding them, and which ipso-facto generates the set of anomalous experiments (failures which can’t be explained by uncontrolled legitimate parameters).

At its best, the historical literature stemming from interest in this issue elucidated the importance of instrument standardization, the complexities of experimental interpretation (Peter Galison’s How Experiments End (1987) is exemplary), or the evolution of “literary technologies” for describing experimental results (Steven Shapin and Simon Schaffer’s Leviathan and the Air Pump (1985)).  At its less-than-best, the literature satisfied itself by tying experimental results and metrics to the rhetoric surrounding them (see my rather cryptic remarks on the “historiography of values” in this post on Schaffer’s works on metrics; see also posts on the Schaffer-Shapiro dispute here and here).

Otto Sibum’s “Reworking the Mechanical Value of Heat: Instruments of Precision and Gestures of Accuracy in Early Victorian England,” Studies in History and Philosophy of Science 26 (1995): 73-106 (paywall), occupies a rather high point within this literature.  In it, Sibum recounts his efforts to replicate James Joule’s (1784-1858) famous “paddle-wheel” experiments (diagrammed above) wherein a falling weight does work on an enclosed volume of water, which then increases in temperature, thereby demonstrating the equivalence of different forms of (what would soon be called) energy.

(See James Prescott Joule, “On the Mechanical Equivalent of Heat,” Philosophical Transactions (1850) — free.)

In broad outlines, Sibum’s paper recapitulated the then-common point that many critical experiments were more difficult to replicate than the experimenters let on.  Sibum notes, in particular, that his own attempts to replicate Joule’s experiments indicated that the vagueries surrounding the successful conduct of precision thermometry made it difficult to arrive at the quantitative equivalent reported by Joule.  Yet, he observes that there “is no literary trace of Joule’s thermometrical skills, nor does he give information about the machinery, possible problems in performance, or his likely assistants” (78) in his publications, his notebooks, or his private correspondence.  Further, these skills did not become the center of disputes over Joule’s work.

Thus, it was only through physically replicating the experiments that these aspects of Joule’s work became detectable.  Sibum states (76):

In order to be able to refer to my ‘local knowledge’ [citing Clifford Geertz’s 1983 collection Local Knowledge: Further Essays in Interpretive Anthropology] I use the term gestural knowledge for the complex of skills and forms of mastery developed in these real-time [experimental] performances.

This “gestural knowledge” turns out to have an intricate structure.   Sibum was unable to replicate Joule’s figures for the mechanical equivalent of heat (82):

My results show the lack of sufficient enculturation in order to accustom myself to the techniques involved in Joule’s trial.  But they gave me a sense for Joule’s meanings for ‘exactness’, ‘accurate thermometrical researches’, ‘to obtain that relation with still greater accuracy’.

Sibum then (with a nice bit of document-based research) goes on to link Joule’s thermometry skills to his background in brewing.  In the first half of the nineteenth century, brewing became an exacting exercise in England, in which the process had to be carefully controlled from beginning to end in order to produce consistent and high quality beers.  State excises further made it so that it had to be possible to carefully monitor the process.  This change in brewing was accompanied by necessary improvements in brewing instrumentation and knowledge, some of which was actually encapsulated in treatises.  However, the use of such instruments and knowledge in the production of particular brews also required learned skills (85):

…for the determination of the appropriate mashing heats, personal experience was necessary in order to produce an invariable product.  The correct temperature did not only depend upon the varying quantity and quality of the malt, and the changing temperature of the atmosphere according to a particular season, but also on the situation of and amount of radiation from the mash tun.  Therefore only long experience and a certain ‘habit of taking his mashing heats’ allowed a successful control of the brewing process.  Brewers and malsters were bearers of this gestural knowledge.  Mashing heats were often secrets of local brewers and depended on their particular scales and modes of production.  The malster’s skill made him one of the most important and respected ‘agents’.

For even such skilled individuals, it was possible for subtle changes in brewing conditions to cause breweries to go “out of order” and produce “foul or bad beer, which could go on for weeks” — but, because such events merely reflected the sensitivity an variability of the brewing process, “it was never regarded as the fault of the malting or brewing agent” (86).

Sibum argues it was just these unfailing skills that were necessary to take proper and highly precise measurements of temperature in an experimental setting where the behavior of thermometers as they change their reading matters, and where even the presence of the experimenter can cause significant changes in temperature measurements.  For these reasons, such experiments could not even be witnessed directly.  As Joule himself noted, experiments done in outside of his cellar laboratory, where others could witness them, “though abundantly sufficient to establish the equivalency of heat to mechanical power, were not adapted to determine the equivalent with very great numerical accuracy, owing to the apparatus having been situated in the open air, and having been in consequence liable to great cooling or heating effects from the atmosphere” (102).

According to Sibum, in bringing his skills to natural philosophy, Joule “had become a performer without an audience” (101).  The “embodied capability, that particular gestural knowledge” honed in his work as a brewer, “was incommunicable.  Over and above that, he could not defend his experimental accuracy by appealing to brewers’ craftsman skill.  On the contrary it was more advantageous to distance himself from this particular gestural collective” (103).  His capability was subsumed in publications under references to precision instrumentation, the unexplained skill required to use it, and tables of the values produced therewith.  Confidence in these figures rested on confidence in his personal skill as an experimenter and in his instruments, and not to the more arcane techniques of brewing.  Recovery of the nature of the skills of precision thermometry and their links with brewing was left only to those who actually attempted to follow Joule’s path.

—

*Gerald Holton, “Subelectrons, Presuppositions, and the Millikan-Ehrenhaft Dispute,” Historical Studies in the Physical Sciences 9 (1978): 161-224