§45. Experiments in Writing
Medicine exemplifies the experimental reasoning Boyle wants to establish in natural philosophy, but that is not the only page he takes from medical tradition. He repeatedly expressed as his own the Hippocratic idea of experience, which includes the recorded and remembered experience (istoria) of a practice community.
An intermittent quality of Hippocratic texts became with him a new style of writing—writing up experiments. Practitioners in the early Royal Society tacitly commended narrative that was “imperfect, hesitant, diffident, halting, and visibly lacking in the rhetorical arts of persuasion.” Boyle gave early lessons on how to strike the right tone. Alluding to his “densely circumstantial prose,” historian of science Steven Shapin explains how “Boyle sought to make historically specific experimental performances vivid in readers' minds and to make it morally warrantable that these things had actually been done as, when, and where described.” Too much consistency is suspicious; incoherently circumstantial narrative becomes a voucher of veracity, and Boyle urges experimental writers to communicate troubles and failures as well as success. We saw this in Hippocratic authors (§4), though Boyle's colleagues seemed to consider it an English trait; for instance, George Starkey, writing to Boyle of alchemical operations: “Now to show yr Honor yt I am an English man I shal adde a disaster.”178Profuse circumstantial detail is a known technique for creating the impression of reality in a text. Boyle is artfully artless, not unlike Montaigne, and wrote two essays dedicated to the topic of experimental failure. “The instances we have given you of the contingencies of experiments, may [i.e., ought to] make you think yourself obliged to try those experiments very carefully, and more than once, upon which you mean to build considerable superstructures either theoretical or practical; and to think it unsafe to rely too much upon single experiments.” He emphasizes repetition of whatever experiment one tries, whether successful or not.
He recounts the occasion of an experimental result entirely unexpected but gratifying for his hypothesis. Nevertheless, because “we do not willingly rely on a single trial of such things, as we know not to have been ever tried before,” he thought it “not amiss for greater security to make the experiment the second time.” A good thing too, as he could not replicate his own result, not then or since.179Boyle’s treatises are innovations in experimental literature. How does one write up experiments? What is the format, where is the model? This literature had to be invented. Textual near relations like books of natural magic or alchemy were discredited and not tending in the direction Boyle wants to take experimental philosophy. He contributed thirty-six papers to the Philosophic Transactions of the Royal Society of London, edited by Boyle’s associate Henry Oldenberg. Commencing publication in 1665, this was the first scientific journal in English, and the volumes became models of experimental literature.180
Initially, many of the experiments were cookbook recipes for marvelous or practical effects. Within the first decade, however, the understanding of experiment narrows to a conscious investigation of phenomena usually involving artificial intervention, though publication of marvels continued. By the third decade of the Transactions, experiments have hypothesistesting and debate-resolving functions, and are recognized as created events designed with specific claims about nature in mind. By the end of the seventeenth century the role of experiment shifted from adjudicating brute claims about nature (Do vacua exist? Is there a spring of the air?), to clarifying and establishing uncertain points in natural philosophy. The work is now understood in terms of problems that motivate experiments, which operate as instruments to advance questions rather than discover secrets.
As more came to depend on forensic credibility, reports began to expand on procedural detail. Beginning with volume 25 (1690), skillfully executed illustrations show important details of experimental apparatus. The value of methodological self-consciousness also rises. Earlier methods were described only in passing, but by the end of the eighteenth century experimenters appreciate that differences of conditions and execution affect results, and they pay attention to variations in equipment that might affect outcomes. It took about a 150 years for the English experimental article to evolve from an uncontested report of an experience to a methodically controlled instrument for advancing a research question.