EXPERIMENTATION AND INSTRUMENTATION

Experiment, William Herschel wrote, is a matter of "putting in action causes and agents over which we have control, and purposely varying their combinations, and noticing what effects take place" (Herschel 1966, p. 76). In this sense, the earliest recorded scientific experiments appeared in biological and medical contexts. In the second century CE, the physician Galen performed detailed animal experiments to find out about the functions of various organs. In the sixteenth century, Andreas Vesalius, pioneer in dissection, carried out elaborate experiments; and William Harvey, notwithstanding his Aristotelian orientation, supported his discovery of the circulation of the blood with painstaking experimental arguments. It is highly plausible that the practice of alchemy also served as an early source of experimentation. From the thirteenth century on, alchemists used laboratory equipment in order to create new agents and were arguing against the overly narrow interpretation of the art-nature divide in Aristotelian philosophy.

A third area where experimentation took place before the scientific revolution was supplied by Ptolemy's optics. Ptolemy, active in the second century CE, formulated an experimental, quantitative law of the refraction of light at the boundary of air and water and performed experiments to investigate binocular vision. In continuing this tradition in the early eleventh century CE, the Arab Ibn al-Haytham (Alhazen) wrote an impressive experimental treatise on optics in which he related in a mathematically demanding way the physics and geometry of light to the anatomy of the eye. Al-Haytham's work was translated into Latin in the thirteenth century and decisively influenced later optical research for a long time. Because of this and similar developments, Crombie saw experimental science of the modern world created by thirteenth-century philosophers of the West transforming Greek geometrical method and uniting it with the experimental habit of the practical arts.

All these different attempts of probing nature through experimental trials certainly contributed to the final emergence of experimentation in the seventeenth century as a self-conscious, methodically controlled and systematically used form of scientific experience. Galileo's new conception of motion, which was based on experiment and measurement from about 1604 on, played an instrumental and decisive role in this (Schmitt 1969). In the second half of the century, scientific academies devoted themselves to the new science and became the primary centers of experimental activity.

From the seventeenth century on, experimentation increasingly meant the implementation of new or improved scientific instruments. Following a suggestion of Thomas S. Kuhn, we can group these instruments mainly into two categories according to their origin in the classical or the Baconian tradition of physical science (Kuhn 1976). The classical sciences comprise those mathematical disciplines like astronomy, geometrical optics, statics, harmonics, and geometry itself, which were first constituted in classical antiquity and experienced their major developments already then. With the exception of harmonics, the close connections of these fields with each other lasted way up into the nineteenth century. The instruments belonging to this tradition were often called "mathematical instruments" and are of a restricted variety: ruler and compass, balance, clock, and geometrical-astronomical devices. They served as aids to "mixed mathematics," which allowed for certain physical attributes in addition to the abstract mathematical ones. To experiment with them mostly meant to confirm a belief that was established beforehand by rational considerations, or to detail a fully established theory in a special respect. Many experiments performed in this tradition proved to be in reality only thought experiments—mental constructions of possible experimental situations whose results were thought to be predictable already from everyday experience. Even Galileo participated sometimes in this attitude.

The second tradition to which we can attribute many of the new instruments of the period is the Baconian one whose disciplines owe their status as sciences mainly to the experimental movement of the seventeenth century and to the practice of "natural histories," including those of the different practical arts that experienced a tremendous re-evaluation at the time. The barrier between the craft and scholarly traditions, which had so far separated the mechanical from the liberal arts, began to break down. To the Baconian sciences belong the studies of heat, electricity, magnetism, chemistry, metallurgy, glass making, and the like. The instruments of these fields were used to investigate nature under previously unobserved or non-existent conditions and were often called "philosophical instruments." During the next decades, the Baconian movement brought forth the telescope, the microscope, the thermometer and the barometer, the air pump, electric charge detectors, the Leyden jar, and many other contrivances. It is interesting to see that these instruments were primarily used in a qualitative way and that a strictly quantitative application came only very late, mainly at the end of the eighteenth or during the early nineteenth century when the two traditions, the classical and the Baconian, started to merge with each other. From about the middle of the seventeenth century on, the Baconian movement had adopted some form of the atomic or corpuscular philosophy and became the official "experimental philosophy" of the Royal Society.

Philosophical Assessments of Experimentation and Instrumentation

In the second book of his Physics, Aristotle had developed a contrast between "physis" and "techne," that is, between natural entities that have an innate principle of change—like plants, animals and humans, but also stones and clouds—and those that are artificially constructed, like bedsteads and clothes. Until the scientific revolution, Aristotelians used this nature-artifact divide as an argument against the epistemological relevance of experimentation. In order to understand nature, they claimed, one must not intervene with her order. Intervention would either invalidate nature's innate principles or play her a trick with mechanical contrivances, but would not lead to any genuine knowledge of natural reality. Instead, one must let nature pursue her own course and purposes and gain knowledge of her principles by closely observing them. The fact that techne or art is declared by Aristotle to be able to complete nature's unfinished processes or to imitate her does not change this state of affairs. To complete nature in regard to the behavior of a natural entity meant to remove all obstacles that might have come in its way; and to imitate nature denoted the general maxim to bring form and matter of an entity in an intricate union as nature does it with her beings.

It seems that the major author in providing a philosophical bridge over the art-nature divide was Francis Bacon (1561–1626). This justifies Kuhn's choice of using Bacon's name for a whole new tradition of experimentation. Bacon argued that art was only a special way of arranging a state of affairs in which nature herself will then produce an intended result. He redefined Aristotle's concept of form and took it as the key to the operational features of a natural being, leaving out the teleological dimension. The discovery of operational rules of an entity can now be identified with the true form or real essence of relations among its simple natures. Consequently, Bacon rejected Aristotle's three other causes besides the formal one and took forms as "nothing more than those laws and determinations of absolute actuality which govern and constitute any simple nature, as heat, light, weight, in every kind of matter and subject that is susceptible of them" (Nov. Org. ii, XVII).

As a result, knowledge of our world cannot, according to Bacon, be read off from its surface, so to say. We can work our way through to the "viscera naturae," or nature's intestines, only by methodical and experimental procedures of induction. Perhaps Bacon's major insight was that simple enumerative induction, as taught by Aristotle, that is, induction without experiment and without the method of exclusion, is not enough to tell essential correlations from accidental ones.

Bacon's procedure of induction was taken as a valuable method of creating new empirical theories and laws way up into the twentieth century. The Baconian tradition culminated during the nineteenth century in John Stuart Mill's elaboration and refinement of Bacon's and Herschel's inductive rules. There is, however, a tendency visible in Mill to take experiment not quite with the same force as Bacon had taken it. For Bacon, experiment is inevitable if one wants to snatch secrets from nature—they never show up by themselves. Yet for Mill, situations are conceivable where observation can serve the same purpose as experiment: "For the purpose of varying the circumstances [in order to find out the real laws] we may have recourse … either to observation or to experiment; we may either find an instance in nature suited to our purposes, or, by an artificial arrangement of circumstances, make one. The value of the instance depends on what it is in itself, not on the mode in which it is obtained: its employment for the purposes of induction depends on the same principles in the one case and in the other, as the uses of money are the same whether it is inherited or acquired. There is, in short, no difference in kind, no real logical distinction, between the two processes of investigation" (System of Logic, III, vii, 2).

The belief that there is no "logical distinction" between observation and experiment became a matter of course for almost all the schools of philosophy of science of the entire twentieth century until the 1980s. It is interesting to see how an excellent nineteenth-century experimentalist, Hermann von Helmholtz, resisted this tendency, although he followed Mill in many other and important respects. His reasons, however, were different from Bacon's: If I can vary the conditions of an event in different respects, he argued, I can be sure that my intervention is the cause of observed change because I know of my will's impulse. If, however, I can only passively observe correlations without any help from me, I can never be sure whether these make up genuine causal relations or only accidental covariation (Helmholtz 1903). Whereas for Bacon it is the coyness of nature that compels humans to experiment, for Helmholtz it is the epistemological limitation of the passive mind that forces them to intervene in nature's course.

One of the strongest and most influential anti-inductive texts ever written is a chapter in Pierre Duhem's Aim and Structure of Physical Theory of 1906, titled "Physical Theory and Experiment." In order to show the general inadequacy of inductivism, Duhem picked the "Newtonian method" to pieces, as it appeared both in the hands of Newton himself as well as with Ampère's electrodynamics. He brilliantly showed that there is no question in Newton's celestial mechanics of any extraction of hypothesis by induction from experimenting, as Newton himself required in the General Scholium, nor in Ampère's mathematical theory of electrodynamic phenomena of any deduction "only from experiment," as stated already in the title of Ampère's treatise of 1827.

As a logical consequence, Duhem concluded that "in the course of its development, a physical theory is free to choose any path it pleases provided that it avoids any logical contradiction; in particular, it is free not to take account of experimental facts." It has to take account of them only "when the theory has reached its complete development" (Duhem 1974, p. 206; Duhem's emphasis). In order that experiment can unfold its true function—the testing of theories— it must be preceded by theory. Duhem intensified the priority of theory when he demanded that "this test by facts should bear exclusively on the conclusions of a theory, for only the latter are offered as an image of reality; the postulates serving as points of departure for the theory and the intermediary steps by which we go from the postulates to the conclusions do not have to be subject to this test."

Duhem's criticism was later taken up and continued by Karl Popper. In exactly the same spirit as Duhem, Popper decreed that "the theoretician puts certain definite questions to the experimenter, and the latter by his experiments tries to elicit a decisive answer to these questions and to no others" (Popper 1959, p. 107). For Popper therefore, it is only the theoretician who shows the experimenter the way, and never the other way around. The only function left for experiment is to liberate us from sterile and false theories. With Popper, experiment has altogether become the handmaiden of theory.

Duhem had even gone one step further than Popper in questioning the capability of experiment to fulfill this critical task of refuting theories as well. Even if a theory is mature enough to be tested, experiment cannot mechanically decide between it and its rival. "An experiment in physics can never condemn an isolated hypothesis but only a whole theoretical group" (p. 183). And it is hardly ever possible to decide trenchantly which of the many assumptions of a theoretical system is doubtful and responsible for the experimental contradiction. "The physicist concerned with remedying a limping theory resembles the doctor and not the watchmaker" (p. 188). A watchmaker, Duhem maintained, can take the broken watch apart and examine each component separately until he finds the defective one. The doctor, however, cannot dissect the patient to find out the problem, but has to guess its seat by inspecting disorders affecting the whole body. And even if all the assumptions of a theoretical group were known to be true except one, the rival group would not have been established as superior. This would be shown only if every possible alternative were conclusively eliminated. But we never know of course what alternatives remain to be discovered.

All these considerations led Duhem to explicitly condemn Bacon's idea of a "crucial experiment." Bacon had suggested that there do exist experiments that conclusively decide between competing theories. They do this in the way of instantiae cruces or "fingerposts" that are set up at crossroads to indicate the several directions. In 1951, W. V. O. Quine joined Duhem in rejecting crucial experiments. He generalized Duhem's argument to all of our empirical tenets. An unexpected unsuitable empirical observation does not only contradict a theoretical system, as Duhem had told us, Quine argued, but all our beliefs and theories: "Our statements about the external world face the tribunal of sense experience not individually but only as a corporate body. … The unit of empirical significance is the whole of science" (Quine 1961, p. 41f.). Quine used this claim for a searching critique of logical empiricism. One consequence of this is that any assumption apparently refuted by observation can be retained as true, so long as we are willing to make appropriate changes elsewhere in the system of our beliefs. This holistic argument for the underdetermination of theories by experience has become known as the "Duhem-Quine thesis."

The series of philosophical arguments to denigrate the role of experiments continued further into the 20th century. The logical empiricist Hans Reichenbach coined the influential distinction between "context of discovery" and "context of justification" which had been developed earlier by the philosophers Alois Riehl, Gottlob Frege and others under different names (Reichenbach 1951). According to this dichotomy, all the actual historical and social circumstances of the creation of a scientific theory, including its experimental generation, if there was one, cannot be used as reasons to justify it. Experiment can be good as a heuristic guide to hit upon a useful theory, but it is neither necessary nor sufficient for the validity of its results. As a result of Reichenbach's division all attention focused on the epistemology of theory and none on discovery and the possibilities of experiment.

Although Thomas S. Kuhn is routinely regarded as major critique of both logical empiricism with its forerunner Duhem and of Popper's critical rationalism, he was surprisingly enough in large agreement with his predecessors as far as the subordinate role of experiment is concerned—at least in his central work The Structure of Scientific Revolutions of 1962/1970. Unlike Reichenbach, however, Kuhn wanted to overcome the separation of discovery and justification, but the admissible discovery part of his logic considered the founding of theories again in overarching paradigms, but not in experiments. In this he followed his teacher Alexandre Koyré and others, who saw the success of modern science in the superiority of mathematically oriented Platonism over Aristotelianism with its "brute, common-sense experience" and over all other experimentally and technologically oriented historical endeavors. For Koyré as for Kuhn a scientific revolution is foremost an "intellectual mutation" (Koyré 1943, p. 400), i.e. a revolution of thought and not of momentous experimental innovation. Paradigms have priority over theories "in their conceptual, observational, and instrumental applications" (Kuhn 1970, p. 43). True experimental research is only possible, if questions to nature are posed in a suitable mathematical language. According to such a view, a history of experimentation could not only be a contingent epiphenomenon of the development of paradigms and would not have much explanatory value. (The contrary view is defended by deSolla Price 1984.) Only when in his later work he began to appreciate the Baconian sciences as an autonomous movement did Kuhn start to appreciate the possibility of a meaningful history of scientific experimentation (Kuhn 1976).

In retrospect, the discussion of experiment in philosophy of science from the late nineteenth century until the 1980s appears as a series of increasingly negative results: We know more and more what experiments don't accomplish and we understand better and better where earlier epistemic pretensions of experimentation find their limits. As a result, we can diagnose an "invisibility of experiment." In the same way as scientific revolutions of a field are, according to Kuhn, normally invisible to the scientific profession of the present, so experiments and their development remain largely invisible to philosophy of science because their exclusive role of testing theories seems ingrained in the ideology of its practitioners.

The New Experimentalism

Since the early 1980s, however, a change has taken place in the attitude of the study of science toward experiment. One can detect a growing awareness of the rich history of experimentation and of the vast variety of its (non-demonstrative) functions. This swing of appreciation is primarily due to detailed work of historians and sociologists of science. It is true that historiography never ceased to deal with experiment, but it had rarely put it into the center of its interest. Socio-historical analysis has now come to concentrate much more on the microstructure of experiment than before and has started to consider all kinds of other sources besides official reports, like diaries and laboratory notebooks. Especially rich sources are Faraday's laboratory notebooks and letters, Ampère's "dossier" in the archive of the Académie des Sciences and Hans Krebs' laboratory diaries and interview protocols (Gooding 1990, Steinle 2005, Holmes 1993, Graßhoff 2000). Historians even went so far as to replicate historical experiments with rebuilt apparatus and to hereby bring to light neglected or otherwise hidden dimensions of experimentation (Heering 2000). Sociologists tried to show that the formulation of experimental results requires special structures of communication in the scientific community and that there is a good deal of negotiation involved until an experimental result is considered as achieved (Shapin and Shaffer 1985, Licoppe 1996; for a discussion see Holmes 1992). The variety of fields from where these case studies come from raise hopes that the traditional concentration on physics in relation to experiment will soon be done with once and for all.

It was Ian Hacking's Representing and Intervening that set the ball rolling in philosophy of science. There are two phrases from Hacking's book that became the slogans of "new experimentalism": "If you can spray them, then they are real" and "Experimentation has a life of its own" (Hacking 1983, pp. 23, 150). The first catchphrase stands for a novel argument in favor of scientific realism. The philosopher's favorite theoretical entity is the electron—never given directly to our senses, but central to modern particle physics. There is an endless debate between scientific realists and their opponents whether explanatory success of a theory is ground for belief in the reality of its theoretically postulated entities. Hacking does not think very highly of this "inference to the best explanation," on which the ordinary scientific realist bases her belief in the reality of the electron. He rather sets high hopes in the fact that if you spray, say, a niobium ball with electrons, it makes a difference in the world: it decreases the charge of the niobium ball. "From that day forth," Hacking confesses, "I've been a scientific realist." In a way, Hacking's argument is a version, adapted to scientific antirealism, of Dr. Johnson's refutation of Bishop Berkeley's metaphysical antirealism concerning matter by kicking a stone. "It is not thinking about the world but changing it that in the end must make us scientific realists."

With the second catchphrase Hacking opposes the alleged theory-domination of experimentation: There actually exists experimental practice, he argues, that is not subordinate to theory and this practice actually proves to be very important. This claim is backed up with many intriguing examples. But liberating experiment from permanent condemnation to the role of theory's handmaiden does not automatically show what other roles it can take on and what the principles of their variations are. About this, Hacking does not say very much. The only other role he addresses in detail is, as he says, the experiment's "chief role": the "creation of phenomena." Some aspects of this role have been brought to light in Steinle's concept of "exploratory" experiments or in Heidelberger's notion of "productive" instruments (Steinle 2005; Heidelberger 1998, 2003).

All in all, Hacking seems to be largely content with a "Baconian fluster of examples of many different relationships between experiment and theory" (Hacking 1983, p. 66). This has surely proven to have been enough to initiate a "Back-to-Bacon movement, in which we attend more seriously to experimental science" (p. 150) as it had been Hacking's intention. But, if neo-Baconianism is sound, it is not enough as an explanation of what happens or should happen with other theoretical commitments of general philosophy of science, like, for example, the theory-ladenness of observation. This doctrine—dear to many philosophers of science for other reasons—comes, at least prima facie, into conflict with Hacking's faith in the priority of experiment.

In the wake of renewed interest in experiment, several substantial studies and edited volumes have appeared. Many of them are divided over the philosophical issue whether experiment can decide between competing theories and thus have an objective meaning or whether social and political factors are in the end responsible for scientific development. There is, for example, Pickering's sociological history of particle physics or Collins's study of gravity wave detection maintaining the social construction of scientific evidence whereas Franklin and Mayo argue for the existence of strategies that secure reliable experimental outcomes and thus of rational belief. It would be wrong, however, to perpetuate the polarization between history, sociology, and philosophy of science. One of the results of taking experiment more seriously is precisely the insight that these dichotomies have to be transcended. An attempt into this direction has been made by Rheinberger who takes "experimental systems" as functional research units, especially of the life sciences (see Hagner and Rheinberger 1998 for a programmatic overview.) They are made up of research objects, theories, experimental arrangements, instruments, as well as disciplinary, social, cultural and institutional constellations that for some time crystallize in a certain stable configuration.

Experimentation and Theory-Ladenness

The idea of theory-ladenness of experience enabled a powerful and effective criticism of logical empiricism. This is the view already encountered with Popper that there are no theory-neutral data and that the meaning of observational terms fundamentally depends upon the theoretical context in which they occur. This view can easily be strengthened to serve as the cornerstone of a constructivist and anti-empiricist account of science: The categories in terms of which we carve up our experience are not read off from the external world but follow from prior theoretical or other commitments of its observers, either individually or socially.

The implications of theory-ladenness for a view of scientific experimentation are straightforward: If observations are theory-laden and if experimentation involves observation of results, then experimentation has to be theory-laden too. Since experiments, according to this view, make sense only in relation to some theoretical background, they cannot play a role that is independent from theory.

Now, the question arises: If new experimentalism is right, do we have to give up the idea of theory-ladenness? It is difficult to imagine a straightforward "yes" as an answer, because the general spirit in which the idea of theory-ladenness has been formulated is largely the same as that of the idea that experimentation has a life of its own. It is the spirit addressed by Hacking at the beginning of his book in which philosophers finally realized that they "long made a mummy of science"—the same spirit which, in the face of history and the reality of the laboratory, denies the "Popper/Carnap common ground." To deny theory-ladenness would to some extent feel like a return to logical empiricism and thereby of mummification, even if the autonomy of experiment is the reward.

Before some kind of dénouement of this question is formulated, let us have a closer look at theory-ladenness as it appeared in the work of its most important originators. One of the first propagators of this outlook was Pierre Duhem who wrote: "An experiment in physics is the precise observation of phenomena accompanied by an interpretation of these phenomena; this interpretation substitutes for the concrete data really gathered by observation abstract and symbolic representations which correspond to them by virtue of the theories admitted by the observer. … The result of an experiment in physics is an abstract and symbolic judgment" (Duhem 1974, p. 147). It would not be enough for an experimental report to state, as a layman would express it, that a piece of iron carrying a mirror oscillates. Instead it should read that the electrical resistance of a coil is measured. This shows that the physicist draws conclusions from experiment only in abstract and symbolic terms "to which you can attach no meaning if you do not know the physical theories admitted by the author." In sciences less advanced than physics like physiology or certain branches of chemistry "where mathematical theory has not yet introduced its symbolic representations" and where causal explanation reigns instead of a causally neutral description, the experimenter can reason "directly on the facts by a method which is only common sense brought to greater attentiveness" (p. 180).

This kind of theory-ladenness by theoretical interpretation, as we can call it, is very often confounded with another sort which was provided by Norwood Russell Hanson in 1958 and which can be called "theory-ladenness by prior belief or knowledge." "Seeing an object x," Hanson wrote, "is to see that it may behave in the ways we know x 's do behave" (Hanson 1958, p. 22). As a result of this, Tycho and Kepler watching the sun at dawn would literally see different things: Tycho who believes in the geocentric theory sees the sun beginning its diurnal circuit, whereas Kepler as defender of heliocentrism sees the earth spinning back into the light of the sun. "Analogously," Hanson wrote, "the physicist sees an X-ray tube, not by first soaking up reflected light and then clamping on interpretations, but just as you see this page before you."

In addition, theory-ladenness in science means "causality-ladenness" for Hanson, being loaded with causal meaning. He does not exclude theory-neutral talk after all, but it only happens in the oculist's office or like circumstances but not in scientific observation or experimentation. This shows that Hanson rejects all of Duhem's points: (1) Seeing an experimental result is not interpreting it; (2) both the layman and the physicist have prior beliefs and therefore both their seeing is theory-laden; and (3) physical theory (as well as common beliefs about the world) is causal theory and not just causally neutral description. Whereas for Hanson any injection of causality into the mere registering of facts is bound to render them theoretical, for Duhem, theory begins with the representation of (causal) relations in an abstract, causally neutral structure.

In Thomas Kuhn's work we find several different conceptions of theory-ladenness that are not always separated clearly. The most frequently used is similar to Hanson's, except that it is not prior knowledge that shapes perception, but paradigm and that it stresses and utilizes the psychology of perception even more than in Hanson: "Something like a paradigm is prerequisite to perception itself. What a man sees depends both upon what he looks at and also upon what his previous visual-conceptual experience has taught him to see" (Kuhn 1970, p. 113).

In order to exhibit his other uses of theory-ladenness, let us have a look at Kuhn's treatment of scientific discovery. Kuhn admits the possibility of "fundamental novelties of fact," that go against a well-established paradigm. Without this possibility, as he himself realizes, science could only develop in a theoretical manner and never by adjustment to facts. "Discovery commences with the awareness of anomaly, i.e., with the recognition that nature has somehow violated the paradigm-induced expectations that govern normal science" (Kuhn 1970, pp. 52–53).

Where, according to Kuhn, does a violation of the paradigm-induced expectations come from? Does it come from a causal process that violates the received view or from a new theoretical interpretation that makes old facts appear in a new light? It seems that in Kuhn, it is almost always the theoretical interpretation, the assimilation to theory, that is decisive for discovery and hardly ever any causal experience. "Assimilating a new sort of fact demands a more than additive adjustment of theory, and until that adjustment is completed—until the scientist has learned to see nature in a different way—the new fact is not quite a new fact at all." That sounds more as if new facts and causal processes were created by new paradigms than the other way around. Lavoisier, we are told, for example, was enabled through his new paradigm "to see in experiments like Priestley's a gas that Priestley had been unable to see there himself" and was "to the end of his life" unable to see p. 56).

The only case where Kuhn explicitly admits that discovery has been effected by a genuinely novel causal experience appears to be the case of the X-rays. "Its story opens on the day that the physicist Roentgen interrupted a normal investigation of cathode rays because he had noticed that a barium platino-cyanide screen at some distance from his shielded apparatus glowed when the discharge was in process" (p. 57). Although Kuhn seems to consider this observation theory-laden, I maintain that, in Duhem's sense, it is not. If it were, Roentgen, by definition of theory-ladenness, would have been able to interpret it in light of the theories of physics he had at his disposal. But here it is exactly the point that his theories deserted him and he could not find a place for this new experience in his customary theoretical structure. For this reason he interrupted his investigation and asked himself why the screen had come to glow. Yet the novel observation is certainly theory-laden in the sense of Hanson, because Roentgen immediately looked for a causal relationship between his apparatus and the glowing of the screen, although this went completely against all his expectations!

Kuhn seems to say that Roentgen would never have paid attention to the glowing screen if he had not disposed of deeply entrenched theories of physics that prohibited such a phenomenon. If this is true then we have here a third sense of the notion of theory-ladenness before us. It frames a psychological hypothesis about the ease with which a phenomenon is detected or paid attention to in the light of a contradicting paradigm: An observation is theory-laden in this sense if it were improbable that an observer would have made it (that an observer would have noticed it or would have attributed any importance to it) without her holding a theory beforehand that created expectations to the contrary. It would be better to drop the term "theory-ladenness" for this case altogether and instead call it "theory-guidance" because the experimental result made sense to Roentgen as an observation in its simple causal structure already without the theoretical background of the theory that guided it or any other one. "Theory-guidance" refers to a psychological disposition how well one is prepared to notice a particular phenomenon in certain situations.

After Roentgen had noticed the anomaly, he conducted various experiments in order to explore the cause of the incident: "Further investigations—they required seven hectic weeks during which Roentgen rarely left the laboratory—indicated that the cause of the glow came in straight lines from the cathode ray tube, that the radiation cast shadows, could not be deflected by a magnet, and much else besides. Before announcing his discovery, Roentgen had convinced himself that his effect was not due to cathode rays but to an agent with at least some similarity to light" (Kuhn 1970, p. 57). This is perhaps the only place in his book where Kuhn uses the term "cause" (or an equivalent) in relation to an experimental investigation. The quotation shows vividly that Roentgen did not conduct his experiments in order to test a theory but to expand our knowledge of causal connections in relation to the scientific instruments and devices involved.

What does our discussion suggest therefore as the most adequate description of Roentgen's early experiments? They were certainly theory-guided in the sense of Kuhn and they were, or immediately became, causality-laden in the sense of Hanson, but not (or not yet) theory-laden in the sense of Duhem (which Kuhn also shares). Kuhn is right when he suggests that only after the phenomena had received an abstract and symbolic representation can we speak of a "discovery" of X-rays. Yet before this interpretation has taken place, we can say that an anomaly has occurred and that it can be replicated in certain ways; not more, but also not less.

If the case of the X-rays is in this way correctly understood, then Kuhn can give in to Hacking without loosing anything essential and admit that experimentation can be, and very often is, autonomous and free from theory. The lesson to learn is to distinguish between two kinds of experiments: those that are causal, but not (yet) embedded in a theoretical structure and those that presuppose the knowledge of such a framework. This emphasis of an autonomous "lower level" in experimentation is not a relapse into positivist observation statements and protocol sentences allegedly giving meaning to theory. The claim rather is that two types of experimentation should conceptually be kept apart: experimentation at the causal level, where the manipulation of instruments and objects under scrutiny takes place, and experimentation taking place at the theoretical level, where the results at the causal level are represented in a theoretical superstructure.

See also Ampère, André Marie; Aristotelianism; Aristotle; Bacon, Francis; Berkeley, George; Carnap, Rudolf; Duhem, Pierre Maurice Marie; Faraday, Michael; Frege, Gottlob; Galen; Galileo Galilei; Harvey, William; Helmholtz, Hermann Ludwig von; Herschel, John; Johnson, Samuel; Kepler, Johannes; Kuhn, Thomas; Lavoisier, Antoine; Logical Positivism; Mill, John Stuart; Newton, Isaac; Philosophy of Science, History of; Platonism and the Platonic Tradition; Popper, Karl Raimund; Priestley, Joseph; Quine, Willard Van Orman; Realism; Reichenbach, Hans; Riehl, Alois; Scientific Method; Thought Experiments in Science; Underdetermination Thesis, Duhem-Quine Thesis.

Bibliography

Bacon, Francis. Novum Organum. London: 1620.

Batens, Dederik, and Jean P. van Bendegem, eds. Theory and Experiment: Recent Insights and New Perspectives on Their Relation. Dordrecht: Reidel, 1988.

Buchwald, Jed Z. Scientific Practice: Theories and Stories of Doing Physics. Chicago: University of Chicago Press, 1995.

Collins, Harry M. Changing Order: Replication and Induction in Scientific Practice. Beverly Hills, CA, and London: Sage, 1985.

Crombie, Alistair C. Robert Grosseteste and the Origins of Experimental Science, 1100–1700. Oxford: Clarendon, 1953.

deSolla Price, Derek J. "Of Sealing Wax and String." Natural History 93 (1) (1984): 48–57.

Duhem, Pierre. The Aim and Structure of Physical Theory. New York: Atheneum, 1974. Translated by Philip P. Wiener from the 2nd ed. Paris: Chevalier & Rivière, 1914.

Franklin, Allan. Experiment, Right or Wrong. Cambridge, U.K.: Cambridge University Press, 1990.

Franklin, Allan. The Neglect of Experiment. Cambridge, U.K.: Cambridge University Press, 1986.

Galavotti, Maria Carla, ed. Observation and Experiment in the Natural and Social Sciences. Dordrecht: Kluwer, 2003.

Galison, Peter. How Experiments End. Chicago: University of Chicago Press, 1987.

Gooding, David. Experiment and the Making of Meaning: Human Agency in Scientific Observation. Dordrecht: Kluwer, 1990.

Gooding, David, Trevor Pinch, and Simon Schaffer, eds. The Uses of Experiment: Studies in the Natural Sciences. Cambridge, U.K.: Cambridge University Press, 1989.

Graßhoff, Gerd, R. Casties, and Kärin Nickelsen. Zur Theorie des Experimentes. Untersuchungen am Beispiel der Entdeckung des Harnstoffzyklus. Bern: Bern Studies, 2000.

Hacking, Ian. Representing and Intervening: Introductory Topics in the Philosophy of Natural Science. Cambridge, U.K.: Cambridge University Press, 1983.

Hacking, Ian. "The Self-Vindication of the Laboratory Sciences." In Science as Practice and Culture, edited by Andrew Pickering, 29–64. Chicago: University of Chicago Press, 1992.

Hagner, Michael, and Hans-Jörg Rheinberger. "Experimental Systems, Objects of Investigation, and Spaces of Representation." In Experimental Essay—Versuche zum Experiment, edited by Michael Heidelberger and Friedrich Steinle, 355–373. Baden-Baden: Nomos, 1998.

Hanson, Norwood Russell. Patterns of Discovery. Cambridge, U.K.: Cambridge University Press, 1958.

Heering, P., Falk Rieß, and C. Sichau, eds. Im Labor der Physikgeschichte. Zur Untersuchung historischer Experimentalpraxis. Oldenburg: BIS-Verlag, 2000.

Heidelberger, Michael. "Die Erweiterung der Wirklichkeit im Experiment." In Experimental Essays—Versuche zum Experiment, edited by Michael Heidelberger and Friedrich Steinle, 71–92. Baden-Baden: Nomos, 1998.

Heidelberger, Michael. "Theory-Ladenness and Scientific Instruments in Experimentation." In The Philosophy of Scientific Experimentation, edited by Hans Radder, 138–151. Pittsburgh: University of Pittsburgh Press, 2003.

Heidelberger, Michael, and Friedrich Steinle, eds. Experimental Essays—Versuche zum Experiment. Baden-Baden: Nomos, 1998.

Helmholtz, Hermann von. "Die Tatsachen in der Wahrnehmung [1878]." In Vorträge und Reden, 213–247. Braunschweig: Vieweg, 1903.

Herschel, John F. W. Preliminary Discourse on the Study of Natural Philosophy (1830). New York: Johnson Reprint Corporation, 1966.

Holmes, Frederic L. "Do We Understand Historically How Experimental Knowledge Is Acquired?" History of Science 30 (1992): 119–136.

Holmes, Frederic L. Hans Krebs: Architect of Intermediary Metabolism 1933–1937. Oxford: Oxford University Press, 1993.

Holmes, Frederic L. "The Logic of Discovery in the Experimental Life Sciences." In Biology and Epistemology, edited by Jane Maienschein, 167–190. Cambridge, U.K.: Cambridge University Press, 2000.

Holmes, Frederic L., Jürgen Renn, and Hans-Jörg Rheinberger, eds. Reworking the Bench: Research Notebooks in the History of Science, New Studies in the History and Philosophy of Science and Technology. Dordrecht: Kluwer, 2003.

Holmes, Frederic L., and Trevor H. Levere, eds. Instruments and Experimentation in the History of Chemistry. Cambridge, MA: MIT Press, 2000.

Koyré, Alexandre. "Galileo and Plato." Journal of the History of Ideas 4 (4) (1943): 400–428.

Kuhn, Thomas S. "Mathematical vs. Experimental Traditions in the Development of Physical Science." Journal of Interdisciplinary History 7 (1) (1976): 1–31. Reprinted in The Essential Tension: Selected Studies in Scientific Thought and Change, 31–65. Chicago: University of Chicago Press, 1977.

Kuhn, Thomas S. The Structure of Scientific Revolutions. 2nd ed. Chicago: University of Chicago Press, 1970.

Latour, Bruno, and Steve Woolgar. Laboratory Life: The Social Construction of Scientific Facts. 2nd ed. Princeton, NJ: Princeton University Press, 1986.

Le Grand, Homer E., ed. Experimental Inquiries: Historical, Philosophical and Social Studies of Experimentation in Science. Dordrecht: Kluwer, 1990.

Licoppe, Christian. La formation de la pratique scientifique: le discours de l'expérience en France et en Angleterre (1630–1820). Paris: Editions La Decouverte, 1996.

Mayo, Deborah G. Error and the Growth of Experimental Knowledge, Science and Its Conceptual Foundations. Chicago: University of Chicago Press, 1996.

Mill, John Stuart. A System of Logic, Ratiocinative and Inductive. London: Parker, 1843.

Pickering, Andrew. Constructing Quarks: A Sociological History of Particle Physics. Chicago: University of Chicago Press, 1984.

Pickering, Andrew. The Mangle of Practice: Time, Agency, and Science. Chicago: University of Chicago Press, 1995.

Pickering, Andrew, ed. Science as Practice and Culture. Chicago: University of Chicago Press, 1992.

Popper, Karl. The Logic of Scientific Discovery. London: Hutchinson, 1959.

Quine, Willard Van Orman. From a Logical Point of View. New York: Harper & Row, 1963.

Radder, Hans, ed. The Philosophy of Scientific Experimentation. Pittsburgh, PA: University of Pittsburgh Press, 2003.

Reichenbach, Hans. The Rise of Scientific Philosophy. Berkeley: University of California Press, 1951.

Schmitt, Charles B. "Experience and Experiment: Comparison of Zabarella's View with Galileo's in De Motu." Studies in the Renaissance 16 (1969): 80–138.

Shapin, Steven, and Simon Schaffer. Leviathan and the Air-Pump: Hobbes, Boyle, and the Experimental Life. Princeton, NJ: Princeton University Press, 1985.

Steinle, Friedrich. Explorative Experimente. Ampère, Faraday und die Ursprünge der Elektrodynamik. Stuttgart: Franz Steiner, 2005.

Michael Heidelberger (2005)