experimental observation theory

• Table of Contents
• New in this Archive
• Chronological
• Editorial Information
• About the SEP
• Editorial Board
• How to Cite the SEP
• Special Characters
• Support the SEP
• PDFs for SEP Friends
• Make a Donation
• SEPIA for Libraries
• Entry Contents

Bibliography

Academic tools.

• Friends PDF Preview
• Author and Citation Info
• Back to Top

Theory and Observation in Science

Scientists obtain a great deal of the evidence they use by observing natural and experimentally generated objects and effects. Much of the standard philosophical literature on this subject comes from 20 th century logical empiricists, their followers, and critics who embraced their issues and accepted some of their assumptions even as they objected to specific views. Their discussions of observational evidence tend to focus on epistemological questions about its role in theory testing. This entry follows their lead even though observational evidence also plays important and philosophically interesting roles in other areas including scientific discovery, the development of experimental tools and techniques, and the application of scientific theories to practical problems.

The issues that get the most attention in the standard philosophical literature on observation and theory have to do with the distinction between observables and unobservables, the form and content of observation reports, and the epistemic bearing of observational evidence on theories it is used to evaluate. This entry discusses these topics under the following headings:

1. Introduction

2. what do observation reports describe, 3. is observation an exclusively perceptual process, 4. how observational evidence might be theory laden, 5. salience and theoretical stance, 6. semantic theory loading, 7. operationalization and observation reports, 8. is perception theory laden, 9. how do observational data bear on the acceptability of theoretical claims, 10. data and phenomena, 11. conclusion, other internet resources, related entries.

Reasoning from observations has been important to scientific practice at least since the time of Aristotle who mentions a number of sources of observational evidence including animal dissection (Aristotle(a) 763a/30–b/15, Aristotle(b) 511b/20–25). But philosophers didn’t talk about observation as extensively, in as much detail, or in the way we have become accustomed to, until the 20 th century when logical empiricists transformed philosophical thinking about it.

The first transformation was accomplished by ignoring the implications of a long standing distinction between observing and experimenting. To experiment is to isolate, prepare, and manipulate things in hopes of producing epistemically useful evidence. It had been customary to think of observing as noticing and attending to interesting details of things perceived under more or less natural conditions, or by extension, things perceived during the course of an experiment. To look at a berry on a vine and attend to its color and shape would be to observe it. To extract its juice and apply reagents to test for the presence of copper compounds would be to perform an experiment. Contrivance and manipulation influence epistemically significant features of observable experimental results to such an extent that epistemologists ignore them at their peril. Robert Boyle (1661), John Herschell (1830), Bruno Latour and Steve Woolgar (1979), Ian Hacking (1983), Harry Collins (1985) Allan Franklin (1986), Peter Galison (1987), Jim Bogen and Jim Woodward (1988), and Hans-Jörg Rheinberger(1997), are some of the philosophers and philosophically minded scientists, historians, and sociologists of science who gave serious consideration to the distinction between observing and experimenting. The logical empiricists tended to ignore it.

A second transformation, characteristic of the linguistic turn in philosophy, was to shift attention away from things observed in natural or experimental settings and concentrate instead on the logic of observation reports. The shift developed from the assumption that a scientific theory is a system of sentences or sentence like structures (propositions, statements, claims, and so on) to be tested by comparison to observational evidence. Secondly it was assumed that the comparisons must be understood in terms of inferential relations. If inferential relations hold only between sentence like structures, it follows that theories must be tested, not against observations or things observed, but against sentences, propositions, etc. used to report observations. (Hempel 1935, 50–51. Schlick 1935)

Friends of this line of thought theorized about the syntax, semantics, and pragmatics of observation sentences, and inferential connections between observation and theoretical sentences. In doing so they hoped to articulate and explain the authoritativeness widely conceded to the best natural, social and behavioral scientific theories. Some pronouncements from astrologers, medical quacks, and other pseudo scientists gain wide acceptance, as do those of religious leaders who rest their cases on faith or personal revelation, and rulers and governmental officials who use their political power to secure assent. But such claims do not enjoy the kind of credibility that scientific theories can attain. The logical empiricists tried to account for this by appeal to the objectivity and accessibility of observation reports, and the logic of theory testing.

Part of what they meant by calling observational evidence objective was that cultural and ethnic factors have no bearing on what can validly be inferred about the merits of a theory from observation reports. So conceived, objectivity was important to the logical empiricists’ criticism of the Nazi idea that Jews and Aryans have fundamentally different thought processes such that physical theories suitable for Einstein and his kind should not be inflicted on German students. In response to this rationale for ethnic and cultural purging of the German educational system the logical empiricists argued that because of its objectivity, observational evidence, rather than ethnic and cultual factors should be used to evaluate scientific theories.(Galison 1990). Less dramatically, the efforts working scientists put into producing objective evidence attest to the importance they attach to objectivity. Furthermore it is possible, in principle at least, to make observation reports and the reasoning used to draw conclusions from them available for public scrutiny. If observational evidence is objective in this sense , it can provide people with what they need to decide for themselves which theories to accept without having to rely unquestioningly on authorities.

Francis Bacon argued long ago that the best way to discover things about nature is to use experiences (his term for observations as well as experimental results) to develop and improve scientific theories (Bacon1620 49ff). The role of observational evidence in scientific discovery was an important topic for Whewell (1858) and Mill (1872) among others in the 19 th century. Recently, Judaea Pearl, Clark Glymour, and their students and associates addressed it rigorously in the course of developing techniques for inferring claims about causal structures from statistical features of the data they give rise to (Pearl, 2000; Spirtes, Glymour, and Scheines 2000). But such work is exceptional. For the most part, philosophers followed Karl Popper who maintained, contrary to the title of one of his best known books, that there is no such thing as a ‘logic of discovery’.(Popper 1959, 31) Drawing a sharp distinction between discovery and justification, the standard philosophical literature devotes most of its attention to the latter.

Theories are customarily represented as collections of sentences, propositions, statements or beliefs, etc., and their logical consequences. Among these are maximally general explanatory and predictive laws (Coulomb’s law of electrical attraction and repulsion, and Maxwellian electromagnetism equations for example), along with lesser generalizations that describe more limited natural and experimental phenomena (e.g., the ideal gas equations describing relations between temperatures and pressures of enclosed gasses, and general descriptions of positional astronomical regularities). Observations are used in testing generalizations of both kinds.

Some philosophers prefer to represent theories as collections of ‘states of physical or phenomenal systems’ and laws. The laws for any given theory are

…relations over states which determine…possible behaviors of phenomenal systems within the theory’s scope. (Suppe 1977, 710)

So conceived, a theory can be adequately represented by more than one linguistic formulation because it is not a system of sentences or propositions. Instead, it is a non-linguistic structure which can function as a semantic model of its sentential or propositional representations. (Suppe 1977, 221–230) This entry treats theories as collections of sentences or sentential structures with or without deductive closure. But the questions it takes up arise in pretty much the same way when theories are represented in accordance with this semantic account.

One answer to this question assumes that observation is a perceptual process so that to observe is to look at, listen to, touch, taste, or smell something, attending to details of the resulting perceptual experience. Observers may have the good fortune to obtain useful perceptual evidence simply by noticing what’s going on around them, but in many cases they must arrange and manipulate things to produce informative perceptible results. In either case, observation sentences describe perceptions or things perceived.

Observers use magnifying glasses, microscopes, or telescopes to see things that are too small or far away to be seen, or seen clearly enough, without them. Similarly, amplification devices are used to hear faint sounds. But if to observe something is to perceive it, not every use of instruments to augment the senses qualifies as observational. Philosophers agree that you can observe the moons of Jupiter with a telescope, or a heart beat with a stethoscope. But minimalist empiricists like Bas Van Fraassen (1980, 16–17) deny that one can observe things that can be visualized only by using electron (and perhaps even) light microscopes. Many philosophers don’t mind microscopes but find it unnatural to say that high energy physicists observe particles or particle interactions when they look at bubble chamber photographs. Their intuitions come from the plausible assumption that one can observe only what one can see by looking, hear by listening, feel by touching, and so on. Investigators can neither look at (direct their gazes toward and attend to) nor visually experience charged particles moving through a bubble chamber. Instead they can look at and see tracks in the chamber, or in bubble chamber photographs.

The identification of observation and perceptual experience persisted well into the 20 th century—so much so that Carl Hempel could characterize the scientific enterprise as an attempt to predict and explain the deliverances of the senses (Hempel 1952, 653). This was to be accomplished by using laws or lawlike generalizations along with descriptions of initial conditions, correspondence rules, and auxiliary hypotheses to derive observation sentences describing the sensory deliverances of interest. Theory testing was treated as a matter of comparing observation sentences describing observations made in natural or laboratory settings to observation sentences that should be true according to the theory to be tested. This makes it imperative to ask what observation sentences report. Even though scientists often record their evidence non-sententially, e.g., in the form of pictures, graphs, and tables of numbers, some of what Hempel says about the meanings of observation sentences applies to non-sentential observational records as well.

According to what Hempel called the phenomenalist account, observation reports describe the observer’s subjective perceptual experiences.

…Such experiential data might be conceived of as being sensations, perceptions, and similar phenomena of immediate experience. (Hempel 1952, 674)

This view is motivated by the assumption that the epistemic value of an observation report depends upon its truth or accuracy, and that with regard to perception, the only thing observers can know with certainty to be true or accurate is how things appear to them. This means that we can’t be confident that observation reports are true or accurate if they describe anything beyond the observer’s own perceptual experience. Presumably one’s confidence in a conclusion should not exceed one’s confidence in one’s best reasons to believe it. For the phenomenalist it follows that reports of subjective experience can provide better reasons to believe claims they support than reports of other kinds of evidence. Furthermore, if C.I. Lewis had been right to think that probabilities cannot be established on the basis of dubitable evidence, (Lewis 1950, 182) observation sentences would have no evidential value unless they report the observer’s subjective experiences. [ 1 ]

But given the expressive limitations of the language available for reporting subjective experiences we can’t expect phenomenalistic reports to be precise and unambiguous enough to test theoretical claims whose evaluation requires accurate, fine- grained perceptual discriminations. Worse yet, if experiences are directly available only to those who have them, there is room to doubt whether different people can understand the same observation sentence in the same way. Suppose you had to evaluate a claim on the basis of someone else’s subjective report of how a litmus solution looked to her when she dripped a liquid of unknown acidity into it. How could you decide whether her visual experience was the same as the one you would use her words to report?

Such considerations led Hempel to propose, contrary to the phenomenalists, that observation sentences report ‘directly observable’, ‘intersubjectively ascertainable’ facts about physical objects

…such as the coincidence of the pointer of an instrument with a numbered mark on a dial; a change of color in a test substance or in the skin of a patient; the clicking of an amplifier connected with a Geiger counter; etc. (ibid.)

Observers do sometmes have trouble making fine pointer position and color discriminations but such things are more susceptible to precise, intersubjectively understandable descriptions than subjective experiences. How much precision and what degree of intersubjective agreement are required in any given case depends on what is being tested and how the observation sentence is used to evaluate it. But all things being equal, we can’t expect data whose acceptability depends upon delicate subjective discriminations to be as reliable as data whose acceptability depends upon facts that can be ascertained intersubjectively. And similarly for non-sentential records; a drawing of what the observer takes to be the position of a pointer can be more reliable and easier to assess than a drawing that purports to capture her subjective visual experience of the pointer.

The fact that science is seldom a solitary pursuit suggests that one might be able to use pragmatic considerations to finesse questions about what observation reports express. Scientific claims—especially those with practical and policy applications—are typically used for purposes that are best served by public evaluation. Furthermore the development and application of a scientific theory typically requires collaboration and in many cases is promoted by competition. This, together with the fact that investigators must agree to accept putative evidence before they use it to test a theoretical claim, imposes a pragmatic condition on observation reports: an observation report must be such that investigators can reach agreement relatively quickly and relatively easily about whether it provides good evidence with which to test a theory (Cf. Neurath 1913). Feyerabend took this requirement seriously enough to characterize observation sentences pragmatically in terms of widespread decidability. In order to be an observation sentence, he said, a sentence must be contingently true or false, and such that competent speakers of the relevant language can quickly and unanimously decide whether to accept or reject it on the basis what happens when they look, listen, etc. in the appropriate way under the appropriate observation conditions (Feyerabend 1959, 18ff).

The requirement of quick, easy decidability and general agreement favors Hempel’s account of what observation sentences report over the phenomenalist’s. But one shouldn’t rely on data whose only virtue is widespread acceptance. Presumably the data must possess additional features by virtue of which it can serve as an epistemically trustworthy guide to a theory’s acceptability. If epistemic trustworthiness requires certainty, this requirement favors the phenomenalists. Even if trustworthiness doesn’t require certainty, it is not the same thing as quick and easy decidability. Philosophers need to address the question of how these two requirements can be mutually satisfied.

Many of the things scientists investigate do not interact with human perceptual systems as required to produce perceptual experiences of them. The methods investigators use to study such things argue against the idea—however plausible it may once have seemed—that scientists do or should rely exclusively on their perceptual systems to obtain the evidence they need. Thus Feyerabend proposed as a thought experiment that if measuring equipment was rigged up to register the magnitude of a quantity of interest, a theory could be tested just as well against its outputs as against records of human perceptions (Feyerabend 1969, 132–137).

Feyerabend could have made his point with historical examples instead of thought experiments. A century earlier Helmholtz estimated the speed of excitatory impulses traveling through a motor nerve. To initiate impulses whose speed could be estimated, he implanted an electrode into one end of a nerve fiber and ran a current into it from a coil. The other end was attached to a bit of muscle whose contraction signaled the arrival of the impulse. To find out how long it took the impulse to reach the muscle he had to know when the stimulating current reached the nerve. But

[o]ur senses are not capable of directly perceiving an individual moment of time with such small duration…

and so Helmholtz had to resort to what he called ‘artificial methods of observation’ (Olesko and Holmes 1994, 84). This meant arranging things so that current from the coil could deflect a galvanometer needle. Assuming that the magnitude of the deflection is proportional to the duration of current passing from the coil, Helmholtz could use the deflection to estimate the duration he could not see ( ibid ). This ‘artificial observation’ is not to be confused e.g., with using magnifying glasses or telescopes to see tiny or distant objects. Such devices enable the observer to scrutinize visible objects. The miniscule duration of the current flow is not a visible object. Helmholtz studied it by looking at and seeing something else. (Hooke (1705, 16–17) argued for and designed instruments to execute the same kind of strategy in the 17 th century.) The moral of Feyerabend’s thought experiment and Helmholtz’s distinction between perception and artificial observation is that working scientists are happy to call things that register on their experimental equipment observables even if they don’t or can’t register on their senses.

Some evidence is produced by processes so convoluted that it’s hard to decide what, if anything has been observed. Consider functional magnetic resonance images (fMRI) of the brain decorated with colors to indicate magnitudes of electrical activity in different regions during the performance of a cognitive task. To produce these images, brief magnetic pulses are applied to the subject’s brain. The magnetic force coordinates the precessions of protons in hemoglobin and other bodily stuffs to make them emit radio signals strong enough for the equipment to respond to. When the magnetic force is relaxed, the signals from protons in highly oxygenated hemoglobin deteriorate at a detectably different rate than signals from blood that carries less oxygen. Elaborate algorithms are applied to radio signal records to estimate blood oxygen levels at the places from which the signals are calculated to have originated. There is good reason to believe that blood flowing just downstream from spiking neurons carries appreciably more oxygen than blood in the vicinity of resting neurons. Assumptions about the relevant spatial and temporal relations are used to estimate levels of electrical activity in small regions of the brain corresponding to pixels in the finished image. The results of all of these computations are used to assign the appropriate colors to pixels in a computer generated image of the brain. The role of the senses in fMRI data production is limited to such things as monitoring the equipment and keeping an eye on the subject. Their epistemic role is limited to discriminating the colors in the finished image, reading tables of numbers the computer used to assign them, and so on.

If fMRI images record observations, it’s hard to say what was observed—neuronal activity, blood oxygen levels, proton precessions, radio signals, or something else. (If anything is observed, the radio signals that interact directly with the equipment would seem to be better candidates than blood oxygen levels or neuronal activity.) Furthermore, it’s hard to reconcile the idea that fMRI images record observations with the traditional empiricist notion that much as they may be needed to draw conclusions from observational evidence, calculations involving theoretical assumptions and background beliefs must not be allowed (on pain of loss of objectively) to intrude into the process of data production. The production of fMRI images requires extensive statistical manipulation based on theories about the radio signals, and a variety of factors having to do with their detection along with beliefs about relations between blood oxygen levels and neuronal activity, sources of systematic error, and so on.

In view of all of this, functional brain imaging differs, e.g., from looking and seeing, photographing, and measuring with a thermometer or a galvanometer in ways that make it uninformative to call it observation at all. And similarly for many other methods scientists use to produce non-perceptual evidence.

Terms like ‘observation’ and ‘observation reports’ don’t occur nearly as much in scientific as in philosophical writings. In their place, working scientists tend to talk about data . Philosophers who adopt this usage are free to think about standard examples of observation as members of a large, diverse, and growing family of data production methods. Instead of trying to decide which methods to classify as observational and which things qualify as observables, philosophers can then concentrate on the epistemic influence of the factors that differentiate members of the family. In particular, they can focus their attention on what questions data produced by a given method can be used to answer, what must be done to use that data fruitfully, and the credibility of the answers they afford.(Bogen 2016)

It is of interest that records of perceptual observation are not always epistemically superior to data from experimental equipment. Indeed it is not unusual for investigators to use non-perceptual evidence to evaluate perceptual data and correct for its errors. For example, Rutherford and Pettersson conducted similar experiments to find out if certain elements disintegrated to emit charged particles under radioactive bombardment. To detect emissions, observers watched a scintillation screen for faint flashes produced by particle strikes. Pettersson’s assistants reported seeing flashes from silicon and certain other elements. Rutherford’s did not. Rutherford’s colleague, James Chadwick, visited Petterson’s laboratory to evaluate his data. Instead of watching the screen and checking Pettersson’s data against what he saw, Chadwick arranged to have Pettersson’s assistants watch the screen while unbeknownst to them he manipulated the equipment, alternating normal operating conditions with a condition in which particles, if any, could not hit the screen. Pettersson’s data were discredited by the fact that his assistants reported flashes at close to the same rate in both conditions (Steuwer 1985, 284–288).

Related considerations apply to the distinction between observable and unobservable objects of investigation. Some data are produced to help answer questions about things that do not themselves register on the senses or experimental equipment. Solar neutrino fluxes are a frequently discussed case in point. Neutrinos cannot interact directly with the senses or measuring equipment to produce recordable effects. Fluxes in their emission were studied by trapping the neutrinos and allowing them to interact with chlorine to produce a radioactive argon isotope. Experimentalists could then calculate fluxes in solar neutrino emission from Geiger counter measurements of radiation from the isotope. The epistemic significance of the neutrinos’ unobservability depends upon factors having to do with the reliability of the data the investigators managed to produce, and its validity as a source of information about the fluxes. It’s validity will depend, among many other things, on the correctness of the investigators ideas about how neutrinos interact with chlorine (Pinch 1985). But there are also unobservables that cannot be detected, and whose features cannot be inferred from data of any kind. These are the only unobservables that are epistemically unavailable. Whether they remain so depends upon whether scientists can figure out how to produce data to study them. The history of particle physics (see e.g. Morrison 2015) and neuro-science (see e.g., Valenstein 2005).

Empirically minded philosophers assume that the evidential value of an observation or observational process depends on how sensitive it is to whatever it is used to study. But this in turn depends on the adequacy of any theoretical claims its sensitivity may depend on. For example we can challenge the use of a thermometer reading, e , to support a description, prediction, or explanation of a patient’s temperature, t , by challenging theoretical claims, C , having to do with whether a reading from a thermometer like this one, applied in the same way under similar conditions, should indicate the patient’s temperature well enough to count in favor of or against t . At least some of the C will be such that regardless of whether an investigator explicitly endorses, or is even aware of them, her use of e would be undermined by their falsity. All observations and uses of observations evidence are theory laden in this sense. (Cf. Chang 2005), Azzouni 2004.) As the example of the thermometer illustrates, analogues of Norwood Hanson’claim that seeing is a theory laden undertaking apply just as well to equipment generated observations.(Hanson 1958, 19). But if all observations and observational processes are theory laden, how can they provide reality based, objective epistemic constraints on scientific reasoning? One thing to say about this is that the theoretical claims the epistemic value of a parcel of observational evidence depends on may be may be quite correct. If so, even if we don’t know, or have no way to establish their correctness, the evidence may be good enough for the uses to which we put it. But this is cold comfort for investigators who can’t establish it. The next thing to say is that scientific investigation is an ongoing process during the course of which theoretical claims whose unacceptability would reduce the epistemic value of a parcel of evidence can be challenged and defended in different ways at different times as new considerations and investigative techniques are introduced. We can hope that the acceptability of the evidence can be established relative to one or more stretches of time even though success in dealing with challenges at one time is no guarantee that all future challenges can be satisfactorily dealt with. Thus as long as scientists continue their work there need be no time at which the epistemic value of of a parcel of evidence can be established once and for all. This should come as no surprise to anyone who is aware that science is fallible. But it is no grounds for skepticism. It can be perfectly reasonable to trust the evidence available at present even though it is logically possible for epistemic troubles to arise in the future.

Thomas Kuhn (1962), Norwood Hanson (1958), Paul Feyerabend (1959) and others cast suspicion on the objectivity of observational evidence in another way by arguing that one can’t use empirical evidence to teat a theory without committing oneself to that very theory. Although some of the examples they use to present their case feature equipment generated evidence, they tend to talk about observation as a perceptual process. Kuhn’s writings contain three different versions of this idea.

K1 . Perceptual Theory Loading. Perceptual psychologists, Bruner and Postman, found that subjects who were briefly shown anomalous playing cards, e.g., a black four of hearts, reported having seen their normal counterparts e.g., a red four of hearts. It took repeated exposures to get subjects to say the anomalous cards didn’t look right, and eventually, to describe them correctly. (Kuhn 1962, 63). Kuhn took such studies to indicate that things don’t look the same to observers with different conceptual resources. (For a more up-to-date discussion of theory and conceptual perceptual loading see Lupyan 2015.) If so, black hearts didn’t look like black hearts until repeated exposures somehow allowed subjects to acquire the concept of a black heart. By analogy, Kuhn supposed, when observers working in conflicting paradigms look at the same thing, their conceptual limitations should keep them from having the same visual experiences (Kuhn 1962, 111, 113–114, 115, 120–1). This would mean, for example, that when Priestley and Lavoisier watched the same experiment, Lavioisier should have seen what accorded with his theory that combustion and respiration are oxidation processes, while Priestley’s visual experiences should have agreed with his theory that burning and respiration are processes of phlogiston release. K2 . Semantical Theory Loading: Kuhn argued that theoretical commitments exert a strong influence on observation descriptions, and what they are understood to mean (Kuhn 1962, 127ff, Longino 1979,38-42). If so, proponents of a caloric account of heat won’t describe or understand descriptions of observed results of heat experiments in the same way as investigators who think of heat in terms of mean kinetic energy or radiation. They might all use the same words (e.g., ‘temperature’) to report an observation without understanding them in the same way. K3 . Salience: Kuhn claimed that if Galileo and an Aristotelian physicist had watched the same pendulum experiment, they would not have looked at or attended to the same things. The Aristotelian’s paradigm would have required the experimenter to measure …the weight of the stone, the vertical height to which it had been raised, and the time required for it to achieve rest (Kuhn 1992, 123)

and ignore radius, angular displacement, and time per swing (Kuhn 1962, 124).

These last were salient to Galileo because he treated pendulum swings as constrained circular motions. The Galilean quantities would be of no interest to an Aristotelian who treats the stone as falling under constraint toward the center of the earth (Kuhn 1962, 123). Thus Galileo and the Aristotelian would not have collected the same data. (Absent records of Aristotelian pendulum experiments we can think of this as a thought experiment.)

Taking K1, K2, and K3 in order of plausibility, K3 points to an important fact about scientific practice. Data production (including experimental design and execution) is heavily influenced by investigators’ background assumptions. Sometimes these include theoretical commitments that lead experimentalists to produce non-illuminating or misleading evidence. In other cases they may lead experimentalists to ignore, or even fail to produce useful evidence. For example, in order to obtain data on orgasms in female stumptail macaques, one researcher wired up females to produce radio records of orgasmic muscle contractions, heart rate increases, etc. But as Elisabeth Lloyd reports, “… the researcher … wired up the heart rate of the male macaques as the signal to start recording the female orgasms. When I pointed out that the vast majority of female stumptail orgasms occurred during sex among the females alone, he replied that yes he knew that, but he was only interested in important orgasms” (Lloyd 1993, 142). Although female stumptail orgasms occuring during sex with males are atypical, the experimental design was driven by the assumption that what makes features of female sexuality worth studying is their contribution to reproduction (Lloyd 1993, 139).

Fortunately, such things don’t always happen. When they do, investigators are often able eventually to make corrections, and come to appreciate the significance of data that had not originally been salient to them. Thus paradigms and theoretical commitments actually do influence saliency, but their influence is neither inevitable nor irremediable.

With regard to semantic theory loading (K2), it’s important to bear in mind that observers don’t always use declarative sentences to report observational and experimental results. They often draw, photograph, make audio recordings, etc. instead or set up their experimental devices to generate graphs, pictorial images, tables of numbers, and other non-sentential records. Obviously investigators’ conceptual resources and theoretical biases can exert epistemically significant influences on what they record (or set their equipment to record), which details they include or emphasize, and which forms of representation they choose (Daston and Galison 2007,115–190 309–361). But disagreements about the epistemic import of a graph, picture or other non-sentential bit of data often turn on causal rather than semantical considerations. Anatomists may have to decide whether a dark spot in a micrograph was caused by a staining artifact or by light reflected from an anatomically significant structure. Physicists may wonder whether a blip in a Geiger counter record reflects the causal influence of the radiation they wanted to monitor, or a surge in ambient radiation. Chemists may worry about the purity of samples used to obtain data. Such questions are not, and are not well represented as, semantic questions to which K2 is relevant. Late 20 th century philosophers may have ignored such cases and exaggerated the influence of semantic theory loading because they thought of theory testing in terms of inferential relations between observation and theoretical sentences.

With regard to sentential observation reports, the significance of semantic theory loading is less ubiquitous than one might expect. The interpretation of verbal reports often depends on ideas about causal structure rather than the meanings of signs. Rather than worrying about the meaning of words used to describe their observations, scientists are more likely to wonder whether the observers made up or withheld information, whether one or more details were artifacts of observation conditions, whether the specimens were atypical, and so on.

Kuhnian paradigms are heterogeneous collections of experimental practices, theoretical principles, problems selected for investigation, approaches to their solution, etc. Connections between components are loose enough to allow investigators who disagree profoundly over one or more theoretical claims to agree about how to design, execute, and record the results of their experiments. That is why neuroscientists who disagreed about whether nerve impulses consisted of electrical currents could measure the same electrical quantities, and agree on the linguistic meaning and the accuracy of observation reports including such terms as ‘potential’, ‘resistance’, ‘voltage’ and ‘current’.

The issues this section touches on are distant, linguistic descendents of issues that arose in connection with Locke’s view that mundane and scientific concepts (the empiricists called them ideas) derive their contents from experience (Locke 1700, 104–121,162–164, 404–408).

Looking at a patient with red spots and a fever, an investigator might report having seen the spots, or measles symptoms, or a patient with measles. Watching an unknown liquid dripping into a litmus solution an observer might report seeing a change in color, a liquid with a PH of less than 7, or an acid. The appropriateness of a description of a test outcome depends on how the relevant concepts are operationalized. What justifies an observer to report having observed a case of measles according to one operationalization might require her to say no more than that she had observed measles symptoms, or just red spots according to another.

In keeping with Percy Bridgman’s view that

…in general, we mean by a concept nothing more than a set of operations; the concept is synonymous with the corresponding sets of operations. (Bridgman 1927, 5)

one might suppose that operationalizations are definitions or meaning rules such that it is analytically true, e.g., that every liquid that turns litmus red in a properly conducted test is acidic. But it is more faithful to actual scientific practice to think of operationalizations as defeasible rules for the application of a concept such that both the rules and their applications are subject to revision on the basis of new empirical or theoretical developments. So understood, to operationalize is to adopt verbal and related practices for the purpose of enabling scientists to do their work. Operationalizations are thus sensitive and subject to change on the basis of findings that influence their usefulness (Feest, 2005).

Definitional or not, investigators in different research traditions may be trained to report their observations in conformity with conflicting operationalizations. Thus instead of training observers to describe what they see in a bubble chamber as a whitish streak or a trail, one might train them to say they see a particle track or even a particle. This may reflect what Kuhn meant by suggesting that some observers might be justified or even required to describe themselves as having seen oxygen, transparent and colorless though it is, or atoms, invisible though they are. (Kuhn 1962, 127ff) To the contrary, one might object that what one sees should not be confused with what one is trained to say when one sees it, and therefore that talking about seeing a colorless gas or an invisible particle may be nothing more than a picturesque way of talking about what certain operationalizations entitle observers to say. Strictly speaking, the objection concludes, the term ‘observation report’ should be reserved for descriptionsthat are neutral with respect to conflicting operationalizations.

If observational data are just those utterances that meet Feyerabend’s decidability and agreeability conditions, the import of semantic theory loading depends upon how quickly, and for which sentences reasonably sophisticated language users who stand in different paradigms can non-inferentially reach the same decisions about what to assert or deny. Some would expect enough agreement to secure the objectivity of observational data. Others would not. Still others would try to supply different standards for objectivity.

The example of Pettersson’s and Rutherford’s scintillation screen evidence (above) attests to the fact that observers working in different laboratories sometimes report seeing different things under similar conditions. It’s plausible that their expectations influence their reports. It’s plausible that their expectations are shaped by their training and by their supervisors’ and associates’ theory driven behavior. But as happens in other cases as well, all parties to the dispute agreed to reject Pettersson’s data by appeal to results of mechanical manipulations both laboratories could obtain and interpret in the same way without compromising their theoretical commitments.

Furthermore proponents of incompatible theories often produce impressively similar observational data. Much as they disagreed about the nature of respiration and combustion, Priestley and Lavoisier gave quantitatively similar reports of how long their mice stayed alive and their candles kept burning in closed bell jars. Priestley taught Lavoisier how to obtain what he took to be measurements of the phlogiston content of an unknown gas. A sample of the gas to be tested is run into a graduated tube filled with water and inverted over a water bath. After noting the height of the water remaining in the tube, the observer adds “nitrous air” (we call it nitric oxide) and checks the water level again. Priestley, who thought there was no such thing as oxygen, believed the change in water level indicated how much phlogiston the gas contained. Lavoisier reported observing the same water levels as Priestley even after he abandoned phlogiston theory and became convinced that changes in water level indicated free oxygen content (Conant 1957, 74–109).

The moral of these examples is that although paradigms or theoretical commitments sometimes have an epistemically significant influence on what observers perceive, it can be relatively easy to nullify or correct for their effects.

Typical responses to this question maintain that the acceptability of theoretical claims depends upon whether they are true (approximately true, probable, or significantly more probable than their competitors) or whether they “save” observable phenomena. They then try to explain how observational data argue for or against the possession of one or more of these virtues.

Truth. It’s natural to think that computability, range of application, and other things being equal, true theories are better than false ones, good approximations are better than bad ones, and highly probable theoretical claims are better than less probable ones. One way to decide whether a theory or a theoretical claim is true, close to the truth, or acceptably probable is to derive predictions from it and use observational data to evaluate them. Hypothetico-Deductive (HD) confirmation theorists propose that observational evidence argues for the truth of theories whose deductive consequences it verifies, and against those whose consequences it falsifies (Popper 1959, 32–34). But laws and theoretical generalization seldom if ever entail observational predictions unless they are conjoined with one or more auxiliary hypotheses taken from the theory they belong to. When the prediction turns to be false, HD has trouble explaining which of the conjuncts is to blame. If a theory entails a true prediction, it will continue to do so in conjunction with arbitrarily selected irrelevant claims. HD has trouble explaining why the prediction doesn’t confirm the irrelevancies along with the theory of interest.

Ignoring details, large and small, bootstrapping confirmation theories hold that an observation report confirms a theoretical generalization if an instance of the generalization follows from the observation report conjoined with auxiliary hypotheses from the theory the generalization belongs to. Observation counts against a theoretical claim if the conjunction entails a counter-instance. Here, as with HD, an observation argues for or against a theoretical claim only on the assumption that the auxiliary hypotheses are true (Glymour 1980, 110–175).

Bayesians hold that the evidential bearing of observational evidence on a theoretical claim is to be understood in terms of likelihood or conditional probability. For example, whether observational evidence argues for a theoretical claim might be thought to depend upon whether it is more probable (and if so how much more probable) than its denial conditional on a description of the evidence together with background beliefs, including theoretical commitments. But by Bayes’ theorem, the conditional probability of the claim of interest will depend in part upon that claim’s prior probability. Once again, one’s use of evidence to evaluate a theory depends in part upon one’s theoretical commitments. (Earman 1992, 33–86. Roush 2005, 149–186)

Francis Bacon (Bacon 1620, 70) said that allowing one’s commitment to a theory to determine what one takes to be the epistemic bearing of observational evidence on that very theory is, if anything, even worse than ignoring the evidence altogether. HD, Bootstrap, Bayesian, and related accounts of conformation run the risk of earning Bacon’s disapproval. According to all of them it can be reasonable for adherents of competing theories to disagree about how observational data bear on the same claims. As a matter of historical fact, such disagreements do occur. The moral of this fact depends upon whether and how such disagreements can be resolved. Because some of the components of a theory are logically and more or less probabilistically independent of one another, adherents of competing theories can often can find ways to bring themselves into close enough agreement about auxiliary hypotheses or prior probabilities to draw the same conclusions from the evidence.

Saving observable phenomena. Theories are said to save observable phenomena if they satisfactorily predict, describe, or systematize them. How well a theory performs any of these tasks need not depend upon the truth or accuracy of its basic principles. Thus according to Osiander’s preface to Copernicus’ On the Revolutions, a locus classicus, astronomers ‘…cannot in any way attain to true causes’ of the regularities among observable astronomical events, and must content themselves with saving the phenomena in the sense of using

…whatever suppositions enable …[them] to be computed correctly from the principles of geometry for the future as well as the past…(Osiander 1543, XX)

Theorists are to use those assumptions as calculating tools without committing themselves to their truth. In particular, the assumption that the planets rotate around the sun must be evaluated solely in terms of how useful it is in calculating their observable relative positions to a satisfactory approximation.

Pierre Duhem’s Aim and Structure of Physical Theory articulates a related conception. For Duhem a physical theory

…is a system of mathematical propositions, deduced from a small number of principles, which aim to represent as simply and completely, and exactly as possible, a set of experimental laws. (Duhem 1906, 19)

‘Experimental laws’ are general, mathematical descriptions of observable experimental results. Investigators produce them by performing measuring and other experimental operations and assigning symbols to perceptible results according to pre-established operational definitions (Duhem 1906, 19). For Duhem, the main function of a physical theory is to help us store and retrieve information about observables we would not otherwise be able to keep track of. If that’s what a theory is supposed to accomplish, its main virtue should be intellectual economy. Theorists are to replace reports of individual observations with experimental laws and devise higher level laws (the fewer, the better) from which experimental laws (the more, the better) can be mathematically derived (Duhem 1906, 21ff).

A theory’s experimental laws can be tested for accuracy and comprehensiveness by comparing them to observational data. Let EL be one or more experimental laws that perform acceptably well on such tests. Higher level laws can then be evaluated on the basis of how well they integrate EL into the rest of the theory. Some data that don’t fit integrated experimental laws won’t be interesting enough to worry about. Other data may need to be accommodated by replacing or modifying one or more experimental laws or adding new ones. If the required additions, modifications or replacements deliver experimental laws that are harder to integrate, the data count against the theory. If the required changes are conducive to improved systematization the data count in favor of it. If the required changes make no difference, the data don’t argue for or against the theory.

It is an unwelcome fact for all of these ideas about theory testing that data are typically produced in ways that make it impossible to predict them from the generalizations they are used to test, or to derive instances of those generalizations from data and non ad hoc auxiliary hypotheses. Indeed, it’s unusual for many members of a set of reasonably precise quantitative data to agree with one another, let alone with a quantitative prediction. That is because precise, publicly accessible data typically cannot be produced except through processes whose results reflect the influence of causal factors that are too numerous, too different in kind, and too irregular in behavior for any single theory to account for them. When Bernard Katz recorded electrical activity in nerve fiber preparations, the numerical values of his data were influenced by factors peculiar to the operation of his galvanometers and other pieces of equipment, variations among the positions of the stimulating and recording electrodes that had to be inserted into the nerve, the physiological effects of their insertion, and changes in the condition of the nerve as it deteriorated during the course of the experiment. There were variations in the investigators’ handling of the equipment. Vibrations shook the equipment in response to a variety of irregularly occurring causes ranging from random error sources to the heavy tread of Katz’s teacher, A.V. Hill, walking up and down the stairs outside of the laboratory. That’s a short list. To make matters worse, many of these factors influenced the data as parts of irregularly occurring, transient, and shifting assemblies of causal influences.

With regard to kinds of data that should be of interest to philosophers of physics, consider how many extraneous causes influenced radiation data in solar neutrino detection experiments, or spark chamber photographs produced to detect particle interactions. The effects of systematic and random sources of error are typically such that considerable analysis and interpretation are required to take investigators from data sets to conclusions that can be used to evaluate theoretical claims.

This applies as much to clear cases of perceptual data as to machine produced records. When 19 th and early 20 th century astronomers looked through telescopes and pushed buttons to record the time at which they saw a moon pass a crosshair, the values of their data points depended, not only upon light reflected from the moon, but also upon features of perceptual processes, reaction times, and other psychological factors that varied non-systematically from time to time and observer to observer. No astronomical theory has the resources to take such things into account. Similar considerations apply to the probabilities of specific data points conditional on theoretical principles, and the probabilities of confirming or disconfirming instances of theoretical claims conditional on the values of specific data points.

Instead of testing theoretical claims by direct comparison to raw data, investigators use data to infer facts about phenomena, i.e., events, regularities, processes, etc. whose instances, are uniform and uncomplicated enough to make them susceptible to systematic prediction and explanation (Bogen and Woodward 1988, 317). The fact that lead melts at temperatures at or close to 327.5 C is an example of a phenomenon, as are widespread regularities among electrical quantities involved in the action potential, the periods and orbital paths of the planets, etc. Theories that cannot be expected to predict or explain such things as individual temperature readings can nevertheless be evaluated on the basis of how useful they they are in predicting or explaining phenomena they are used to detect. The same holds for the action potential as opposed to the electrical data from which its features are calculated, and the orbits of the planets in contrast to the data of positional astronomy. It’s reasonable to ask a genetic theory how probable it is (given similar upbringings in similar environments) that the offspring of a schizophrenic parent or parents will develop one or more symptoms the DSM classifies as indicative of schizophrenia. But it would be quite unreasonable to ask it to predict or explain one patient’s numerical score on one trial of a particular diagnostic test, or why a diagnostician wrote a particular entry in her report of an interview with an offspring of a schizophrenic parents (Bogen and Woodward, 1988, 319–326).

The fact that theories are better at predicting and explaining facts about or features of phenomena than data isn’t such a bad thing. For many purposes, theories that predict and explain phenomena would be more illuminating, and more useful for practical purposes than theories (if there were any) that predicted or explained members of a data set. Suppose you could choose between a theory that predicted or explained the way in which neurotransmitter release relates to neuronal spiking (e.g., the fact that on average, transmitters are released roughly once for every 10 spikes) and a theory which explained or predicted the numbers displayed on the relevant experimental equipment in one, or a few single cases. For most purposes, the former theory would be preferable to the latter at the very least because it applies to so many more cases. And similarly for theories that predict or explain something about the probability of schizophrenia conditional on some genetic factor or a theory that predicted or explained the probability of faulty diagnoses of schizophrenia conditional on facts about the psychiatrist’s training. For most purposes, these would be preferable to a theory that predicted specific descriptions in a case history.

In view of all of this, together with the fact that a great many theoretical claims can only be tested directly against facts about phenomena, it behooves epistemologists to think about how data are used to answer questions about phenomena. Lacking space for a detailed discussion, the most this entry can do is to mention two main kinds of things investigators do in order to draw conclusions from data. The first is causal analysis carried out with or without the use of statistical techniques. The second is non-causal statistical analysis.

First, investigators must distinguish features of the data that are indicative of facts about the phenomenon of interest from those which can safely be ignored, and those which must be corrected for. Sometimes background knowledge makes this easy. Under normal circumstances investigators know that their thermometers are sensitive to temperature, and their pressure gauges, to pressure. An astronomer or a chemist who knows what spectrographic equipment does, and what she has applied it to will know what her data indicate. Sometimes it’s less obvious. When Ramon y Cajal looked through his microscope at a thin slice of stained nerve tissue, he had to figure out which if any of the fibers he could see at one focal length connected to or extended from things he could see only at another focal length, or in another slice.

Analogous considerations apply to quantitative data. It was easy for Katz to tell when his equipment was responding more to Hill’s footfalls on the stairs than to the electrical quantities is was set up to measure. It can be harder to tell whether an abrupt jump in the amplitude of a high frequency EEG oscillation was due to a feature of the subjects brain activity or an artifact of extraneous electrical activity in the laboratory or operating room where the measurements were made. The answers to questions about which features of numerical and non-numerical data are indicative of a phenomenon of interest typically depend at least in part on what is known about the causes that conspire to produce the data.

Statistical arguments are often used to deal with questions about the influence of epistemically relevant causal factors. For example, when it is known that similar data can be produced by factors that have nothing to do with the phenomenon of interest, Monte Carlo simulations, regression analyses of sample data, and a variety of other statistical techniques sometimes provide investigators with their best chance of deciding how seriously to take a putatively illuminating feature of their data.

But statistical techniques are also required for purposes other than causal analysis. To calculate the magnitude of a quantity like the melting point of lead from a scatter of numerical data, investigators throw out outliers, calculate the mean and the standard deviation, etc., and establish confidence and significance levels. Regression and other techniques are applied to the results to estimate how far from the mean the magnitude of interest can be expected to fall in the population of interest (e.g., the range of temperatures at which pure samples of lead can be expected to melt).

The fact that little can be learned from data without causal, statistical, and related argumentation has interesting consequences for received ideas about how the use of observational evidence distinguishes science from pseudo science, religion, and other non-scientific cognitive endeavors.First, scientists aren’t the only ones who use observational evidence to support their claims; astrologers and medical quacks use them too. To find epistemically significant differences, one must carefully consider what sorts of data they use, where it comes from, and how it is employed. The virtues of scientific as opposed to non-scientific theory evaluations depend not only on its reliance on empirical data, but also on how the data are produced, analyzed and interpreted to draw conclusions against which theories can be evaluated. Secondly, it doesn’t take many examples to refute the notion that adherence to a single, universally applicable “scientific method” differentiates the sciences from the non-sciences. Data are produced, and used in far too many different ways to treat informatively as instance of any single method. Thirdly, it is usually, if not always, impossible for investigators to draw conclusions to test theories against observational data without explicit or implicit reliance on theoretical principles. This means that when counterparts to Kuhnian questions about theory loading and its epistemic significance arise in connection with the analysis and interpretation of observational evidence, such questions must be answered by appeal to details that vary from case to case.

Grammatical variants of the term ‘observation’ have been applied to impressively different perceptual and non-perceptual process and to records of the results they produce. Their diversity is a reason to doubt whether general philosophical accounts of observation, observables, and observational data can tell epistemologists as much as local accounts grounded in close studies of specific kinds of cases. Furthermore, scientists continue to find ways to produce data that can’t be called observational without stretching the term to the point of vagueness.

It’s plausible that philosophers who value the kind of rigor, precision, and generality to which l logical empiricists and other exact philosophers aspired could do better by examining and developing techniques and results from logic, probability theory, statistics, machine learning, and computer modeling, etc. than by trying to construct highly general theories of observation and its role in science. Logic and the rest seem unable to deliver satisfactory, universally applicable accounts of scientific reasoning. But they have illuminating local applications, some of which can be of use to scientists as well as philosophers.

• Aristotle(a), Generation of Animals in Complete Works of Aristotle (Volume 1), J. Barnes (ed.), Princeton: Princeton University Press, 1995, pp. 774–993
• Aristotle(b), History of Animals in Complete Works of Aristotle (Volume 1), J. Barnes (ed.), Princeton: Princeton University Press, 1995, pp. 1111–1228.
• Azzouni, J., 2004, “Theory, Observation, and Scientific Realism,” British Journal for the Philosophy of Science , 55(3): 371-92.
• Bacon, Francis, 1620, Novum Organum with other parts of the Great Instauration , P. Urbach and J. Gibson (eds. and trans.), La Salle: Open Court, 1994.
• Bogen, J., 2016, “Empiricism and After,”in P. Humphreys (ed.), Oxford Handbook of Philosophy of Science , Oxford: Oxford Univesity Press, 779-795.
• Bogen, J, and Woodward, J., 1988, “Saving the Phenomena,” Philosophical Review , XCVII (3): 303–352.
• Boyle, R., 1661, The Sceptical Chymist , Montana: Kessinger (reprint of 1661 edition).
• Bridgman, P., 1927, The Logic of Modern Physics , New York: Macmillan.
• Chang, H., 2005, “A Case for Old-fashioned Observability, and a Reconstructive Empiricism,” Philosophy of Science , 72(5): 876–887.
• Collins, H. M., 1985 Changing Order , Chicago: University of Chicago Press.
• Conant, J.B., 1957, (ed.) “The Overthrow of the Phlogiston Theory: The Chemical Revolution of 1775–1789,” in J.B.Conant and L.K. Nash (eds.), Harvard Studies in Experimental Science , Volume I, Cambridge: Harvard University Press, pp. 65–116).
• Duhem, P., 1906, The Aim and Structure of Physical Theory , P. Wiener (tr.), Princeton: Princeton University Press, 1991.
• Earman, J., 1992, Bayes or Bust? , Cambridge: MIT Press.
• Feest, U., 2005, “Operationism in psychology: what the debate is about, what the debate should be about,” Journal of the History of the Behavioral Sciences , 41(2): 131–149.
• Feyerabend, P.K., 1959, “An Attempt at a Realistic Interpretation of Expeience,” in P.K. Feyerabend, Realism, Rationalism, and Scientific Method (Philosophical Papers I), Cambridge: Cambridge University Press, 1985, pp. 17–36.
• Feyerabend, P.K., 1969, “Science Without Experience,” in P.K. Feyerabend, Realism, Rationalism, and Scientific Method (Philosophical Papers I), Cambridge: Cambridge University Press, 1985, pp. 132–136.
• Franklin, A., 1986, The Neglect of Experiment , Cambridge: Cambridge University Press.
• Galison, P., 1987, How Experiments End , Chicago: University of Chicago Press.
• Galison, P., 1990, “Aufbau/Bauhaus: logical positivism and architectural modernism,” Critical Inquiry , 16 (4): 709–753.
• Galison, P., and Daston, L., 2007, Objectivity , Brooklyn: Zone Books.
• Glymour, C., 1980, Theory and Evidence , Princeton: Princeton University Press.
• Hacking, I, 1983, Representing and Intervening , Cambridge: Cambridge University Press.
• Hanson, N.R., 1958, Patterns of Discovery , Cambridge, Cambridge University Press.
• Hempel, C.G., 1935, “On the Logical Positivists’ Theory of Truth,” Analysis , 2 (4): 50–59.
• Hempel, C.G., 1952, “Fundamentals of Concept Formation in Empirical Science,” in Foundations of the Unity of Science , Volume 2, O. Neurath, R. Carnap, C. Morris (eds.), Chicago: University of Chicago Press, 1970, pp. 651–746.
• Herschel, J. F. W., 1830, Preliminary Discourse on the Study of Natural Philosophy , New York: Johnson Reprint Corp., 1966.
• Hooke, R., 1705, “The Method of Improving Natural Philosophy,” in R. Waller (ed.), The Posthumous Works of Robert Hooke , London: Frank Cass and Company, 1971.
• Jeffrey, R.C., 1983, The Logic of Decision , Chicago: University Press.
• Kuhn, T.S., The Structure of Scientific Revolutions , 1962, Chicago: University of Chicago Press, reprinted,1996.
• Latour, B., and Woolgar, S., 1979, Laboratory Life, The Construction of Scientific Facts , Princeton: Princeton University Press, 1986.
• Lewis, C.I., 1950, Analysis of Knowledge and Valuation , La Salle: Open Court.
• Lloyd, E.A., 1993, “Pre-theoretical Assumptions In Evolutionary Explanations of Female Sexuality,”, Philosophical Studies , 69: 139–153.
• Lupyan, G., 2015, “Cognitive Penetrability of Perception in the Age of Prediction – Predictive Systems are Penetrable Systems,” Review of Philosophical Psychology , 6(4): 547–569. doi:10.1007/s13164-015-0253-4
• Longino, H., 1979, “Evidence and Hypothesis: An Analysis of Evidential Relations,” Philosophy of Science , 46(1): 35-56.
• Morrison, M., 2015, Reconstructing Reality , New York: Oxford University Press.
• Neurath, O., 1913, “The Lost Wanderers of Descartes and the Auxilliary Motive,” in O. Neurath, Philosophical Papers , Dordrecht: D. Reidel, 1983, pp. 1–12.
• Olesko, K.M. and Holmes, F.L., 1994, “Experiment, Quantification and Discovery: Helmholtz’s Early Physiological Researches, 1843–50,” in D. Cahan, (ed.), Hermann Helmholtz and the Foundations of Nineteenth Century Science , Berkeley: UC Press, pp. 50–108)
• Osiander, A., 1543, “To the Reader Concerning the Hypothesis of this Work,” in N. Copernicus On the Revolutions , E. Rosen (tr., ed.), Baltimore: Johns Hopkins University Press, 1978, p. XX.
• Pearl, J., 2000, Causality , Cambridge: Cambridge University Press.
• Pinch, T., 1985, “Towards an Analysis of Scientific Observation: The Externality and Evidential Significance of Observation Reports in Physics,” in Social Studies of Science , 15, pp. 3–36.
• Popper, K.R.,1959, The Logic of Scientific Discovery , K.R. Popper (tr.), New York: Basic Books.
• Rheinberger, H. J., 1997, Towards a History of Epistemic Things: Synthesizing Proteins in the Test Tube , Stanford: Stanford University Press.
• Roush, S., 2005, Tracking Truth , Cambridge: Cambridge University Press.
• Schlick, M., 1935, “Facts and Propositions,” in Philosophy and Analysis , M. Macdonald (ed.), New York: Philosophical Library, 1954, pp. 232–236.
• Spirtes, C., Glymour, C., and Scheines, R., 2000, Causation, Prediction, and Search , Cambridge: MIT Press.
• Steuer, R.H., “Artificial Distintegration and the Cambridge-Vienna Controversy,” in P. Achinstein and O. Hannaway (eds.), Observation, Experiment, and Hypothesis in Modern Physical Science , Cambridge: MIT Press, 1985, 239–307)
• Suppe, F., 1977, in F. Suppe (ed.) The Structure of Scientific Theories , Urbana: University of Illinois Press.
• Valenstein, E.S., 2005, The War of the Soups and the Sparks , New York: Columbia University Press.
• Van Fraassen, B.C, 1980, The Scientific Image , Oxford: Clarendon Press.
• Whewell, W., 1858, Novum Organon Renovatum , Book II, in William Whewell Theory of Scientfic Method , R.E. Butts (ed.), Indianapolis: Hackett Publishing Company, 1989, pp. 103–249.

How to cite this entry . Preview the PDF version of this entry at the Friends of the SEP Society . Look up this entry topic at the Indiana Philosophy Ontology Project (InPhO). Enhanced bibliography for this entry at PhilPapers , with links to its database.

• Confirmation , by Franz Huber, in the Internet Encyclopedia of Philosophy .
• Transcript of Katzmiller v. Dover Area School District (on the teaching of intelligent design).

Bacon, Francis | Bayes' Theorem | constructive empiricism | Duhem, Pierre | empiricism: logical | epistemology: Bayesian | Lewis, Clarence Irving | Locke, John | -->logical positivism --> | physics: experiment in | science: and pseudo-science

Copyright © 2017 by James Bogen < rtjbog @ comcast . net >

Support SEP

Mirror sites.

View this site from another server:

• Info about mirror sites

The Stanford Encyclopedia of Philosophy is copyright © 2016 by The Metaphysics Research Lab , Center for the Study of Language and Information (CSLI), Stanford University

Library of Congress Catalog Data: ISSN 1095-5054

An official website of the United States government

Official websites use .gov A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS A lock ( Lock Locked padlock icon ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

• Publications
• Account settings
• Advanced Search
• Journal List

Understanding Observational Learning: An Interbehavioral Approach

Mitch j fryling, cristin johnston, linda j hayes.

• Author information
• Copyright and License information

Correspondence concerning this article should be addressed to Mitch Fryling, The Chicago School of Professional Psychology; 617 W. 7th St., 8th Floor, Los Angeles, CA 90017. (e-mail: [email protected] ).

Observational learning is an important area in the field of psychology and behavior science more generally. Given this, it is essential that behavior analysts articulate a sound theory of how behavior change occurs through observation. This paper begins with an overview of seminal research in the area of observational learning, followed by a consideration of common behavior analytic conceptualizations of these findings. The interbehavioral perspective is then outlined, shedding light on some difficulties with the existing behavior analytic approaches. The implications of embracing the interbehavioral perspective for understanding the most complex sorts of behavior, including those involved in observational learning are considered.

Keywords: observational learning, interbehaviorism, interbehavioral psychology, stimulus substitution, rule-governed behavior

Research in observational learning represents a critical development in the history of psychology. Indeed, the research and scholarly work conducted by Bandura and colleagues set the occasion for the social cognitive perspective of learning ( Bandura, 1986 ), which seemed to challenge the possibility that all behavior could be accounted for by respondent and operant processes alone. Toward this, the social cognitive perspective focused more explicitly on both modeling and cognition, and their role in understanding behavior. Meanwhile, behavior analysts have continued to contend that observational learning can be explained through processes of generalized imitation, conditioned reinforcement, and rule-governed behavior (e.g., Catania, 2007 ; Pear, 2001 ; Pierce & Cheney, 2008 ). However, these contentions become increasingly difficult when we take a closer look at the psychological event of interest in observational learning. Further, while behavior analysts have continued to conduct research in the area of observational learning, relatively little progress has been made toward developing a theoretical understanding of this work. The primary aim of the current paper is to consider the general findings of the observational learning research within a thoroughly naturalistic, behavioral perspective. Of course, verbal processes play an important role in understanding observational learning, and thus, they are given both general and specific treatment throughout. In pursuing this work, J. R. Kantor's philosophy of interbehaviorism and scientific system of interbehavioral psychology are reviewed. The potential benefits of embracing the interbehavioral perspective with respect to understanding observational learning and complex behavior more generally are considered.

OBSERVATIONAL LEARNING

In the 1960s and 70s Albert Bandura and his colleagues became well known for their social psychology research in the area of observational learning. Indeed, several of the early experiments in this area are very well known, and considered hallmarks in the field of psychology and behavior science (e.g., Bandura & McDonald, 1963 ; Bandura, Ross, & Ross, 1963 ). These studies were pursued for a variety of reasons; partially to undermine the value of common psychoanalytic ( Bandura & Huston, 1961 ; Bandura, Ross, et al., 1963 ) and developmental theories ( Bandura & McDonald, 1963 ), and also to evaluate the role of observation as a primary determinant of behavior change. Early studies examined the role of modeling 1 on the acquisition of aggression ( Bandura, Ross, & Ross, 1963 ) and moral judgment ( Bandura & McDonald, 1963 ), for example, and provided a foundation upon which the social cognitive theory was built. Importantly, this theory is often considered to extend beyond behavioral theories, questioning the possibility that behaviorism alone could provide a comprehensive understanding of learning. Given the importance of this research, we will now provide a brief overview of some of the general findings of studies on observational learning. It is important to note that our review is admittedly less than comprehensive, and that our primary aim is to describe some common themes within this literature.

The Role of Modeling

An early and longstanding aim of the observational learning literature is to understand the role of modeling in behavior change (e.g., Bandura & Huston, 1961 ; Bandura & McDonald, 1963 ; Bandura, Ross, & Ross, 1961 ). For example, an early study examined how the incidental behaviors of an experimenter might be acquired in the context of learning another task (Bandura & Huston). The important conclusion of these studies is that behavior change can and does occur through observation, even when such observation is incidental, occurring in the context of other activities. While this finding seems rather simple, it has significant implications for how we conceptualize learning. As we will discuss in the coming paragraphs, this general finding may present specific conceptual challenges for behavioral theories of learning.

The role of consequences

Specific emphasis was also placed on the role of consequences in the observational learning literature (e.g., Bandura, 1965 ; Bandura, Grusec, & Menlove, 1966 ; Bandura & McDonald, 1963 , Bandura, Ross, & Ross, 1963 ). Experiments that added to our understanding of the role of consequences generally compared behavior change between children who either observed a model who was rewarded, a model who was punished, or a control condition (e.g., observing non-aggressive play or observing no consequences). Generally, less behavior change is observed when a child observes a model being punished (e.g., Bandura, Ross, & Ross, 1963 ). 2

Interestingly, there is often no difference between conditions involving rewards and conditions involving no consequences at all. For example, Bandura and McDonald (1963) compared the effects of three different variables on the acquisition of moral judgment responses. In this study, the three variables involved three different groups of adult/child dyads: group one involved both the model and child's target judgments be reinforced, group two involved the model's behavior being reinforced but not the child's, and group three involved no model and only child reinforcement. Importantly, in the model/child groups trials alternated between the model and the child. Groups one and two demonstrated more behavior change than group three at a 1–3 week post-treatment assessment. Thus, the researchers concluded that modeling was the significant factor involved in the acquisition of the moral judgment repertoire. 3 Other experiments also found no difference between the reward and no consequence groups, while the model punished group continued to yield different results (e.g., Bandura, 1965 ).

Along similar lines, other studies seemed to raise questions about the potentially detrimental effects of incentives on the acquisition of behavior. For example, at the beginning of one experiment ( Bandura, Grusec, & Menlove, 1966 ) half of the participants were placed into an incentive condition where they were told that they would be given candy treats for correctly demonstrating what they learned after watching a movie. More specifically, after watching a film, children in both conditions were asked to demonstrate what they observed on the movie. Generally, the researchers found that children in the incentive condition did slightly worse than those in the no incentive condition, raising questions about the benefits of incentives on learning (see Bandura, et al., p. 505). 4

At this point we must note that the terms reward , reinforcement , and operant conditioning are used rather loosely within this literature. From a behavior analytic perspective, a stimulus change can only be classified as a reinforcer if it increases the future frequency of the class of behavior it was made contingent upon (e.g., Cooper, Heron, & Heward, 2007 ). Given this, the majority of stimulus changes called “rewards” or “reinforcers” in the observational learning literature do not technically meet the criteria to be classified as reinforcers, or as being involved in the process of reinforcement or operant conditioning in general. Nevertheless, we can say that consequences seem to play some role in observational learning. Again, there are studies suggesting that there are no differences between observation with reinforcement and observation with no consequence at all, leaving us more confident that if consequences have a role, aversive consequences seem to play a large part. Given these important concerns, however, these findings need to be interpreted with caution.

The Role of Verbal Behavior

As this line of researched progressed, increasing attention was paid to the role of cognitive factors, often described with the terms coding and rehearsal . Generally, coding can be thought of as describing what is observed in some way, whereas rehearsal can be thought of as practicing what was observed. For example, Bandura, Grusec, & Menlove. (1966) examined the effects of describing the activity of the model (“coding”) on the acquisition of observed behavior. Of specific interest, this study was fueled by motivation to discredit behavior analysts who failed to account for “delayed reproduction of modeling behavior” (p. 499), which was assumed to necessarily involve some sort of cognitive activity. In this study three groups of children all viewed a video; one group was asked to “verbalize every action of the model as it is being performed” (p. 501), the second group to “count 1 and a 2, and a 3, and a 4, and a 5” (p. 501) repeatedly while watching the video, and a third group observed without any instruction. The researchers found that those individuals who verbally described every action of the model were the most successful when tested for behavior change at a later time. Importantly, this study highlights the early recognition of “cognitive” factors in observational learning.

In an effort to elaborate upon this sort of research, Bandura and Jeffrey (1973) examined the role of “coding and rehearsal” on the acquisition of observed behavior. The researchers found that participants who “symbolically coded” (i.e., developed number or letter coding systems) the model's actions, and also immediately rehearsed (i.e., practiced) those codes had the best outcomes. Neither coding without symbolic rehearsal or symbolic rehearsal without coding was found to be sufficient. Put differently, developing a coded description of the models actions and practicing that description were both found to be important factors in the acquisition of observed behavior. Interestingly, physically practicing (“motor rehearsal”) the observed behavior was found to be less important. This seemed to support a growing distinction between different aspects of an individual's repertoire and the various processes that contribute to their existence (see below).

Learning and performance

Related to the role of verbal behavior, Bandura and colleagues began to notice a difference between the observers imitative performance at a later time compared to their ability to describe what was observed when asked. The ability to describe what was observed was viewed as a measure of learning, while engaging in the observed behavior at a later time was viewed as performance. For example, Bandura, Ross, & Ross (1963) found that children in both the aggressive-reward (participants observed a model be rewarded for engaging in a sequence of responses) and aggressive-punished (participants observed a model be punished for engaging in a sequence of responses) groups were able to describe the observed sequences of behavior, despite differences in imitative behavior change. Similarly, Bandura (1965) found that differences between group measures on imitation of observed behavior were removed on an “acquisition index,” where children were told they would get a reward for telling the experimenter what the model did. These findings further highlighted the role of verbal behavior in the process of learning from observation, including the various ways in which such learning from observation might be measured. That is, one way of measuring learning from observation is through imitation of the observed response at a later time, while another is through descriptions of the observed behavior. As these repertoires seemed to be influenced by different factors, Bandura and colleagues began to distinguish between them more and more.

Theoretical Developments

Throughout the above studies Bandura and colleagues began to articulate a theoretical model of observational learning. Fueled by findings that individuals might be able to describe observed behavior at a later time, even if they did not actually engage in the behavior themselves during a testing condition (e.g., Bandura, 1965 ; Bandura, Ross, & Ross, 1963 ), Bandura and colleagues began to distinguish between learning and performance (also see Greer, Singer-Dudek, & Gautreaux, 2006 ). Specifically, Bandura and colleagues noted that verbal processes were more likely to influence learning, 5 whereas consequences were more likely to influence the extent to which the individual's behavior changed through observation (i.e., that they actually engaged in the observed behavior). Indeed, theoretical accounts of observational learning highlight this distinction (e.g., Bandura & Jeffrey, 1973 ; Greer, Singer-Dudek, & Gautreaux, 2006 ).

Bandura and colleagues assumed that learning from observation occurred via an input-output, cognitive model. Specifically, Bandura and Jeffrey (1973) described four processes that account for learning from observation: attentional, retention, motor reproduction, and motivational. Bandura and Jeffery (1973) say, “Within this framework acquisition of modeled patterns is primarily controlled by attention and retention processes. Whereas performance of observationally learned responses is regulated by motor reproduction and incentive processes” (p. 122).

Attentional processes were described as cognitive abilities that “regulate sensory registration of modeled actions” and retention processes were those that took “transitory influences and converted to enduring internal guides for memory representation” ( Bandura & Jeffery, 1973 , p. 122). Motor reproduction processes are those that move component actions stored in memory into overt action resembling that of the modeled behaviors. Finally, motivational processes determine whether or not those behaviors emerge as overt action.

According to the authors, this model not only explains how a modeled response can be imitated immediately after it is observed, but can also explain how this behavior can be reproduced later under many different circumstances. Bandura and Jeffrey (1973) conclude, “After modeled activities have been transformed into images and readily utilizable verbal symbols, these memory codes can function as guides for subsequent reproduction” (p. 123). The authors also concluded that participants who engage in transforming modeled actions into either descriptive words or visual images achieve higher levels of observational learning than those who did not.

As a result of these and other experiments, Bandura theorized that observational learning was an integral part of human development, which accounted for the development of the personality ( Bandura & Walters, 1963 ), as well as social and antisocial behaviors in children ( Bandura, 1973 ). Importantly, this research shows that humans can learn without directly experiencing the consequences of their own actions. Thus, if behavior analysts aim to develop a comprehensive account of learning it must include an adequate description of these instances. In particular, behavior analysts must account for the acquisition of novel behavior in the absence of contingent reinforcement for the individual engaging in those responses, and also articulate the role of verbal behavior in observational learning.

In summary, the studies conducted by Bandura and colleagues seemed to question the role of rewards on the behavior of the observer. Importantly, Bandura believed that reinforcement history alone was not sufficient, and that the observation of a model was the most critical factor. Moreover, learning from observation was viewed to be a result of other processes, of which “verbal coding” was one. These general findings seemed to devalue the comprehensiveness of the behavioral position, and set the stage for the social cognitive perspective. However, it is crucial that we reiterate the fact that Bandura and colleagues often misused the terms reinforcer and reinforcement , and thus, it is difficult to draw valid conclusions about the role of consequences from this line of research. What can be said is that observational learning is an important area for behavior science to consider.

Bandura found limitations with the operant interpretation of behavior, albeit a less than thoroughly informed understanding of it. Observational learning seems to defy traditional discriminative stimulus—response—reinforcer analyses, even when more contemporary concepts (e.g., the motivating operation) are considered. Specifically, novel responses occur in observational learning models, responses that have obviously never been reinforced. Added to this, delayed responding is common, and such responding presents conceptual challenges to traditional behavioral concepts (e.g., Bandura, Grusec, & Menlove, 1966 , p. 499). As mentioned earlier, it is perhaps not surprising that Bandura's work may be considered by some to be an extension or move beyond the behavioral position. The limitations of Bandura's work not withstanding, Bandura and colleagues raised several important issues regarding the role of observation and verbal behavior in behavior change processes.

Still, Bandura's model relies upon the existence of hypothetical entities that do not exist in the spatiotemporal event matrix comprising the natural world. In other words, Bandura's theoretical constructs are not derived from events, and as such cannot be found and thereby can never actually be studied (see Kantor, 1957 ; Smith, 2007 ). Rather, they are inferences derived from a thoroughly mentalistic, dualistic worldview. Behavior analysts have long held that embracing such constructs can only distract workers from a scientific analysis (e.g., Skinner, 1953 ). It isn't surprising, then, that behavior analysts have proposed an alternative conceptualization of observational learning. In the following section we provide an overview of the behavior analytic position on observational learning.

THE BEHAVIOR ANALYTIC POSITION

The behavior analytic account of observational learning rests squarely upon the process of generalized imitation ( Baer, Peterson, & Sherman, 1967 ; Baer & Sherman, 1964 ; Pierce & Cheney, 2008 ). This is a familiar process, where the organism is asked to imitate several responses of the model (e.g., “do this” while the model is touching their nose), and after multiple exemplars have been successfully trained, the organism is asked to engage in a response which has never been modeled before. Generalized imitation is said to occur when the organism engages in a response that has never been modeled or reinforced in the past; that is, when imitation has “generalized” to new behaviors. Furthermore, it is assumed that the social community shapes up delays in imitative responses, and thus, it is said that “all instances of modeling and imitation involve the absence of the Sd” ( Pierce & Cheney, 2008 , p. 252). For example, a child might watch their favorite TV show, and at a much later time repeat a phrase from the show, perhaps while sitting in the car, and their parent might say “yes, that's what you heard on TV!”. In other words, the organism is said to learn to imitate observed behavior in the absence of any particular stimulus, and perhaps at a much later point in time. In this sense, the organism may be said to “emit” behaviors, which typically fall under the purview of generalized imitation.

Importantly, conditioned reinforcement hypotheses are also central to the behavior analytic conceptualization of observational learning and imitation in general. In this sense, behaviors that closely resemble the observed behavior of models are presumed to have a history of reinforcement, and thus, behaving in a manner which is similar to the model may become conditioned reinforcer itself. This sort of conceptualization seems to be particularly helpful toward the behavior analytic understanding of delayed imitation (see Gladstone & Cooley, 1975 ; Rosales-Ruiz & Baer, 1997 ).

Behavior analysts have also provided an account of the verbal coding that is said to participate in observational learning. For example, behavior analysts propose that individuals derive self-rules when they observe their environment (e.g., Hayes, Barnes-Holmes, & Roche, 2001 ; Hayes, Zettle, & Rosenfarb, 1989 ; Poppen, 1989 ). It is assumed that society teaches the organism to tact ( Skinner, 1957 ) relationships in their environment, and that these descriptions exert tremendous control over behavior. Indeed, it is suggested that a large amount of rule-following behavior is reinforced throughout the organisms lifetime, and when combined with a history of tact repertoires being reinforced, individuals both derive self-rules (i.e., tact if-then relations in their environment) and subsequently engage in a great deal of rule-following with respect to those rules.

For example, a child might observe a teacher praising another child for accurately matching a Spanish flashcard to the corresponding English flashcard (“Good job matching perro with dog!”). Two days later, the child who observed the incident may be asked to “match same” when given that same Spanish flashcard, and correctly place it on the corresponding English flashcard. From the behavior analytic perspective it may be assumed that the child already has a generalized imitative repertoire, so they are imitating the child they observed at a later point in time (see conditioned reinforcement hypotheses above). Furthermore, the child may or may not have tacted the observed relationship when it occurred (rule-stating), and engaged in rule-following behavior when she interacted with the card at a later time. Both of these possibilities are consistent with the behavior analytic position. Importantly, the behavior analytic position does not require the individual to engage in rule-stating and following for observational learning to occur. Related to the latter, a recent series of studies conducted by Greer and colleagues seems to support the notion that observational learning may occur without rule-following. For example, individuals have acquired the ability to learn new words through experiences that do not involve observing consequences of another, and stimuli have been conditioned as reinforcers through the observation of others interacting with them, both of which do not require analyses of rule-governed behavior (see Greer & Ross, 2008 , Greer & Speckman, 2009 ).

It must be noted that many of these issues are at the center of current controversy, debate, and development in the field of behavior analysis. For example, the perspectives of joint control (e.g., Lowenkron, 1998 ) naming ( Horne & Lowe, 1996 ), relational frame theory ( Hayes, Barnes-Holmes, & Roche, 2001 ), and verbal behavior development (e.g., Greer & Ross, 2008 ; Greer & Speckman, 2009 ) all seem to account for the type of phenomena we have commented on herein. Given the importance of these issues, this is a good sign. We primarily mention this to acknowledge the current fact that there is not a behavior analytic position on many of these issues. Nevertheless, missteps may occur while we are on our journey to account for such phenomena, missteps that could have more or less dangerous implications for behavior analysis as an enterprise. It is our perspective that the interbehavioral position may be a rather useful foundation for workers as we continue on this journey (see Morris, Higgins, & Bickel, 1982 ).

Generally speaking, the behavior analytic conceptualization of observational learning relies on generalized imitation, conditioned reinforcement, and a range of verbal processes, depending on ones theoretical preference. These processes seem to account for the fact that imitative responses which have never been reinforced occur at a later time, and also for the role of verbal behavior in observational learning. The fact that there are a number of different perspectives on many of these issues may be considered a sign of progress and growth within behavior analysis, but at the same time highlights the need for further system building in this area. In the following sections we take a closer look at the behavior analytic position through the lens of interbehavioral psychology. Before doing so, we briefly introduce the reader to the interbehavioral position, as it is relatively less familiar to most behavior analysts.

THE INTERBEHAVIORIAL POSITION

From the perspective of interbehavioral psychology the event of interest is always a thoroughly naturalistic, psychological event. Specifically, this event is always the stimulus function ( sf ) ←→response function ( rf ) interaction ( Kantor, 1958 ). Moreover, this interaction always participates in a multifactored, inter related field. This field is conceptualized by the following formula: PE  =  C ( k , sf , rf , hi , st , md ); where PE is the psychological event, C is the interrelationship of all of the participating factors, k is the unique organization of all factors, sf is the stimulus function, rf is the response function, hi is the interbehavioral history, st is setting factors, and md is the medium of contact. Importantly, this is one event, one interbehavioral field. When one factor is changed the entire field is altered. This is to say none of the above factors are viewed as independent, dependent, or having causal status. Rather, all of the factors are equal participants in the one, integrated whole (see Smith, 2006 ).

Of particular relevance to our discussion of observational learning and complex behavior in general is the explicit distinction between stimulus objects and stimulus functions made within Kantor's system (e.g., Kantor, 1924 , pp. 47–48; Parrott, 1983a , 1983b , 1986 ). In other words, the stimulating action of stimulus objects is differentiated from the formal properties of those objects in Kantor's system. Kantor has suggested that the borrowing of the terms stimulus and response from biology, where stimulus and response functions are at least relatively more determined by their structural properties, has perhaps contributed to the failure to distinguish between object and functional properties in the domain of psychology ( Kantor, 1958 , p. 68). For example, in Kantor's system a picture as a stimulus object would be explicitly distinguished from its psychological functions, such that accounting for seeing something in the absence of the thing seen (as when looking at a picture “reminds you” of the time or place it was taken) is not difficult (see Parrott, 1983a , 1983b , 1986 ; Skinner, 1974 ). The process by which this happens is central to understanding complex behavior, including those that typically fall within the purview of observational learning, and we will now describe this process in more detail.

Kantor suggested that association conditions are fundamental psychological processes (1921, 1924). The term association is used here to refer to spatiotemporal relationships; that is, to relationships among various factors that occur in the environment together in space and time. To be clear, these factors are associated in the environment , and not within the organism. Further, it is not the organism who is associating; rather, the environment is where all associating takes place. Association conditions may involve stimuli and responses, stimuli and stimuli, settings and stimuli, settings and reactions, settings and settings, and reactions and reactions (including implicit and nonimplicit variations thereof; Kantor, 1924 , pp. 321–322).

Stimulus Substitution

Stimulus substitution is the outcome of a history of an organism interacting with various association conditions ( Kantor, 1924 , 1958 ; Parrott, 1983a , 1983b , 1986 ). That is, given an organisms history of interacting with spatiotemporal relationships ( A -coffee shop←→ B -Peter), stimulus objects may have the stimulational properties of other objects, even when those other objects are no longer physically present. This is how you might see Peter when you enter a coffee shop you frequented with him, even when he isn't physically there. In this example, stimulus A (coffee shop) and B (Peter) occurred together in space and time, and an organism interacted with that relationship, such that B becomes A ( B [ A ]) and A becomes B ( A [ B ]), psychologically speaking (see Hayes, 1992a ). This process is of particular importance to understanding complex behavior of various sorts. Furthermore, this is how interbehaviorists are able to conceptualize the past and present as one, avoiding both mentalistic and reductionistic practices which place the past within the organism in one way or another (see Hayes, 1992b ).

Added to this, through processes of generalization, stimuli that share physical features of those that participated in spatiotemporal association conditions may also develop substitute stimulus functions. For example, a coffee shop that is physically similar to the coffee shop you went to with your friend Peter might also substitute for Peter. Specifically, you might see Peter in the presence of a coffee shop that is physically similar to the shop you frequented with him. That is to say, substitute stimulus functions also generalize to stimuli which have never actually participated in spatiotemporal association conditions, but which are physically similar to stimuli which have, and thereby involve similar stimulus functions. This type of process may become particularly subtle, and is likely to be involved in a range of complex behaviors, including imagining and dreaming.

At this point it is important to address one potential misunderstanding with the interbehavioral perspective, specifically with respect to association conditions and the development of substitute stimulus functions. 6 We are suggesting that all stimuli which occur together in space and time, and which the organism interacts with, may develop substitute stimulus functions of one another. That is, it is possible for all stimuli to develop substitute stimulus functions of any other stimulus, given the appropriate interbehavioral history. Indeed, as an individual's interbehavioral history becomes more and more elaborate, one might imagine how all stimuli could develop substitute stimulus functions of all other stimuli, such that everything might become one, psychologically speaking. However, recall that the stimulus function←→response function interaction is always a participant in an exceptionally unique, complex, multifactored field. Indeed, Kantor stated “Each interaction is always absolutely specific. What the reacting organism and the stimulus object do in each interaction constitutes a distinctly unique relational happening” (1977, p. 38). Thus, while a specific stimulus object may indeed substitute for a wide range of things given an appropriate interbehavioral history, specific substitute stimulus functions are always actualized (or not) in a unique interbehavioral field. For example, a glass of sangria might substitute for a particular friend in a specific multifactored field (you might see your friend and remember drinking sangria together), whereas that same glass of sangria might substitute for the music of a live band in a different multifactored field (you might hear the music that was playing at a restaurant where you drank sangria in the past). As this example demonstrates, while there may be a wide range of potential substitute stimulus functions for every stimulus object, in each and every specific psychological event, particular substitute stimulus functions are actualized.

Thus far we have briefly introduced some important features of interbehavioral psychology, which we find to be particularly relevant to our understanding of observational learning. From the interbehavioral perspective, individuals observe (i.e., interact with) spatiotemporal association conditions in the environment (e.g., a child putting scrap paper in the recycling bin and this being followed by praise), such that at a later time the stimulus objects involved might substitute for the prior observation (e.g., the scrap paper might have the stimulus functions of praise in the previous observation). In other words, the scrap paper develops the stimulational properties of the observed relations; it substitutes for them. Psychologically speaking, the scrap paper is those relations (see Hayes, 1992a , 1992b ).

The role of verbal behavior must also be considered in the context of our analysis thus far. Generally speaking, one outcome of interacting with an observed relationship is being able to describe it. In other words, describing an observed relationship requires the organism to interact with it, and thus, descriptions are a particularly strong indication that the relations assumed to be observed have indeed actually been contacted. However, from our perspective verbal behavior, including rules more generally, does not explain observational learning. This is to say, whether or not the organism describes the observed relationship does not explain behavior change at a later time; however, not surprisingly, it is likely to be correlated with it, as it assures the organism has interacted with the observed relation. Moreover, to the extent that rule-statements substitute for a history of reinforcement, they may further enhance any learning by observation. Importantly, in this sense verbal behavior does not “mediate” responding. Its participation in the process of observational learning, however, seems to be worth considering. In doing so, it is important that verbal behavior not be given any causal or special sort of status. Observational learning certainly can, and does occur in the absence of verbal behavior, as is the case in animal research within this area (e.g., Biederman, Robertson, & Vanayan, 1986 ; Meyers, 1970 ; Reiss, 1972 ).

Our contention that verbal behavior not be given any causal status within the conceptualization of observational learning may seem to be at odds with a number of popular perspectives in behavior analysis. For example, a growing body of research on naming (e.g., Miguel, Petursdottir, Carr, & Michael, 2008 ), joint control (e.g., Lowenkron, 1998 ), and generalized imitation (e.g., Horne & Erjavec, 2007 ) seems to support the idea that verbal behavior is mediational. Again, as stated above, we do not deny that verbal behavior is likely to be helpful in a number of circumstances, but caution against giving it any sort of special status. That is, verbal behavior may, but importantly also may not, participate in learning from observation. In this sense, verbal behavior need not be considered “meditational.” Our perspective on this matter seems to be both parsimonious and comprehensive. That is, it does not employ any unnecessary assumptions or constructs, and accounts for observational learning that occurs with and without verbal behavior. 7

We hope we have made it clear that observational learning isn't puzzling from an interbehavioral perspective. Stimulus substitution offers a straight forward, naturalistic, and parsimonious way to conceptualize complex processes, including those involved in observational learning. Importantly, the interbehavioral perspective also avoids some shortcomings found with the behavior analytic interpretation of observational learning. In the following section we outline and address these issues specifically.

Review of the Behavior Analytic Perspective

As described earlier, the behavior analytic conceptualization of observational learning rests on the processes of generalized imitation, conditioned reinforcement, rule-governed behavior, and verbal processes more generally. From our perspective these analyses fail to fully articulate the nature of stimulation in the psychological event. Again, from the interbehavioral perspective the psychological event is always the stimulus function←→response function interaction. The generalized imitation analysis leaves us questioning the nature of the stimulus interacted with. In other words, it is not clear what the stimulus is. This problem is further underscored by the suggestion that generalized imitation involves responding in the absence of a discriminative stimulus ( Pierce & Cheney, 2008 , p. 252). Given our assumption that psychological events always involve sf ←→ rf interactions, as participants in multifactored fields, this account is problematic. The process of deriving and following self-rules leaves us in a similar situation. Again, we are left questioning the nature of the stimulus interacted with. That is, it unclear what the organism is interacting with when he/she derives a self-rule, and similarly, when he/she follows such a rule. Again, given our assumptions about the psychological event, both of these analyses require further consideration of the stimulus involved.

Added to the concerns described above, behavior analytic conceptualizations also fail to explicitly articulate the location of the stimulus. In other words, it is unclear where the stimulus interacted with is located. Failing to fully describe the nature and location of the stimulus leaves the door open for common mentalistic explanations to thrive. In the case of generalized imitation we find ourselves saying that the response is “in the repertoire” of the organism, because the stimulus is private, covert, or biological in nature (also see Hayes & Fryling, 2009 ). Alternatively, the organism may be said to “derive” or “relate” with respect to participating verbal processes. In other words, we either avoid attempting to specify the stimulus, place it within the organism, or, alternatively, suggest that it is available only to those involved in other scientific disciplines, namely biology. 8 In each of these cases, we fail to provide a thoroughly psychological account of the event we are interested in, leaving our job unfinished. As has been the case throughout history, where our work is left unfinished, both dualistic and reductionistic workers are quick to complete the job. While it may be argued that much of the contemporary work in the area of complex behavior does in fact avoid many of the concerns we have described, a failure to be explicit about these important issues can only result in long-term confusion, and a possible resurfacing of mentalistic thinking.

The behavior analytic community continues to be interested in the important processes involved in observational learning (e.g., Alvero & Austin, 2004 ; Bruzek & Thompson, 2007 ; Greer & Singer-Dudek, 2008 ; Greer, Singer-Dudek, Longano, & Zrino, 2008 ; Moore & Fisher, 2007 ; Ramirez & Rehfeldt, 2009 ; Rehfeldt, Latimore, & Stromer, 2003 ). Added to this, there are some interesting reasons to believe that this process has important clinical value when compared to other procedures (see Hayes, Kohlenberg, & Melancohn, 1989 ). What is needed is a thoroughly naturalistic conceptualization of observational learning, one that avoids all mentalism (i.e., no intermediate steps within the organism). As we have described, the interbehavioral perspective offers us just that, a clear, consistent, and thoroughly naturalistic conceptualization of observational learning. Moreover, it is one that does not require any additional constructs to explain complex processes, remaining comprehensive all the while.

It is our perspective that the position described in this paper may be integrated with contemporary research and scholarship in behavior analysis. This is especially so when we make clear distinctions between investigative constructs and events, as is advocated by interbehaviorists (see Fryling & Hayes, 2009 ; Kantor, 1957 ; Smith, 2007 ). Kantor (1958) has suggested that investigative constructs are acceptable within the context of the investigative subsystem of science, but that these constructs should not be confused with the constructions of the subject matter and philosophy more generally. That is, the constructs we employ to understand various interrelations among factors participating in psychological events should never be confused to be representations of the subject matter as a whole, as being explanatory of one another, or as having more or less causal status. For example, both operant and respondent processes can be conceptualized within the more global processes of association and subsequent outcomes of stimulus substitution. Contemporary research in behavior analysis requires us to emphasize specific aspects to the interbehavioral position, particularly with respect to the role of the context (unique multifactored fields), and the actualization of specific substitute stimulus functions. In this regard, the research on relational responding is particularly stimulating. In this line of research a multitude of historical association conditions are manipulated in unique ways, under various contextual conditions, and the development or “emergence” of a wide range of events is then tested. When these interesting outcomes are conceptualized as unique sorts of substitute stimulation, operating in historical, multifactored fields, their explanations remain wholly consistent and naturalistic. We think most contemporary research and scholarship in behavior analysis can and should be integrated with the interbehavioral perspective. Importantly, such integration might serve to coordinate the efforts of various workers in the field, and ultimately maximize on our productivity as a scientific enterprise.

The limitations of Bandura's work not withstanding, the process of learning from observation is interesting and relevant to a comprehensive analysis of behavior. Indeed, if one values such comprehensiveness, our most basic concepts and principles must be relevant to, and provide an account of observational learning. Moreover, this comprehensiveness is only valuable when it is achieved within the context of validity (internal consistency) and significance (external consistency within the greater field of the sciences; see Clayton, Hayes, & Swain, 2005 ; Kantor, 1958 ). The interbehavioral perspective is particularly valuable in this regard. Kantor's conceptualization of the psychological event, with all of its fullness, provides an avenue by which the most complex sorts of behavior, including those involved in observational learning, might be fully integrated into a natural science approach to the analysis of behavior.

Cristin Johnston is affiliated with Spectrum Center, Oakland, CA.

The term modeling is used synonymously with observation and demonstration in this context. In other words, when something has been modeled the individual has observed a demonstration of the response and factors surrounding it.

See Greer et al., 2004 for a description of related studies on peer tutoring, where it was the observation of corrections, and not simply of reinforcement, that resulted in observational learning.

Of note, the researchers acknowledged the possibility that their positive statements may not have been the most optimal reinforcers, and thus, it is possible that the modeling plus reinforcement condition would have been superior had more powerful reinforcers been used ( Bandura & McDonald, 1963 , p. 281).

The idea that rewards distract individuals from learning seems to be related to the concerns raised by Alfie Kohn (1999) .

In this literature the term learning is used to describe the individual's ability to describe observed behavior at a later time.

For example, some have criticized interbehaviorism for its “loose form of associationism” (e.g., Hayes, Barnes-Holmes, & Roche, 2001 , p. 8).

A number of socially significant behaviors involve language, and we are not questioning the interest in it for the purposes of understanding how to promote such behaviors (e.g., categorization). However, we are arguing that language not be given special status in the conceptualization of observational learning.

Here, it is important to note that even when biological factors are observed (and indeed, they increasingly are) they are never observed to be engaging in the psychological event of interest. That is to say, we can never observe the brain or any biological component of the organism engaging in the behavior we are most interested in (see Kantor, 1947 ). Confusions between what is measured and what ones says they measuring are common in science (see Kantor, 1957 ; Smith, 2007 ), and are especially likely when there is a failure to fully articulate the boundary conditions between individual scientific disciplines.

• Alvero A.M, Austin J. The effects of conducting behavioral observations on the behavior of the observer. Journal of Applied Behavior Analysis. 2004;37:457–468. doi: 10.1901/jaba.2004.37-457. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Baer D.M, Peterson R.F, Sherman J.A. The development of imitation by reinforcing similarity to a model. Journal of the Experimental Analysis of Behavior. 1967;10:405–416. doi: 10.1901/jeab.1967.10-405. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Baer D.M, Sherman J.A. Reinforcement control of generalized imitation in young children. Journal of Experimental Child Psychology. 1964;1:37–49. [ Google Scholar ]
• Bandura A. Influence of models' reinforcement contingencies on the acquisition of imitative responses. Journal of Personality and Social Psychology. 1965;1:589–595. doi: 10.1037/h0022070. [ DOI ] [ PubMed ] [ Google Scholar ]
• Bandura A. Aggression: A social learning analysis. Oxford, England: Prentice-Hall; 1973. [ Google Scholar ]
• Bandura A. Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice-Hall; 1986. [ Google Scholar ]
• Bandura A, Grusec J.E, Menlove F.L. Observational learning as a function of symbolization and incentive set. Child Development. 1966;37:499–506. [ PubMed ] [ Google Scholar ]
• Bandura A, Huston A.C. Identification as a process of incidental learning. Journal of Abnormal and Social Psychology. 1961;63:311–318. doi: 10.1037/h0040351. [ DOI ] [ PubMed ] [ Google Scholar ]
• Bandura A, Jeffrey R.W. Role of symbolic coding and rehearsal processes in observational learning. Journal of Personality and Social Psychology. 1973;26:122–130. [ Google Scholar ]
• Bandura A, McDonald F.J. Influence of social reinforcement and the behavior of models in shaping children's judgment. The Journal of Abnormal and Social Psychology. 1963;67:274–281. doi: 10.1037/h0044714. [ DOI ] [ PubMed ] [ Google Scholar ]
• Bandura A, Ross D, Ross S.A. Transmission of aggression through imitation of aggressive models. Journal of Abnormal and Social Psychology. 1961;63:575–582. doi: 10.1037/h0045925. [ DOI ] [ PubMed ] [ Google Scholar ]
• Bandura A, Ross D, Ross S.A. Vicarious reinforcement and imitative learning. Journal of Abnormal and Social Psychology. 1963;67:601–607. doi: 10.1037/h0045550. [ DOI ] [ PubMed ] [ Google Scholar ]
• Bandura A, Walters R.H. Social learning and personality development. New York: Holt Rinehart and Winston; 1963. [ Google Scholar ]
• Biederman G.B, Robertson H.A, Vanayan M. Observational learning of Two visual discriminations by pigeons: A within-subjects design. Journal of the Experimental Analysis of Behavior. 1986;46:45–49. doi: 10.1901/jeab.1986.46-45. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Bruzek J.L, Thompson R.H. Antecedent effects of observing peer play. Journal of Applied Behavior Analysis. 2007;40:327–331. doi: 10.1901/jaba.2007.102-06. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Catania A.C. Learning, interim 4th edition. Cornwall-on-Hudson, NY: Sloan Publishing; 2007. [ Google Scholar ]
• Clayton M.C, Hayes L.J, Swain M.A. The nature and value of scientific system building: The case of interbehaviorism. The Psychological Record. 2005;55:335–359. [ Google Scholar ]
• Cooper J.O, Heron T.E, Heward W.L. Applied behavior analysis (2nd ed.) Upper Saddle River, NJ: Pearson; 2007. [ Google Scholar ]
• Fryling M.J, Hayes L.J. Psychological constructs and events: An alliance with Smith. The Psychological Record. 2009;59:133–142. [ Google Scholar ]
• Gladstone B.W, Cooley J. Behavioral similarity as a reinforcer for preschool children. Journal of the Experimental Analysis of Behavior. 1975;23:357–368. doi: 10.1901/jeab.1975.23-357. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Greer R.D, Keohane D, Meincke K, Gautreaux G, Pereira J.A, Chavez-Brown M, Yuan L. Key instructional components of effective peer tutoring for tutors, tutees, and peer observers. In: Moran D.J, Malott R.W, editors. Evidence-based educational methods. New York: Elsevier/Academic Press; 2004. pp. 295–334. (Eds.) [ Google Scholar ]
• Greer R.D, Ross D.E. Verbal behavior analysis: Inducing and expanding complex communication in children with severe language delays. Boston: Allyn & Bacon; 2008. [ Google Scholar ]
• Greer R.D, Singer-Dudek J. The emergence of conditioned reinforcement from observation. Journal of the Experimental Analysis of Behavior. 2008;89:15–29. doi: 10.1901/jeab.2008.89-15. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Greer R.D, Singer-Dudek J, Gautreaux G. Observational learning. International Journal of Psychology. 2006;41(6):486–499. [ Google Scholar ]
• Greer R.D, Singer-Dudek J, Longano J, Zrino M. The emergence of praise as conditioned reinforcement as a function of observation in preschool and school age children. Revista Mexicana de Psicologia. 2008;25(1):5–26. [ Google Scholar ]
• Greer R.D, Speckman J. The integration of speaker and listener responses: A theory of verbal development. The Psychological Record. 2009;59:449–488. [ Google Scholar ]
• Hayes L.J. Equivalence as process. In: Hayes S.C, Hayes L.J, editors. Understanding verbal relations. Reno, NV: Context Press; 1992a. pp. 97–108. (Eds.) [ Google Scholar ]
• Hayes L.J. The psychological present. The Behavior Analyst. 1992b;15:139–145. doi: 10.1007/BF03392596. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Hayes L.J, Fryling M.J. Overcoming the pseudo-problem of privacy in the analysis of behavior. Behavior & Philosophy. 2009;37:39–57. [ Google Scholar ]
• Hayes S.C, Barnes-Holmes D, Roche B, editors. Relational Frame Theory: A post-Skinnerian account of human language and cognition. New York: Plenum; 2001. (Eds.) [ DOI ] [ PubMed ] [ Google Scholar ]
• Hayes S.C, Kohlenberg B.S, Melancon S.M. Avoiding and altering rule-control as a strategy of clinical intervention. In: Hayes S.C, editor. Rule-governed behavior: Cognition, contingencies, and instructional control. New York: Plenum; 1989. pp. 359–385. (Ed.) [ Google Scholar ]
• Hayes S.C, Zettle R.D, Rosenfarb I. Rule following. In: Hayes S.C, editor. Rule-governed behavior: Cognition, contingencies, and instructional control. New York: Plenum; 1989. pp. 191–220. (Ed.) [ Google Scholar ]
• Horne P.J, Erjavec M. Do infants show generalized imitation of gestures. Journal of the Experimental Analysis of Behavior. 2007;87:63–87. doi: 10.1901/jeab.2007.11-06. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Horne P.J, Lowe C.F. On the origins of naming and other symbolic behavior. Journal of the Experimental Analysis of Behavior. 1996;65:185–241. doi: 10.1901/jeab.1996.65-185. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Kantor J.R. Association as a fundamental process of objective psychology. The Psychological Review. 1921;28(6):385–424. [ Google Scholar ]
• Kantor J.R. Principles of psychology (I) Chicago: The Principia Press; 1924. [ Google Scholar ]
• Kantor J.R. Problems of physiological psychology. Chicago: The Principia Press; 1947. [ Google Scholar ]
• Kantor J.R. Constructs and events in psychology: Philosophy: Banished and recalled. The Psychological Record. 1957;7:55–60. [ Google Scholar ]
• Kantor J.R. Interbehavioral psychology. Chicago: The Principia Press; 1958. [ Google Scholar ]
• Kantor J.R. Psychological linguistics. Chicago: The Principia Press; 1977. [ Google Scholar ]
• Kohn A. Punished by rewards: The trouble with gold stars, incentive plans, A's, praise, and other bribes. New York: Houghton Mifflin; 1999. [ Google Scholar ]
• Lowenkron B. Some logical functions of joint control. Journal of the Experimental Analysis of Behavior. 1998;69:327–354. doi: 10.1901/jeab.1998.69-327. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Meyers W.A. Observational learning in monkeys. Journal of the Experimental Analysis of Behavior. 1970;14:225–235. doi: 10.1901/jeab.1970.14-225. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Miguel C.F, Petursdottir A.I, Carr J.E, Michael J. The role of naming in stimulus categorization by preschool children. Journal of the Experimental Analysis of Behavior. 2008;89:383–405. doi: 10.1901/jeab.2008-89-383. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Moore J.W, Fisher W.W. The effects of videotape modeling on staff Acquisition of functional analysis methodology. Journal of Applied Behavior Analysis. 2007;40:197–202. doi: 10.1901/jaba.2007.24-06. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Morris E.K, Higgins S.T, Bickel W.K. The influence of Kantor's interbehavioral psychology on behavior analysis. The Behavior Analyst. 1982;5:158–173. doi: 10.1007/BF03392384. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Parrott L.J. Similarities and differences between Skinner's radical behaviorism and Kantor's interbehaviorism. Mexican Journal of Behavior Analysis. 1983a;9(2):95–115. [ Google Scholar ]
• Parrott L.J. Systemic foundations for the concept of “private events.”. In: Smith N.W, Mountjoy P.T, Ruben D.H, editors. Reassessment in psychology: The interbehavioral alternative. Lanham, MD: University Press of America; 1983b. pp. 251–268. (Eds.) [ Google Scholar ]
• Parrott L.J. On the role of postulation in the analysis of inapparent events. In: Reese H.W, Parrott L.J, editors. Behavior science: Philosophical, methodological, and empirical advances. Hillsdale, NJ: L. Erlbaum Associates; 1986. pp. 35–60. (Eds.) [ Google Scholar ]
• Pear J.J. The science of learning. Philadelphia, PA: Psychology Press; 2001. [ Google Scholar ]
• Pierce D.W, Cheney C.D. Behavior analysis and learning (4th ed.) New York: Psychology Press; 2008. [ Google Scholar ]
• Poppen R.L. Some clinical implications of rule-governed behavior. In: Hayes S.C, editor. Rule-governed behavior: Cognition, contingencies, and instructional control. New York: Plenum; 1989. pp. 325–357. (Ed.) [ Google Scholar ]
• Ramirez J, Rehfeldt R.A. Observational learning and the emergence of symmetry relations in teaching Spanish vocabulary words to typically developing children. Journal of Applied Behavior Analysis. 2009;42:801–805. doi: 10.1901/jaba.2009.42-801. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rehfeldt R.A, Latimore D, Stromer R. Observational learning and the formation of classes of reading skills by individuals with autism and other developmental disabilities. Research in Developmental Disabilities. 2003;24:333–358. doi: 10.1016/s0891-4222(03)00059-3. [ DOI ] [ PubMed ] [ Google Scholar ]
• Reiss D. Vicarious conditioned acceleration: Successful observational learning of an aversive Pavolovian stimulus contingency. Journal of the Experimental Analysis of Behavior. 1972;18:181–186. doi: 10.1901/jeab.1972.18-181. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Rosales-Ruiz J, Baer D.M. Behavioral cusps: A developmental and pragmatic concept for behavior analysis. Journal of Applied Behavior Analysis. 1997;30:533–544. doi: 10.1901/jaba.1997.30-533. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ]
• Skinner B.F. Science and human behavior. New York: Free Press; 1953. [ Google Scholar ]
• Skinner B.F. Verbal behavior. New York: Appleton-Century-Crofts; 1957. [ Google Scholar ]
• Skinner B.F. About behaviorism. New York: Knopf; 1974. [ Google Scholar ]
• Smith N.W. The interbehavioral field. In: Midgley B.D, Morris E.K, editors. Modern perspectives on J. R. Kantor and interbehaviorism. Reno, NV: Context Press; 2006. pp. 87–110. (Eds.) [ Google Scholar ]
• Smith N.W. Events and constructs. The Psychological Record. 2007;57:169–186. [ Google Scholar ]
• View on publisher site
• PDF (115.2 KB)
• Collections

Similar articles

Cited by other articles, links to ncbi databases.

• Download .nbib .nbib
• Format: AMA APA MLA NLM

Add to Collections

Kolb’s Learning Styles and Experiential Learning Cycle

Saul McLeod, PhD

Editor-in-Chief for Simply Psychology

BSc (Hons) Psychology, MRes, PhD, University of Manchester

Saul McLeod, PhD., is a qualified psychology teacher with over 18 years of experience in further and higher education. He has been published in peer-reviewed journals, including the Journal of Clinical Psychology.

Learn about our Editorial Process

Olivia Guy-Evans, MSc

Associate Editor for Simply Psychology

BSc (Hons) Psychology, MSc Psychology of Education

Olivia Guy-Evans is a writer and associate editor for Simply Psychology. She has previously worked in healthcare and educational sectors.

On This Page:

David Kolb published his learning styles model in 1984, from which he developed his learning style inventory.

Kolb’s experiential learning theory works on two levels: a four-stage learning cycle and four separate learning styles. Much of Kolb’s theory concerns the learner’s internal cognitive processes.

Kolb states that learning involves the acquisition of abstract concepts that can be applied flexibly in a range of situations. In Kolb’s theory, the impetus for the development of new concepts is provided by new experiences.

“Learning is the process whereby knowledge is created through the transformation of experience” (Kolb, 1984, p. 38).

The Experiential Learning Cycle

Kolb’s experiential learning style theory is typically represented by a four-stage learning cycle in which the learner “touches all the bases”:

The terms “Reflective Cycle” and “Experiential Learning Cycle” are often used interchangeably when referring to this four-stage learning process. The main idea behind both terms is that effective learning occurs through a continuous cycle of experience, reflection, conceptualization, and experimentation.

• Concrete Experience – the learner encounters a concrete experience. This might be a new experience or situation, or a reinterpretation of existing experience in the light of new concepts.
• Reflective Observation of the New Experience – the learner reflects on the new experience in the light of their existing knowledge. Of particular importance are any inconsistencies between experience and understanding.
• Abstract Conceptualization – reflection gives rise to a new idea, or a modification of an existing abstract concept (the person has learned from their experience).
• Active Experimentation – the newly created or modified concepts give rise to experimentation. The learner applies their idea(s) to the world around them to see what happens.

Effective learning is seen when a person progresses through a cycle of four stages: of (1) having a concrete experience followed by (2) observation of and reflection on that experience which leads to (3) the formation of abstract concepts (analysis) and generalizations (conclusions) which are then (4) used to test a hypothesis in future situations, resulting in new experiences.

Kolb (1984) views learning as an integrated process, with each stage mutually supporting and feeding into the next. It is possible to enter the cycle at any stage and follow it through its logical sequence.

However, effective learning only occurs when a learner can execute all four stages of the model. Therefore, no one stage of the cycle is effective as a learning procedure on its own.

The process of going through the cycle results in the formation of increasingly complex and abstract ‘mental models’ of whatever the learner is learning about.

Learning Styles

Kolb’s learning theory (1984) sets out four distinct learning styles, which are based on a four-stage learning cycle (see above). Kolb explains that different people naturally prefer a certain single different learning style.

Various factors influence a person’s preferred style. For example, social environment, educational experiences, or the basic cognitive structure of the individual.

Whatever influences the choice of style, the learning style preference itself is actually the product of two pairs of variables, or two separate “choices” that we make, which Kolb presented as lines of an axis, each with “conflicting” modes at either end.

A typical presentation of Kolb’s two continuums is that the east-west axis is called the Processing Continuum (how we approach a task), and the north-south axis is called the Perception Continuum (our emotional response, or how we think or feel about it).

Kolb believed that we cannot perform both variables on a single axis simultaneously (e.g., think and feel). Our learning style is a product of these two choice decisions.

It’s often easier to see the construction of Kolb’s learning styles in terms of a two-by-two matrix. Each learning style represents a combination of two preferred styles.

The matrix also highlights Kolb’s terminology for the four learning styles; diverging, assimilating, and converging, accommodating:

Knowing a person’s (and your own) learning style enables learning to be orientated according to the preferred method.

That said, everyone responds to and needs the stimulus of all types of learning styles to one extent or another – it’s a matter of using emphasis that fits best with the given situation and a person’s learning style preferences.

Here are brief descriptions of the four Kolb learning styles:

Diverging (feeling and watching – CE/RO)

These people are able to look at things from different perspectives. They are sensitive. They prefer to watch rather than do, tending to gather information and use imagination to solve problems. They are best at viewing concrete situations from several different viewpoints.

Kolb called this style “diverging” because these people perform better in situations that require ideas-generation, for example, brainstorming. People with a diverging learning style have broad cultural interests and like to gather information.

They are interested in people, tend to be imaginative and emotional, and tend to be strong in the arts. People with the diverging style prefer to work in groups, to listen with an open mind and to receive personal feedback.

Assimilating (watching and thinking – AC/RO)

The assimilating learning preference involves a concise, logical approach. Ideas and concepts are more important than people.

These people require good, clear explanations rather than a practical opportunity. They excel at understanding wide-ranging information and organizing it in a clear, logical format.

People with an assimilating learning style are less focused on people and more interested in ideas and abstract concepts.  People with this style are more attracted to logically sound theories than approaches based on practical value.

This learning style is important for effectiveness in information and science careers. In formal learning situations, people with this style prefer readings, lectures, exploring analytical models, and having time to think things through.

Converging (doing and thinking – AC/AE)

People with a converging learning style can solve problems and will use their learning to find solutions to practical issues. They prefer technical tasks, and are less concerned with people and interpersonal aspects.

People with a converging learning style are best at finding practical uses for ideas and theories. They can solve problems and make decisions by finding solutions to questions and problems.

People with a converging learning style are more attracted to technical tasks and problems than social or interpersonal issues. A converging learning style enables specialist and technology abilities.

People with a converging style like to experiment with new ideas, to simulate, and to work with practical applications.

Accommodating (doing and feeling – CE/AE)

The Accommodating learning style is “hands-on,” and relies on intuition rather than logic. These people use other people’s analysis, and prefer to take a practical, experiential approach. They are attracted to new challenges and experiences, and to carrying out plans.

They commonly act on “gut” instinct rather than logical analysis. People with an accommodating learning style will tend to rely on others for information than carry out their own analysis. This learning style is prevalent within the general population.

Educational Implications

Both Kolb’s (1984) learning stages and the cycle could be used by teachers to critically evaluate the learning provision typically available to students, and to develop more appropriate learning opportunities.

Educators should ensure that activities are designed and carried out in ways that offer each learner the chance to engage in the manner that suits them best.

Also, individuals can be helped to learn more effectively by the identification of their lesser preferred learning styles and the strengthening of these through the application of the experiential learning cycle.

Ideally, activities and material should be developed in ways that draw on abilities from each stage of the experiential learning cycle and take the students through the whole process in sequence.

Kolb, D. A. (1976). The Learning Style Inventory: Technical Manual . Boston, MA: McBer.

Kolb, D.A. (1981). Learning styles and disciplinary differences, in: A.W. Chickering (Ed.) The Modern American College (pp. 232–255). San Francisco, LA: Jossey-Bass.

Kolb, D. A. (1984). Experiential learning: Experience as the source of learning and development (Vol. 1). Englewood Cliffs, NJ: Prentice-Hall.

Kolb, D. A., & Fry, R. (1975). Toward an applied theory of experiential learning. In C. Cooper (Ed.), Studies of group process (pp. 33–57). New York: Wiley.

Kolb, D. A., Rubin, I. M., & McIntyre, J. M. (1984). Organizational psychology: readings on human behavior in organizations . Englewood Cliffs, NJ: Prentice-Hall.

Further Reading

• How to Write a Psychology Essay
• David Kolb’s Website
• Pashler, H., McDaniel, M., Rohrer, D., & Bjork, R. (2008). Learning styles: Concepts and evidence. Psychological science in the public interest, 9(3) , 105-119.
• What? So What? Now What? Reflective Model

You are using an outdated browser. Please upgrade your browser to improve your experience.

Health & Nursing

Courses and certificates.

• Bachelor's Degrees
• View all Business Bachelor's Degrees
• Business Management – B.S. Business Administration
• Healthcare Administration – B.S.
• Human Resource Management – B.S. Business Administration
• Information Technology Management – B.S. Business Administration
• Marketing – B.S.
• Accounting – B.S. Business Administration
• Finance – B.S.
• Supply Chain and Operations Management – B.S.
• Communications – B.S.
• User Experience Design – B.S.
• Accelerated Information Technology Bachelor's and Master's Degree (from the School of Technology)
• Health Information Management – B.S. (from the Leavitt School of Health)
• View all Business Degrees

Master's Degrees

• View all Business Master's Degrees
• Master of Business Administration (MBA)
• MBA Information Technology Management
• MBA Healthcare Management
• Management and Leadership – M.S.
• Accounting – M.S.
• Marketing – M.S.
• Human Resource Management – M.S.
• Master of Healthcare Administration (from the Leavitt School of Health)
• Data Analytics – M.S. (from the School of Technology)
• Information Technology Management – M.S. (from the School of Technology)
• Education Technology and Instructional Design – M.Ed. (from the School of Education)

Certificates

• Supply Chain
• Accounting Fundamentals
• Digital Marketing and E-Commerce

Bachelor's Preparing For Licensure

• View all Education Bachelor's Degrees
• Elementary Education – B.A.
• Special Education and Elementary Education (Dual Licensure) – B.A.
• Special Education (Mild-to-Moderate) – B.A.
• Mathematics Education (Secondary)– B.S.
• Science Education (Secondary Chemistry) – B.S.
• Science Education (Secondary Physics) – B.S.
• Science Education (Secondary Biological Sciences) – B.S.
• Science Education (Secondary Earth Science)– B.S.
• View all Education Degrees

Bachelor of Arts in Education Degrees

• Educational Studies – B.A.

Master of Science in Education Degrees

• View all Education Master's Degrees
• Curriculum and Instruction – M.S.
• Educational Leadership – M.S.
• Education Technology and Instructional Design – M.Ed.

Master's Preparing for Licensure

• Teaching, Elementary Education – M.A.
• Teaching, English Education (Secondary) – M.A.
• Teaching, Mathematics Education (Secondary) – M.A.
• Teaching, Science Education (Secondary) – M.A.
• Teaching, Special Education (K-12) – M.A.

Licensure Information

• State Teaching Licensure Information

Master's Degrees for Teachers

• Mathematics Education (K-6) – M.A.
• Mathematics Education (Middle Grade) – M.A.
• Mathematics Education (Secondary) – M.A.
• English Language Learning (PreK-12) – M.A.
• Endorsement Preparation Program, English Language Learning (PreK-12)
• Science Education (Middle Grades) – M.A.
• Science Education (Secondary Chemistry) – M.A.
• Science Education (Secondary Physics) – M.A.
• Science Education (Secondary Biological Sciences) – M.A.
• Science Education (Secondary Earth Science)– M.A.
• View all Technology Bachelor's Degrees
• Cloud Computing – B.S.
• Computer Science – B.S.
• Cybersecurity and Information Assurance – B.S.
• Data Analytics – B.S.
• Information Technology – B.S.
• Network Engineering and Security – B.S.
• Software Engineering – B.S.
• Accelerated Information Technology Bachelor's and Master's Degree
• Information Technology Management – B.S. Business Administration (from the School of Business)
• User Experience Design – B.S. (from the School of Business)
• View all Technology Master's Degrees
• Cybersecurity and Information Assurance – M.S.
• Data Analytics – M.S.
• Information Technology Management – M.S.
• MBA Information Technology Management (from the School of Business)
• Full Stack Engineering
• Front End Web Developer with CodeSignal
• Java Developer with CodeSignal
• ServiceNow Application Developer
• Data Analytics Skills
• Artificial Intelligence (Single Course)

3rd Party Certifications

• IT Certifications Included in WGU Degrees
• View all Technology Degrees
• View all Health & Nursing Bachelor's Degrees
• Nursing (RN-to-BSN online) – B.S.
• Nursing (Prelicensure) – B.S. (Available in select states)
• Health Information Management – B.S.
• Health and Human Services – B.S.
• Psychology – B.S.
• Health Science – B.S.
• Public Health – B.S.
• Healthcare Administration – B.S. (from the School of Business)
• Nursing Leadership
• Nursing Education—Post-Master's Certificate
• Nursing Leadership and Management—Post-Master's Certificate
• Family Nurse Practitioner—Post-Master's Certificate
• Psychiatric Mental Health Nurse Practitioner —Post-Master's Certificate
• View all Health & Nursing Degrees
• View all Nursing & Health Master's Degrees
• Nursing – Education (BSN-to-MSN Program) – M.S.
• Nursing – Leadership and Management (BSN-to-MSN Program) – M.S.
• Nursing – Nursing Informatics (BSN-to-MSN Program) – M.S.
• Nursing – Family Nurse Practitioner (BSN-to-MSN Program) – M.S. (Available in select states)
• Nursing – Psychiatric Mental Health Nurse Practitioner (BSN-to-MSN Program) – M.S. (Available in select states)
• Nursing – Education (RN-to-MSN Program) – M.S.
• Nursing – Leadership and Management (RN-to-MSN Program) – M.S.
• Nursing – Nursing Informatics (RN-to-MSN Program) – M.S.
• Master of Healthcare Administration
• Master of Public Health
• MBA Healthcare Management (from the School of Business)
• Business Leadership (with the School of Business)
• Supply Chain (with the School of Business)
• Accounting Fundamentals (with the School of Business)
• Digital Marketing and E-Commerce (with the School of Business)
• Java Developer (with CodeSignal and the School of Technology)
• Front End Web Developer (with CodeSignal and the School of Technology)
• Full Stack Engineering (with the School of Technology)
• ServiceNow Application Developer (with the School of Technology)
• Data Analytics Skills (with the School of Technology)
• Artificial Intelligence (Single Course with the School of Technology)
• Nursing Leadership (with the Leavitt School of Health)
• Single Courses

Apply for Admission

Admission requirements.

• New Students
• WGU Returning Graduates
• WGU Readmission
• Enrollment Checklist
• Accessibility
• Accommodation Request
• School of Education Admission Requirements
• School of Business Admission Requirements
• School of Technology Admission Requirements
• Leavitt School of Health Admission Requirements

Additional Requirements

• Computer Requirements
• No Standardized Testing
• Clinical and Student Teaching Information

Transferring

• FAQs about Transferring
• Transfer to WGU
• Transferrable Certifications
• Request WGU Transcripts
• International Transfer Credit
• Tuition and Fees
• Financial Aid
• Scholarships

Other Ways to Pay for School

• Tuition—School of Business
• Tuition—School of Education
• Tuition—School of Technology
• Tuition—Leavitt School of Health
• Your Financial Obligations
• Tuition Comparison
• Applying for Financial Aid
• State Grants
• Consumer Information Guide
• Responsible Borrowing Initiative
• Higher Education Relief Fund

FAFSA Support

• Net Price Calculator
• FAFSA Simplification
• See All Scholarships
• Military Scholarships
• State Scholarships
• Scholarship FAQs

Payment Options

• Payment Plans
• Corporate Reimbursement
• Current Student Hardship Assistance
• Military Tuition Assistance

WGU Experience

• How You'll Learn
• Scheduling/Assessments
• Accreditation
• Student Support/Faculty
• Military Students
• Part-Time Options
• Virtual Military Education Resource Center
• Student Outcomes
• Return on Investment
• Students and Gradutes
• Career Growth
• Student Resources
• Communities
• Testimonials
• Career Guides
• Skills Guides
• Online Degrees
• All Degrees
• Explore Your Options

Admissions & Transfers

• Admissions Overview

Tuition & Financial Aid

Student Success

• Prospective Students
• Current Students
• Military and Veterans
• Commencement
• Careers at WGU
• Advancement & Giving
• Partnering with WGU

Experiential Learning Theory

• Professional Development
• See More Tags

What is experiential learning?

Learning by doing. This is the basis for the experiential learning theory. Experiential learning focuses on the idea that the best ways to learn things is by actually having experiences. Those experiences then stick out in your mind and help you retain information and remember facts. 

For teachers, creating opportunities for students to have experiences based on the things they are learning about is key. Teachers can help create environments where students can learn and have experiences at the same time. 

If you’re a current teacher, or studying to become one, it’s important to get a degree that will give you qualifications and knowledge for your career, and help prepare you to be licensed. Additionally, it’s key to understand how different students learn and understand how different learning theories impact education. Teachers who understand learning theories can better optimize their classroom and help more students learn in ways that work for them. Being a successful teacher means focusing on how best to help students succeed. 

Learn more about the experiential learning theory and how teachers can use it to help their students. 

Kolb’s experiential learning theory

David Kolb is best known for his work on the experiential learning theory or ELT. Kolb published this model in 1984, getting his influence from other great theorists including John Dewey, Kurt Lewin, and Jean Piaget. The experiential learning theory works in four stages—concrete learning, reflective observation, abstract conceptualization, and active experimentation. The first two stages of the cycle involve grasping an experience, the second two focus on transforming an experience. Kolb argues that effective learning is seen as the learner goes through the cycle, and that they can enter into the cycle at any time.

Concrete learning is when a learner gets a new experience, or interprets a past experience in a new way. 

Reflective observation comes next, where the learner reflects on their experience personally. They use the lens of their experience and understanding to reflect on what this experience means.

Abstract conceptualization happens as the learner forms new ideas or adjusts their thinking based on the experience and their reflection about it.

Active experimentation is where the learner applies the new ideas to the world around them, to see if there are any modifications to be made. This process can happen over a short period of time, or over a long span of time. 

Kolb went on to explain that learners will have their own preferences for how they enter the cycle of experiential learning, and that these preferences boil down to a learning cycle.

Kolb's experiential learning cycle model.

The experiential learning cycle rests on the idea that each person has a specific type of learning tendencies, and they are thus dominant in certain stages of experiential learning. For example, some learners will be more dominant in concrete learning and reflective observation, while others will be dominant in abstract conceptualization and active experimentation. 

The four learning styles are:

Diverging. The diverging learning style is full of learners who look at things with a unique perspective. They want to watch instead of do, and they also have a strong capacity to imagine. These learners usually prefer to work in groups, have broad interests in cultures and people, and more. They usually focus on concrete learning and reflective observation, wanting to observe and see the situation before diving in. 

Assimilating. This learning style involves learners getting clear information. These learners prefer concepts and abstracts to people, and explore using analytic models. These learners focus on abstract conceptualization and reflective observation in the experiential learning style.

Converging. Converging learners solve problems. They apply what they’ve learned to practical issues, and prefer technical tasks. They are also known to experiment with new ideas, and their learning focuses on abstract conceptualization and active experimentation.

Accommodating: These learners prefer practicality. They enjoy new challenges and use intuition to help solve problems. These learners utilize concrete learning and active experimentation when they learn.

Experiential learning examples.

There are many ways that experiential learning is used every day. Some examples include:

Going to the zoo to learn about animals through observation, instead of reading about them.

Growing a garden to learn about photosynthesis instead of watching a movie about it.

Hoping on a bicycle to try and learn to ride, instead of listening to your parent explain the concept

Benefits of experiential learning.

There are many benefits of experiential learning for teachers and students, including:

Opportunity to immediately apply knowledge. Experiential learning can allow students to immediately apply things they are learning to real-world experiences. This helps them retain the information better.

Promotion of teamwork. Experiential learning often involves working in a team, so learning in this setting allows students to practice teamwork.

Improved motivation. Students are more motivated and excited about learning in experiential settings. Experiments are exciting and fun for students, and they will be passionate about learning.

Opportunity for reflection. Students using the experiential model are able to spend time reflecting about what they are experiencing and learning. This is valuable as they are able to better retain information when they can think about what’s happening to them.

Real world practice. Students can greatly benefit from learning that helps them prepare for the real world. Experiential learning is focused on using real situations to help students learn, so they are then better prepared for their future.

Experiential learning activities to include in the classroom.

It’s important for current and aspiring teachers to work to include experiential learning opportunities in their classroom. There are many ways teachers can work to include these learning activities in their class including:

Field trips

Art projects

Science experiments

Mock cities and trials

Role playing

Reflection and journaling

Internship opportunities

Interactive classroom games

Students can greatly benefit from experiential learning inside their classroom. If you’re a teacher or studying to become one , this learning theory can help  you connect with your students more effectively. Utilizing projects and experiences inside the classroom will help students learn more effectively and enjoy their learning experiences. 

Ready to Start Your Journey?

HEALTH & NURSING

Recommended Articles

Take a look at other articles from WGU. Our articles feature information on a wide variety of subjects, written with the help of subject matter experts and researchers who are well-versed in their industries. This allows us to provide articles with interesting, relevant, and accurate information. 

{{item.date}}

{{item.preTitleTag}}

{{item.title}}

The university, for students.

• Student Portal
• Alumni Services

Most Visited Links

• Business Programs
• Student Experience
• Diversity, Equity, and Inclusion
• Student Communities

Observational Experiential Learning: Theoretical Support for Observer Roles in Health Care Simulation

• PMID: 31945168
• DOI: 10.3928/01484834-20191223-03

Background: Confusion remains about the use of the observer role in simulation. Observational learning is an emerging form of brain-based learning that is applicable to experiential learning and simulation, warranting the further exploration of theoretical foundations. This article describes how observational experiential learning theoretically supports the use of observer roles in simulation.

Method: Constructs and concepts from experiential learning and observational learning theories were explored in tandem with brain-based learning evidence from different disciplines.

Results: Observational experiential learning was developed by merging these theories together in simulation and debriefing to support both observer and participant roles for learning outcomes.

Conclusion: Observational experiential learning incorporates experiential learning, social learning, and social cognitive theories to support the use of the observer role. Educators should consider strategies to foster attention and motivation through prebriefing, debriefing, and observational brain-based learning protocols. [J Nurs Educ. 2020;59(1):7-14.].

Copyright 2020, SLACK Incorporated.

• Education, Nursing / methods*
• Models, Educational*
• Observation*
• Patient Simulation*
• Problem-Based Learning / methods*

Observation Versus Experiment: An Adequate Framework for Analysing Scientific Experimentation?

• Open access
• Published: 07 May 2016
• Volume 48 , pages 71–95, ( 2017 )

Cite this article

You have full access to this open access article

• Saira Malik 1  

16k Accesses

12 Citations

4 Altmetric

Explore all metrics

Observation and experiment as categories for analysing scientific practice have a long pedigree in writings on science. There has, however, been little attempt to delineate observation and experiment with respect to analysing scientific practice; in particular, scientific experimentation, in a systematic manner. Someone who has presented a systematic account of observation and experiment as categories for analysing scientific experimentation is Ian Hacking. In this paper, I present a detailed analysis of Hacking’s observation versus experiment account. Using a range of cases from various fields of scientific enquiry, I argue that the observation versus experiment account is not an adequate framework for delineating scientific experimentation in a systematic manner.

Similar content being viewed by others

Why Thought Experiments do have a Life of Their Own: Defending the Autonomy of Thought Experimentation Method

Science and Experiment

Generative and Demonstrative Experiments

Avoid common mistakes on your manuscript.

1 Introduction

“They [the Greeks] observed but did not experiment”. Footnote 1

This quote from Desmond Lee, the famous translator of Aristotle’s scientific works, identifies the two categories that form the bedrock of modern scientific practice. Footnote 2

This quote also identifies well one of the principal markers used to delineate modernity. It is now a well-worn axiom that what distinguishes Western modernity—and implicit in this, Western hegemony—is the phenomenon of the Scientific Revolution in the West. What marks out the scientific practices of the Scientific Revolution and thereafter in the West from other scientific enterprises in the past—in the popular imagination—is supposed to be ‘experiment’. It is posited that this is the hallmark of the Scientific Revolution—what went before is ‘observation’. Footnote 3 What Desmond Lee appears to be doing here is setting up a binary of ‘observation versus experiment’ rather than ‘observation and experiment’. Many decades later, the distinguished philosopher of science Ian Hacking echoes Lee by positing, “Observation and experiment are not one thing [,] nor even opposite poles of a smooth continuum”. Footnote 4 The casting of experiment in opposition to observation as Hacking and Lee do, rather than in addition to it, is a very modern turn.

Experiment—as experimentum (and its cognates) in Latin and empeiria or peira in Greek—has a continuity of usage as a category for scientific learning finding its genesis in the works of Hippocrates, Aristotle and Pliny (Pomata 2011 , 45–46). Footnote 5 Observation, as a scientific category, does not enjoy the same continuity as the essays in Histories of Scientific Observation , edited by Lorraine Daston and Elizabeth Lunbeck, show. Footnote 6

In fact, what may be thought of as scientific observational practices were subsumed under a myriad of terms in Latin: experientia , experimentum , contemplatio , consideratio —and the least used— observatio . Where Greek is concerned, there is no equivalent for observation ( observatio )— teresis —in the scientific canon of Hippocrates and Aristotle (Pomata 2011 , 45). It is only in the seventeenth century that observation and experiment as scientific categories- respectively as observatio and experimentum —become established as well as conjoined (Daston 2011 , 81–113). Despite this distinguishment, the terms remained conjoined: “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 complement of one another” (Daston 2011 , 86). The difference implied between the two then, on the eve of the nineteenth century, was that experiment implied intervention and manipulation whereas observation did not (Daston 2011 , 86). In many cases even this implied distinction was subsumed for others, including notables such as Robert Boyle and Robert Hooke, who appear to make no distinction between observation and experiment as long as both were dedicated to the cause of knowledge acquisition of the natural world (Anstey 2014 , 105).

It is in the nineteenth century that one sees observation being cast in opposition to experiment rather than in addition to it (Daston and Lunbeck 2011 , 3). During this period one can see the reconfiguration of vision insofar as it becomes detached from a referent and thus abstracted, leading to the inevitable subjectivity of the observer (Crary 1992 ). Footnote 7 This leads, as Jonathan Crary explains, to the ‘social remaking of the observer’ (Crary 2001 , 4). The observer goes from a passive receiver of the external world to an active producer of it (Crary 2001 , 95–97). This nineteenth century reconfiguration of vision and the observer has been made clear not just in Jonathan Crary’s work on the camera obscura and works of art, but also in Christoph Hoffmann’s work on scientific practices, the senses and instrumentation (astronomical, in particular) during the same period (Hoffmann 2006 ) where the author shows that any qualitative distinction between the observer and instrumentation fades away. Footnote 8

In light of the importance of observation and experiment as categories in scientific practice, particularly scientific experimentation, it is surprising that relatively little attention has been paid to them as a binary within the modern academy of philosophy of science Footnote 9 —in spite of philosophers of experiment such as Hans Radder and David Gooding calling for such. Footnote 10 An exception is Ian Hacking—in Representing and Intervening (Hacking 1983 ). Footnote 11 In this essay, I scrutinise Hacking’s account of observation and experiment in order to assess its efficacy as an adequate account for delineating scientific experimentation. I show that there are significant weaknesses in Hacking’s account when used to analyse a range of cases from different fields of scientific enquiry.

2 Hacking: Observation Versus Experiment

In Representing and Intervening (Hacking 1983 ), Ian Hacking makes a category distinction between experiment and observation. He states, ‘Observation and experiment are not one thing nor even opposite poles of a smooth continuum’ (Hacking 1983 , 173). According to Hacking, ‘Much of the discussion about observation, observation statements and observability is due to our positivist heritage’ (Hacking 1983 , 168). He thinks the need to make these distinctions at all, and to take them seriously, is a task very much confined to professional philosophers. According to Hacking, these distinctions do not worry scientists. He gives the example of Francis Bacon to show what he means (Hacking 1983 , 168–169).

Francis Bacon does not mention the term once in his discussion of the inductive sciences despite the term being in circulation during his time. Observation at this time was restricted in its use—used mainly for observations made of the heavenly bodies via telescopes. That is, the use of the term observation in the natural sciences was associated with the use of instrumentation. Instead of observation, Bacon uses the term ‘prerogative instances’. In his Novum Organum of 1620, he lists 27 different ‘prerogative instances’: these include a range of activities which today one may refer to as scientific practices: experiments, tests to distinguish between hypotheses, notable observations, some are made with devices that ‘aid the immediate actions of the senses’. The latter includes microscopes as well as telescopes, rods, astrolabes and similar devices. He calls devices that aid the senses ‘evoking devices’, devices that ‘reduce the non-sensible to the sensible; that is, make manifest, things not directly perceptible, by means of others which are’ (Hacking 1983 , 168–169).

Bacon recognises the difference between what is directly perceptible and that which is hidden from the senses and needs to be ‘evoked’. He recognises it and does not give it much significance—for Bacon, the difference is not important. For Bacon, there is no difference between directly seeing the sun overhead at noon and seeing a planet via a telescope at night.

Hacking states it is only later, in the nineteenth century, that the difference between things that are directly perceptible and those that are hidden from the senses and have to be ‘evoked’ becomes important. It becomes important because the very notion of ‘seeing’ undergoes a transformation. In the nineteenth century, ‘to see’ is to see the surface—and only the surface—and all knowledge must be derived via this way. This marks the beginnings of positivism and phenomenology. Positivism needs to distinguish between inference and seeing with the unaided eye (Hacking 1983 , 169). Thus, unlike Bacon, there is a difference between seeing the sun overhead at noon and seeing a planet at night via a telescope. For the positivist, the planet seen via a telescope can only be inferred—it is not an observation. According to Hacking, this marks the start of the distinction made between observation and theory in the philosophy of science and is articulated by someone like Bas van Fraassen ( 1980 ). This view has come to be contested in two ways: one that emphasises the scope of observation and the other of theory. Grover Maxwell is a good exemplar of the former view (Maxwell  1962 ) while Paul Feyerabend is an example of the latter (Hacking 1983 , 172–173).

Hacking deals with Maxwell thus (Hacking 1983 , 170). Maxwell makes a historically contingent argument. He suggests that what may be unobservable at some particular time may subsequently become observable—or in Bacon’s language be ‘evoked’—with the development of adequate instrumentation and/or the expansion of the capacity of existing instrumentation. For example, in the case of visual perception, there is a continuum that starts with seeing through a vacuum, through the atmosphere, through a simple microscope and, at present, finishes with seeing through the current batch of advanced microscopes. In this way, what in previous generations would have been unobservable—and according to positivists only inferred, and thus theoretical—becomes observable with the development of appropriate instrumentation. For example, prior to Louis Pasteur, the notion of microbial entities responsible for disease was considered theoretical. However, with the advent of microscopy these entities became observable. Other examples include genes on chromosomes, cell bodies in cells and the fine structure of metals. In all these cases the entities were regarded as theoretical until the development of adequate instruments rendered them observable. For Maxwell there is no significant difference between knowledge gained directly through the senses and that gained indirectly with the aid of instrumentation—Bacon’s ‘evoking’ devices.

The second type of critique of the positivist stance is based on the notion that the distinction between observation and theory is redundant. That is, all observations—whether made directly via the senses or not—are theoretical—that is, there are no pure observations. All observations are ‘theory laden’ to coin Norwood Hanson’s term from his Patterns of Discovery (Hanson 1958 , 19). Hanson states, ‘seeing is a “theory laden” undertaking’.

Paul Feyerabend agrees with Hanson but goes even further (Hacking 1983 , 172–173). For Feyerabend, there is no difference between observation and theory. In fact, he has rejected the term ‘theory laden’ on the grounds that there can be no observation without theory. He states, ‘Nobody will deny that such distinctions [between observation and theory] can be made, but nobody will put great weight on them, for they do not play any decisive role in the business of science’ (Hacking 1983 , 173). He comments on the everyday practices of science, ‘observational reports, experimental results, “factual statements”, either contain theoretical assumptions or assert them by the manner in which they are used (Hacking 1983 , 173).

Hacking chooses to align himself with Grover Maxwell rather than Feyerabend; and is particularly scathing of Feyerabend’s [lack of] understanding of scientific practice exemplified by the statement, ‘observational reports, experimental results, “factual statements”, either contain theoretical assumptions or assert them by the manner in which they are used’. Hacking explains why using two historical examples: the work of Albert Michelson and Edward Morley along with that of William Herschel.

The work of Michelson and Morley is well known to historians of the physical sciences (Hacking 1983 , 174). It is famous because, on reflection, it refuted the existence of ‘electromagnetic aether’ and led to the establishment of the special theory of relativity. Hacking focuses on the scientific practices of Michelson and Morley and what these mean with respect to Feyerabend’s comment, ‘observational reports, experimental results, “factual statements”, either contain theoretical assumptions or assert them by the manner in which they are used’. The published ‘report’ of the experiment of 1887 was 12 pages long. The ‘observations’ made were done so for a total of a couple of hours over 4 days in July. The ‘results’ of the experiment remain controversial: Michelson believed that this work showed that the earth’s motion was independent relative to the [presumed] aether. Hacking goes on to identity the components which (in his view) contributed towards the impact of this work—in its own time and up to the 1920s. These components include, inter alia , the making and re-making of apparatus, getting the apparatus to actually work and, most importantly—knowing when the apparatus was working. Interestingly, the most important result of this work, according to Hacking, had less to do with aether and more to do with the transformation of measurement. Hacking concludes, “In short, ‘Feyerabend’s factual statements, observation reports, and experimental results’ are not even the same kind of thing. To lump them together is to make it impossible to notice anything about what goes on in experimental science” (Hacking 1983 , 174).

Hacking then proceeds to show that Feyerabend’s notion that all observations carry theory is false. Hacking uses the historical case study of William Herschel (d. 1822) as a rebuttal to Feyerabend’s, ‘all observations carry theory’.

William Herschel was an astronomer, who, in the year 1800, is attributed to have discovered radiant heat whilst conducting his astronomical work with his telescope (Hacking 1983 , 176). On using different coloured filters in his telescope, Herschel realised that different colour filters gave off different amounts of heat. Herschel, in the reporting of his work in the Philosophical Transactions of the Royal Society for the year 1800 states, “When I used some of them I felt a sensation of heat, though I had but little light, while others gave me much light with scarce any sensation of heat”. It was this incidental observation to his principal work on the sun which led Herschel in a new experimental direction and the discovery of radiant heat: that the sun emits both visible and invisible rays and that human sight is sensitive to only the visible rays. This incidental observation led him to conduct a whole series of experiments investigating the transmission, reflection and refraction of these rays (Hacking 1983 , 177). Hacking concludes, “Feyarabend says that observations reports, etc., always contain or assert theoretical assumptions. This assertion is hardly worth debating because it is obviously false” (Hacking 1983 , 174).

Thus, Hacking’s notion of observation appears to be very much aligned with Grover Maxwell, and with Francis Bacon’s ‘evoking devices’. Hacking’s anti-positivist stance on observation becomes even clearer when considering his position on observation of sub-atomic particles via indirect methods, and finally, with his view of observation of entities via a microscope.

On observation of sub-atomic particles using indirect methods, Hacking is in agreement with Dudley Shapere (Shapere 1982 ). Shapere uses the discourse of ‘observing the interior of the sun or another star’ as his starting point in his argument to show what is meant by ‘to see’ in modern science—and in the process shows how far we have journeyed along the path of Bacon’s ‘evoking devices’. Shapere analyses the solar neutrino experiment in which physicists claim that the core of the sun (or any star) can be directly observed via the detection of neutrinos. Footnote 12 Shapere shows the various layers of detection—’seeing’—involved in the ‘direct observation’ of neutrinos emitted from the core of the sun. Shapere argues that despite what appears to be a complicated series of events entailed in the detection of neutrinos, it is justifiable to term this process as ‘direct observation’ (Shapere 1982 , 492).

Hacking suggests that it is the fact that the theories underlying the detection mechanism are not entwined with the subject matter under investigation is what gives credence to the claim in the solar neutrino experiment that the “stellar core of the sun can be directly observed” (Hacking 1983 , 185). Footnote 13 For Hacking therefore what would count as an observation would include the detection of electrons in a bubble-chamber, as the theory used in the manufacture and operation of the bubble-chamber does not directly use theory about electrons. Footnote 14 For Hacking, this also holds for the use of microscopes—that is, that the theories, assumptions and norms on which microscopes are built and used (from simple light microscopes to electron-scanning and X-ray diffraction ones) are independent of the subject matter being studied (Hacking 1983 , 186–209) and for him therefore what is seen using these instruments counts as an observation.

Hacking is very much wedded to an anti-positivist stance on what constitutes observation—very much in the tradition of Francis Bacon and his “evoking devices making manifest, things which are not directly perceptible, by means of others which are”. Hacking, thus, aligns himself with working scientists for whom ‘to see’ includes detection methods ranging from the simple microscope to its X-ray diffraction and electron scanning versions—things Bacon could not have imagined when he composed his Novum Organum . What Hacking thus means by observation is detection.

Where experiment is concerned—again—Hacking appears to be in support of Bacon (Hacking 1983 , 246–250) as the following citations from Bacon show, “The secrets of nature reveal themselves more readily under the vexation of art than when they go their own way” (Hacking 1983 , 246), “shake out the folds of nature” and to “twist the lion’s tail” (Hacking 1983 , 246). Hacking says this alludes to “Bacon’s good sense” (Hacking 1983 , 250). However, for Hacking an experiment is not just a case of intervention, or ‘to twist the lion’s tail’, as it is for Bacon as we see below.

What is Hacking saying an experiment is? He stipulates very clearly—It is the “creation of phenomena” (Hacking 1983 , 220). This is made more emphatic in “Experiment is the creation of phenomena” (Hacking 1983 , 229). What does Hacking mean? First—phenomena: Hacking says that he agrees with scientists as to what is meant by phenomena,

A phenomenon is noteworthy . A phenomenon is discernable . A phenomenon is commonly an event or process of certain type that occurs regularly under defined circumstances. When we know the regularity exhibited in a phenomenon we express it in a law-like generalization. The very fact of such a regularity is sometimes called the phenomenon. (Hacking 1983 , 221)

This description fits very closely with the etymology of the (Greek) term phenomenon: a thing, an event or process that can be seen. However, phenomenon, Hacking points out, has quite a different sense in philosophy.

Phenomenon has a long history in its philosophical usage (Hacking 1983 , 220–221) quite different to its etymological roots. Phenomenon, for philosophers—both ancient and modern—has come to indicate something related to the senses. For many ancients, phenomena were in opposition to reality insofar as phenomena—perceived via the senses—were subject to change (Hacking 1983 , 221). The fact of phenomena being the subject of change led to the juxtaposition of phenomena to noumena: phenomena were only appearances of things whereas noumena were things as they actually were. Kant took up this distinction and proposed that only phenomena could be known—the noumena could not. With the advent of positivism, phenomena came to indicate sense-data—things that are “private, personal sensations” (Hacking 1983 , 221)—rendered as ‘phenomenalism’, and according to one of its principal proponents, J. S. Mill, “things are only the permanent possibilities of sensation, and that the external world is constructed out of actual and possible sense-data” (Hacking 1983 , 221).

Hacking breaks from the way philosophers have come to use the term phenomenon, and aligns himself with the scientists. He says,

My use of the word ‘phenomenon’ is like that of the physicists. It must be kept as separate as possible from the philosophers’ phenomenalism, phenomenology and private, fleeting sense-data. A phenomenon, for me, is something public, regular, possibly law-like, but perhaps exceptional. I pattern my use of the word [phenomenon] after physics and astronomy. (Hacking 1983 , 222)

Hacking illustrates what he means by using the ‘Hall effect’ from the field of modern physics as an exemplar.

Edwin (E. J.) Hall’s work on the relationship between a magnetic field and electric potential is referred to as the Hall effect (Hacking 1983 , 224–225). In the late 1870s, Hall, under the supervision of Henry Rowland at John Hopkins University, had been expanding on some of James Clerk Maxwell’s ideas from his Treatise on Electricity and Magnetism . In the Treatise Maxwell had proposed that, where a conductor carrying an electric current was under the influence of a magnetic field, the magnetic field acts on the conductor rather than the current. Hall proposed that if this were the case then there should be two possible outcomes: either the resistance of the conductor would be affected by the magnetic field or that an electric potential across the field would be produced. Hall discarded the first possibility as he failed to observe any effect by the magnetic field on the resistance of the conductor. However, the second possibility bore fruition: he was successful at measuring an electric potential across the magnetic field. He obtained an electric potential when he placed a gold leaf at right angles to the magnetic field and electric current. Hall says,

It seemed hardly safe, even then, to believe that a new phenomenon has been discovered, but now after nearly a fortnight has elapsed, and the experiment has been many times and under various circumstances successfully repeated … it is perhaps not too early to declare that the magnet does have an effect on the electric current or at least an effect on the circuit never before expressly observed or proved. (Hacking 1983 , 225)

Hacking tells us that by the 1880s it was common for physicists to call a phenomenon an effect: as in the Compton effect, Footnote 15 the Zeeman effect Footnote 16 and the photoelectric effect Footnote 17 (Hacking 1983 , 224). He states,

Phenomena and effects are in the same line of business: noteworthy discernable regularities. The words ‘phenomena’ and ‘effects’ can often serve as synonyms, yet they point in different directions. Phenomena remind us, in that semiconscious repository of language, of events that can be recorded by the gifted observer who does not intervene in the world but who watches the stars. Effects remind us of the great experiments after whom, in general, we name the effects: the men and women, the Compton and Curie, who intervened in the course of nature, to create a regularity which, at least at first, can be seen as regular (or anomalous) only against the further background of theory. (Hacking 1983 , 225)

Here Hacking starts by telling us phenomena and effects have similar aims—they yield “noteworthy discernable regularities”—albeit he draws a difference between them in so far as the kinds of activities they are: phenomena as ‘events’ noted by those who do not ‘intervene in the world’ while effects are things which “remind us of the great experiments” done by those “who intervened in the course of nature, to create a regularity…”.

If we now turn to consider what Hacking means by ‘creation’ in his stipulation of experiment—‘creation of phenomena’, we find that Hacking is conferring a very constricted meaning to creation.

Hacking (again) uses Hall’s work to illustrate what he means by creation. He says, “Hall’s effect does not exist outside of certain kinds of apparatus” (Hacking 1983 , 226). This is made more emphatic in, “Hall’s effect did not exist until, with great ingenuity, he [Hall] had discovered how to isolate, purify, create it in the laboratory” (Hacking 1983 , 226). To give even more emphasis, Hacking cites another example: the ‘Josephson effect’ referring to the work of Brian Josephson in the 1960s. Again, the example Hacking chooses is from modern physics and, in this case, concerns the subject of electrical conduction by super-conductors. Footnote 18 He says, “The Josephson effect did not exist in nature until people created the apparatus” (Hacking 1983 , 229). For Hacking, it is these effects—bounded by the apparatus in which they can be demonstrated in laboratory conditions that appear to fulfil his criteria for what qualifies as experiment.

Hacking explains what he means by his statement, “the Hall effect does not exist outside of certain kinds of apparatus” (Hacking 1983 , 226). He asks rhetorically, “Does not a current passing through a conductor, at right angles [ sic ] to a magnetic field, produce a potential, anywhere in nature?”, answering ambivalently, “Yes and no”. According to Hacking, if there were such an event in nature, which occurred in isolation of any other processes, then it could be said that the Hall effect occurs in nature. However, it is only in laboratory conditions that the Hall effect can be produced independent of any other processes. It is with this explanation that it becomes clear that, for Hacking, in order for a phenomenon to be created—it needs to be produced in isolation, or what he calls “in a pure state” (Hacking 1983 , 226).

Hacking’s commentary on the work of Edwin Hall tells us what Hacking means when he stipulates “experiment is the creation of phenomena”. What we see is that Hacking’s stipulation of experiment as ‘creation of phenomena’ becomes highly constricted because of his insistence on the fact that the phenomena under consideration needs to be produced in a ‘pure state’ or in isolation. His repeated emphasis on ‘creation’ gives emphasis on the importance of this aspect in his stipulation of experiment. The emphasis on ‘pure state’ is underlined from the highly selective way he chooses his case studies in support of his position. He cites many examples but chooses either not to deal with them in any sustained way or dismisses them—on occasion, flippantly (Hacking 1983 , 227–228). Amongst the many examples Hacking cites, he chooses to focus only on cases from modern physics, such as the Hall and Josephson effects. This appears to be a deliberate strategy on his part as the following illustrates.

Hacking introduces the medical work of Claude Bernard (published as Introduction to the Study of Experimental Medicine in 1865) as a potential case study to show the distinction between experiment and observation (Hacking 1983 , 173). Hacking states,

Consider Dr Beauchamp [ sic ] who, in the Anglo-American war of 1812 [ sic ], had the fortune to observe, over an extended period of time, the workings of the digestive tract of a man with a dreadful stomach wound. Was that an experiment or just a sequence of fateful observations in almost unique circumstances? (Hacking 1983 , 173)

In this example, Hacking not only makes a couple of errors in transposing historical details from Bernard, Footnote 19 but more importantly, considerably truncates the details of William Beaumont’s study on the digestive physiology of the human stomach. Footnote 20 Hacking finishes by choosing not to engage with this case from medical physiology, saying, “I do not want to pursue such points” (Hacking 1983 , 174).

In contrast, I think it worth pursuing this case from medical physiology, as well as some others from different fields of scientific enquiry, in order to assess how Hacking’s observation versus experiment account maps onto cases from a range of scientific experimentation.

First, returning to Beaumont’s story. William Beaumont himself believed that the work he was doing was an experimental investigation of human digestion (Beaumont 1833 , 5–6). However, more important is how this case fits with Hacking’s account as he chooses not to address this question himself. First—a very brief overview of Beaumont’s work on digestion.

Beaumont, in his capacity as a doctor, treated a patient suffering from a gunshot wound, which had caused damage to his left lung and stomach (Beaumont 1833 , 10). The patient recovered but with a very unusual outcome: the stomach lining did not heal in a uniform way. Instead it formed a fistula with an exterior valve. Beaumont used this valve as the access point in conducting a series of investigations on digestion (Beaumont 1833 , 11–23). He used a pipetting technique to both put substances into the stomach, as well as to draw them out. In this way he examined the digestion of various substances in the stomach.

If one were to use the Hacking’s criteria for experiment, the ‘creation of phenomena’—Beaumont’s work within the stomach falls short of qualifying as experiment. That is, although the digestion process qualifies as a phenomenon (a discernable change) as well as an effect (requires activity and intervention on the part of the investigator); it does not fulfil Hacking’s criteria for creation—that is, the effect is not occurring in isolation of other processes. Footnote 21 However, Beaumont goes on to perform a series of investigations looking at the digestive action of the ‘gastric juice’ in isolation (Beaumont 1833 , 73–101). Footnote 22 This series of investigations would appear to fulfil Hacking’s criteria for experiment as Beaumont sets up apparatus (however rudimentary) which gives rise to an effect—in isolation of others.

Thus, if we use Hacking’s criteria for experiment in respect to William Beaumont’s work, the outcome would be that only part of Beaumont’s work qualifies as experiment—the in vitro part. That is, the part done within the apparatus that Beaumont sets up to investigate digestion outside the stomach. The in vivo part of Beaumont’s work, that is, the work done on the stomach directly, fails to qualify as experiment as the effects occurring are not in isolation of other processes. This ignores the crucial point in William Beaumont’s investigations: the in vitro part (experiment) is contingent on the work that William Beaumont has previously done on the stomach. Beaumont would never have set up the apparatus part of his work if he had not already done the work on the stomach.

According to Hacking, thus, if Beaumont’s in vivo work is not experiment then is it a series of observations? If so, Hacking has told us that, “[o]bservation and experiment are not one thing nor even opposite poles of a smooth continuum” (Hacking 1983 , 173). This statement would imply that, in Beaumont’s case, the work done in vivo and that done in vitro are not part of ‘a smooth continuum’. This is not reasonable given that the work done in vitro is continuous with the work done in the stomach. The limitations of Hacking’s framework are not only confined to this case from physiology.

In evolutionary biology, the name of Henry Kettlewell is well known. Kettlewell’s work on moths in the 1950s was important in understanding the process of natural selection. Footnote 23 Kettlewell used three kinds of moths: typical ( Biston betularia ), intermediate ( Biston insularia ) and dark ( Biston carbonaria ). In Britain, the typical moth had been prevalent in most areas prior to industrialization. However, the proportion of the typical variety in relation to the other two types changed during the twentieth century. Kettlewell showed that this change was due to the change in colour of the landscape. Kettlewell first showed that the different kinds of moths were more or less conspicuous depending on the colour of the background on which they were settled. He did this by using volunteers to rank the degree of conspicuousness of each type of moth on different colour backgrounds. In the next stage, he put all three types of moths in a cage with different colour bark on which they could settle. He then introduced birds (predator to moths) into the cage. He found that the rate at which the moths were eaten depended on the colour of the bark on which they were settled. As three different kinds of moths were used along with different colour barks, the data analysis was very complex in this part of the study. The third part of his study was done in native conditions. Kettlewell released all three kinds of moths in both polluted (dark background) and unpolluted (lighter background) areas and tracked how many survived. This last part of the investigation depended on previously marked moths that had survived being recaptured in traps. Kettlewell showed that the dark species of moths survived better in a polluted (dark) environment than the lighter colour varieties whereas the lighter typical species survived better in the less polluted (light) environment compared with the darker varieties. He showed this was due to the colour of the landscape.

What do we see when we map Hacking’s observation versus experiment account onto this case? The observation part of the account can be done in a straightforward manner. Hacking has told us that the observation part is a source of detection—in this case the numerical values related to what kind of moth species is conspicuous on which colour bark, the numerical values related to different species surviving predation in the cage, the kind of bait used for re-capture in native conditions. However, what, according to Hacking, is the experiment part? The phenomenon under study here—natural selection—is not being produced in isolation of other processes and therefore, in Hacking’s account, cannot—or should not—be included in his category of experiment.

We see this anomalous consequence of Hacking’s account in fields of scientific enquiry other than the two (physiology, evolutionary biology) mentioned already.

In the field of study of animal behavior and psychology, the work of Harry Harlow is well known amongst those working on attachment theory. Footnote 24 Harlow did his experimental work on rhesus monkeys (macaques) during the 1950s and 60s. Footnote 25 His work on isolated infant monkeys had shown that the infants formed a close attachment to the soft materials in their cages (diapers, bedding) whereas those infants who had their mothers in the cage did not form this attachment. Harlow conducted a series of experiments to measure degrees of attachment of an infant monkey to the quality of a carer.

Eight new-born monkeys were separated from their mothers immediately after birth. Each was placed in a cage with two ‘surrogate mothers’—one surrogate was made of wire with a box face while the other surrogate was made of soft cloth with a quasi-monkey face. Milk was dispensed from each surrogate. Harlow measured the time that each infant monkey spent with each surrogate over a period of some months. He found that the infant monkeys spent more time with the cloth-covered surrogate than with the wire one. He then withdrew milk dispensation from the cloth surrogate. He found that the total time that the infants spent with the cloth surrogate was still much greater than that time spent with the wire surrogate—the infants would only go to the wire surrogate to feed when hungry—as soon as their hunger abated, they returned to the cloth covered surrogate. Harlow concluded from these particular experiments that infant monkeys had requirements (social, cognitive, emotional) beyond those of (just) nutrition (milk) in their early years.

In the field of geology, the work of Nevil Maskelyne and colleagues gave an initial indication of the density of the earth (Danson 2009 ). Footnote 26 Their work was based on the notion that a pendulum, placed near a mountain in a uniform gravitational field, would shift from the true vertical. This shift could be measured against a reference such as a fixed star and—given Newton’s proposal that force is proportional to the mass of an object—the density of the earth could be calculated. Isaac Newton himself, in the Principia had indicated that this should be possible but had discarded the idea as he believed the instrumentation of the day would not be able to detect the small changes in the shift of the pendulum. Footnote 27

Just over a century later in 1772, Maskelyne, the Royal Astronomer to George III, believed that the instrumentation at the Royal Observatory in Greenwich was up to the task that Newton had set. The French astronomer Pierre Bouguer had carried out Newton’s proposal of using a mountain in South America some decades earlier—but had not met with much success on account of numerous technical obstacles (Danson 2009 , 40–42; 97–98).

Maskelyne met with greater success at a mountain in central Scotland, Schiehallion (chosen for its symmetry). The investigation was divided into two stages. The first entailed measurement of the deflection of the pendulum with respect to positions of fixed stars for which two observatories were built—one on the north side and one on the south. The measurements taken were in the astronomical measure of arc minutes. The other stage of the investigation involved the survey of the mountain in order to measure its volume. These measurements were expressed in terms of height (feet/inches). The work took until 1778 to complete and the final density of the earth was computed to within 20 % of that calculated by Henry Cavendish some twenty odd years later using a torsion balance to measure the attraction between two lead spheres.

Staying within geology and in Scotland, James Hutton’s extensive investigations on soil erosion helped significantly shape understanding of landscape formation (Dean 1992 ). Footnote 28 Over a span of decades, Hutton made extensive surveys and measurements of various areas of Britain as well as France, Belgium and Holland. Much of this work was subsequently the starting point for Charles Lyell and Charles Darwin in their work on geology (Rudwick 2005a ). Footnote 29 The first outline of Hutton’s work was circulated as Abstract of a Dissertation Concerning the System of Earth, its Duration and Stability in 1786. His work consisted of analysis of rock strata and analysis (chemical, thermogenic) of different kinds of rock formations (granite and gneiss, sediment[ary] and volcanic [igneous] as well as the identification and recording of the frequency of the occurrence of fossils in these different rock strata. Footnote 30 The records of his results consisted of temperature readings at which different kinds of rocks changed appearance, the recording of what these changes entailed, the recording of which (if any) rock kinds reacted with different kinds of chemicals, extensive classification and frequency tabulations of fossil finds, numerous drawings of fossil finds and rock strata.

I now want to turn to physics—the principal focus of study for Hacking. In the early part of the twentieth century, Robert Millikan conducted a series of investigations to establish that the charge of the electron was quantized (had a discrete fundamental value) and occurred in situ as multiples of this value rather than a continuum as had been previously proposed by Thomas Edison, amongst others (Holton 1978 ). Footnote 31

The received narrative of Millikan’s investigations is presented as an ingenious use of the cloud chamber developed by Charles Wilson (Franklin 1986 , 216). In Wilson’s original, within a sealed container, ions act as loci around which water droplets can form. Wilson used a sealed container filled with air and water vapour at the point of condensing—a supersaturated environment—which he produced using a vacuum pump for first compressing and then expanding air inside the sealed container (‘chamber’). Any charged particle in the container containing this supersaturated mixture causes ionization as it moves. This ionization acts as loci around which vapour (‘cloud’) forms as a consequence of condensation. The movement (or fall) of this ‘cloud’ in this ‘chamber’ under gravity can be detected via a viewer (short focal distance telescope) and the visible ionization path measured (by calibrating the eyepiece of the telescope). If an electric field is applied (vertically) across the chamber (in the form of two charged plates—positive at the top and negative at the bottom with a DC voltage applied to each plate via a battery)—then the change in the rate at which the cloud moves/falls under gravity can be detected. Measurement of the velocities of the fall of the cloud under just gravity and then with a known voltage should determine the charge on the electron.

J. J. Thompson had attempted to measure the charge on the electron in this way but had tried to measure the charge of the whole cloud and had met with little success—owing in the main to practical obstacles (Goodstein 2001 , 54).

Millikan, in attempting the same as Thompson, found that applying a much greater electric field across the charged plates resulted not in the cloud being suspended, as had been predicted, but most of the cloud dispersing, leaving only a few drops suspended between the plates. Millikan deduced that working with individual droplets would overcome many of the logistical and numerical obstacles that Thompson had faced in working with a whole cloud (ibid.).

Millikan (and his graduate students) set about repeating Thompson’s work with single droplets of water but found no success as the single water drops tended to evaporate quickly, making reliable measurements impossible. They thus set about adapting Wilson’s cloud chamber, as well as Thompson’s method, over a period of some years. The appearance of simplicity in Millikan’s final investigative set-up belies the various stages it took for the investigation to mature.

The first issue they had to overcome was that of evaporation. They did this by replacing water drops with substances whose evaporation rate would have a negligible effect on their measurements. The first substance they used was oil with a low vapour pressure that would easily form a spray (they produced the oil drops as a spray with a perfume atomizer using watch oil bought at minimal cost at a local market). Although Millikan’s published work dealt with the results from work done with the oil drops, Millikan and his group had done the same investigations with glycerine and mercury. Evaporation issues were only the first of many obstacles they had to overcome to arrive at a working system, including inter alia : temperature within the chamber affecting viscosity of the air, allowing for the evaporation of the oil (as well as glycerine and mercury)—however minimal, the motion of the air inside the chamber, the fluctuation of the charge applied by the battery source (Franklin 1981 ). Footnote 32

Their final set up (which led to Millikan’s published work on the quantization of the charge on the electron in 1910 and 1913) ran as follows. Within a sealed container Millikan et al. placed two charged plates 16 mm apart which were connected to a DC supply (battery). Above the top plate was an aperture through which the atomizer could spray droplets into the container. The top charged plate had a small aperture through which oil (glycerine, mercury) droplets could drop under gravity. In the space between the two plates were three apertures: one for the short focal telescope to view the drops, one for a light source in order to be able to see the drops and the other for an X-ray source to induce ionization of the air. The actual measurements were made in units of time—in seconds (range 11–19 s)—taken for an oil drop to move across a known distance of 10.21 mm (Millikan 1913 ). Footnote 33 The voltage (when used) was set at 5 kV. Differences in the time measured for an oil drop to move across the given distance (10.21 mm) under (just) gravity and then under the given current (5 kV) allowed Millikan to calculate the charge and, with repeated measurements under varying conditions, deduce that the charge was quantized. Footnote 34

As with Beaumont and Kettlewell, how does Hacking’s observation versus experiment account map onto the cases outlined above from various fields of scientific enquiry?

As with Beaumont and Kettlewell, the observation part of Hacking’s account—a means of detection—can be identified easily in the mentioned cases:

In Harlow’s work with infant monkeys, his measurements of time spent with each surrogate;

In Maskelyne’s investigations on the density of the earth, measurements of arc minutes to measure the shift in the pendulum from the true vertical with respect to fixed star positions and those of height in terms of feet and inches in order to measure volume;

In Hutton’s case, the results of chemical and heat changes of different kinds of rock formations, drawings of fossils and rock strata as well as frequency tabulation in the form of integer numbers related to recording of numbers of fossils found in various rock formations;

In Millikan’s case, the measurements of the time taken for an oil drop to traverse the distance of 10.21 mm.

Again as with the cases of Beaumont and Kettlewell, it is much more difficult to see how Hacking’s category of experiment fits with these cases. In none of them is it clear to see where the ‘creation of phenomena’ with its emphasis on ‘pure state’, as we saw with the Hall or Josephson Effects, lies:—emotional attachment for Harlow, density for Maskelyne, landscape erosion for Hutton—even in physics—Millikan’s quantization of charge.

Hacking’s observation versus experiment account—as a means of delineating scientific experimentation as part of practice —thus appears not very helpful when faced with cases from a range of fields of scientific enquiry as those described. Even within modern physics—as Millikan’s work shows—Hacking’s account has limited use. Footnote 35

4 Experiment and Observation as Processes

David Gooding notes that it is in facing real accounts of scientific experimentation that what he calls “the familiar distinction between observation and experiment” collapses; calling the distinction an “artifact of the disembodied, reconstructed character of retrospective accounts” (Gooding 1992 , 68). We should then perhaps not be surprised that Hacking’s observation versus experiment framework does not survive intact when put to the test in a range of cases of scientific experimentation Footnote 36 such as those described.

Hacking’s account, in its attempt to reify and stipulate the notion of experiment, fails to capture the range and complexity of actions (mental and physical) entailed in what is indicated by experiment in scientific practice. If we return to the examples of scientific experimentation described above, in all cases—some undertaken over decades—the investigations consisted of the accumulation of parts: in Beaumont’s case, his in vivo as well as his in vitro work; in Kettlewell’s case, his struggles in finding the appropriate control landscapes; in Harlow’s case, trials with different kinds of ‘soft material’ used as a surrogate with which the infant monkeys could identify; Maskelyne’s case consisted of two distinct parts—the astronomical measurements made in the two observatories and the land survey of Schiehallion which followed the astronomical part and took nearly 2 years to complete (due to weather conditions); Hutton’s work on investigating rock strata and formations, and their relationship with the age of the earth, took decades and consisted of two distinct parts—analysis of the rock strata and work with fossils; Millikan’s experiment—which appears straightforward—went though a number of stages as it was optimized for substances (different kinds of oils, glycerine and mercury) and conditions (such as temperature, air viscosity) and calibration (different scales used to measure distance).

Looking at these examples of scientific work, should we refer to them as experiments or a series of experiments? Gooding proposes a potentially helpful way of thinking about this question. Gooding asks us not to talk about experiment but experimentation and think of it as a process Footnote 37 ( 1992 , 65–67). Hasok Chang too, using a different lexicon, asks us to think of experiment as a series of activities which themselves are composed of processes ( 2011 , 208–210). Footnote 38 The idea of process fits well with the range of examples described above. However, does viewing experimentation as a process help us in delineating experiment from observation as categories distinct from each other as Hacking does in his account? We have seen that observation, for Hacking, is a means of detection. However, this too, more often than not, tends to be a process. If we take just one case from amongst those that Hacking categorizes as observation this becomes apparent. One example (cited earlier) Hacking uses as an example of an observation is of the detection of solar neutrinos ( 1983 , 182). The detection of solar neutrinos runs thus. Footnote 39

Solar neutrinos are produced as a by-product of nuclear fusion in the core of the sun (Pinch 1985 , 5). As they are highly unreactive, they can pass through the outer layers of the sun and pass through the earth’s atmosphere (predominantly) in the state in which they were produced in the sun’s interior. The fact that they are highly unreactive of course makes them very difficult to detect. In the 1960s, Raymond Davis Jr. developed the methodology for detection of solar neutrinos of a particular kind (pp or proton–proton). A 100–400 k gallon container of dry cleaning fluid (perchloroethylene) was buried over a mile underground (in a disused mineshaft). The chlorine in perchloroethylene contains traces of a radioactive chlorine isotope (37) with which the solar neutrinos are able to react. The reaction between the chlorine isotope and solar neutrinos gives rise to the production of a radioactive argon isotope (37). This argon isotope is allowed to accumulate over typically a month (not longer as the half-life of the argon isotope (37) is 35 days). Other isotopes of argon (36 or 38) are added which aids the argon isotope (37) to bind to the helium gas, which is flushed through the container to remove the argon isotope (37). The helium containing the argon isotope (37) is then passed through pre-cooled charcoal, which collects the argon isotope (37). It is the decay of the argon isotope (37), which can be detected via a pre-calibrated Geiger counter.

It is apparent that this process of detection of solar neutrinos is exactly that—a process—with a multitude of different manipulations, practices and interpretations. In fact, very similar to the practices and processes of experimentation described in the cases above, and act for Gooding’s claim that in the face of real cases of scientific practice, to (try to) distinguish between observation and experiment is futile. In Hacking’s account, experiment is defined in a verbal phrase—‘creation of phenomena’—based on activity and ‘endless different tasks’ (Hacking 1983 , 230). However, observation too in this account entails the same as a means of detection and in most fields of scientific enquiry requires ‘endless different tasks’—one could replace the case of the solar neutrinos described above by numerous others including inter alia : other sub-atomic particle decay experiments in physics, chain reactions in chemistry (organic and inorganic), cascade and chain reactions in biochemistry. Both observation and experiment in practice involve undertaking various activities, manipulations, interventions and interpretations.

Hasok Chang has proposed that the pursuit of a systematic analysis of activities entailed in scientific practice is a worthy goal (Chang 2011 ). He proposes a “philosophical grammar of scientific practice” (ibid, 206) where he tentatively draws a taxonomy of what he says are only some of the “epistemic activities” entailed in scientific practice including, inter alia , Describing, Explaining, Hypothesizing, Testing, Observing, Measuring, Classifying, Representing, Modelling, Simulating, Synthesizing, Analyzing, Causing, Abstracting, Idealizing. David Gooding too has made an attempt to describe scientific practice (Gooding 1990 , 1992 ) albeit in diagrammatic form—in what he calls “experimental maps” (Gooding 1992 , 67), rather than discursively as Chang does. However, both are interested in considering the nature (in Chang’s case) and ordering (in Gooding’s case) of the multitude of epistemic activities entailed in scientific practice; rather than in differentiating between them as Hacking appears to be doing in his experiment versus observation account, as a means of categorization in an ‘either/or’ way. It is therefore not surprising that Hacking’s account, based as it is on casting observation and experiment as polarities, rather than seeing them as parts of a continuum within the process of experimentation, is not able to adequately account for cases of scientific experimentation other than from the very narrow area of physics on which he chooses to focus, such as high-energy lasers and such like.

The categorical distinction Hacking makes between observation and experiment would seem to rely in the main on his very particular definition of experiment—‘creation of phenomena’ and the many issues arising out of his stipulation of ‘creation’ as demonstrated earlier. Footnote 40 If one disregards Hacking’s stipulation of ‘creation’ in his definition for experiment, then it is difficult to see how a category distinction can be maintained between observation and experiment. As we saw earlier, both encompass generation of data so this could not act as an adequate marker.

If one were to broaden Hacking’s notion of experiment to consider another candidate as a marker for making a distinction between observation and experiment, then one of the more obvious ones is intervention. Lorraine Daston, in her account of practices of observation in the period 1600–1800, gives a glimpse of the various views circulating around the projected distinction between observation and experiment during this period (Daston 2011 , 85–87). Amongst these views, many gave importance to intervention (or its synonyms) as an important marker for distinguishing between observation and experiment. However, even then (that is, before the use of increasingly complex instrumentation became ubiquitous in scientific experimentation and practice in the modern age) some could see ambiguities arising. Gottfried Wilhelm Liebniz notes, “there are certain experiments that would be better called observations, in which one considers rather than produces the work” (ibid. 86).

This attempt to cast ‘intervention’ as a potential marker to distinguish between observation and experiment as categories, of course, pre-dates the crucial nineteenth century shift towards the dissipation of any qualitative difference between ‘seeing’ with help such as with instrumentation with its associated range of interventions, and that without—as the works of Clary, Hoffmann and Schikore amongst others have shown. Footnote 41 Looking back to the case of William Beaumont and his work on human digestion we see that: both his in vivo and in vitro work needs intervention (of some kind) to be satisfactorily completed making it impossible to distinguish (in any consistent and coherent way) between what should be observation and what experiment. The case of observation of solar neutrinos also makes very clear, with its numerous and complex manipulations, that intervention is not a reasonable candidate for acting as a category distinguisher between observation and experiment. The category distinction Hacking makes between observation and experiment thus rests very much on his narrow definition of experiment—‘creation of phenomena’—with the anomalous consequences that arise when this definition is used across various instances of scientific experimentation as the cases earlier demonstrate.

Other accounts than those of Chang and Gooding have also been advanced to analyse scientific experimentation; although interestingly—but perhaps unsurprisingly in light of our discussion thus far, very few use Hacking’s nomenclature of observation/experiment. Like Gooding and Chang, most believe that scientific experimentation should be viewed as a continuous process rather than one entailing discrete parts—and the terminology used underlines this sense of continuousness. Friedrich Steinle and Richard Burian have coined the term ‘exploratory experimentation’ which gives the same sense of the continuousness of the experimental process as do Chang and Gooding in their work: Steinle working on the early history of electromagnetism (Steinle 1997 , 2002 ), and Burian in his work on molecular biology (Burian 1997 , 2007 ). Footnote 42

Steinle, in analysing the experimental work of Oersted, Ampere and Faraday draws a distinction between two kinds of experiments: those designed with the specific aim of tracing particular effects which were expected because of the field of knowledge within electromagnetism at the time, and those experiments set up where the investigators had, what Steinle calls, “no theory—or—even more fundamentally—no conceptual framework” (Steinle 1997 , S65). Richard Burian, too, has used the term ‘exploratory experimentation’. Burian first used the term in his analysis of the work of Jean Brachet’s experiments on the localization and functioning of nucleic acids (Burian 1997 ). Burian examines the research of Brachet’s on the distribution of nucleic acids across cell life cycles. Burian shows that Brachet was not guided by theoretical considerations about how the nucleic acids may be distributed across the lifetime of cells in various organisms. This was very much in contrast to Brachet’s peer, Francis Crick, working on the same subject matter, who was much more theoretically inclined which greatly influenced the kinds of experiments he chose to undertake (ibid., 40–41). Burian therefore also uses the term in the same sense as Steinle insofar as to distinguish a particular kind of experimentation from theory. Since this inception, ‘exploratory experimentation’ has gradually gained more definitive structure: for example, it is clear that it is not the case that exploratory experimentation is free from theory—rather, the question is how theory influences the experimental process thus leading to a distinction between ‘theory-directed’ and ‘theory-informed’ (Waters 2007 , 277); leading to calls for the creation of its own sub-structure that can account for historical cases more adequately than it does in its present form (O’Malley 2007 ).

Another term, ‘experimental system’, has been used within writing on epistemology of experimentation, which conveys the same sense of ‘continuousness’ as does exploratory experimentation. The term ‘experimental system’ was first used by Hans-Jörg Rheinberger to refer to the experimental research on protein synthesis ( 1997 ). Rheinberger describes experimental systems as “systems of manipulation designed to give unknown answers to questions that the experimenters themselves are not yet clearly to ask” (ibid., 28). As with the term ‘exploratory experimentation’, Rheinberger uses ‘experimental system’ in order to distinguish it from a theory-dominated approach, arguing that experimental work in biology always “begins with the choice of a system rather than with the choice of theoretical framework” (ibid., 25). Other similar terms to experimental systems have been used to indicate scientific experimentation as a process with the sense of continuousness embedded at their centre such as ‘manipulable systems’ (Turnball and Stokes 1990 ) and ‘production systems’ (Kohler 1991 ).

All these terms (exploratory experimentation, experimental system, manipulable system, production system) and their respective accounts emerge with the aim of distinguishing them from theory-dominated accounts such as hypothesis testing. None of these accounts seek to do what Hacking does with his stipulation of experiment: distinguish between different kinds of activities and interventions within the process of scientific experimentation. Hasok Chang’s aim in delineating the various kinds of activities and interventions involved in scientific practice uses a descriptive rather than a stipulative approach (Chang 2011 ). James Woodward uses the terms observation and experiment as distinct categories but with the aim of defending what he calls a “manipulationist account of causation” rather than in an attempt to delineate scientific experimentation as process (Woodward 2003 , 88). Footnote 43

However, elsewhere, James Woodward, together with James Bogen, has sought to put forward an account which seeks to specifically delineate the process of scientific experimentation. It completely abandons the vocabulary of observation and experiment and uses data and phenomena instead. Footnote 44 Bogen and Woodward initially put forward their data phenomena account in 1988 (Bogen and Woodward 1988 ). Footnote 45

Bogen and Woodward tell us that data should be thought of as that which provides evidence for the existence of phenomena (Bogen and Woodward 1988 , 305). Data can (usually) be detected. However, data (usually) cannot be predicted. Phenomena, on the other hand, can only be detected through the use of data (Bogen and Woodward 1988 , 306). Examples of data include bubble chamber photographs, patterns of discharge in electronic particle detectors and records of reaction times and error rates in psychological experiments. These instances of data provide evidence for the following phenomena respectively: weak neutral currents, decay of the proton and chunking effects in human short-term memory.

Bogen and Woodward analyse a number of examples to illustrate the distinction between data and phenomena (Bogen and Woodward 1988 , 308–322). Examples they use to show what they mean include the melting point of lead (from chemistry) and weak neutral currents (from physics).

Bogen and Woodward analyse the following statement about the melting point of lead to show what they mean by their data phenomenon distinction: ‘lead melts at 327.5 ± 0.1 degrees centigrade’. However, this is not what actually happens. It is not possible to determine the melting point of lead by taking a single thermometer reading. Footnote 46 It is necessary to take a series of measurements. Even if systematic errors are reduced, there will be variations in the thermometer readings such as to give a scatter of results that all differ from each other, even if potential sources of error are minimized. The figure 327.5 represents the mean of the scatter of thermometer readings while the figure 0.1 represents the standard deviation.

Within Bogen and Woodward’s account, the thermometer readings fall within the category data while the calculated melting point, 327.5 degree centigrade fall within the category phenomena. It is the latter, phenomena, which becomes the object of systematic scientific explanation. Thus, in the case of the melting point of lead, the figure 327.5 degree centigrade becomes the object of explanation in terms of the molecular structure of metals. This would be expressed in terms such as metallic bonding mechanisms and type of co-ordination.

The data too can become the object of scientific explanation. However, the terms in which explanations regarding data would be made would be different from those made for phenomena. Explanations related to data would include discussion of the accuracy of the thermometer, the purity of the lead sample used, the point at which the thermometer is taken (when the sample of lead starts to melt, at mid-way, when the sample has all melted), the reliability of the heating mechanism and such like. These terms and considerations are very different to those related to discussions in terms of molecular structure. In Bogen and Woodward’s account, thus, data are distinguished from phenomena by the fact that the terms in which phenomena are explained are distinct from the terms in which data are explained.

Another difference between data and phenomena in Bogen and Woodward’s account relates to phenomena possessing regular characteristics, which are detectable from very different kinds of data (or evidence). Bogen and Woodward use the following example to show what they mean.

The evidence for the existence of the phenomenon of weak neutral currents came from two different kinds of investigations. One was at CERN in Switzerland and the other at the NAL (National Accelerator Laboratory) in the US. The data from CERN comprised of bubble chamber photographs (where the detection method depended on the formation of bubbles) while that from NAL consisted of patterns of discharge in particle detectors (where the detection method registered the passage of charged tracks by electronic means). These two very different kinds of data—from very different kinds of apparatus—provided the evidence for the same phenomenon: the weak neutral current.

The terms of explanation for the phenomenon, the weak neutral current, comprise the interaction of the Z particle with the weak force—this is common in both cases: from the data from CERN as well as the very different data from the NAL. However, the terms of explanation of the two different data sets have very little in common: the data set from CERN comprises of terms consisting of, inter alia, the nature of the neutron beam, the shielding chamber, the size of the chamber as well as the type of liquid used in the chamber. The data from the NAL, however, required explanation in terms of, inter alia, the strength of the magnetic field, the characteristics of the calorimeter used to stop, absorb and measure a particle’s energy, and the nature of the tracking device.

For Bogen and Woodward, phenomena are “in the world, as belonging to the natural order itself and not just to the way we talk about or conceptualize that order” ( 1988 , 321). This may include, “particular objects, objects with features, events, processes, and states” (Bogen and Woodward 1988 , 321). To Bogen and Woodward, the key feature of phenomena is that they be the objects of general scientific explanation, rather than the particular explanations, which are the characteristic feature of data, and from which they are distinct ( 1988 , 322). Data are highly localized and idiosyncratic and demand explanations that are framed in very different terms to that of phenomena for which they act as evidence (Bogen and Woodward 1988 , 319).

Mapping the data phenomena account onto the cases cited earlier would thus yield the following outcomes. For Beaumont, the data relates to the results of digestion from both the in vivo and in vitro parts of his investigation and the explanatory terms in which they are framed relate to degrees of acidity, temperature readings and measurements of time, while the terms in which the explanatory terms for the phenomena are framed include peristaltic movement, the anatomy and composition of gastric cell types and the physical topography of the stomach with respect to the rest of the gastrointestinal tract. For Kettlewell, knowledge about data would relate to what kind of moth is conspicuous on which colour bark, the numbers of different kinds of moths surviving exposure to predation in the cage, what kind of bait is used to trap surviving moths in native conditions. The phenomenon is accounted for by discussion in terms of the changing colour of the landscape owing to pollution and degrees of conspicuousness to predators. For Harlow, the data would be framed in terms such as time spent with each kind of surrogate while discussion of phenomena would be framed in terms of emotional bonding, cognitive support and imitation along with the need for physical contact with an animate like material. For Maskelyne, the data would be framed in terms of arc minutes for the astronomical measurements and feet/inches for the physical survey of the mountain while the phenomena concerned (density of the earth) was expressed in terms of a numerical value (4500 kg/m 3 ). For Hutton, the data was framed in terms of chemical, temperature and field measurements and anatomical differentiation in fossil records while the phenomena was framed in terms of soil erosion and the influence of the physical elements (wind, water) on this erosion as indicative of changing climate and its correlation with fossil deposits. For Millikan, the data would be framed in terms of time taken for an oil drop to travel a distance of 10.21 mm, the viscosity of the oil, the temperature of the cloud chamber while the phenomena was framed in terms of the discrete nature of the charge carried by an electron, its interactions with other parts of the atom, the nature of these interactions and the value of the charge itself (1.5924 × 10 −19  C).

Although both Hacking and Bogen and Woodward aim, in each of their accounts of delineating scientific experimentation, to use a criteria based approach, the criteria they use are very different.

It is worth noting some points of conceptual overlap, as well as departure, between the two accounts—notwithstanding the different lexicon of each. There is considerable congruence between Hacking’s category of observation (or results(s) of observation) and Bogen and Woodward’s category of data—the outcomes when mapping each account onto the mentioned cases makes this clear: Beaumont’s time measurements for digestion, Kettlewell’s survival rates, Harlow’s record of time spent with each kind of surrogate, Maskelyne’s astronomical and survey measurements in arc minutes and units of height, Hutton’s complex and varied array of temperature, chemical, maps and diagrammatic records and Millikan’s measurements of time taken for an oil drop to traverse 10.21 mm. It is when we turn to consider the second part of each account—Hacking’s experiment (creation of phenomena) and Bogen and Woodward’s phenomena—that the conceptual overlap starts to dissipate. At first glance, both use phenomena in a similar way. For Hacking, it is a discernable regularity, “[a] phenomenon is noteworthy . A phenomenon is discernable . A phenomenon is commonly an event or process of certain type that occurs regularly under defined circumstances” (Hacking 1983 , 221, emphasis in original). For Bogen and Woodward, a phenomenon has “stable, repeatable characteristics which will be detectable by means of a variety of different procedures, which may yield quite different kinds of data” (Bogen and Woodward 1988 , 317). However, the common use of vocabulary—both in terms of naming and description—should not prevent us from noting the different ways each is conceptualized in each different account—as Bogen and Woodward themselves note (ibid, 306), suggesting that although there are certain similarities between their notion of phenomena and that of Hacking’s, they find Hacking’s notion limited insofar as Hacking’s notion “is not correct as a general characterization of phenomena” and continue on to say that the “features which Hacking ascribes to phenomena are more characteristic of data”. Others too have noted the ambiguity in Hacking’s description of the relationship between experiment and phenomena. Footnote 47

Bogen and Woodward here identify the principal limitation in Hacking’s observation versus experiment account as a means for (systematically) delineating scientific experimentation as practice: in the observation versus experiment account (the results of) observation and experiment both are ways of generating data. It is therefore perhaps not surprising that earlier we saw quite anomalous and ambiguous outcomes on mapping Hacking’s account on to cases of scientific experimentation from a range of fields of enquiry.

Bogen and Woodward use an explanation,—or what Rheinberger has called ‘epistemic object’—based, Footnote 48 criteria. Hacking, however, uses a narrowly construed criteria, centred on (kinds of) action/activity/intervention in his stipulation of experiment as ‘creation of phenomena’. This stipulative approach, as we have seen, has limited value when used in practice across a whole range of fields of enquiry.

5 Concluding Remarks

We have seen from our discussion that the observation versus experiment account has significant weaknesses as a means of delineating scientific experimentation within scientific practice—across a range of cases from various fields of scientific enquiry. This would suggest that the experiment versus observation framework—where observation and experiment are cast as polarities, rather than as complements of each other—as Hooke and Boyle did—is not a sound basis on which to make value judgments.

See Desmond Lee's Introduction to his translation of Aristotle's Meteorologica .

See ‘Introduction' by Lorraine Daston and Elizabeth Lunbeck in Histories of Scientific Observation ; in particular page 3.

This, of course, belies the considerable academic scholarship by historians of science that exists on the nature and characteristics of scientific practice, in particular, scientific experimentation, in pre-modern cultures that has shown the significant limitations of this position; Greek, Latin, Arabic and Chinese to name just a few. See Lloyd ( 2004 , 2006 ) on Greek and Chinese science and references therein. For Arabic science, see Sabra ( 1996 ). For Latin, see Lindberg ( 2007 ). For an example from the exact sciences, see the case of geometrical optics: for Greek, see Smith ( 1996 ), and for Arabic, see Sabra ( 2003 ). For the case of medicine, see Pormann and Savage-Smith ( 2007 ).

See Hacking ( 1983 , 173).

In particular, see footnote 12 in Pomata ( 2011 ).

See Park ( 2011 , 15–44), Pomata ( 2011 , 45–80) and Daston ( 2011 , 81–113).

See pp. 148–149 in particular.

See also Schickore ( 2007 ) for the case of the microscope. Also see Daston and Galison ( 2007 ).

Those who have been interested in the detailed historical accounts of particular experiments include Galison ( 1982 , 1983 ), Pickering ( 1981 ), Gooding ( 1982 ), Worral ( 1982 ), Wheaton ( 1983 ), Stuewer ( 1975 ) and Franklin ( 1986 ). Others have been concerned with the role of experiment in knowledge acquisition such as Gooding ( 2000 ), Kuhn ( 1976 ), Dear ( 1995 ) and Tiles ( 1993 ). Some have been interested in the philosophy of scientific experimentation (Radder 2003a , b ) which takes into account the nexus that experimentation provides for the meeting of theory, technology and modelling amongst others. Others have been concerned with the relationship between theory, observing and experimentation such as Latour and Woolgar ( 1986 ), Collins ( 1985 ), Galison ( 1987 ), Bogen and Woodward ( 1988 ) and Rheinberger ( 1997 ). Philsophers of science interested in observation include Shapere ( 1982 ) and Fodor ( 1983 ).

See Radder ( 2003b , 15) and Gooding ( 1992 , 68).

Hacking's primary aim in Representing and Intervening (Hacking 1983 ), however, lies in the juxtaposition of experiment to theory rather than an analysis of experiment relative to observation per se. Although Hacking takes up the subject of experiment again in some of his later work, there he is more concerned with other matters. He deals with the anti-realist position (see Hacking 1989 , for a response, see Shapere 1993 ) or with trying to defend the stability of laboratory practice (see Hacking 1992 ).

Also see Pinch ( 1985 ).

Brigitte Falkenburg has proposed that this position has limited value as theories of entities such as neutrinos, their detectors and the way information is transmitted from the source are all inextricably linked (see Falkenburg 2000 ).

For a detailed explanation see Galison ( 1985 ).

The Compton effect refers to the scattering of X-rays by electrons in work done by Arthur Compton in the 1920s.

The Zeeman effect refers to the splitting of the energy levels of an atom when it is placed in a magnetic field. Pieter Zeeman and Hendrik Lorentz did this work in the 1890s.

The photoelectric effect refers to the detection of a current when light is shone on some metals and is taken as an indication of the emission of electrons.

Many substances act as superconductors at temperatures near to absolute zero. Brian Josephson (in 1962) predicted that a weak current (subsequently named a ‘super-current’) would flow between two superconductors that were separated by a thin sheet of electrical insulation. Philip Anderson and John Rowell confirmed Josephson’s prediction a year later in 1963.

Dr. Beauchamp should read Dr. Beaumont (see Bernard 1957 , 8). The work was conducted during the 1820s, not a decade earlier as stated (see Bernard 1957 , 8).

See Beaumont ( 1833 ).

Such as neurological processes which control the mechanical and nerve impulse activities of the stomach.

In particular, see p. 82.

For a synopsis of Henry Kettlewell's study on moths, see Franklin ( 2012 ). Kettlewell's work was published in Heredity ( 1955 , 1956 , 1958 ). David Rudge has worked extensively on the history of Kettlewell's work, see Rudge ( 2005a , b , 2006 , 2009 , 2010 ). He has also dealt extensively with the issue of statistical error in Kettlewell's numerical analysis (Rudge 2001 , 2005a , b ) and the issue of validity of control experiments ( 1999 ) in which he deals in particular with Joel Hagen's critique of Kettlewell's use of controls (Hagen 1999 ); for an overview of the issue of the use of controls on Kettlewell's experiments, see Brandon ( 1999 ). The validity of the controls Kettlewell used relate to the geographical areas in which he performed the experiments (Birmingham, UK and Dorset, UK).

‘Attachment theory' relates to the notion that non-material provision from a (primary) carer is significant in the cognitive formation and development of higher mammals.

See Prior and Glaser ( 2006 ), Ainsworth ( 1991 ), Blum ( 2002 ); see a review of the latter at: http://primate.uchicago.edu/2004PC.pdf (accessed 8 Mar 2015).

See Chapters 11–15. See also Smallwood ( 2009 ).

See the edition of Andrew Motte's translation of Newton's Principia: The Mathematical Principles of Natural Philosophy , pp. 527–528.

Also see Repcheck ( 2004 ). For reception of Hutton's work amongst his contemporaries, see Dean ( 1973 ). For a synopsis of Hutton's biography see his entry in the Dictionary of Scientific Biography .

See also Rudwick ( 1985 , 2004 , 2005b ).

A visitor to Hutton's home in Edinburgh remarked, “his study is so full of fossils and chemical apparatus of various kinds that there is barely room to sit down”.

See also Franklin ( 1981 ), Barnes et al. ( 1996 ), and Goodstein ( 2001 ). Also see Niaz ( 2005 ) for an appraisal of the studies of Holton, Franklin, Barnes et al. and Goodstein.

Also see Franklin ( 1986 , 215–224).

See in particular pp. 124–125.

Millikan's conclusions were contested amongst specialists in the field for more than a decade after publication of this work; see Holton ( 1978 ) in particular; for a defence of Millikan, see Goodstein ( 2001 ).

In defence of Hacking, his principal aim in Representing and Intervening in making his observation experiment distinction is in service of other philosophical ends such as entity realism. Further, within its own time, Hacking's drawing of a polarity between observation and experiment served the purpose of challenging the hitherto identification of experiment with observation (as a perceptual rather than a detection form). One may therefore reasonably posit that the criteria Hacking puts forward as his description of experiment should not be applied rigidly. However, I believe he appears quite committed to his stipulation of experiment as ‘creation of phenomena' in a formalistic way—he emphasizes the ‘creation' part of ‘creation of phenomena' in his discussion at length explicitly and reiterates this commitment by the examples with which he chooses to engage at length—certain kinds of cases from physics (Hall effect, Josephson effect) while consciously stepping away from others such as those such as the work of William Beaumont which, as we see above, are not easily receptive to the observation versus experiment account. In addition, Hacking underlines his commitment to ‘creation' in his description of experiment as ‘creation of phenomena' as well as ‘purity' of said in his later work (Hacking 1992 , 37; here Hacking uses the photoelectric effect as an exemplar). I think therefore it is not unreasonable to take Hacking at his own (repeated) word. If one does do that, then it appears from our discussion that Hacking's stipulation of experiment as ‘creation of phenomena', and his emphasis on ‘pure state', gives rise to anomalies in a range of cases of scientific practices as shown.

Hans Radder uses ‘scientific experimentation' in a way which reflects the importance of the processual nature of scientific work and practices, (Radder 2003b , 15).

Hacking too has used the term ‘experimentation', ‘Experimentation has many lives of its own' ( 1983 , 165). However, Hacking uses ‘experimentation' in contrast to theory, saying “…, let us not pretend that the various phenomenological laws of solid state physics required a theory—any theory—before they were known. Experimentation has many lives of its own” (ibid.). In contrast, David Gooding uses ‘experimentation' as a process qua process.

Also see Rouse (1996).

This case has been analyzed in detail by Shapere ( 1982 ) and Pinch ( 1985 ) as well as dealt with in summary by Bogen and Woodward ( 1988 , 316).

Others too have noted the ambiguities arising out of the very particular way Hacking stipulates his category of experiment, See Feest ( 2011 , 63–64).

See earlier references to each of these authors.

Rose-Mary Sargent too uses the term but for descriptive rather than analytical purposes (Steinle 1997 , S71).

See also Woodward ( 2013 ) for a review of the topic where he deals with the different positions on the subject matter, including his own.

In so doing, they avoid linguistic oddities such as, ‘What has been shown as well is that, in actual practice, making scientific observations often includes doing genuine experiments' (Radder 2003b , 15).

Since then it has been re-stated by Woodward on a number of occasions (Woodward 1989 , 2000 , 2011 ). In these revised versions, however, Woodward has been more concerned with dealing with the relationship between this account and its relationship with scientific theory. The data phenomena account has been contested on various grounds. These contestations have tended to focus on two areas. First, whether it is reasonable to draw a distinction between data and phenomena at all whereas both should be viewed as patterns within data sets (Glymour 2000 ). Further, even if one were to draw a distinction between them, how one does this—in particular, the role of assumptions in this process (McAllister 1997 ). Woodward ( 2011 , 175–176) amongst others (Apel 2011 , 27–31), have responded to these points in recent years. The other area of focus has been the relationship of data and phenomena within Bogen and Woodward's account to theory (Schindler 2007 , 2011 ); in particular, as it relates to the influence of theory on observation and its implications for reliability. Woodward counters this view in detail (Woodward 2011 , 172–174) by suggesting that the charge that data in the data phenomena account is assumed to be independent of ‘additional assumptions' or ‘theory free' is unfounded. However, where used to delineate scientific practice qua practice, it appears reasonably robust—as even its detractors concede (Schindler 2011 , 54).

For details of how the melting point of lead is determined under laboratory conditions, see Bogen and Woodward ( 1988 , 309–310).

See Feest ( 2011 , 63–64).

See Rheinberger ( 1997 , 28).

Ainsworth, M. D. S. (1991). Attachments and other affectional bonds across the life-cycle. In B. Caldwell & H. Riccuiti (Eds.), Review of child development research (pp. 1–94). Chicago: The University of Chicago Press.

Google Scholar  

Anstey, P. (2014). Philosophy of experiment in early modern England: The case of Bacon, Boyle and Hooke. Early Science and Medicine, 19 , 103–132.

Article   Google Scholar  

Apel, J. (2011). On the meaning and the epistemological relevance of the notion of scientific phenomena. Synthese, 182 , 23–38.

Aristotle. (1962). Meteorologica . Introduction and Translation by H. D. P. Lee. London: Heinemann, Loeb Classical Library Series.

Barnes, B., Bloor, D., & Henry, J. (1996). Scientific knowledge: A sociological analysis . Chicago: The University of Chicago Press.

Beaumont, W. (1833). Experiments and observations on the gastric juice and the physiology of digestion . Boston: F. P. Allen.

Bernard, C. (1957). An introduction to the study of experimental medicine . New York: Dover Publications Inc.

Blum, D. (2002). Love at Goon Park: Harry Harlow and the science of affection . Cambridge, MA: Perseus.

Bogen, J., & Woodward, J. (1988). Saving the phenomena. The Philosophical Review, 97 , 303–352.

Brandon, R. N. (1999). Introduction. Biology and Philosophy, 14 , 1–7.

Burian, R. (1997). Exploratory experimentation and the role of histochemical techniques in the work of Jean Brachet, 1938–1952. History and Philosophy of the Life Sciences, 19 , 27–45.

Burian, R. (2007). On micro RNA and the need for exploratory experimentation in post-genomic molecular biology. History and Philosophy of the Life Sciences, 29 , 285–312.

Chang, H. (2011). The philosophical grammar of scientific practice. International Studies in the Philosophy of Science, 25 , 205–221.

Collins, H. (1985). Changing order . Chicago: The University of Chicago Press.

Crary, J. (1992). Techniques of the observer: On vision and modernity in the nineteenth century . Cambridge, MA: MIT Press.

Crary, J. (2001). Suspensions of perception: Attention spectacle and modern culture . Cambridge, MA: MIT Press.

Danson, E. (2009). Weighing the world: The quest to measure the Earth . Oxford: Oxford University Press.

Daston, L. (2011). The empire of observation. In L. Daston & E. Lunbeck (Eds.), Histories of scientific observation (pp. 81–113). Chicago: The University of Chicago Press.

Daston, L., & Galison, P. (2007). Objectivity . New York: Zone Books.

Daston, L., & Lunbeck, E. (2011). Introduction. In L. Daston & E. Lunbeck (Eds.), Histories of scientific observation (pp. 1–9). Chicago: The University of Chicago Press.

Dean, D. R. (1973). James Hutton and his public, 1785–1802. Annals of Science, 30 , 89–105.

Dean, D. R. (1992). James Hutton and the history of geology . Ithaca: Cornell University Press.

Dear, P. (1995). Discipline and experience: The mathematical way in the scientific revolution . Chicago: The University of Chicago Press.

Book   Google Scholar  

Falkenburg, B. (2000). How to observe quarks. In E. Agazzi & M. Pauri (Eds.), The reality of the unobservable (pp. 329–341). Dordrecht: Kluwer.

Chapter   Google Scholar  

Feest, U. (2011). What exactly is stabilized when phenomena are stabilized? Synthese, 182 , 57–71.

Fodor, J. (1983). Observation reconsidered. Philosophy of Science, 51 , 23–43.

Franklin, A. D. (1981). Millikan’s published and unpublished data on oil drops. Historical Studies in the Physical Sciences, 11 , 185–201.

Franklin, A. (1986). The neglect of experiment . Cambridge: Cambridge University Press.

Franklin, A. (2012). Experiment in physics. In Stanford encyclopedia of philosophy . http://plato.stanford.edu . accessed on 14 May 2014.

Galison, P. (1982). Theoretical predispositions in experimental physics: Einstein and the gyromagnetic experiments, 1915–1925. Historical Studies in the Physical Sciences, 12 , 285–323.

Galison, P. (1983). How the first neutral current experiments ended. Reviews of Modern Physics, 55 , 477–509.

Galison, P. (1985). Bubble chambers and the experimental workplace. In P. Achinstein & O. Hannaway (Eds.), Observation, experiment and physical science (pp. 309–373). Cambridge, MA: MIT Press.

Galison, P. (1987). How experiments end . Chicago: The University of Chicago Press.

Glymour, B. (2000). Data and phenomena: A distinction reconsidered. Erkenntis, 52 , 29–37.

Gooding, D. (1982). Empiricism in practice: Teleology, economy, and observation in Faraday’s physics. Isis, 73 , 46–67.

Gooding, D. (1990). Experiment and the making of meaning . Boston: Kluwer Academic.

Gooding, D. (1992). Putting agency back into experiment. In A. Pickering (Ed.), Science as practice and culture (pp. 65–112). Chicago: The University of Chicago Press.

Gooding, D. (2000). Experiment. In W. H. Newton-Smith (Ed.), A companion to the philosophy of science (pp. 117–126). Oxford: Oxford University Press.

Goodstein, D. (2001). In defence of Robert Andrews Millikan. American Scientist, 89 , 54–60.

Hacking, I. (1983). Representing and intervening . Cambridge: Cambridge University Press.

Hacking, I. (1989). Extragalactic reality: The case of gravitational lensing. Philosophy of Science, 56 , 555–581.

Hacking, I. (1992). The self-vindication of the laboratory sciences. In A. Pickering (Ed.), Science as practice and culture (pp. 29–64). Chicago: The University of Chicago Press.

Hagen, J. B. (1999). Retelling experiments: H. B. D. Kettlewell’s studies of industrial melanism in peppered moths. Biology and Philosophy, 14 , 39–54.

Hanson, N. R. (1958). Patterns of discovery . Cambridge: Cambridge University Press.

Hoffmann, C. (2006). Unter Beobachtung: Naturforschung in der Zeit der Sinnesapparate . Gottingen: Wallstein.

Holton, G. (1978). Subelectrons, presuppositions, and the Millikan–Ehrenhaft dispute. Historical Studies in the Physical Sciences, 9 , 161–224.

Kettlewell, H. (1955). Selection experiments on industrial melanism in the Lepidoptera. Heredity, 9 , 323–342.

Kettlewell, H. (1956). Further selection experiments on industrial melanism in the Lepidoptera. Heredity, 10 , 287–301.

Kettlewell, H. (1958). A survey of the frequencies of Biston betularia (L.) and its melanic forms in Great Britain. Heredity, 12 , 51–72.

Kohler, R. (1991). Systems of production: Drosophila, neurospora and biochemical genetics. Historical Studies in the Physical and Biological Sciences, 22 , 87–130.

Kuhn, T. S. (1976). Mathematical vs. experimental traditions in the development of physical sciences. Journal of Interdisciplinary History, VII (I), 1–31.

Latour, B., & Woolgar, S. (1986). Laboratory life: The construction of scientific facts . Princeton: Princeton University Press.

Lindberg, D. C. (2007). The beginnings of western science . Chicago: The University of Chicago Press.

Lloyd, G. E. R. (2004). Ancient worlds, modern reflections: Philosophical perspectives on Greek and Chinese science and culture . Oxford: Oxford University Press.

Lloyd, G. E. R. (2006). Principles and practices in ancient Greek and Chinese science . Aldershot: Variorum.

Maxwell, G. (1962). The ontological status of theoretical entities. In H. Feigi & G. Maxwell (Eds.), Scientific explanation, space and time: Minnesota studies in the philosophy of science (pp. 181–192). Minnesota: University of Minnesota Press.

McAllister, J. (1997). Phenomena and patterns in data sets. Erkenntis, 47 , 217–228.

Millikan, R. A. (1913). On the elementary charge and the Avogadro constant. The Physical Review, II , 109–143.

Newton, I. (1846). Newton’s principia. The mathematical principles of natural philosophy . Trans. A. Motte. New York: Daniel Adee.

Niaz, M. (2005). An appraisal of the controversial nature of the oil drop experiment: Is closure possible? British Journal for the Philosophy of Science, 56 , 681–702.

O’Malley, M. A. (2007). Exploratory experimentation and scientific practice: metagenomics and the proteorhodopsin case. History and Philosophy of the Life Sciences, 29 , 337–360.

Park, K. (2011). Observation in the margins, 500–1500. In L. Daston & E. Lunbeck (Eds.), Histories of scientific observation (pp. 15–44). Chicago: The University of Chicago Press.

Pickering, A. (1981). The hunting of the quark. Isis, 72 , 216–236.

Pinch, T. (1985). Towards an analysis of scientific observation: The externality and evidential significance of observational reports in physics. Social Studies of Science, 15 , 3–36.

Pomata, G. (2011). Observation rising: Birth of an epistemic genre, 1500–1650. In L. Daston & E. Lunbeck (Eds.), Histories of scientific observation (pp. 45–80). Chicago: The University of Chicago Press.

Pormann, P., & Savage-Smith, E. (2007). Medieval Islamic medicine . Washington D.C.: Georgetown University Press.

Prior, V., & Glaser, D. (2006). Understanding attachment and attachment disorders: Theory, evidence and practice . London: Jessica Kingsley.

Radder, H. (Ed.). (2003a). The philosophy of scientific experimentation . Pittsburgh: University of Pittsburgh Press.

Radder, H. (2003b). Towards a more developed philosophy of scientific experimentation. In H. Radder (Ed.), The philosophy of scientific experimentation (pp. 1–18). Pittsburgh: University of Pittsburgh Press.

Repcheck, J. (2004). The man who found time . London: Perseus.

Rheinberger, H.-J. (1997). Towards a history of epistemic things: Synthesizing proteins in the test tube . Stanford: Stanford University Press.

Rudge, D. W. (1999). Taking the peppered moth with a grain of salt. Biology and Philosophy, 14 , 9–37.

Rudge, D. W. (2001). Kettlewell from an Error Statistician’s Point of View. Perspectives on Science, 9 , 59–77.

Rudge, D. W. (2005a). Did Kettlewell commit Fraud? Re-examining the evidence. Public Understanding of Science, 14 , 249–268.

Rudge, D. W. (2005b). The beauty of Kettlewell’s classic experimental demonstration of natural selection. Bio-Science, 55 , 369–375.

Rudge, D. W. (2006). H. D. P. Kettlewell’s research 1937–1953: The influence of E. B. Ford, E. A. Cockayne and P. M. Sheppard. History and Philosophy of the Life Sciences, 28 , 359–388.

Rudge, D. W. (2009). H. B. D. Kettlewell’s research 1934–1961: The influence of J. W. Heslop Harrison. Transactions of the American Philosophical Society, 99 , 243–270.

Rudge, D. W. (2010). Tut-tut tutt, not so fast: Did Kettlewell really test Tutt’s explanation of industrial melanism? History and Philosophy of the Life Sciences, 28 , 359–388.

Rudwick, M. J. S. (1985). The Great Devonian controversy: The shaping of scientific knowledge among gentlemanly specialists . Chicago: The University of Chicago Press.

Rudwick, M. J. S. (2004). The new science of geology: Studies in the earth sciences in the age of revolution . London: Ashgate.

Rudwick, M. J. S. (2005a). Lyell and Darwin, geologists: Studies in the earth sciences in the age of reform . London: Ashgate.

Rudwick, M. J. S. (2005b). Bursting the limits of time: the reconstruction of geohistory in the age of revolution . Chicago: The University of Chicago Press.

Sabra, A. I. (1996). Situating Arabic science: Locality versus essence. Isis, 87 , 654–670.

Sabra, A. I. (2003). Ibn al-Haytham’s revolutionary project in optics: The achievement and the obstacle. In J. P. Hogendijk & A. I. Sabra (Eds.), The enterprise of science in Islam (pp. 85–118). Boston, MA: The MIT Press.

Schickore, J. (2007). The microscope and the eye: A history of reflections, 1740–1870 . Chicago: The University of Chicago Press.

Schindler, S. (2007). Rehabilitating theory: Refusal of the ‘bottom-up’ construction of scientific phenomena. Studies in the History and Philosophy of Science, 38 , 160–184.

Schindler, S. (2011). Bogen and Wooodward’s data-phenomena distinction, forms of theory-ladenness, and the reliability of data. Synthese, 182 , 39–55.

Shapere, D. (1982). The concept of observation in science and philosophy. Philosophy of Science, 49 , 485–525.

Shapere, D. (1993). Astronomy and antirealism. Philosophy of Science, 60 , 134–150.

Smallwood, J. R. (2009). John Playfair on Schiehallion, 1801–1811. In C. L. E. Lewis & S. J. Knell (Eds.), The making of the Geological Society of London (pp. 279–298). London: The Geological Society.

Smith, A. M. (1996). Ptolemy’s theory of visual perception . Philadelphia: The American Philosophical Society.

Steinle, F. (1997). Entering new fields: Exploratory uses of experimentation. Philosophy of Science (Proceedings), 64 , S65–S74.

Steinle, F. (2002). Experiments in history and philosophy of science. Perspectives on Science, 10 , 408–432.

Stuewer, R. (1975). The Compton effect . New York: Science History Publications.

Tiles, J. E. (1993). Experiment as intervention. British Journal for the Philosophy of Science, 44 , 463–475.

Turnball, D., & Stokes, T. (1990). Manipulable systems and laboratory strategies in a biomedical institute. In H. E. Le Grand (Ed.), Experimental enquiries (pp. 167–192). Dordrecht: Kluwer.

van Fraassen, B. C. (1980). The scientific image . Oxford: Oxford University Press.

Waters, C. K. (2007). The nature and context of exploratory experimentation: An introduction to three case studies of exploratory research. History and Philosophy of the Life Sciences, 29 , 275–284.

Wheaton, B. (1983). The tiger and the shark . Cambridge: Cambridge University Press.

Woodward, J. (1989). Data and phenomena. Synthese, 79 , 393–472.

Woodward, J. (2000). Data, phenomena, and reliability. Philosophy of Science, 67 , S163–S179.

Woodward, J. (2003). Experimentation, causal inference, and instrumental realism. In H. Radder (Ed.), The philosophy of scientific experimentation (pp. 87–118). Pittsburg: University of Pittsburg Press.

Woodward, J. (2011). Data and phenomena: A restatement and defense. Synthese, 182 , 165–179.

Woodward, J. (2013). Causation and manipulability. In E. N. Zalta (Ed) The Stanford encyclopedia of philosophy . http://plato.stanford.edu/archives/win2013/entries/causation-mani/ .

Worral, J. (1982). The pressure of light: The strange case of the vacillating crucial experiment. Studies in History and Philosophy of Science, 13 , 133–171.

Download references

Acknowledgments

I would like to thank Emilie Savage-Smith and Nick Jardine for their comments during the early gestation of this work. I am also very grateful to the anonymous referees, as well as the Editors, for their very helpful comments during review.

Author information

Authors and affiliations.

Cardiff University, Cardiff, CF10 3EU, UK

Saira Malik

You can also search for this author in PubMed   Google Scholar

Corresponding author

Correspondence to Saira Malik .

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License ( http://creativecommons.org/licenses/by/4.0/ ), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Reprints and permissions

About this article

Malik, S. Observation Versus Experiment: An Adequate Framework for Analysing Scientific Experimentation?. J Gen Philos Sci 48 , 71–95 (2017). https://doi.org/10.1007/s10838-016-9335-y

Download citation

Published : 07 May 2016

Issue Date : March 2017

DOI : https://doi.org/10.1007/s10838-016-9335-y

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Philosophy of experiment
• Ian Hacking
• Observation
• Scientific experimentation
• Scientific practice
• Find a journal
• Publish with us
• Track your research