§61. Observation and Its Others

The first ideas about observation arise in ancient medical empiricism and its argument for the superiority of experience over theory. The new prestige of experiments from the seventeenth century is a second context for evaluating observation.

Aristotle had no special term for observation and no theory of its con­tribution to science. However, “observe” (terein) is a keyword of Epicurean methodology, which relies on observations to discover the unobserved. The “observations” are perceptual evidence that is to be taken exactly as given, though there is more to Epicurean observation than perception, for memory is a component. Scholars explain that the primary meaning of terein in Greek is “to keep” or “watch,” and Epicurus does not neglect this connotation. Perceptions become observations when they are duly “kept” in awareness as we investigate and philosophize, when we attend to and remember them, re­flect and combine them, and in that way transform them into evidence for explanations. This is ancient empiricism’s deviation from Plato and Aristotle. The initial concepts of science are a mnemic synthesis of observations, and science consists in using them in inferences to the unobserved.⁹⁷

In medieval science the term observatio is restricted to astronomy and meteorology, with no place in natural history before the sixteenth century. A largely complete Latin Hippocrates was not available until 1525. Among the newly available texts was Epidemics, which gave prominence to the method of case histories. The work contains forty-two fully developed histo­ries, which have been described as among the best examples of Greek empir­ical science.

For their own use these physicians created a genre of medical observationes, collections of observations made and checked by physicians in the course of medical practice, cases newly gathered from the field and not from books. Acknowledging the new genre, Petrus Ramus in 1569 began the prac­tice of translating istoria in Aristotle as observatio. It has been said that the observationes genre traveled easily. It traveled geographically, socially, and across disciplines, from medicine to natural history and astronomy. The orig­inally medical concept of observatio joined the discourse on experience and its value in science. These works are unencumbered by learned references, and studiously avoid theoretical points. The new format became a medium of international exchange in the Republic of Letters, and was taken over by the first scientific periodicals.

By the beginning of the seventeenth century observation had become a distinct term and style, contributing to the revaluation of experience in nat­ural philosophy. Its new prominence also revived the old tension between observation and hypothesis. The collection of observations, energetically taken up by a kind of medical radical empiricism, revived the Hippocratic legacy of the physician contributing to a body of collective knowledge, and was called on to settle the boundaries of medical science, which was founded on observed fact and repeated experience without hypothesis or theory.

Hippocratics honored such contributions and wrote them into the physician’s credo, though there is something new in this neo-Hippocratic ob­servation. Observation and being a good observer become a kind of voca­tion eliciting ethical dedication. Before it was a scientific term, observatio had the sense of observing the law, observing rules, being observant about rites. Its first home was in the monastery. In the Renaissance, monastic and med­ical lineages converge. Observation becomes a mode of observance, being observant becomes a way of life, empiricism becomes an ethical technics of the self.

Observation settled into the canon of scientific methods in the seven­teenth century, when a new question arose about its relation to experi­ment. Initially, perhaps a legacy of the medieval synonymy of experimenta and experientia, observation and experiment tended to be used indiscrim­inately. Leibniz distinguished them in terms of an active or passive agent. “There are certain experiments that would be better called observations, in which one considers rather than produces the work.” Observing is passive, experimenting should not be. Boyle had already understood observation and experiment in this way. Observations are made “of what nature does, without being over-ruled by the power and skill of man,” while experiments are made when “nature is guided, and as it were, mastered by art.” D’Alembert was in­terested in how observation and experiment combine and complement each other. “Observation, by the curiosity it inspires and the gaps that it leaves, leads to experiment; experiment returns to observation by the same curiosity that seeks to fill and close the gaps still more; thus one can regard experiment and observation as in some fashion the consequence and the complement of one another.” This was already Bacon’s idea, and suggests his influence on the philosophy of the Encyclopedie."

Claude Bernard also finds complementarity. “Observation shows and ex­periment teaches.” Georges Cuvier more aggressively advances the value of ex­periment over observation.

John F. W Herschel, another prominent nineteenth-century experimentalist, deconstructs the dichotomy. Every experiment involves observation, and there is something experimental about any truly searching observation. He finds observation, especially new observations with new or improved instruments, among the best helps to experimental innovation and discovery. “It is to the dis­covery of improved methods of observation that the chiefprogress of those parts of science which depend on exact determinations is owing,” he says, mentioning the balance, galvanometer, and methods for measuring refraction. Modern efforts at cosmological measurement are usually referred to as “experiments.” One of the first, the Cosmic Background Explorer (COBE) satellite (1992), was itself an experiment in observation, testing its uttermost limits, observing traces of events that occurred in the cosmic era of decoupling, at 10^-35 seconds after the Big Bang. The instrumentation of the Laser Interferometer Gravity-Wave Observatory (LIGO) detects changes of 10^-16 cm, which is one hundred mil­lionth the diameter of the hydrogen atom.

The Hippocratic worry about observation and theory came back in the twen­tieth century with logical empiricism. This version of positivism reduced the interaction of theory and experience to a logical relation among propositions. Experiments are black boxes generating propositions that bear logical relations to statements of theory. Instruments are irrelevant and experimental agency is invisible. There is nothing for experimenters to do except vocalize protocol sentences (§66). When we open the black box, however, what we learn about experiments destroys the positivist concept of their evidence (§67). Passive ob­servation occurs only in finished science, when the meaning of the experience is uncontested. Nothing new is disclosed to an eye that can only gaze. If that eye is not attached to a hand, it might as well not look. Eye, hand, and intelligence are inseparable from experiments and their capacity to generate data.¹⁰²

A truly non-hypothetical observation seems unlikely, as it proves difficult to separate what is observed and what is believed about it, our “hypotheses.” Science requires theories to discover new ways to observe; for instance, how to observe the sun’s core by collecting neutrinos. An emission of solar neutrinos was a testable implication of a hypothesis about the nuclear physics of stellar burning. “Is there any hope of observing these solar neutrinos?” a scientist won­dered. “If Dr. Davis obtains a positive result, that is, if he actually detects some neutrinos from the sun, he will have given us for the first time a direct proof of the occurrence of nuclear reactions in the stars.” That is not an observation Galileo could have made. We can make it because we have the theory and the instruments to allow solar neutrinos to leave a trace (ifthey exist), to make an in­dicative sign, to speak, as it were, and make their presence known. Observation no longer means perception.

Careful, systematic observation no less than the artificial conditions of exper­iment enable latent, cryptic, reticent things that are not normally indicated to leave a trace, which becomes our sign of their existence. Practically any worth­while observation is to a degree experimental or at least searching, uncertain, and perplexed. Observations become experiments, useless without hypotheses. We can be sure that these distinctions are superfluous when it can be said, “The hypothesis that predicted the existence of a planet Neptune was tested by the ex­periment of directing a telescope towards a certain predicated region of the sky” The pure gaze of the Renaissance has lost its luster, and the polemic that made us choose loyalty to observations or experiments or theories is no longer compel­ling or even intelligible.¹⁰⁴