Craft Skills and Tacit Knowledge in Organic Chemistry

All of this suggests that a certain convergence of chemical toward physical hypothetico-deductive methodology took place around mid-century, attributable in substantial part to the emergence of the theory of structure. There is more to the methodological question than that, however. Physics and chemistry were distinguished by a host of different qualities, values, exemplars, habits of thought, details of practice, and so on, and the result of all of these particulars was that the two sciences maintained quite distinct cultures. Especially on the European continent, physics gradually became less concrete, more abstract, and more firmly based on an axiomatized mathematical foundation. The influence of technology was evident in increasingly complex instrumentation and apparatus, and the culture of precision measurements combined with rigorous error analysis imbued the practice of physics ever more strongly.^[24]

None of these qualities characterized chemistry by and large, particularly not the organic field that became so dominant in the middle and later decades of the century. Today as a hundred years ago, most organic chemists have little need for higher mathematics. As far as precision measurements are concerned, they need to know how to weigh out stoichiometric quantities, measure densities, record melting and boiling points, and determine the correct atomic ratios in combustion

analyses,^[25] but these modest calculational and metrical demands cannot compare with those of the physicists. There was technology in the chemical laboratory, as in the physical laboratory, but it consisted mostly of simple materials in simple combinations and could not compare with the complex optical and electrical apparatus being used by the physicists.

What the chemists had, and needed in the fullest measure, were craft and observational skills. Chemistry has sometimes been compared to cookery, and the simile is apt in many respects. Substantial manual dexterity and technical know-how—for glass-blowing, devising and assembling apparatus, sealing connections, cutting rubber and cork, heating, cooling, pouring, mixing, grinding, and on and on—were imperative qualities. There was a standard repertoire of procedures, such as distillation, filtration, solvent extraction, titration, and recrystallization, that quickly became well-practiced habits for the novice. Precise observation was vital—color, viscosity, clarity, smell, taste, texture, crystal size and shape, and so on. An excellent memory was a virtual necessity to deal with the thousands of compounds regularly encountered.

There was much truth, therefore, in the perception that organic chemists were like naturalists exploring exotic ecosystems, only in the chemists' case, the object of study had often just been created for the first time by the investigator himself. When the discipline of physical chemistry was organized in the 1880s, the leading promoters advertised their field as the "chemistry of the future" and battled for social and institutional support with the well-entrenched organikers over the next two or three decades. Organic chemists resented having their lovely science depicted as a mere compounding of novelties, and their more laboratory-oriented values were even shared by a few of the physical chemists themselves. "Of course a man must specialize," W. D. Bancroft said. "He must be a chemist rather than a physicist." About the same time, he confessed privately, "I abominate exact measurements myself." But as a physical chemist, Bancroft was in a distinct minority.^[26]

It must be emphasized that the organic chemist's laboratory repertoire was not a set of simile mechanical operations that all learned and performed equally. Cognitive activity, both conscious and unconscious, must accompany all craft operations. The ultimate quality and success of a reaction or of a recipe, even of a single operation in the process, is closely tied to what is happening in the chemist's or the cook's mind. Consequently, both the manual and the intellectual sides of the craft had to be mastered. This was a good reason for Kolbe's (and many others') emphasis on practical laboratory work as the centerpiece of

chemical education. Precepts, ideas, and data could be learned in the classroom, but the craft skills and tacit knowledge of the professional could only be assimilated side by side with the master and his assistants.

There was a real art to be learned, and this is not simply a trope. Everyone with experience in the laboratory, including the greatest masters, knows of struggles with reactions that just will not go, products that emerge in gummy resins and refuse to give distinct crystals, yields that evaporate to a pittance, and product mixtures that resist every attempt at separation. The virtuosi, however, know how to overcome and circumvent problems and are often able somehow to coax success out of failure. The other side of the coin is the rapture of success, when the dazzling crystals of product suddenly and mysteriously blossom in the recrystallization flask. The ineffable "master's touch"—that of a Wöhler, a Bunsen, or a Kolbe—is the most difficult quality to teach and may well be impossible to impart.

There was an even deeper cognitive side to the chemist's craft. The epistemological technique of transdiction —inferring invisible submicroscopic details from macroscopic observables—was habitual with chemists long before physicists developed a similar art. Ever since chemical atomism arose in the early years of the nineteenth century, chemists were comfortable in routinely inferring the atomistic compositions of molecules from macroscopic gravimetric measurements. There was, of course, an epistemological debate, but most chemists refused to worry excessively about fine philosophical distinctions, and the radical phenomenalism that Gerhardt defended for a time never really became popular.^[27] As the century progressed, transdiction became ever more elaborate, and inferences to composition were supplemented by inferences to molecular structure. When the chemist adds methyl iodide to an etherial solution of potassium ethoxide in a Williamson synthesis, for example, her mind's eye is focused on the activity at the molecular level, watching for signs of smooth assimilation of the two reactants, and hoping that she has sufficiently dried the solvent. Examples of such mental habits can be extracted from the host of reactions discussed in this book.

In general, reactions are assessed by analysis of the product—the payoff of an experiment—which also involves transdictive procedures and assumptions. The object of the game here is separating the components of a reaction mixture and purifying the substance on which interest is focused. Procedures for accomplishing these goals were developed, mostly empirically and many of them during the eighteenth century, and assessment criteria were well standardized. Every chemist knows, and knew in the period with which we are concerned, that con-

stant sharp boiling and melting points are a primary indicator of purity and that admixture of impurities tends to lower and spread out these transition points. Every chemist knows, and knew, that attractive, well-formed, uniform crystals are another positive indicator. If such signs are not observed, then the chemist must try again—another distillation, sublimation, recrystallization, or repetition or modification of the reaction itself.

Elemental analysis is the final step, but only after product purity is attained. If a product matches the crystal appearance and sharp melting point of a known substance, and its percentage of carbon, hydrogen, and oxygen comes within the experimental error of that calculated for the relevant formula, then a match is achieved and the collegial community will probably be satisfied. If the purity indicators are good and the chemist has reason to expect a novel product, then the percentage composition of the observed product is again compared to the predicted formula, using accepted values for atomic weights. Such an identification was usually accompanied by at least a brief examination of chemical and physical properties and characteristic reactions, as well as preparation and characterization of simple derivatives such as salts of acidic or basic compounds, and so on.

Often the point of the organic chemist's investigation was achieved by establishing the very existence of a new compound—such as Williamson's asymmetric ethers. For other research, it was necessary to create and enumerate isomers corresponding to a single formula in order to test an idea. In still other cases, synthetic or analytical reactions needed to be explored to establish or refute conjectures, such as those concerning structural details of a molecule or genetic relationships among compounds. Synthesis was often a goal per se, especially of natural products. Finally, in some cases it was the particular properties of a new substance that provided the point of interest.

Although there certainly were distinctions in the details of laboratory practice in different universities and different countries, the generic chemical procedures, standards, and goals just outlined were quite uniform throughout the world during the period with which we are concerned. Thus, in principle, consensus was attainable regarding the general course of individual reactions and the existence of novel organic compounds. However, there were many potential points of challenge throughout the series of operations just described. Disputes over the "facts" were certainly not uncommon, and we have seen a number of them here, but what is striking is how infrequently experiments by members of the profession were successfully challenged. The story of Kolbe's chimerical salylic acid, which is related in chapter 12, is notable in its relative singularity.

I have gone through this material, at risk of trying the patience of chemically sophisticated readers, because I believe it is helpful for other readers to understand something of the mental equipment, operational repertoire, and epistemological habits of thought to properly appreciate and evaluate the historical action in this narrative. One theme of this book is pedagogy; chemistry students start out at the layman's level and have to acquire all of that equipment during their apprenticeship. We began in the first chapter with the guild model for the German university, and regarding chemistry as both a cognitive and a manual craft fits well in that model. It also needs to be stressed once more that, although chemistry had much in common with physics as a natural or exact science, it had a culture that was quite distinct in many ways, and acquisition of this culture was also part of the educational process.