Scientific Objectivity

First published Mon Aug 25, 2014; substantive revision Fri Oct 30, 2020

Scientific objectivity is a property of various aspects of science. Itexpresses the idea that scientific claims, methods, results—andscientists themselves—are not, or should not be, influenced byparticular perspectives, value judgments, community bias or personalinterests, to name a few relevant factors. Objectivity is oftenconsidered to be an ideal for scientific inquiry, a good reason forvaluing scientific knowledge, and the basis of the authority ofscience in society.

Many central debates in the philosophy of science have, in one way oranother, to do with objectivity: confirmation and the problem ofinduction; theory choice and scientific change; realism; scientificexplanation; experimentation; measurement and quantification;statistical evidence; reproducibility; evidence-based science;feminism and values in science. Understanding the role of objectivityin science is therefore integral to a full appreciation of thesedebates. As this article testifies, the reverse is true too: it isimpossible to fully appreciate the notion of scientific objectivitywithout touching upon many of these debates.

The ideal of objectivity has been criticized repeatedly in philosophyof science, questioning both its desirability and its attainability.This article focuses on the question of how scientific objectivityshould bedefined, whether the ideal of objectivity isdesirable, and to what extent scientists canachieveit.

1. Introduction

Objectivity is a value. To call a thing objective implies that it hasa certain importance to us and that we approve of it. Objectivitycomes in degrees. Claims, methods, results, and scientists can be moreor less objective, and, other things being equal, the more objective,the better. Using the term “objective” to describesomething often carries a special rhetorical force with it. Theadmiration of science among the general public and the authorityscience enjoys in public life stems to a large extent from the viewthat science is objective or at least more objective than other modesof inquiry. Understanding scientific objectivity is therefore centralto understanding the nature of science and the role it plays insociety.

If what is so great about science is its objectivity, then objectivityshould be worth defending. The close examinations of scientificpractice that philosophers of science have undertaken in the pastfifty years have shown, however, that several conceptions of the idealof objectivity are either questionable or unattainable. The prospectsfor a science providing a non-perspectival “view fromnowhere” or for proceeding in a way uninformed by human goalsand values are fairly slim, for example.

This article discusses several proposals to characterize the idea andideal of objectivity in such a way that it is both strong enough to bevaluable, and weak enough to be attainable and workable in practice.We begin with a natural conception of objectivity:faithfulness to facts. We motivate the intuitiveappeal of this conception, discuss its relation to scientific methodand discuss arguments challenging both its attainability as well asits desirability. We then move on to a second conception ofobjectivity asabsence of normative commitments andvalue-freedom, and once more we contrast arguments in favorof such a conception with the challenges it faces. A third conceptionof objectivity which we discuss at length is the idea ofabsence of personal bias.

Finally there is the idea that objectivity is anchored inscientific communities and their practices. Afterdiscussing threecase studies from economics, socialscience and medicine, we address theconceptual unity ofscientific objectivity: Do the various conceptions have acommon valid core, such as promoting trust in science or minimizingrelevant epistemic risks? Or are they rivaling and only looselyrelated accounts? Finally we present some conjectures about whataspects of objectivity remain defensible and desirable in the light ofthe difficulties we have encountered.

2. Objectivity as Faithfulness to Facts

The basic idea of this first conception of objectivity is thatscientific claims are objective in so far as they faithfully describefacts about the world. The philosophical rationale underlying thisconception of objectivity is the view that there are facts “outthere” in the world and that it is the task of scientists todiscover, analyze, and systematize these facts.“Objective” then becomes a success word: if a claim isobjective, it correctly describes some aspect of the world.

In this view, science is objective to the degree that it succeeds atdiscovering and generalizing facts, abstracting from the perspectiveof the individual scientist. Although few philosophers have fullyendorsed such a conception of scientific objectivity, the idea figuresrecurrently in the work of prominent twentieth-century philosophers ofscience such as Carnap, Hempel, Popper, and Reichenbach.

2.1 The View From Nowhere

Humans experience the world from a perspective. The contents of anindividual’s experiences vary greatly with his perspective,which is affected by his personal situation, and the details of hisperceptual apparatus, language and culture. While the experiencesvary, there seems to be something that remains constant. Theappearance of a tree will change as one approaches itbut—according to common sense and most philosophers—thetree itself doesn’t. A room may feel hot or cold for differentpersons, but its temperature is independent of their experiences. Theobject in front of me does not disappear just because the lights areturned off.

These examples motivate a distinction between qualities that vary withone’s perspective, and qualities that remain constant throughchanges of perspective. The latter are the objective qualities. ThomasNagel explains that we arrive at the idea of objective qualities inthree steps (Nagel 1986: 14). The first step is to realize (orpostulate) that our perceptions are caused by the actions of thingsaround us, through their effects on our bodies. The second step is torealize (or postulate) that since the same qualities that causeperceptions in us also have effects on other things and can existwithout causing any perceptions at all, their true nature must bedetachable from their perspectival appearance and need not resembleit. The final step is to form a conception of that “truenature” independently of any perspective. Nagel calls thatconception the “view from nowhere”, Bernard Williams the“absolute conception” (Williams 1985 [2011]). Itrepresents the world as it is, unmediated by human minds and other“distortions”.

This absolute conception lies at the basis of scientific realism (fora detailed discussion, see the entry onscientific realism) and it is attractive in so far as it provides a basis for arbitratingbetween conflicting viewpoints (e.g., two different observations).Moreover, the absolute conception provides a simple and unifiedaccount of the world. Theories of trees will be very hard to come byif they use predicates such as “height as seen by anobserver” and a hodgepodge if their predicates track the habitsof ordinary language users rather than the properties of the world. Tothe extent, then, that science aims to provide explanations fornatural phenomena, casting them in terms of the absolute conceptionwould help to realize this aim. A scientific account cast in thelanguage of the absolute conception may not only be able to explainwhy a tree is as tall as it is but also why we see it in one way whenviewed from one standpoint and in a different way when viewed fromanother. As Williams (1985 [2011: 139]) puts it,

[the absolute conception] nonvacuously explain[s] how it itself, andthe various perspectival views of the world, are possible.

A third reason to find the view from nowhere attractive is that if theworld came in structures as characterized by it and we did have accessto it, we could use our knowledge of it to ground predictions (which,to the extent that our theories do track the absolute structures, willbe borne out). A fourth and related reason is that attempts tomanipulate and control phenomena can similarly be grounded in ourknowledge of these structures. To attain any of the fourpurposes—settling disagreements, explaining the world,predicting phenomena, and manipulation and control—the absoluteconception is at best sufficient but not necessary. We can, forinstance, settle disagreements by imposing the rule that the personwith higher social rank or greater experience is always right. We canexplain the world and our image of it by means of theories that do notrepresent absolute structures and properties, and there is no need toget things (absolutely) right in order to predict successfully.Nevertheless, there is something appealing in the idea that factualdisagreements can be settled by the very facts themselves, thatexplanations and predictions grounded in what’s really thererather than in a distorted image of it.

No matter how desirable, our ability to use scientific claims torepresent facts about the world depends on whether these claims canunambiguously be established on the basis of evidence, and of evidencealone. Alas, the relation between evidence and scientific hypothesisis not straightforward.Subsection 2.2 andsubsection 2.3 will look at two challenges of the idea that even the best scientificmethod will yield claims that describe an aperspectival view fromnowhere.Section 5.2 will deal with socially motivated criticisms of the view fromnowhere.

2.2 Theory-Ladenness and Incommensurability

According to a popular picture, all scientific theories are false andimperfect. Yet, as we add true and eliminate false beliefs, our bestscientific theories become moretruthlike (e.g., Popper 1963,1972). If this picture is correct, then scientific knowledge grows bygradually approaching the truth and it will become more objective overtime, that is, more faithful to facts. However, scientific theoriesoften change, and sometimes several theories compete for the place ofthe best scientific account of the world.

It is inherent in the above picture of scientific objectivity thatobservations can, at least in principle, decide between competingtheories. If they did not, the conception of objectivity asfaithfulness would be pointless to have as we would not be in aposition to verify it. This position has been adopted by Karl R.Popper, Rudolf Carnap and other leading figures in (broadly)empiricist philosophy of science. Many philosophers have argued thatthe relation between observation and theory is way more complex andthat influences can actually run both ways (e.g., Duhem 1906 [1954];Wittgenstein 1953 [2001]). The most lasting criticism, however, wasdelivered by Thomas S. Kuhn (1962 [1970]) in his book “TheStructure of Scientific Revolutions”.

Kuhn’s analysis is built on the assumption that scientistsalways view research problems through the lens of a paradigm, definedby set of relevant problems, axioms, methodological presuppositions,techniques, and so forth. Kuhn provided several historical examples infavor of this claim. Scientific progress—and the practice ofnormal, everyday science—happens within a paradigm that guidesthe individual scientists’ puzzle-solving work and that sets thecommunity standards.

Can observations undermine such a paradigm, and speak for a differentone? Here, Kuhn famously stresses thatobservations are“theory-laden” (cf. also Hanson 1958): theydepend on a body of theoretical assumptions through which they areperceived and conceptualized. This hypothesis has two importantaspects.

First, themeaning of observational concepts is influenced bytheoretical assumptions and presuppositions. For example, the concepts“mass” and “length” have different meanings inNewtonian and relativistic mechanics; so does the concept“temperature” in thermodynamics and statistical mechanics(cf. Feyerabend 1962). In other words, Kuhn denies that there is atheory-independent observation language. The “faithfulness toreality” of an observation report is always mediated by atheoreticalüberbau, disabling the role of observationreports as an impartial, merely fact-dependent arbiter betweendifferent theories.

Second, not only the observational concepts, but also theperception of a scientist depends on the paradigm she isworking in.

Practicing in different worlds, the two groups of scientists [who workin different paradigms, J.R./J.S.] see different things when they lookfrom the same point in the same direction. (Kuhn 1962 [1970: 150])

That is, our own sense data are shaped and structured by a theoreticalframework, and may be fundamentally distinct from the sense data ofscientists working in another one. Where a Ptolemaic astronomer likeTycho Brahe sees a sun setting behind the horizon, a Copernicanastronomer like Johannes Kepler sees the horizon moving up to astationary sun. If this picture is correct, then it is hard to assesswhich theory or paradigm is more faithful to the facts, that is, moreobjective.

The thesis of the theory-ladenness of observation has also beenextended to theincommensurability of different paradigms orscientific theories, problematized independently by Thomas S.Kuhn (1962 [1970]) and Paul Feyerabend (1962). Literally, this conceptmeans “having no measure in common”, and it figuresprominently in arguments against a linear and standpoint-independentpicture of scientific progress. For instance, the Special Theory ofRelativity appears to be more faithful to the facts and therefore moreobjective than Newtonian mechanics because it reduces, for low speeds,to the latter, and it accounts for some additional facts that are notpredicted correctly by Newtonian mechanics. This picture isundermined, however, by two central aspects of incommensurability.First, not only do the observational concepts in both theories differ,but the principles for specifying their meaning may be inconsistentwith each other (Feyerabend 1975: 269–270). Second, scientificresearch methods and standards of evaluation change with the theoriesor paradigms. Not all puzzles that could be tackled in the oldparadigm will be solved by the new one—this is the phenomenon of“Kuhn loss”.

A meaningful use of objectivity presupposes, according to Feyerabend,to perceive and to describe the world from a specific perspective,e.g., when we try to verify the referential claims of a scientifictheory. Onlywithin a peculiar scientific worldview, theconcept of objectivity may be applied meaningfully. That is,scientific method cannot free itself from the particular scientifictheory to which it is applied; the door to standpoint-independence islocked. As Feyerabend puts it:

our epistemic activities may have a decisive influence even upon themost solid piece of cosmological furniture—they make godsdisappear and replace them by heaps of atoms in empty space. (1978:70)

Kuhn and Feyerabend’s theses about theory-ladenness ofobservation, and their implications for the objectivity of scientificinquiry have been much debated afterwards, and have often beenmisunderstood in a social constructivist sense. Therefore Kuhn laterreturned to the topic of scientific objectivity, of which he gives hisown characterization in terms of the shared cognitive values of ascientific community. We discuss Kuhn’s later view insection 3.1. For a more thorough coverage, see the entries ontheory and observation in science,the incommensurability of scientific theories andThomas S. Kuhn.

2.3 Underdetermination, Values, and the Experimenters’ Regress

Scientific theories are tested by comparing their implications withthe results of observations and experiments. Unfortunately, neitherpositive results (when the theory’s predictions are borne out inthe data) nor negative results (when they are not) allow unambiguousinferences about the theory. A positive result can obtain even thoughthe theory is false, due to some alternative that makes the samepredictions. Finding suspect Jones’ fingerprints on the murderweapon is consistent with his innocence because he might have used itas a kitchen knife. A negative result might be due not to thefalsehood of the theory under test but due to the failing of one ormore auxiliary assumptions needed to derive a prediction from thetheory. Testing, let us say, the implications of Newton’s lawsfor movements in our planetary system against observations requiresassumptions about the number of planets, the sun’s and theplanets’ masses, the extent to which the earth’satmosphere refracts light beams, how telescopes affect the results andso on. Any of these may be false, explaining an inconsistency. Thelocus classicus for these observations is PierreDuhem’sThe Aim and Structure of Physical Theory (Duhem1906 [1954]). Duhem concluded that there was no “crucialexperiment”, an experiment that conclusively decides between twoalternative theories, in physics (1906 [1954: 188ff.]), and thatphysicists had to employ their expert judgment or what Duhem called“good sense” to determine what an experimental resultmeans for the truth or falsehood of a theory (1906 [1954:216ff.]).

In other words, there is a gap between the evidence and the theorysupported by it. It is important to note that the alleged gap is moreprofound than the gap between the premisses ofany inductiveargument and its conclusion, say, the gap between “All hithertoobserved ravens have been black” and “All ravens areblack”. The latter gap could be bridged by an agreed upon ruleof inductive reasoning. Alas, all attempts to find an analogous rulefor theory choice have failed (e.g., Norton 2003). Variousphilosophers, historians, and sociologists of science have respondedthat theory appraisal is “a complex form of valuejudgment” (McMullin 1982: 701; see also Kuhn 1977; Hesse 1980;Bloor 1982).

Insection 3.1 below we will discuss the nature of the value judgments in moredetail. For now the important lesson is that if these philosophers,historians, and sociologists are correct, the “faithfulness tofacts” ideal is untenable. As the scientific image of the worldis a joint product of the facts and scientists’ value judgments,that image cannot be said to be aperspectival. Science does not eschewthe human perspective. There are of course ways to escape thisconclusion. If, as John Norton (2003; ms.—see Other InternetResources) has argued, it is material facts that power and justifyinductive inferences, and not value judgments, we can avoid thenegative conclusion regarding the view from nowhere. Unsurprisingly,Norton is also critical of the idea that evidence generallyunderdetermines theory (Norton 2008). However, there are good reasonsto mistrust Norton’s optimism regarding the ineliminability ofvalues and other non-factual elements in inductive inferences (Reiss2020).

There is another, closely related concern. Most of the earlier criticsof “objective” verification or falsification focused onthe relation between evidence and scientific theories. There is asense in which the claim that this relation is problematic is not sosurprising. Scientific theories contain highly abstract claims thatdescribe states of affairs far removed from the immediacy of senseexperience. This is for a good reason: sense experience is necessarilyperspectival, so to the extent to which scientific theories are totrack the absolute conception, they must describe a world differentfrom that of sense experience. But surely, one might think, theevidence itself is objective. So even if we do have reasons to doubtthat abstract theories faithfully represent the world, we should standon firmer grounds when it comes to the evidence against which we testabstract theories.

Theories are seldom tested against brute observations, however. Simplegeneralizations such as “all swans are white” are directlylearned from observations (say, of the color of swans) but they do notrepresent the view from nowhere (for one thing, the view from nowheredoesn’t have colors). Genuine scientific theories are testedagainst experimental facts or phenomena, which are themselvesunobservable to the unaided senses. Experimental facts or phenomenaare instead established using intricate procedures of measurement andexperimentation.

We therefore need to ask whether the results of scientificmeasurements and experiments can be aperspectival. In an importantdebate in the 1980s and 1990s some commentators answered that questionwith a resounding “no”, which was then rebutted by others.The debate concerns the so-called “experimenter’sregress” (Collins 1985). Collins, a prominent sociologist ofscience, claims that in order to know whether an experimental resultis correct, one first needs to know whether the apparatus producingthe result is reliable. But one doesn’t know whether theapparatus is reliable unless one knows that it produces correctresults in the first place and so on and so onad infinitum.Collins’ main case concerns attempts to detect gravitationalwaves, which were very controversially discussed among physicists inthe 1970s.

Collins argues that the circle is eventually broken not by the“facts” themselves but rather by factors having to do withthe scientist’s career, the social and cognitive interests ofhis community, and the expected fruitfulness for future work. It isimportant to note that in Collins’s view these factors do notnecessarily make scientific results arbitrary. But what he does argueis that the experimental results do not represent the world accordingto the absolute conception. Rather, they are produced jointly by theworld, scientific apparatuses, and the psychological and sociologicalfactors mentioned above. The facts and phenomena of science aretherefore necessarily perspectival.

In a series of contributions, Allan Franklin, aphysicist-turned-philosopher of science, has tried to show that whilethere are indeed no algorithmic procedures for establishingexperimental facts, disagreements can nevertheless be settled byreasoned judgment on the basis ofbona fideepistemological criteria such as experimental checks and calibration,elimination of possible sources of error, using apparatuses based onwell-corroborated theory and so on (Franklin 1994, 1997). Collinsresponds that “reasonableness” is a social category thatis not drawn from physics (Collins 1994).

The main issue for us in this debate is whether there are any reasonsto believe that experimental results provide an aperspectival view onthe world. According to Collins, experimental results areco-determined by the facts as well as social and psychologicalfactors. According to Franklin, whatever else influences experimentalresults other than facts is not arbitrary but instead based onreasoned judgment. What he has not shown is that reasoned judgmentguarantees that experimental results reflect the facts alone and aretherefore aperspectival in any interesting sense. Another importantchallenge for the aperspectival account comes from feministepistemology and other accounts that stress the importance of theconstruction of scientific knowledge through epistemic communities.These accounts are reviewed insection 5.

3. Objectivity as Absence of Normative Commitments and the Value-Free Ideal

In the previous section we have presented arguments against the viewof objectivity as faithfulness to facts and an impersonal “viewfrom nowhere”. An alternative view is that science is objectiveto the extent that it isvalue-free. Why would we identifyobjectivity with value-freedom or regard the latter as a prerequisitefor the former? Part of the answer is empiricism. If science is in thebusiness of producing empirical knowledge, and if differences aboutvalue judgments cannot be settled by empirical means, values shouldhave no place in science. In the following we will try to make thisintuition more precise.

3.1 Epistemic and Contextual Values

Before addressing what we will call the “value-freeideal”, it will be helpful to distinguish four stages at whichvalues may affect science. They are: (i) the choice of a scientificresearch problem; (ii) the gathering of evidence in relation to theproblem; (iii) the acceptance of a scientific hypothesis or theory asan adequate answer to the problem on the basis of the evidence; (iv)the proliferation and application of scientific research results(Weber 1917 [1949]).

Most philosophers of science would agree that the role of values inscience is contentious only with respect to dimensions (ii) and (iii):thegathering of evidence and theacceptanceof scientific theories. It is almost universally acceptedthat the choice of a research problem is often influenced by interestsof individual scientists, funding parties, and society as a whole.This influence may make science more shallow and slow down itslong-run progress, but it has benefits, too: scientists will focus onproviding solutions to those intellectual problems that are consideredurgent by society and they may actually improve people’s lives.Similarly, the proliferation and application of scientific researchresults is evidently affected by the personal values of journaleditors and end users, and little can be done about this. The realdebate is about whether or not the “core” of scientificreasoning—the gathering of evidence and the assessment andacceptance scientific theories—is, and should be,value-free.

We have introduced the problem of the underdetermination of theory byevidence above. The problem does not stop, however, at values beingrequired for filling the gap between theory and evidence. A furthercomplication is that these values can conflict with each other.Consider the classical problem of fitting a mathematical function to adata set. The researcher often has the choice between using a complexfunction, which makes the relationship between the variables lesssimple but fits the data moreaccurately, orpostulating asimpler relationship that is lessaccurate. Simplicity and accuracy are both importantcognitive values, and trading them off requires a careful valuejudgment. However, philosophers of science tend to regardvalue-ladenness in this sense as benign.Cognitivevalues (sometimes also called “epistemic” or“constitutive” values) such as predictive accuracy, scope,unification, explanatory power, simplicity and coherence with otheraccepted theories are taken to be indicative of the truth of a theoryand therefore provide reasons for preferring one theory over another(McMullin 1982, 2009; Laudan 1984; Steel 2010). Kuhn (1977) evenclaims that cognitive values define the shared commitments of science,that is, the standards of theory assessment that characterize thescientific approach as a whole. Note that not every philosopherentertains the same list of cognitive values: subjective differencesin ranking and applying cognitive values do not vanish, a point Kuhnmade emphatically.

In most views, the objectivity and authority of science is notthreatened by cognitive values, but only bynon-cognitive orcontextual values.Contextual values are moral, personal, social, political and culturalvalues such as pleasure, justice and equality, conservation of thenatural environment and diversity. The most notorious cases ofimproper uses of such values involve travesties of scientificreasoning, where the intrusion of contextual values led to anintolerant and oppressive scientific agenda with devastating epistemicand social consequences. In the Third Reich, a large part ofcontemporary physics, such as the theory of relativity, was condemnedbecause its inventors were Jewish; in the Soviet Union, biologistNikolai Vavilov was sentenced to death (and died in prison) becausehis theories of genetic inheritance did not match Marxist-Leninistideology. Both states tried to foster a science that was motivated bypolitical convictions (“Deutsche Physik” in Nazi Germany,Lysenko’s Lamarckian theory of inheritance and denial ofgenetics), leading to disastrous epistemic and institutionaleffects.

Less spectacular, but arguably more frequent are cases where researchis biased toward the interests of the sponsors, such as tobaccocompanies, food manufacturers and large pharmaceutic firms (e.g.,Resnik 2007; Reiss 2010). Thispreference bias,defined by Wilholt (2009) as the infringement of conventionalstandards of the research community with the aim of arriving at aparticular result, is clearly epistemically harmful. Especially forsensitive high-stakes issues such as the admission of medical drugs orthe consequences of anthropogenic global warming, it seems desirablethat research scientists assess theories without being influenced bysuch considerations. This is the core idea of the

Value-Free Ideal (VFI): Scientists should strive tominimize the influence of contextual values on scientific reasoning,e.g., in gathering evidence and assessing/accepting scientifictheories.

According to the VFI, scientific objectivity is characterized byabsence of contextual values and by exclusive commitment to cognitivevalues in stages (ii) and (iii) of the scientific process. See Dorato(2004: 53–54), Ruphy (2006: 190) or Biddle (2013: 125) foralternative formulations.

For value-freedom to be a reasonable ideal, it must not be a goalbeyond reach and be attainable at least to some degree. This claim isexpressed by the

Value-Neutrality Thesis (VNT): Scientistscan—at least in principle—gather evidence andassess/accept theories without making contextual value judgments.

Unlike the VFI, the VNT is not normative: its subject is whether thejudgments that scientists make are, or could possibly be, free ofcontextual values. Similarly, Hugh Lacey (1999) distinguishes threeprincipal components or aspects of value-free science: impartiality,neutrality and autonomy.Impartiality means thattheories are solely accepted or appraised in virtue of theircontribution to the cognitive values of science, such as truth,accuracy or explanatory power. This excludes the influence ofcontextual values, as stated above.Neutrality meansthat scientific theories make no value statements about the world:they are concerned with what there is, not with what there should be.Finally, scientificautonomy means that thescientific agenda is shaped by the desire to increase scientificknowledge, and that contextual values have no place in scientificmethod.

These three interpretations of value-free science can be combined witheach other, or used individually. All of them, however, are subject tocriticisms that we examine below. Denying the VNT, or theattainability of Lacey’s three criteria for value-free science,poses a challenge for scientific objectivity: one can either concludethat the ideal of objectivity should be rejected, or develop aconception of objectivity that differs from the VFI.

3.2 Acceptance of Scientific Hypotheses and Value Neutrality

Lacey’s characterization of value-free science and the VNT wereonce mainstream positions in philosophy of science. Their widespreadacceptance was closely connected to Reichenbach’s famousdistinction betweencontext of discovery andcontext of justification. Reichenbach first made thisdistinction with respect to the epistemology of mathematics:

the objective relation from the given entities to the solution, andthe subjective way of finding it, are clearly separated for problemsof a deductive character […] we must learn to make the samedistinction for the problem of the inductive relation from facts totheories. (Reichenbach 1938: 36–37)

The standard interpretation of this statement marks contextual values,which may have contributed to the discovery of a theory, as irrelevantforjustifying the acceptance of a theory, and for assessinghow evidence bears on theory—the relation that is crucial forthe objectivity of science. Contextual values are restricted to amatter of individual psychology that may influence the discovery,development and proliferation of a scientific theory, but not itsepistemic status.

This distinction played a crucial role in post-World War II philosophyof science. It presupposes, however, a clear-cut distinction betweencognitive values on the one hand and contextual values on the other.While this may beprima facie plausible for disciplines suchas physics, there is an abundance of contextual values in the socialsciences, for instance, in the conceptualization and measurement of anation’s wealth, or in different ways to measure the inflationrate (cf. Dupré 2007; Reiss 2008). More generally, three majorlines of criticism can be identified.

First, Helen Longino (1996) has argued that traditional cognitivevalues such as consistency, simplicity, breadth of scope andfruitfulness are not purely cognitive or epistemic after all, and thattheir use imports political and social values into contexts ofscientific judgment. According to her, the use of cognitive values inscientific judgments is not always, not even normally, politicallyneutral. She proposes to juxtapose these values with feminist valuessuch as novelty, ontological heterogeneity, mutuality of interaction,applicability to human needs and diffusion of power, and argues thatthe use of the traditional value instead of its alternative (e.g.,simplicity instead of ontological heterogeneity) can lead to biasesand adverse research results. Longino’s argument here isdifferent from the one discussed insection 3.1. It casts the very distinction between cognitive and contextual valuesinto doubt.

The second argument against the possibility of value-free science issemantic and attacks the neutrality of scientific theories: fact andvalue are frequently entangled because of the use of so-called“thick” ethical concepts in science (Putnam2002)—i.e., ethical concepts that have mixed descriptive andnormative content. For example, a description such as “dangeroustechnology” involves a value judgment about the technology andthe risks it implies, but it also has a descriptive content: it isuncertain and hard to predict whether using that technology willreally trigger those risks. If the use of such terms, where facts andvalues are inextricably entangled, is inevitable in scientificreasoning, it is impossible to describe hypotheses and results in avalue-free manner, undermining the value-neutrality thesis.

Indeed, John Dupré has argued that thick ethical terms areineliminable from science, at least certain parts of it (Dupré2007). Dupré’s point is essentially that scientifichypotheses and results concern us because they are relevant to humaninterests, and thus they will necessarily be couched in a languagethat uses thick ethical terms. While it will often be possible totranslate ethically thick descriptions into neutral ones, thetranslation cannot be made without losses, and these losses obtainprecisely because human interests are involved (seesection 6.2 for a case study from social science). According to Dupré,then, many scientific statements are value-free only because theirtruth or falsity does not matter to us:

Whether electrons have a positive or a negative charge and whetherthere is a black hole in the middle of our galaxy are questions ofabsolutely no immediate importance to us. The only human intereststhey touch (and these they may indeed touch deeply) are cognitiveones, and so the only values that they implicate are cognitive values.(2007: 31)

A third challenge to the VNT, and perhaps the most influential one,was raised first by Richard Rudner in his influential article“The Scientist Qua Scientist Makes Value Judgments”(Rudner 1953). Rudner disputes the core of the VNT and the context ofdiscovery/justification distinction: the idea that the acceptance of ascientific theory can in principle be value-free. First, Rudner arguesthat

no analysis of what constitutes the method of science would besatisfactory unless it comprised some assertion to the effect that thescientist as scientistaccepts or rejects hypotheses.(1953: 2)

This assumption stems from industrial quality control and otherapplication-oriented research. In such contexts, it is often necessaryto accept or to reject a hypothesis (e.g., the efficacy of a drug) inorder to make effective decisions.

Second, he notes that no scientific hypothesis is ever confirmedbeyond reasonable doubt—some probability of error alwaysremains. When we accept or reject a hypothesis, there is always achance that our decision is mistaken. Hence, our decision is also“a function of theimportance, in the typically ethicalsense, of making a mistake in accepting or rejecting ahypothesis” (1953: 2): we are balancing the seriousness of twopossible errors (erroneous acceptance/rejection of the hypothesis)against each other. This corresponds to type I and type II error instatistical inference.

The decision to accept or reject a hypothesis involves a valuejudgment (at least implicitly) because scientists have to judge whichof the consequences of an erroneous decision they deem more palatable:(1) some individuals die of the side effects of a drug erroneouslyjudged to be safe; or (2) other individuals die of a condition becausethey did not have access to a treatment that was erroneously judged tobe unsafe. Hence, ethical judgments and contextual values necessarilyenter the scientist’s core activity of accepting and rejectinghypotheses, and the VNT stands refuted. Closely related arguments canbe found in Churchman (1948) and Braithwaite (1953). Hempel (1965:91–92) gives a modified account of Rudner’s argument bydistinguishing between judgments ofconfirmation, which arefree of contextual values, and judgments ofacceptance. Sinceeven strongly confirming evidence cannot fully prove a universalscientific law, we have to live with a residual “inductiverisk” in inferring that law. Contextual values influencescientific methods by determining the acceptable amount of inductiverisk (see also Douglas 2000).

But how general are Rudner’s objections? Apparently, his resultholds true of applied science, but not necessarily of fundamentalresearch. For the latter domain, two major lines of rebuttals havebeen proposed. First, Richard Jeffrey (1956) notes that lawlikehypotheses in theoretical science (e.g., the gravitational law inNewtonian mechanics) are characterized by their general scope and notconfined to a particular application. Obviously, a scientist cannotfine-tune her decisions to their possible consequences in a widevariety of different contexts. So she should just refrain from theessentially pragmatic decision to accept or reject hypotheses. Byrestricting scientific reasoning to gathering and interpretingevidence, possibly supplemented by assessing the probability of ahypothesis, Jeffrey tries to save the VNT in fundamental scientificresearch, and the objectivity of scientific reasoning.

Second, Isaac Levi (1960) observes that scientists commit themselvesto certain standards of inference when they become a member of theprofession. This may, for example, lead to the statistical rejectionof a hypothesis when the observed significance level is smaller than5%. These community standards may eliminate any room for contextualethical judgment on behalf of the scientist: they determine when sheshould accept a hypothesis as established. Value judgments may beimplicit in how a scientific community sets standards of inference(comparesection 5.1), but not in the daily work of anindividual scientist (cf.Wilholt 2013).

Both defenses of the VNT focus on the impact of values in theorychoice, either by denying that scientists actually choose theories(Jeffrey), or by referring to community standards and restricting theVNT to the individual scientist (Levi). Douglas (2000: 563–565)points out, however, that the “acceptance” of scientifictheories is only one of several places for values to enter scientificreasoning, albeit an especially prominent and explicit one. Manydecisions in the process of scientific inquiry may conceal implicitvalue judgments: the design of an experiment, the methodology forconducting it, the characterization of the data, the choice of astatistical method for processing and analyzing data, theinterpretational process findings, etc. None of these methodologicaldecisions could be made without consideration of the possibleconsequences that could occur. Douglas gives, as a case study, aseries of experiments where carcinogenic effects of dioxin exposure onrats were probed. Contextual values such as safety and risk aversionaffected the conducted research at various stages: first, in theclassification of pathological samples as benign or cancerous (overwhich a lot of expert disagreement occurred), second, in theextrapolation from the high-dose experimental conditions to the morerealistic low-dose conditions. In both cases, the choice of aconservative classification or model had to be weighed against theadverse consequences for society that could result fromunderestimating the risks (see also Biddle 2013).

These diagnoses cast a gloomy light on attempts to divide scientificlabor between gathering evidence and determining the degree ofconfirmation (value-free) on the one hand and accepting scientifictheories (value-laden) on the other. The entire process ofconceptualizing, gathering and interpreting evidence is so entangledwith contextual values that no neat division, as Jeffrey envisions,will work outside the narrow realm of statistical inference—andeven there, doubts may be raised (see section 4.2).

Philip Kitcher (2011a: 31–40; see also Kitcher 2011b) gives analternative argument, based on his idea of “significanttruths”. There are simply too many truths that are of nointerest whatsoever, such as the total number of offside positions ina low-level football competition. Science, then, doesn’t aim attruthsimpliciter but rather at something more narrow: truthworth pursuing from the point of view of our cognitive, practical andsocial goals. Any truth that is worth pursuing in this sense is whathe calls a “significant truth”. Clearly, it is valuejudgments that help us decide whether or not any given truth issignificant.

Kitcher goes on to observing that the process of scientificinvestigation cannot neatly be divided into a stage in which theresearch question is chosen, one in which the evidence is gathered andone in which a judgment about the question is made on the basis of theevidence. Rather, the sequence is multiply iterated, and at eachstage, the researcher has to decide whether previous results warrantpursuit of the current line of research, or whether she should switchto another avenue. Such choices are laden with contextual values.

Values in science also interact, according to Kitcher, in anon-trivial way. Assume we endorse predictive accuracy as an importantgoal of science. However, there may not be a convincing strategy toreach this goal in some domain of science, for instance because thatdomain is characterized by strong non-linear dependencies. In thiscase, predictive accuracy might have to yield to achieving othervalues, such as consistency with theories in neighbor domains.Conversely, changing social goals lead to re-evaluations of scientificknowledge and research methods.

Science, then, cannot be value-free because no scientist ever worksexclusively in the supposedly value-free zone of assessing andaccepting hypotheses. Evidence is gathered and hypotheses are assessedand accepted in the light of their potential for application andfruitful research avenues. Both cognitive and contextual valuejudgments guide these choices and are themselves influenced by theirresults.

3.3 Science, Policy and the Value-Free Ideal

The discussion so far has focused on the VNT, the practicalattainability of the VFI, but little has been said about whether avalue-free science is desirable in the first place. This subsectiondiscusses this topic with special attention to informing and advisingpublic policy from a scientific perspective. While the VFI, and manyarguments for and against it, can be applied to science as a whole,the interface of science and public policy is the place where theintrusion of values into science is especially salient, and where itis surrounded by the greatest controversy. In the 2009“Climategate” affair, leaked emails from climatescientists raised suspicions that they were pursuing a particularsocio-political agenda that affected their research in an improperway. Later inquiries and reports absolved them from charges ofmisconduct, but the suspicions alone did much to damage the authorityof science in the public arena.

Indeed, many debates at the interface of science and public policy arecharacterized by disagreements on propositions that combine a factualbasis with specific goals and values. Take, for instance, the viewthat growing transgenic crops carries too much risk in terms ofbiosecurity, or addressing global warming by phasing out fossilenergies immediately. The critical question in such debates is whetherthere are theses \(T\) such that one side in the debate endorses\(T\), the other side rejects it, the evidence is shared, and bothsides have good reasons for their respective positions.

According to the VFI, scientists should uncover an epistemic,value-free basis for resolving such disagreements and restrict thedissent to the realm of value judgments. Even if the VNT should turnout to be untenable, and a strict separation to be impossible, the VFImay have an important function forguiding scientificresearch and for minimizing the impact of values on an objectivescience. In the philosophy of science, one camp of scholars defendsthe VFI as a necessary antidote to individual and institutionalinterests, such as Hugh Lacey (1999, 2002), Ernan McMullin (1982) andSandra Mitchell (2004), while others adopt a critical attitude, suchas Helen Longino (1990, 1996), Philip Kitcher (2011a) and HeatherDouglas (2009). These criticisms we discuss mainly refer to thedesirability or the conceptual (un)clarity of the VFI.

First, it has been argued that the VFI is not desirable at all.Feminist philosophers (e.g., Harding 1991; Okruhlik 1994; Lloyd 2005)have argued that science often carries a heavy androcentric values,for instance in biological theories about sex, gender and rape. Thecharge against these values is not so much that they are contextualrather than cognitive, but that they are unjustified. Moreover, ifscientists did follow the VFI rigidly, policy-makers would pay evenless attention to them, with a detrimental effect on the decisionsthey take (Cranor 1993). Given these shortcomings, the VFI has to berethought if it is supposed to play a useful role for guidingscientific research and leading to better policy decisions.Section 4.3 andsection 5.2 elaborate on this line of criticism in the context of scientificcommunity practices, and a science in the service of society.

Second, the autonomy of science often fails in practice due to thepresence of external stakeholders, such as funding agencies andindustry lobbies. To save the epistemic authority of science, Douglas(2009: 7–8) proposes to detach it from its autonomy byreformulating the VFI and distinguishing betweendirect andindirect roles of values in science. Contextual values maylegitimately affect the assessment of evidence by indicating theappropriate standard of evidence, the representation of complexprocesses, the severity of consequences of a decision, theinterpretation of noisy datasets, and so on (see also Winsberg 2012).This concerns, above all, policy-related disciplines such as climatescience or economics that routinely perform scientific risk analysesfor real-world problems (cf. also Shrader-Frechette 1991). Valuesshould, however, not be “reasons in themselves”, that is,evidence or defeaters for evidence (direct role, illegitimate) and as“helping to decide what should count as asufficientreason for a choice” (indirect role, legitimate). Thisprohibition for values to replace or dismiss scientific evidence iscalleddetached objectivity by Douglas, but it iscomplemented by various other aspects that relate to a reflectivebalancing of various perspectives and the procedural, social aspectsof science (2009: ch. 6).

That said, Douglas’ proposal is not very concrete when it comesto implementation, e.g., regarding the way diverse values should bebalanced. Compromising in the middle cannot be the solution (Weber1917 [1949]). First, no standpoint is, just in virtue of being in themiddle, evidentially supported vis-à-vis more extremepositions. Second, these middle positions are also, from a practicalpoint of view, the least functional when it comes to advisingpolicy-makers.

Moreover, the distinction between direct and indirect roles of valuesin science may not be sufficiently clear-cut to police the legitimateuse of values in science, and to draw the necessary borderlines.Assume that a scientist considers, for whatever reason, theconsequences of erroneously accepting hypothesis \(H\) undesirable.Therefore he uses a statistical model whose results are likely tofavor ¬\(H\) over \(H\). Is this a matter of reasonableconservativeness? Or doesn’t it amount to reasoning to aforegone conclusion, and to treating values as evidence (cf. Elliott2011: 320–321)?

The most recent literature on values and evidence in science presentsus with a broad spectrum of opinions. Steele (2012) and Winsberg(2012) agree that probabilistic assessments of uncertainty involvecontextual value judgments. While Steele defends this point byanalyzing the role of scientists as policy advisors, Winsberg pointsto the influence of contextual values in the selection andrepresentation of physical processes in climate modeling. Betz (2013)argues, by contrast, that scientists can largely avoid makingcontextual value judgments if they carefully express the uncertaintyinvolved with their evidential judgments, e.g., by using a scaleranging from purely qualitative evidence (such as expert judgment) toprecise probabilistic assessments. The issue of value judgments atearlier stages of inquiry is not addressed by this proposal; however,disentangling evidential judgments and judgments involving contextualvalues at the stage of theory assessment may be a good thing initself.

Thus, should we or should we not worried about values in scientificreasoning? While the interplay of values and evidential considerationsneed not be pernicious, it is unclear why itadds to thesuccess or the authority of science. How are we going to ensure thatthe permissive attitude towards values in setting evidential standardsetc. is not abused? In the absence of a general theory about whichcontextual values are beneficial and which are pernicious, the VFImight as well be as a first-order approximation to a sound,transparent and objective science.

4. Objectivity as Freedom from Personal Biases

This section deals with scientific objectivity as a form ofintersubjectivity—as freedom from personal biases. According tothis view, science is objective to the extent that personal biases areabsent from scientific reasoning, or that they can be eliminated in asocial process. Perhaps all science is necessarily perspectival.Perhaps we cannot sensibly draw scientific inferences without a hostof background assumptions, which may include assumptions about values.Perhaps all scientists are biased in some way. But objectivescientific results do not, or so the argument goes, depend onresearchers’ personal preferences or experiences—they arethe result of a process where individual biases are gradually filteredout and replaced by agreed upon evidence. That, among other things, iswhat distinguishes science from the arts and other human activities,and scientific knowledge from a fact-independent social construction(e.g., Haack 2003).

Paradigmatic ways to achieve objectivity in this sense are measurementand quantification. What has been measured and quantified has beenverified relative to a standard. The truth, say, that the Eiffel Toweris 324 meters tall is relative to a standard unit and conventionsabout how to use certain instruments, so it is neither aperspectivalnor free from assumptions, but it is independent of the person makingthe measurement.

We will begin with a discussion of objectivity, so conceived, inmeasurement, discuss the ideal of “mechanical objectivity”and then investigate to what extent freedom from personal biases canbe implemented in statistical evidence and inductiveinference—arguably the core of scientific reasoning, especiallyin quantitatively oriented sciences. Finally, we discussFeyerabend’s radical criticism of a rational scientific methodthat can be mechanically applied, and his defense of the epistemic andsocial benefits of personal “bias” and idiosyncrasy.

4.1 Measurement and Quantification

Measurement is often thought to epitomize scientific objectivity, mostfamously captured in Lord Kelvin’s dictum

when you cannot express it in numbers, your knowledge is of a meagreand unsatisfactory kind; it may be the beginning of knowledge, but youhave scarcely, in your thoughts, advanced to the stage ofscience, whatever the matter may be. (Kelvin 1883, 73)

Measurement can certainly achieve some independence of perspective.Yesterday’s weather in Durham UK may have been “reallyhot” to the average North Eastern Brit and “verycold” to the average Mexican, but they’ll both accept thatit was 21°C. Clearly, however, measurement does not result in a“view from nowhere”, nor are typical measurement resultsfree from presuppositions. Measurement instruments interact with theenvironment, and so results will always be a product of both theproperties of the environment we aim to measure as well as theproperties of the instrument. Instruments, thus, provide aperspectival view on the world (cf. Giere 2006).

Moreover, making sense of measurement results requires interpretation.Consider temperature measurement. Thermometers function by relating anunobservable quantity, temperature, to an observable quantity,expansion (or length) of a fluid or gas in a glass tube; that is,thermometers measure temperature by assuming that length is a functionof temperature: length = \(f\)(temperature). The function \(f\) is notknowna priori, and it cannot be tested either (because itcould in principle only be tested using averidicalthermometer, and the veridicality of the thermometer is just what isat stake here). Making a specific assumption, for instance that \(f\)is linear, solves that problem by fiat. But this“solution” does not take us very far because differentthermometric substances (e.g., mercury, air or water) yield differentresults for the points intermediate between the two fixed points0°C and 100°C, and so they can’t all expandlinearly.

According to Hasok Chang’s account of early thermometry (Chang2004), the problem was eventually solved by using a “principleof minimalist overdetermination”, the goal of which was to finda reliable thermometer while making as few substantial assumptions(e.g., about the form for \(f\)) as possible. It was argued that if athermometer was to be reliable, different tokens of the samethermometer type should agree with each other, and the results of airthermometers agreed the most. “Minimal” doesn’t meanzero, however, and indeed this procedure makes an importantpresupposition (in this case a metaphysical assumption about theone-valuedness of a physical quantity). Moreover, the procedureyielded at best a reliable instrument, not necessarily one that wasbest at tracking the uniquely real temperature (if there is such athing).

What Chang argues about early thermometry is true of measurements moregenerally: they are always made against a backdrop of metaphysicalpresuppositions, theoretical expectations and other kinds of belief.Whether or not any given procedure is regarded as adequate depends toa large extent on the purposes pursued by the individual scientist orgroup of scientists making the measurements. Especially in the socialsciences, this often means that measurement procedures are laden withnormative assumptions, i.e., values.

Julian Reiss (2008, 2013) has argued that economic indicators such asconsumer price inflation, gross domestic product and the unemploymentrate are value-laden in this sense. Consumer-price indices, forinstance, assume that if a consumer prefers a bundle \(x\) over analternative \(y\), then \(x\) is better for her than \(y\), which isas ethically charged as it is controversial. National income measuresassume that nations that exchange a larger share of goods and serviceson markets are richer than nations where the same goods and servicesare provided by the government or within households, which too isethically charged and controversial.

While not free of assumptions and values, the goal of many measurementprocedures remains to reduce the influence of personal biases andidiosyncrasies. The Nixon administration, famously, indexed socialsecurity payments to the consumer-price index in order to eliminatethe dependence of security recipients on the flimsiest of partypolitics: to make increases automatic instead of a result of politicalnegotiations (Nixon 1969). Lorraine Daston and Peter Galison refer tothis asmechanical objectivity. They write:

Finally, we come to the full-fledged establishment of mechanicalobjectivity as the ideal of scientific representation. What we find isthat the image, as standard bearer of is objectivity is tied to arelentless search to replace individual volition and discretion indepiction by the invariable routines of mechanical reproduction.(Daston and Galison 1992: 98)

Mechanical objectivity reduces the importance of human contributionsto scientific results to a minimum, and therefore enables science toproceed on a large scale where bonds of trust between individuals canno longer hold (Daston 1992). Trust in mechanical procedures thusreplaces trust in individual scientists.

In his bookTrust in Numbers, Theodore Porter pursues thisline of thought in great detail. In particular, on the basis of casestudies involving British actuaries in the mid-nineteenth century, ofFrench state engineers throughout the century, and of the US ArmyCorps of Engineers from 1920 to 1960, he argues for two causal claims.First, measurement instruments and quantitative procedures originatein commercial and administrative needs and affect the ways in whichthe natural and social sciences are practiced, not the other wayaround. The mushrooming of instruments such as chemical balances,barometers, chronometers was largely a result of social pressures andthe demands of democratic societies. Administering large territoriesor controlling diverse people and processes is not always possible onthe basis of personal trust and thus “objectiveprocedures” (which do not require trust in persons) took theplace of “subjective judgments” (which do). Second, heargues that quantification is a technology of distrust and weakness,and not of strength. It is weak administrators who do not have thesocial status, political support or professional solidarity to defendtheir experts’ judgments. They therefore subject decisions topublic scrutiny, which means that they must be made in a publiclyaccessible form.

This is the situation in which scientists who work in areas where thescience/policy boundary is fluid find themselves:

The National Academy of Sciences has accepted the principle thatscientists should declare their conflicts of interest and financialholdings before offering policy advice, or even information to thegovernment. And while police inspections of notebooks remainexceptional, the personal and financial interests of scientists andengineers are often considered material, especially in legal andregulatory contexts.
Strategies of impersonality must be understood partly as defensesagainst such suspicions […]. Objectivity means knowledge thatdoes not depend too much on the particular individuals who author it.(Porter 1995: 229)

Measurement and quantification help to reduce the influence ofpersonal biases and idiosyncrasies and they reduce the need to trustthe scientist or government official, but often at a cost.Standardizing scientific procedures becomes difficult when theirsubject matters are not homogeneous, and few domains outsidefundamental physics are. Attempts to quantify procedures for treatmentand policy decisions that we find in evidence-based practices arecurrently transferred to a variety of sciences such as medicine,nursing, psychology, education and social policy. However, they oftenlack a certain degree of responsiveness to the peculiarities of theirsubjects and the local conditions to which they are applied (see alsosection 5.3).

Moreover, the measurement and quantification of characteristics ofscientific interest is only half of the story. We also want todescribe relations between the quantities and make inferences usingstatistical analysis. Statistics thus helps to quantify furtheraspects of scientific work. We will now examine whether or notstatistical analysis can proceed in a way free from personal biasesand idiosyncrasies—for more detail, see the entry onphilosophy of statistics.

4.2 Statistical Evidence

The appraisal of scientific evidence is traditionally regarded as adomain of scientific reasoning where the ideal of scientificobjectivity has strong normative force, and where it is alsowell-entrenched in scientific practice. Episodes such asGalilei’s observations of the Jupiter moons, Lavoisier’scalcination experiments, and Eddington’s observation of the 1919eclipse are found in all philosophy of science textbooks because theyexemplify how evidence can be persuasive and compelling to scientistswith different backgrounds. The crucial question is therefore: can weidentify an “objective” concept of scientific evidencethat is independent of the personal biases of the experimenter andinterpreter?

Inferential statistics—the field that investigates the validityof inferences from data to theory—tries to answer this question.It is extremely influential in modern science, pervading experimentalresearch as well as the assessment and acceptance of our mostfundamental theories. For instance, a statistical argument helped toestablish the recent discovery of the Higgs Boson. We now compare themain theories of statistical evidence with respect to the objectivityof the claims they produce. They mainly differ with respect to therole of an explicitly subjective interpretation of probability.

4.2.1 Bayesian Inference

Bayesian inference quantifies scientific evidence by means ofprobabilities that are interpreted as a scientist’s subjectivedegrees of belief. The Bayesian thus leaves behind Carnap’s(1950) idea that probability is determined by a logical relationbetween sentences. For example, the prior degree of belief inhypothesis \(H\), written \(p(H)\), can in principle take any value inthe interval \([0,1]\). Simultaneously held degrees of belief indifferent hypotheses are, however, constrained by the laws ofprobability. After learning evidence E, the degree of belief in \(H\)is changed from its prior probability \(p(H)\) to the conditionaldegree of belief \(p(H \mid E)\), commonly called the posteriorprobability of \(H\). Both quantities can be related to each other bymeans ofBayes’ Theorem.

These days, the Bayesian approach is extremely influential inphilosophy and rapidly gaining ground across all scientificdisciplines. For quantifying evidence for a hypothesis, Bayesianstatisticians almost uniformly use theBayes factor,that is, the ratio of prior to posterior odds in favor of ahypothesis. The Bayes factor in favor of hypothesis \(H\) against itsnegation \(\neg\)\(H\) in the light of evidence \(E\) can be writtenas

\[\tag{3}\label{eqn:BF} BF (E) := \frac{p(H \mid E)}{p(\neg H \mid E)} \cdot \frac{p(\neg H)}{p(H)} = \frac{p(E \mid H)}{p(E \mid \neg H)},\]

or in other words, as the likelihood ratio between \(H\) and\(\neg\)\(H\). The Bayes factor reduces to the likelihoodistconception of evidence (Royall 1997) for the case of two competingpoint hypotheses. For further discussion of Bayesian measures ofevidence, see Good (1950), Sprenger and Hartmann (2019: ch. 1) and theentry onconfirmation and evidential support.

Unsurprisingly, the idea to measure scientific evidence in terms ofsubjective probability has met resistance. For example, thestatistician Ronald A. Fisher (1935: 6–7) has argued thatmeasuring psychological tendencies cannot be relevant for scientificinquiry and sustain claims to objectivity. Indeed, how shouldscientific objectivity square with subjective degree of belief?Bayesians have responded to this challenge in various ways:

Howson (2000) and Howson and Urbach (2006) consider the objectionmisplaced. In the same way that deductive logic does not judge thecorrectness of the premises but just advises you what to infer fromthem,Bayesian inductive logic provides rationalrules for representing uncertainty and making inductive inferences.Choosing the premises (e.g., the prior distributions)“objectively” falls outside the scope of Bayesiananalysis.
Convergence or merging-of-opinion theorems guaranteethat under certain circumstances, agents with very different initialattitudes who observe the same evidence will obtain similar posteriordegrees of belief in the long run. However, they are asymptoticresults without direct implications for inference with real-lifedatasets (see also Earman 1992: ch. 6). In such cases, the choice ofthe prior matters, and it may be beset with idiosyncratic bias andmanifest social values.
Adopting a more modest stance, Sprenger (2018) accepts that Bayesianinference does not achieve the goal of objectivity in the sense ofintersubjective agreement (concordant objectivity), or being free ofpersonal values, bias and subjective judgment. However, he argues thatcompeting schools of inference such as frequentist inference face thisproblem to the same degree, perhaps even worse. Moreover, somefeatures of Bayesian inference (e.g., the transparency about priorassumptions) fit recent, socially oriented conceptions of objectivitythat we discuss insection 5.

A radical Bayesian solution to the problem of personal bias is toadopt a principle that radically constrains an agent’s rationaldegrees of belief, such as thePrinciple of MaximumEntropy (MaxEnt—Jaynes 1968; Williamson 2010).According to MaxEnt, degrees of belief must be probabilistic and insync with empirical constraints, but conditional on these constraints,they must be equivocal, that is, as middling as possible. This latterconstraint amounts to maximizing the entropy of the probabilitydistribution in question. The MaxEnt approach eliminates varioussources of subjective bias at the expense of narrowing down the rangeof rational degrees of belief. An alternative objective Bayesiansolution consists in so-called“objectivepriors”: prior probabilities that do not represent anagent’s factual attitudes, but are determined by principles ofsymmetry, mathematical convenience or maximizing the influence of thedata on the posterior (e.g., Jeffreys 1939 [1980]; Bernardo 2012).

Thus, Bayesian inference, which analyzes statistical evidence from thevantage point of rational belief, provides only a partial answer tosecuring scientific objectivity from personal idiosyncrasy.

4.2.2 Frequentist Inference

The frequentist conception of evidence is based on the idea of thestatistical test of a hypothesis. Under the influenceof the statisticians Jerzy Neyman and Egon Pearson, tests were oftenregarded as rational decision procedures that minimize the relativefrequency of wrong decisions in a hypothetical series of repetitionsof a test (hence the name “frequentism”). Rudner’sargument insection 3.2 has pointed out the limits of this conception of hypothesis tests:the choice of thresholds for acceptance and rejection (i.e., theacceptable type I and II error rates) may reflect contextual valuejudgments and personal bias. Moreover, the losses associated witherroneously accepting or rejecting that hypothesis depend on thecontext of application which may be unbeknownst to theexperimenter.

Alternatively, scientists can restrict themselves to a purelyevidential interpretation of hypothesis tests and leave decisions topolicy-makers and regulatory agencies. The statistician and biologistR.A. Fisher (1935, 1956) proposed what later became the orthodoxquantification of evidence in frequentist statistics. Suppose a“null” or default hypothesis \(H_0\) denotes that anintervention has zero effect. If the observed data are“extreme” under \(H_0\)—i.e., if it was highlylikely to observe a result that agrees better with \(H_0\)—thedata provide evidence against the null hypothesis and for the efficacyof the intervention. The epistemological rationale is connected to theidea of severe testing (Mayo 1996): if the intervention wereineffective, we would, in all likelihood, have found data that agreebetter with the null hypothesis. The strength of evidence against\(H_0\) is equal to the\(p\)-value: the lower it is,the stronger evidence \(E\) speaks against the null hypothesis\(H_0\).

Unlike Bayes factors, this concept of statistical evidence does notdepend on personal degrees of belief. However, this does notnecessarily mean that \(p\)-values are more objective. First,\(p\)-values are usually classified as “non-significant”(\(p > .05\)), “significant” (\(p < .05\)),“highly significant”, and so on. Not only that thesethresholds and labels are largely arbitrary, they also promotepublication bias: non-significant findings are oftenclassified as “failed studies” (i.e., the efficacy of theintervention could not be shown), rarely published and end up in theproverbial “file drawer”. Much valuable research issuppressed. Conversely, significant findings may often occur when thenull hypothesis is actually true, especially when researchers havebeen “hunting for significance”. In fact, researchers havean incentive to keep their \(p\)-values low: the stronger theevidence, the more convincing the narrative, the greater theimpact—and the higher the chance for a good publication andcareer-relevant rewards. Moving the goalpost by“p-hacking” outcomes—for example by eliminatingoutliers, selective reporting or restricting the analysis to asubgroup—evidently biases the research results and compromisesthe objectivity of experimental research.

In particular, suchquestionable research practices(QRP) increase the type I error rate, which measures the rateat which false hypotheses are accepted, substantially over its nominal5% level and contribute to publication bias (Bakker et al. 2012).Ioannidis (2005) concludes that “most published researchfindings are false”—they are the combined result of a lowbase rate of effective causal interventions, the file drawer effectand the widespread presence of questionable research practices. Thefrequentist logic of hypothesis testing aggravates the problem becauseit provides a framework where all these biases can easily enter(Ziliak and McCloskey 2008; Sprenger 2016). These radical conclusionsare also confirmed by empirical findings: in many disciplinesresearchers fail to replicate findings by other scientific teams. Seesection 5.1 for more detail.

Summing up our findings, neither of the two major frameworks ofstatistical inference manages to eliminate all sources of personalbias and idiosyncrasy. The Bayesian considers subjective assumptionsto be an irreducible part of scientific reasoning and sees no harm inmaking them explicit. The frequentist conception of evidence based on\(p\)-values avoids these explicitly subjective elements, but at theprice of a misleading impression of objectivity and frequent abuse inpractice. A defense of frequentist inference should, in our opinion,stress that the relatively rigid rules for interpreting statisticalevidence facilitate communication and assessment of research resultsin the scientific community—something that is harder to achievefor a Bayesian. We now turn from specific methods for stating andinterpreting evidence to a radical criticism of the idea that there isa rational scientific method.

4.3 Feyerabend: The Tyranny of the Rational Method

In his writings of the 1970s,Paul Feyerabend launched a profound attack on the rationality and objectivity ofscientific method. His position is exceptional in the philosophicalliterature since traditionally, the threat for objective andsuccessful science is located in contextual rather than epistemicvalues. Feyerabend turns this view upside down: it is the“tyranny” of rational method, and the emphasis onepistemic rather than contextual values that prevents us from having ascience in the service of society. Moreover, he welcomes a diversityof different personal, also idiosyncratic perspectives, thus denyingthe idea that freedom from personal “bias” isepistemically and socially beneficial.

The starting point of Feyerabend’s criticism of rational methodis the thesis that strict epistemic rules such as those expressed bythe VFI only suppress an open exchange of ideas, extinguish scientificcreativity and prevent a free and truly democratic science. In hisclassic “Against Method” (1975: chs. 8–13),Feyerabend elaborates on this criticism by examining a famous episodein the history of science. When the Catholic Church objected toGalilean mechanics, it had the better arguments by the standards ofseventeenth-century science. Their conservatism in their position wasscientifically backed: Galilei’s telescopes were unreliable forcelestial observations, and many well-established phenomena (no fixedstar parallax, invariance of laws of motion) could not yet beexplained in the heliocentric system. With hindsight, Galilei managedto achieve groundbreaking scientific progress just because hedeliberately violated rules of scientific reasoning. HenceFeyerabend’s dictum “Anything goes”: no methodologywhatsoever is able to capture the creative and often irrational waysby which science deepens our understanding of the world. Goodscientific reasoning cannot be captured by rational method, as Carnap,Hempel and Popper postulated.

The drawbacks of an objective, value-free and method-bound view onscience and scientific method are not only epistemic. Such a viewnarrows down our perspective and makes us less free, open-minded,creative, and ultimately, less human in our thinking (Feyerabend 1975:154). It is therefore neither possible nor desirable to have anobjective, value-free science (cf. Feyerabend 1978: 78–79). As aconsequence, Feyerabend sees traditional forms of inquiry about ourworld (e.g., Chinese medicine) on a par with their Westerncompetitors. He denounces appeals to “objective” standardsas rhetorical tools for bolstering the epistemic authority of a smallintellectual elite (=Western scientists), and as barely disguisedstatements of preference for one’s own worldview:

there is hardly any difference between the members of a“primitive” tribe who defend their laws because they arethe laws of the gods […] and a rationalist who appeals to“objective” standards, except that the former know whatthey are doing while the latter does not. (1978: 82)

In particular, when discussing other traditions, we often project ourown worldview and value judgments into them instead of making animpartial comparison (1978: 80–83). There is no purely rationaljustification for dismissing other perspectives in favor of theWestern scientific worldview—the insistence on our Westernapproach may be as justified as insisting on absolute space and timeafter the Theory of Relativity.

The Galilei example also illustrates that personal perspective andidiosyncratic “bias” need not be bad for science.Feyerabend argues further that scientific research is accountable tosociety and should be kept in check by democratic institutions, andlaymen in particular. Their particular perspectives can help todetermine the funding agenda and to set ethical standards forscientific inquiry, but also be useful for traditionally value-freetasks such as choosing an appropriate research method and assessingscientific evidence. Feyerabend’s writings on this issue weremuch influenced by witnessing the Civil Rights Movement in the U.S.and the increasing emancipation of minorities, such as Blacks, Asiansand Hispanics.

All this is not meant to say that truth loses its function as anormative concept, nor that all scientific claims are equallyacceptable. Rather, Feyerabend advocates anepistemicpluralism that accepts diverse approaches to acquiringknowledge. Rather than defending a narrow and misleading ideal ofobjectivity, science should respect the diversity of values andtraditions that drive our inquiries about the world (1978:106–107). This would put science back into the role it hadduring the scientific revolution or the Enlightenment: as a liberatingforce that fought intellectual and political oppression by thesovereign, the nobility or the clergy. Objections to this view arediscussed at the end ofsection 5.2.

5. Objectivity as a Feature of Scientific Communities and Their Practices

This section addresses various accounts that regard scientificobjectivity essentially as a function of social practices in scienceand the social organization of the scientific community. All theseaccounts reject the characterization of scientific objectivity as afunction of correspondence between theories and the world, as afeature of individual reasoning practices, or as pertaining toindividual studies and experiments (see also Douglas 2011). Instead,they evaluate the objectivity of acollective of studies, aswell as the methods and community practices that structure and guidescientific research. More precisely, they adopt a meta-analyticperspective for assessing the reliability of scientific results(section 5.1), and they construct objectivity from a feministperspective: as an open interchange of mutual criticism, or as beinganchored in the “situatedness” of our scientific practicesand the knowledge we gain (section 5.2).

5.1 Reproducibility and the Meta-Analytic Perspective

The collectivist perspective is especially useful when an entirediscipline enters a stage of crisis: its members become convinced thata significant proportion of findings are not trustworthy. Acontemporary example of such a situation is thereplicationcrisis, which was briefly mentioned in the previous sectionand concerns thereproducibility of scientific knowledgeclaims in a variety of different fields (most prominently: psychology,biology, medicine). Large-scale replication projects have noticed thatmany findings which we considered as an integral part of scientificknowledge failed to replicate in settings that were designed to mimicthe original experiment as closely as possible (e.g., Open ScienceCollaboration 2015). Successful attempts at replicating anexperimental result have long been argued to provide evidence offreedom from particular kinds of artefacts and thus thetrustworthiness of the result. Compare theentry on experiment in physics. Likewise, failure to replicate indicates that either the originalfinding, the result of the replication attempt, or both, arebiased—though see John Norton’s (ms., ch. 3—seeOther Internet Resources) arguments that the evidential value of(failed) replications crucially depends on researchers’ materialbackground assumptions.

When replication failures in a discipline are particularlysignificant, one may conclude that the published literature lacksobjectivity—at a minimum the discipline fails to inspire trustthat its findings are more than artefacts of the researchers’efforts. Conversely, when observed effects can be replicated infollow-up experiments, a kind of objectivity is reached that goesbeyond the ideas of freedom from personal bias, mechanicalobjectivity, and subject-independent measurement, discussed insection 4.1.

Freese and Peterson (2018) call this ideastatisticalobjectivity. It grounds in the view that even the mostscrupulous and diligent researchers cannot achieve full objectivityall by themselves. The term “objectivity” instead appliesto a collection or population of studies, withmeta-analysis (a formal method for aggregating theresults from ranges of studies) as the “apex ofobjectivity” (Freese and Peterson 2018, 304; see also Stegenga2011, 2018). In particular, aggregating studies from differentresearchers may provide evidence of systematic bias and questionableresearch practices (QRP) in the published literature. This diagnosticfunction of meta-analysis for detecting violations of objectivity isenhanced by statistical techniques such as the funnel plot and the\(p\)-curve (Simonsohn et al. 2014).

Apart from this epistemic dimension, research on statisticalobjectivity also has an activist dimension: methodologists urgeresearchers to make publicly available essential parts of theirresearch before the data analysis starts, and to make their methodsand data sources more transparent. For example, it is conjectured thatthe replicability (and thus objectivity) of science will increase bymaking all data available online, by preregistering experiments, andby using the registered reports model for journal articles (i.e., thejournal decides on publication before data collection on the basis ofthe significance of the proposed research as well as the experimentaldesign). The idea is that transparency about the data set and theexperimental design will make it easier to stage a replication of anexperiment and to assess its methodological quality. Moreover,publicly committing to a data analysis plan beforehand will lower therate of QRPs and of attempts toaccommodate data tohypotheses rather than making proper predictions.

All in all, statistical objectivity moves the discussion ofobjectivity to the level of population of studies. There, it takes upand modifies several conceptions of objectivity that we have seenbefore: most prominently, freedom of subjective bias, which isreplaced with collective bias and pernicious conventions, and thesubject-independent measurement of a physical quantity, which isreplaced by reproducibility of effects.

5.2 Feminist and Standpoint Epistemology

Traditional notions of objectivity as faithfulness to facts or freedomof contextual values have also been challenged from a feministperspective. These critiques can be grouped in three major researchprograms: feminist epistemology, feminist standpoint theory andfeminist postmodernism (Crasnow 2013). The program offeministepistemology explores the impact of sex and gender on theproduction of scientific knowledge. More precisely, feministepistemology highlights the epistemic risks resulting from thesystematic exclusion of women from the ranks of scientists, and theneglect of women as objects of study. Prominent case studies are theneglect of female orgasm in biology, testing medical drugs on maleparticipants only, focusing on male specimen when studying the socialbehavior of primates, and explaining human mating patterns by means ofimaginary neolithic societies (e.g., Hrdy 1977; Lloyd 1993, 2005). Seealso theentry on feminist philosophy of biology.

Often but not always, feminist epistemologists go beyond pointing outwhat they regard as androcentric bias and reject the value-free idealaltogether—with an eye on the social and moral responsibility ofscientific inquiry. They try to show that a value-laden science canalso meet important criteria for being epistemically reliable andobjective (e.g., Anderson 2004; Kourany 2010). A classicalrepresentative of such efforts is Longino’s (1990)contextual empiricism. She reinforces Popper’sinsistence that “the objectivity of scientific statements liesin the fact that they can be inter-subjectively tested” (1934[2002]: 22), but unlike Popper, she conceives scientific knowledgeessentially as a social product. Thus, our conception of scientificobjectivity must directly engage with the social process thatgenerates knowledge. Longino assigns a crucial function to socialsystems of criticism in securing the epistemic success of science.Specifically, she develops an epistemology which regards a method ofinquiry as “objective to the degree that it permitstransformative criticism” (Longino 1990: 76).For an epistemic community to achieve transformative criticism, theremust be:

avenues for criticism: criticism is an essential partof scientific institutions (e.g., peer review);
shared standards: the community must share a set ofcognitive values for assessing theories (more on this insection 3.1);
uptake of criticism: criticism must be able totransform scientific practice in the long run;
equality of intellectual authority: intellectualauthority must be shared equally among qualified practitioners.

Longino’s contextual empiricism can be understood as adevelopment of John Stuart Mill’s view that beliefs should neverbe suppressed, independently of whether they are true or false. Eventhe most implausible beliefs might be true, and even if they arefalse, they might contain a grain of truth which is worth preservingor helps to better articulate true beliefs (Mill 1859 [2003: 72]). Theunderlying intuition is supported by recent empirical research on theepistemic benefits of a diversity of opinions and perspectives (Page2007). By stressing the social nature of scientific knowledge, and theimportance of criticism (e.g., with respect to potential androcentricbias and inclusive practice), Longino’s account fits into thebroader project of feminist epistemology.

Standpoint theory undertakes a more radical attack ontraditional scientific objectivity. This view develops Marxist ideasto the effect that epistemic position is related to, and a product of,social position. Feminist standpoint theory builds on these ideas butfocuses on gender, racial and other social relations. Feministstandpoint theorists and proponents of“situatedknowledge” such as Donna Haraway (1988), Sandra Harding(1991, 2015a, 2015b) and Alison Wylie (2003) deny the internalcoherence of a view from nowhere: all human knowledge is at basehuman knowledge and therefore necessarily perspectival. Butthey argue more than that. Not only is perspectivality the humancondition, it is also a good thing to have. This is becauseperspectives, especially the perspectives of underprivileged classesand groups in society, come along with epistemic benefits. These ideasare controversial but they draw attention to the possibility thatattempts to rid science of perspectives might not only be futile butalso costly: they prevent scientists from having the epistemicbenefits certain standpoints afford and from developing knowledgefor marginalized groups in society. The perspectival stancecan also explain why criteria for objectivity often vary with context:the relative importance of epistemic virtues is a matter of goals andinterests—in other words, standpoint.

By endorsing a perspectival stance, feminist standpoint theory rejectsclassical elements of scientific objectivity such as neutrality andimpartiality (seesection 3.1 above). This is a notable difference to feminist epistemology, whichis in principle (though not always in practice) compatible withtraditional views of objectivity. Feminist standpoint theory is also apolitical project. For example, Harding (1991, 1993) demands thatscientists, their communities and their practices—in otherwords, the ways through which knowledge is gained—beinvestigated as rigorously as the object of knowledge itself. Thisidea she refers to as“strongobjectivity” replaces the “weak” conceptionof objectivity in the empiricist tradition: value-freedom,impartiality, rigorous adherence to methods of testing and inference.Like Feyerabend, Harding integrates a transformation of epistemicstandards in science into a broader political project of renderingscience more democratic and inclusive. On the other hand, she isexposed to similar objections (see also Haack 2003). Isn’t itgrossly exaggerated to identify class, race and gender as importantfactors in the construction of physical theories? Doesn’t thefeminist approach—like social constructivistapproaches—lose sight of the particular epistemic qualities ofscience? Should non-scientists really have as much authority astrained scientists? To whom does the condition of equally sharedintellectual authority apply? Nor is it clear—especially intimes of fake news and filter bubbles—whether it is always agood idea to subject scientific results to democratic approval. Thereis no guarantee (arguably there are few good reasons to believe) thatdemocratized or standpoint-based science leads to more reliabletheories, or better decisions for society as a whole.

6. Issues in the Special Sciences

So far everything we discussed was meant to apply across all or atleast most of the sciences. In this section we will look at a numberof specific issues that arise in the social sciences, in economics,and in evidence-based medicine.

6.1 Max Weber and Objectivity in the Social Sciences

There is a long tradition in the philosophy of social sciencemaintaining that there is a gulf in terms of both goals as well asmethods between the natural and the social sciences. This tradition,associated with thinkers such as the neo-Kantians Heinrich Rickert andWilhelm Windelband, the hermeneuticist Wilhelm Dilthey, thesociologist-economist Max Weber, and the twentieth-centuryhermeneuticists Hans-Georg Gadamer and Michael Oakeshott, holds thatunlike the natural sciences whose aim it is to establish natural lawsand which proceed by experimentation and causal analysis, the socialsciences seek understanding (“Verstehen”) ofsocial phenomena, the interpretive examination of the meaningsindividuals attribute to their actions (Weber 1904 [1949]; Weber 1917[1949]; Dilthey 1910 [1986]; Windelband 1915; Rickert 1929; Oakeshott1933; Gadamer 1960 [1989]). See also the entries onhermeneutics andMax Weber.

Understood this way, social science lacks objectivity in more than onesense. One of the more important debates concerning objectivity in thesocial sciences concerns the role value judgments play and,importantly, whether value-laden research entails claims about thedesirability of actions. Max Weber held that the social sciences arenecessarily value laden. However, they can achieve some degree ofobjectivity by keeping out the social researcher’s views aboutwhether agents’ goals are commendable. In a similar vein,contemporary economics can be said to be value laden because itpredicts and explains social phenomena on the basis of agents’preferences. Nevertheless, economists are adamant that economists arenot in the business of telling people what they ought to value. Moderneconomics is thus said to be objective in the Weberian sense of“absence of researchers’values”—a conception that we discussed in detailinsection 3.

In his widely cited essay “‘Objectivity’ in SocialScience and Social Policy” (Weber 1904 [1949]), Weber arguedthat the idea of an aperspectival social science was meaningless:

There is no absolutely objective scientific analysis of […]“social phenomena” independent of special and“one-sided” viewpoints according to which expressly ortacitly, consciously or unconsciously they are selected, analyzed andorganized for expository purposes. (1904 [1949: 72])
All knowledge of cultural reality, as may be seen, is always knowledgefrom particular points of view. (1904 [1949:. 81])

The reason for this is twofold. First, social reality is too complexto admit of full description and explanation. So we have to select.But, perhaps in contraposition to the natural sciences, we cannot justselect those aspects of the phenomena that fall under universalnatural laws and treat everything else as “unintegratedresidues” (1904 [1949: 73]). This is because, second, in thesocial sciences we want to understand social phenomena in theirindividuality, that is, in their unique configurations that havesignificance for us.

Values solve a selection problem. They tell us what research questionswe ought to address because they inform us about the culturalimportance of social phenomena:

Only a small portion of existing concrete reality is colored by ourvalue-conditioned interest and it alone is significant to us. It issignificant because it reveals relationships which are important touse due to their connection with our values. (1904 [1949: 76])

It is important to note that Weber did not think that social andnatural science were different in kind, as Dilthey and others did.Social science too examines the causes of phenomena of interest, andnatural science too often seeks to explain natural phenomena in theirindividual constellations. The role of causal laws is different in thetwo fields, however. Whereas establishing a causal law is often an endin itself in the natural sciences, in the social sciences laws play anattenuated and accompanying role as mere means to explain culturalphenomena in their uniqueness.

Nevertheless, for Weber social science remains objective in at leasttwo ways. First, once research questions of interest have beensettled, answers about the causes of culturally significant phenomenado not depend on the idiosyncrasies of an individual researcher:

But it obviously does not follow from this that research in thecultural sciences can only have results which are“subjective” in the sense that they are valid for oneperson and not for others. […] For scientific truth isprecisely what is valid for all who seek the truth. (Weber 1904 [1949:84], emphasis original)

The claims of social science can therefore be objective in our thirdsense (see section 4). Moreover, by determining that a given phenomenon is “culturallysignificant” a researcher reflects on whether or not a practiceis “meaningful” or “important”, and notwhether or not it is commendable: “Prostitution is a culturalphenomenon just as much as religion or money” (1904 [1949: 81]).An important implication of this view came to the fore in theso-called “Werturteilsstreit” (quarrel concerningvalue judgments) of the early 1900s. In this debate, Weber maintainedagainst the “socialists of the lectern” around GustavSchmoller the position that social scientists qua scientists shouldnot be directly involved in policy debates because it was not the aimof science to examine the appropriateness of ends. Given a policygoal, a social scientist could make recommendations about effectivestrategies to reach the goal; but social science was to be value-freein the sense of not taking a stance on the desirability of the goalsthemselves. This leads us to our conception of objectivity as freedomfrom value judgments.

6.2 Contemporary Rational Choice Theory

Contemporary mainstream economists hold a view concerning objectivitythat mirrors Max Weber’s (see above). On the one hand, it isclear that value judgments are at the heart of economic theorizing.“Preferences” are a key concept of rational choice theory,the main theory in contemporary mainstream economics. Preferences areevaluations. If an individual prefers \(A\) to \(B\), shevalues \(A\) higher than \(B\) (Hausman 2012). Thus, to theextent that economists predict and explain market behavior in terms ofrational choice theory, they predict and explain market behavior in away laden with value judgments.

However, economists are not themselves supposed to take a stance aboutwhether or not whatever individuals value is also“objectively” good in a stronger sense:

[…] that an agent is rational from [rational choicetheory]’s point of view does not mean that the course of actionshe will choose is objectively optimal. Desires do not have to alignwith any objective measure of “goodness”: I may want torisk swimming in a crocodile-infested lake; I may desire to smoke ordrink even though I know it harms me. Optimality is determined by theagent’s desires, not the converse. (Paternotte 2011:307–8)

In a similar vein, Gul and Pesendorfer write:

However, standard economics has no therapeutic ambition, i.e., it doesnot try to evaluate or improve the individual’s objectives.Economics cannot distinguish between choices that maximize happiness,choices that reflect a sense of duty, or choices that are the responseto some impulse. Moreover, standard economics takes no position on thequestion of which of those objectives the agent should pursue. (Guland Pesendorfer 2008: 8)

According to the standard view, all that rational choice theorydemands is that people’s preferences are (internally)consistent; it has no business in telling people what they ought toprefer, whether their preferences are consistent with external normsor values. Economics is thus value-laden, but laden with the values ofthe agents whose behavior it seeks to predict and explain and not withthe values of those who seek to predict and explain this behavior.

Whether or not social science, and economics in particular, can beobjective in this—Weber’s and the contemporaryeconomists’—sense is controversial. On the one hand, thereare some reasons to believe that rational choice theory (which is atwork not only in economics but also in political science and othersocial sciences) cannot be applied to empirical phenomena withoutreferring to external norms or values (Sen 1993; Reiss 2013).

On the other hand, it is not clear that economists and other socialscientists qua social scientists shouldn’t participate in adebate about social goals. For one thing, trying to do welfareanalysis in the standard Weberian way tends to obscure rather than toeliminate normative commitments (Putnam and Walsh 2007). Obscuringvalue judgments can be detrimental to the social scientist as policyadviser because it will hamper rather than promote trust in socialscience. For another, economists are in a prime position to contributeto ethical debates, for a variety of reasons, and should thereforetake this responsibility seriously (Atkinson 2001).

6.3 Evidence-based Medicine and Social Policy

The same demands calling for “mechanical objectivity” inthe natural sciences and quantification in the social and policysciences in the nineteenth century and mid-twentieth century areresponsible for a recent movement in biomedical research, which, evenmore recently, have swept to contemporary social science and policy.Early proponents of so-called “evidence-based medicine”made their pursuit of a downplay of the “human element” inmedicine plain:

Evidence-based medicine de-emphasizes intuition, unsystematic clinicalexperience, and pathophysiological rationale as sufficient grounds forclinical decision making and stresses the examination of evidence fromclinical research. (Guyatt et al. 1992: 2420)

To call the new movement “evidence-based” is a misnomerstrictly speaking, as intuition, clinical experience andpathophysiological rationale can certainly constitute evidence. Butproponents of evidence-based practices have a much narrower concept ofevidence in mind: analyses of the results of randomized controlledtrials (RCTs). This movement is now very strong in biomedicalresearch, development economics and a number of areas of socialscience, especially psychology, education and social policy, andespecially in the English speaking world.

The goal is to replace subjective (biased, error-prone, idiosyncratic)judgments by mechanically objective methods. But, as in other areas,attempting to mechanize inquiry can lead to reduced accuracy andutility of the results.

Causal relations in the social and biomedical sciences hold on accountof highly complex arrangements of factors and conditions. Whether forinstance a substance is toxic depends on details of the metabolicsystem of the population ingesting it, and whether an educationalpolicy is effective on the constellation of factors that affect thestudents’ learning progress. If an RCT was conductedsuccessfully, the conclusion about the effectiveness of the treatment(or toxicity of a substance) under test is certain for the particulararrangement of factors and conditions of the trial (Cartwright 2007).But unlike the RCT itself, many of whose aspects can be (relatively)mechanically implemented, applying the result to a new setting(recommending a treatment to a patient, for instance) always involvessubjective judgments of the kind proponents of evidence-basedpractices seek to avoid—such as judgments about the similarityof the test to the target or policy population.

On the other hand, RCTs can be regarded as “debiasingprocedure” because they prevent researchers from allocatingtreatments to patients according to their personal interests, so thatthe healthiest (or smartest or…) subjects get theresearcher’s favorite therapy. While unbalanced allocations cancertainly happen by chance, randomization still provides some warrantthat the allocation was not doneon purpose with a view topromoting somebody’s interests.A priori, theexperimental procedure is thus more impartial with respect to theinterests at stake. It has thus been argued that RCTs in medicine,while no guarantor of the best outcomes, were adopted by the U.S. Foodand Drugs Administration (FDA) to different degrees during the 1960sand 1970s in order to regain public trust in its decisions abouttreatments, which it had lost due to the thalidomide and otherscandals (Teira and Reiss 2013; Teira 2010). It is important tonotice, however, that randomization is at best effective with respectto one kind of bias, viz. selection bias. Important other epistemicconcerns are not addressed by the procedure but should not be ignored(Worrall 2002).

7. The Unity and Disunity of Scientific Objectivity

In sections 2–5, we have encountered various concepts ofscientific objectivity and their limitations. This prompts thequestion of how unified (or disunified) scientific objectivity is as aconcept: Is there something substantive shared by all of theseanalyses? Or is objectivity, as Heather Douglas (2004) puts it, an“irreducibly complex” concept?

Douglas defendspluralism about scientificobjectivity and distinguishes three areas of application ofthe concept: (1) interaction of humans with the world, (2) individualreasoning processes, (3) social processes in science. Within eacharea, there are various distinct senses which are again irreducible toeach other and do not have a common core meaning. This does not meanthat the senses are unrelated; they share a complex web ofrelationships and can also support each other—for example,eliminating values from reasoning may help to achieve proceduralobjectivity. For Douglas, reducing objectivity to a single coremeaning would be a simplification without benefits; instead of acomplex web of relations between different senses of objectivity wewould obtain an impoverished concept out of touch with scientificpractice. Similar arguments and pluralist accounts can be found inMegill (1994), Janack (2002) and Padovani et al. (2015)—see alsoAxtell (2016).

It has been argued, however, that pluralist approaches give up tooquickly on the idea that the different senses of objectivity share oneor several important common elements. As we have seen in section4.1 and5.1, scientific objectivity andtrust in science areclosely connected. Scientific objectivity is desirable because to theextent that science is objective we have reasons trust scientists,their results and recommendations (cf. Fine 1998: 18). Thus, perhapswhat is unifying among the difference senses of objectivity is thateach sense describes a feature of scientific practice that is able toinspire trust in science.

Building on this idea, Inkeri Koskinen has recently argued that it isin fact not trust but reliance that we are after (Koskinenforthcoming). Trust is something that can be betrayed, but onlyindividuals can betray whereas objectivity pertains to institutions,practices, results, etc. We call scientific institutions, practices,results, etc. objective to the extent that we have reasons to rely onthem. The analysis does not stop here, however. There is a distinctview about objectivity that is behind Daston and Galison’shistorical epistemology of the concept and has been defended by IanHacking: that objectivity is not a—positive—virtue butrather the absence of this or that vice (Hacking 2015: 26). Speakingof objectivity in imaging, for instance, Daston and Galison write thatthe goal is to

let the specimen appear without that distortion characteristic of theobserver’s personal tastes, commitments, or ambitions. (Dastonand Galison 2007: 121)

Koskinen picks up this idea ofobjectivity as absence ofvice and argues that it is specifically the aversion ofepistemic risks for which the term is reserved. Epistemicrisks comprise “any risk of epistemic error that arises anywhereduring knowledge practices’ (Biddle and Kukla 2017: 218) such asthe risk of having mistaken beliefs, the risk of errors in reasoningand risks related to operationalization, concept formation, and modelchoice. Koskinen argues that only those epistemic risks that relate tofailings of scientists as human beings are relevant to objectivity(Koskinen forthcoming: 13):

For instance, when the results of an experiment are incorrect becauseof malfunctioning equipment, we do not worry aboutobjectivity—we just say that the results should not be takeninto account. [...] So it is only when the epistemic risk is relatedto our own failings, and is hard to avert, that we start talking aboutobjectivity. Illusions, subjectivity, idiosyncrasies, and collectivebiases are important epistemic risks arising from our imperfections asepistemic agents.

Koskinen understands her account as a response to Hacking’s(2015) criticism that we should stop talking about objectivityaltogether. According to Hacking, “objectivity” is an“elevator” or second-level word, similar to“true” or “real”—“Instead ofsaying that the cat is on the mat, we move up one story and and saythat it is true that the cat is on the mat” (2015: 20). Herecommends to stick to ground-level questions and worry about whetherspecific sources of error have been controlled. (A similar eliminationrequest with respect to the labels “objective” and“subjective” in statistical inference has been advanced byGelman and Hennig (2017).) In focussing on averting specific epistemicrisks, Koskinen’s account does precisely that. Koskinen arguesthat a unified account of objectivity as averting epistemic riskstakes into account Hacking’s negative stance and explains at thesame time important features of the concept—for example, whyobjectivity does not imply certainty and why it varies withcontext.

The strong point of this account is that none of the threats to apeculiar analysis puts scientific objectivity at risk. We can (and infact, we do) rely on scientific practices that represent the worldfrom a perspective and where non-epistemic values affect outcomes anddecisions. What is left open by Koskinen’s account is thenormative question of what a scientist who cares about her experimentsand inferences being objective should actually do. That is, thephilosophical ideas we have reviewed in this section stay mainly onthe descriptive level and do not give an actual guideline for workingscientists. Connecting the abstract philosophical analysis today-to-day work in science remains an open problem.

8. Conclusions

So is scientific objectivity desirable? Is it attainable? That, as wehave seen, depends crucially on how the term is understood. We havelooked in detail at four different conceptions of scientificobjectivity: faithfulness to facts, value-freedom, freedom frompersonal biases, and features of community practices. In each case,there are at least some reasons to believe that either science cannotdeliver full objectivity in this sense, or that it would not be a goodthing to try to do so, or both. Does this mean we should give up theidea of objectivity in science?

We have shown that it is hard to define scientific objectivity interms of a view from nowhere, value freedom, or freedom from personalbias. It is a lot harder to say anything positive about the matter.Perhaps it is related to a thorough critical attitude concerningclaims and findings, as Popper thought. Perhaps it is the fact thatmany voices are heard, equally respected and subjected to acceptedstandards, as Longino defends. Perhaps it is something elsealtogether, or a combination of several factors discussed in thisarticle.

However, one should not (as yet) throw out the baby with thebathwater. Like those who defend a particular explication ofscientific objectivity, the critics struggle to explain what makesscience objective, trustworthy and special. For instance, ourdiscussion of the value-free ideal (VFI) revealed that alternatives tothe VFI are as least as problematic as the VFI itself, and that theVFI may, with all its inadequacies, still be a useful heuristic forfostering scientific integrity and objectivity. Similarly, althoughentirely “unbiased” scientific procedures may beimpossible, there are many mechanisms scientists can adopt forprotecting their reasoning against undesirable forms of bias, e.g.,choosing an appropriate method of statistical inference, beingtransparent about different stages of the research process andavoiding certain questionable research practices.

Whatever it is, it should come as no surprise that finding a positivecharacterization of what makes science objective is hard. If we knewan answer, we would have done no less than solve the problem ofinduction (because we would know what procedures or forms oforganization are responsible for the success of science). Work on thisproblem is an ongoing project, and so is the quest for understandingscientific objectivity.

Bibliography

Anderson, Elizabeth, 2004, “Uses of Value Judgments inScience: A General Argument, with Lessons from a Case Study ofFeminist Research on Divorce”,Hypatia, 19(1):1–24. doi:10.1111/j.1527-2001.2004.tb01266.x
Atkinson, Anthony B., 2001, “The Strange Disappearance ofWelfare Economics”,Kyklos, 54(2‐3):193–206. doi:10.1111/1467-6435.00148
Axtell, Guy, 2016,Objectivity, Cambridge: PolityPress.
Bakker, Marjan, Annette van Dijk, and Jelte M. Wicherts, 2012,“The Rules of the Game Called Psychological Science”,Perspectives on Psychological Science, 7(6): 543–554.doi:10.1177/1745691612459060
Bernardo, J.M., 2012, “Integrated Objective BayesianEstimation and Hypothesis Testing”, inBayesian Statistics9: Proceedings of the Ninth Valencia Meeting, J.M. Bernardo etal. (eds.), Oxford: Oxford University Press, 1–68.
Betz, Gregor, 2013, “In Defence of the Value FreeIdeal”,European Journal for Philosophy of Science,3(2): 207–220. doi:10.1007/s13194-012-0062-x
Biddle, Justin B., 2013, “State of the Field: TransientUnderdetermination and Values in Science”,Studies inHistory and Philosophy of Science Part A, 44(1): 124–133.doi:10.1016/j.shpsa.2012.09.003
Biddle, Justin B. and Rebecca Kukla, 2017, “The Geography ofEpistemic Risk”, inExploring Inductive Risk: Case Studiesof Values in Science, Kevin C. Elliott and Ted Richards (eds.),New York: Oxford University Press, 215–238.
Bloor, David, 1982, “Durkheim and Mauss Revisited:Classification and the Sociology of Knowledge”,Studies inHistory and Philosophy of Science Part A, 13(4): 267–297.doi:10.1016/0039-3681(82)90012-7
Braithwaite, R. B., 1953,Scientific Explanation,Cambridge: Cambridge University Press.
Carnap, Rudolf, 1950 [1962],Logical Foundations ofProbability, second edition, Chicago: University of ChicagoPress.
Cartwright, Nancy, 2007, “Are RCTs the GoldStandard?”,BioSocieties, 2(1): 11–20.doi:10.1017/S1745855207005029
Chang, Hasok, 2004,Inventing Temperature: Measurement andScientific Progress, Oxford: Oxford University Press.doi:10.1093/0195171276.001.0001
Churchman, C. West, 1948,Theory of ExperimentalInference, New York: Macmillan.
Collins, H. M., 1985,Changing Order: Replication andInduction in Scientific Practice, Chicago, IL: University ofChicago Press.
–––, 1994, “A Strong Confirmation of theExperimenters’ Regress”,Studies in History andPhilosophy of Science Part A, 25(3): 493–503.doi:10.1016/0039-3681(94)90063-9
Cranor, Carl F., 1993,Regulating Toxic Substances: APhilosophy of Science and the Law, New York: Oxford UniversityPress.
Crasnow, Sharon, 2013, “Feminist Philosophy of Science:Values and Objectivity: Feminist Philosophy of Science”,Philosophy Compass, 8(4): 413–423.doi:10.1111/phc3.12023
Daston, Lorraine, 1992, “Objectivity and the Escape fromPerspective”,Social Studies of Science, 22(4):597–618. doi:10.1177/030631292022004002
Daston, Lorraine and Peter Galison, 1992, “The Image ofObjectivity”,Representations, 40(special issue: SeeingScience): 81–128. doi:10.2307/2928741
–––, 2007,Objectivity, Cambridge, MA:MIT Press.
Dilthey, Wilhelm, 1910 [1981],Der Aufbau der geschichtlichenWelt in den Geisteswissenschaften, Frankfurt am Main:Suhrkamp.
Dorato, Mauro, 2004, “Epistemic and Nonepistemic Values inScience”, in Machamer and Wolters 2004: 52–77.
Douglas, Heather E., 2000, “Inductive Risk and Values inScience”,Philosophy of Science, 67(4): 559–579.doi:10.1086/392855
–––, 2004, “The Irreducible Complexity ofObjectivity”,Synthese, 138(3): 453–473.doi:10.1023/B:SYNT.0000016451.18182.91
–––, 2009,Science, Policy, and theValue-Free Ideal, Pittsburgh, PA: University of PittsburghPress.
–––, 2011, “Facts, Values, andObjectivity”, Jarvie and Zamora Bonilla 2011:513–529.
Duhem, Pierre Maurice Marie, 1906 [1954],La théoriephysique. Son objet et sa structure, Paris: Chevalier et Riviere;translated by Philip P. Wiener,The Aim and Structure of PhysicalTheory, Princeton, NJ: Princeton University Press, 1954.
Dupré, John, 2007, “Fact and Value”, inKincaid, Dupré, and Wylie 2007: 24–71.
Earman, John, 1992,Bayes or Bust? A Critical Examination ofBayesian Confirmation Theory, Cambridge/MA: The MITPress.
Elliott, Kevin C., 2011, “Direct and Indirect Roles forValues in Science”,Philosophy of Science, 78(2):303–324. doi:10.1086/659222
Feyerabend, Paul K., 1962, “Explanation, Reduction andEmpiricism”, in H. Feigl and G. Maxwell (ed.),ScientificExplanation, Space, and Time, (Minnesota Studies in thePhilosophy of Science, 3), Minneapolis, MN: University of MinnesotaPress, pp. 28–97.
–––, 1975,Against Method, London:Verso.
–––, 1978,Science in a Free Society,London: New Left Books.
Fine, Arthur, 1998, “The Viewpoint of No-One inParticular”,Proceedings and Addresses of the AmericanPhilosophical Association, 72(2): 7. doi:10.2307/3130879
Fisher, Ronald Aylmer, 1935,The Design of Experiments,Edinburgh: Oliver and Boyd.
–––, 1956,Statistical Methods andScientific Inference, New York: Hafner.
Franklin, Allan, 1994, “How to Avoid theExperimenters’ Regress”,Studies in History andPhilosophy of Science Part A, 25(3): 463–491.doi:10.1016/0039-3681(94)90062-0
–––, 1997, “Calibration”,Perspectives on Science, 5(1): 31–80.
Freese, Jeremy and David Peterson, 2018, “The Emergence ofStatistical Objectivity: Changing Ideas of Epistemic Vice and Virtuein Science”,Sociological Theory, 36(3): 289–313.doi:10.1177/0735275118794987
Gadamer, Hans-Georg, 1960 [1989],Wahrheit und Methode,Tübingen : Mohr. Translated asTruth and Method,2^nd edition, Joel Weinsheimer and Donald G. Marshall(trans), New York, NY: Crossroad, 1989.
Gelman, Andrew and Christian Hennig, 2017, “BeyondSubjective and Objective in Statistics”,Journal of theRoyal Statistical Society: Series A (Statistics in Society),180(4): 967–1033. doi:10.1111/rssa.12276
Giere, Ronald N., 2006,Scientific Perspectivism,Chicago, IL: University of Chicago Press.
Good, Irving John, 1950,Probability and the Weighing ofEvidence, London: Charles Griffin.
Gul, Faruk and Wolfgang Pesendorfer, 2008, “The Case forMindless Economics”, inThe Foundations of Positive andNormative Economics: a Handbook, Andrew Caplin and AndrewSchotter (eds), New York, NY: Oxford University Press, pp.3–39.
Guyatt, Gordon, John Cairns, David Churchill, Deborah Cook, BrianHaynes, Jack Hirsh, Jan Irvine, Mark Levine, Mitchell Levine, JimNishikawa, et al., 1992, “Evidence-Based Medicine: A NewApproach to Teaching the Practice of Medicine”,JAMA: TheJournal of the American Medical Association, 268(17):2420–2425. doi:10.1001/jama.1992.03490170092032
Haack, Susan, 2003,Defending Science—Within Reason:Between Scientism and Cynicism, Amherst, NY: PrometheusBooks.
Hacking, Ian, 1965,Logic of Statistical Inference,Cambridge: Cambridge University Press.doi:10.1017/CBO9781316534960
–––, 2015, “Let’s Not Talk AboutObjectivity”, in Padovani, Richardson, and Tsou 2015:19–33. doi:10.1007/978-3-319-14349-1_2
Hanson, Norwood Russell, 1958,Patterns of Discovery: AnInquiry into the Conceptual Foundations of Science, Cambridge:Cambridge University Press.
Haraway, Donna, 1988, “Situated Knowledges: The ScienceQuestion in Feminism and the Privilege of Partial Perspective”,Feminist Studies, 14(3): 575–599.doi:10.2307/3178066
Harding, Sandra, 1991,Whose Science? Whose Knowledge?Thinking from Women’s Lives, Ithaca, NY: Cornell UniversityPress.
–––, 1993, “Rethinking StandpointEpistemology: What is Strong Objectivity?”, inFeministEpistemologies, Linda Alcoff and Elizabeth Potter (ed.), NewYork, NY: Routledge, 49–82.
–––, 2015a,Objectivity and Diversity:Another Logic of Scientific Research, Chicago: University ofChicago Press.
–––, 2015b, “After Mr. Nowhere: What Kindof Proper Self for a Scientist?”,Feminist PhilosophyQuarterly, 1(1): 1–22. doi:10.5206/fpq/2015.1.2
Hausman, Daniel M., 2012,Preference, Value, Choice, andWelfare, New York: Cambridge University Press.doi:10.1017/CBO9781139058537
Hempel, Carl G., 1965,Aspects of Scientific Explanation,New York: The Free Press.
Hesse, Mary B., 1980,Revolutions and Reconstructions in thePhilosophy of Science, Bloomington, IN: University of IndianaPress.
Howson, Colin, 2000,Hume’s Problem: Induction and theJustification of Belief, Oxford: Oxford University Press.
Howson, Colin and Peter Urbach, 1993,Scientific Reasoning:The Bayesian Approach, second edition, La Salle, IL: OpenCourt.
Hrdy, Sarah Blaffer, 1977,The Langurs of Abu: Female and MaleStrategies of Reproduction, Cambridge, MA: Harvard UniversityPress.
Ioannidis, John P. A., 2005, “Why Most Published ResearchFindings Are False”,PLoS Medicine, 2(8): e124.doi:10.1371/journal.pmed.0020124
Janack, Marianne, 2002, “Dilemmas of Objectivity”,Social Epistemology, 16(3): 267–281.doi:10.1080/0269172022000025624
Jarvie, Ian C. and Jesús P. Zamora Bonilla (eds.), 2011,The SAGE Handbook of the Philosophy of Social Sciences,London: SAGE.
Jaynes, Edwin T., 1968, “Prior Probabilities”,IEEE Transactions on Systems Science and Cybernetics, 4(3):227–241. doi:10.1109/TSSC.1968.300117
Jeffrey, Richard C., 1956, “Valuation and Acceptance ofScientific Hypotheses”,Philosophy of Science, 23(3):237–246. doi:10.1086/287489
Jeffreys, Harold, 1939 [1980],Theory of Probability,third edition, Oxford: Oxford University Press.
Kelvin, Lord (William Thomson), 1883, “Electrical Units ofMeasurement”, Lecture to the Institution of Civil Engineers on 3May 1883, reprinted in 1889,Popular Lectures and Addresses,Vol. I, London: MacMillan and Co., p. 73.
Kincaid, Harold, John Dupré, and Alison Wylie (eds.), 2007,Value-Free Science?: Ideals and Illusions, Oxford: OxfordUniversity Press. doi:10.1093/acprof:oso/9780195308969.001.0001
Kitcher, Philip, 2011a,Science in a Democratic Society,Amherst, NY: Prometheus Books.
–––, 2011b,The Ethical Project,Cambridge, MA: Harvard University Press.
Koskinen, Inkeri, forthcoming, “Defending a Risk Account ofScientific Objectivity”,The British Journal for thePhilosophy of Science, first online: 3 August 2018.doi:10.1093/bjps/axy053
Kourany, Janet A., 2010,Philosophy of Science afterFeminism, Oxford: Oxford University Press.
Kuhn, Thomas S., 1962 [1970],The Structure of ScientificRevolutions, second edition, Chicago: University of ChicagoPress.
–––, 1977, “Objectivity, Value Judgment,and Theory Choice”, in hisThe Essential Tension. SelectedStudies in Scientific Tradition and Change, Chicago: Universityof Chicago Press: 320–39.
Lacey, Hugh, 1999,Is Science Value-Free? Values andScientific Understanding, London: Routledge.
–––, 2002, “The Ways in Which the SciencesAre and Are Not Value Free”, inIn the Scope of Logic,Methodology and Philosophy of Science: Volume Two of the 11thInternational Congress of Logic, Methodology and Philosophy ofScience, Cracow, August 1999, Peter Gärdenfors, JanWoleński, and Katarzyna Kijania-Placek (eds.), Dordrecht:Springer Netherlands, 519–532.doi:10.1007/978-94-017-0475-5_9
Laudan, Larry, 1984,Science and Values: An Essay on the Aimsof Science and Their Role in Scientific Debate, Berkeley/LosAngeles: University of California Press.
Levi, Isaac, 1960, “Must the Scientist Make ValueJudgments?”,The Journal of Philosophy, 57(11):345–357. doi:10.2307/2023504
Lloyd, Elisabeth A., 1993, “Pre-Theoretical Assumptions inEvolutionary Explanations of Female Sexuality”,Philosophical Studies, 69(2–3): 139–153.doi:10.1007/BF00990080
–––, 2005,The Case of the Female Orgasm:Bias in the Science of Evolution, Cambridge, MA: HarvardUniversity Press.
Longino, Helen E., 1990,Science as Social Knowledge: Valuesand Objectivity in Scientific Inquiry, Princeton, NY: PrincetonUniversity Press.
–––, 1996, “Cognitive and Non-CognitiveValues in Science: Rethinking the Dichotomy”, inFeminism,Science, and the Philosophy of Science, Lynn Hankinson Nelson andJack Nelson (eds.), Dordrecht: Springer Netherlands, 39–58.doi:10.1007/978-94-009-1742-2_3
Machamer, Peter and Gereon Wolters (eds.), 2004,Science,Values and Objectivity, Pittsburgh: Pittsburgh UniversityPress.
Mayo, Deborah G., 1996,Error and the Growth of ExperimentalKnowledge, Chicago & London: The University of ChicagoPress.
McMullin, Ernan, 1982, “Values in Science”,PSA:Proceedings of the Biennial Meeting of the Philosophy of ScienceAssociation 1982, 3–28.
–––, 2009, “The Virtues of a GoodTheory”, inThe Routledge Companion to Philosophy ofScience, Martin Curd and Stathis Psillos (eds), London:Routledge.
Megill, Allan, 1994, “Introduction: Four Senses ofObjectivity”, inRethinking Objectivity, Allan Megill(ed.), Durham, NC: Duke University Press, 1–20.
Mill, John Stuart, 1859 [2003],On Liberty, New Haven andLondon: Yale University Press.
Mitchell, Sandra D., 2004, “The Prescribed and ProscribedValues in Science Policy”, in Machamer and Wolters 2004:245–255.
Nagel, Thomas, 1986,The View From Nowhere, New York, NY:Oxford University Press.
Nixon, Richard, 1969, “Special Message to the Congress onSocial Security”, 25 September 1969. [Nixon 1969 available online]
Norton, John D., 2003, “A Material Theory ofInduction”,Philosophy of Science, 70(4):647–670. doi:10.1086/378858
–––, 2008, “Must Evidence UnderdetermineTheory?”, inThe Challenge of the Social and the Pressure ofPractice, Martin Carrier, Don Howard and Janet Kourany (eds),Pittsburgh, PA: Pittsburgh University Press: 17–44.
Oakeshott, Michael, 1933,Experience and Its Modes,Cambridge: Cambridge University Press.
Okruhlik, Kathleen, 1994, “Gender and the BiologicalSciences”,Canadian Journal of Philosophy SupplementaryVolume, 20: 21–42. doi:10.1080/00455091.1994.10717393
Open Science Collaboration, 2015, “Estimating theReproducibility of Psychological Science”,Science,349(6251): aac4716. doi:10.1126/science.aac4716
Padovani, Flavia, Alan Richardson, and Jonathan Y. Tsou (eds.),2015,Objectivity in Science: New Perspectives from Science andTechnology Studies, (Boston Studies in the Philosophy and Historyof Science 310), Cham: Springer International Publishing.doi:10.1007/978-3-319-14349-1
Page, Scott E., 2007,The Difference: How the Power ofDiversity Creates Better Groups, Firms, Schools, and Societies,Princeton, NJ: Princeton University Press.
Paternotte, Cédric, 2011, “Rational ChoiceTheory”, inThe SAGE Handbook of The Philosophy of SocialSciences, Jarvie and Zamora Bonilla 2011: 307–321.
Popper, Karl. R., 1934 [2002],Logik der Forschung,Vienna: Julius Springer. Translated asLogic of ScientificDiscovery, London: Routledge.
–––, 1963,Conjectures and Refutations: TheGrowth of Scientific Knowledge, New York: Harper.
–––, 1972,Objective Knowledge: AnEvolutionary Approach, Oxford: Oxford University Press.
Porter, Theodore M., 1995,Trust in Numbers: The Pursuit ofObjectivity in Science and Public Life, Princeton, NJ, PrincetonUniversity Press.
Putnam, Hilary, 2002,The Collapse of the Fact/Value Dichotomyand Other Essays, Cambridge, MA: Harvard University Press.
Putnam, Hilary and Vivian Walsh, 2007, “A Response toDasgupta”,Economics and Philosophy, 23(3):359–364. doi:10.1017/S026626710700154X
Reichenbach, Hans, 1938, “On Probability andInduction”,Philosophy of Science, 5(1): 21–45.doi:10.1086/286483
Reiss, Julian, 2008,Error in Economics: The Methodology ofEvidence-Based Economics, London: Routledge.
–––, 2010, “In Favour of a MillianProposal to Reform Biomedical Research”,Synthese,177(3): 427–447. doi:10.1007/s11229-010-9790-7
–––, 2013,Philosophy of Economics: AContemporary Introduction, New York, NY: Routledge.
–––, 2020, “What Are the Drivers ofInduction? Towards a Material Theory+”,Studies in Historyand Philosophy of Science Part A 83: 8–16.
Resnik, David B., 2007,The Price of Truth: How Money Affectsthe Norms of Science, Oxford: Oxford University Press.
Rickert, Heinrich, 1929,Die Grenzen dernaturwissenschaftlichen Begriffsbildung. Eine logische Einleitung indie historischen Wissenschaften, 6th edition, Tübingen: MohrSiebeck. First edition published in 1902.
Royall, Richard, 1997,Scientific Evidence: A LikelihoodParadigm, London: Chapman & Hall.
Rudner, Richard, 1953, “The Scientistqua ScientistMakes Value Judgments”,Philosophy of Science, 20(1):1–6. doi:10.1086/287231
Ruphy, Stéphanie, 2006, “‘Empiricism All theWay down’: A Defense of the Value-Neutrality of Science inResponse to Helen Longino’s Contextual Empiricism”,Perspectives on Science, 14(2): 189–214.doi:10.1162/posc.2006.14.2.189
Sen, Amartya, 1993, “Internal Consistency of Choice”,Econometrica, 61(3): 495–521.
Shrader-Frechette, K. S., 1991,Risk and Rationality,Berkeley/Los Angeles: University of California Press.
Simonsohn, Uri, Leif D. Nelson, and Joseph P. Simmons, 2014,“P-Curve: A Key to the File-Drawer.”,Journal ofExperimental Psychology: General, 143(2): 534–547.doi:10.1037/a0033242
Sprenger, Jan, 2016, “Bayesianism vs. Frequentism inStatistical Inference”, inOxford Handbook on Philosophy ofProbability, Alan Hájek and Christopher Hitchcock (eds),Oxford: Oxford University Press.
–––, 2018, “The Objectivity of SubjectiveBayesianism”,European Journal for Philosophy ofScience, 8(3): 539–558. doi:10.1007/s13194-018-0200-1
Sprenger, Jan and Stephan Hartmann, 2019,Bayesian Philosophyof Science, Oxford: Oxford University Press.doi:10.1093/oso/9780199672110.001.0001
Steel, Daniel, 2010, “Epistemic Values and the Argument fromInductive Risk”,Philosophy of Science, 77(1):14–34. doi:10.1086/650206
Steele, Katie, 2012, “The Scientistqua PolicyAdvisor Makes Value Judgments”, Philosophy of Science, 79(5):893–904. doi:10.1086/667842
Stegenga, Jacob, 2011, “Is Meta-Analysis the PlatinumStandard of Evidence?”,Studies in History and Philosophy ofScience Part C: Studies in History and Philosophy of Biological andBiomedical Sciences, 42(4): 497–507.doi:10.1016/j.shpsc.2011.07.003
–––, 2018,Medical Nihilism, Oxford:Oxford University Press. doi:10.1093/oso/9780198747048.001.0001
Teira, David, 2010, “Frequentist versus Bayesian ClinicalTrials”, inPhilosophy of Medicine, Fred Gifford (ed.),(Handbook of the Philosophy of Science 16), Amsterdam: Elsevier,255–297. doi:10.1016/B978-0-444-51787-6.50010-6
Teira, David and Julian Reiss, 2013, “Causality,Impartiality and Evidence-Based Policy”, inMechanism andCausality in Biology and Economics, Hsiang-Ke Chao, Szu-TingChen, and Roberta L. Millstein (eds.), (History, Philosophy and Theoryof the Life Sciences 3), Dordrecht: Springer Netherlands,207–224. doi:10.1007/978-94-007-2454-9_11
Weber, Max, 1904 [1949], “Die‘Objektivität’ sozialwissenschaftlicher undsozialpolitischer Erkenntnis”,Archiv fürSozialwissenschaft und Sozialpolitik, 19(1): 22–87.Translated as “‘Objectivity’ in Social Science andSocial Policy”, in Weber 1949: 50–112.
–––, 1917 [1949], “Der Sinn der‘Wertfreiheit’ der soziologischen und ökonomischenWissenschaften”. Reprinted inGesammelte Aufsätze zurWissenschaftslehre, Tübingen: UTB, 1988, 451–502.Translated as “The Meaning of ‘Ethical Neutrality’in Sociology and Economics” in Weber 1949: 1–49.
–––, 1949,The Methodology of the SocialSciences, Edward A. Shils and Henry A. Finch (trans/eds), NewYork, NY: Free Press.
Wilholt, Torsten, 2009, “Bias and Values in ScientificResearch”,Studies in History and Philosophy of Science PartA, 40(1): 92–101. doi:10.1016/j.shpsa.2008.12.005
–––, 2013, “Epistemic Trust inScience”,The British Journal for the Philosophy ofScience, 64(2): 233–253. doi:10.1093/bjps/axs007
Williams, Bernard, 1985 [2011],Ethics and the Limits ofPhilosophy, Cambridge, MA: Harvard University Press. ReprintedLondon and New York, NY: Routledge, 2011.
Williamson, Jon, 2010,In Defence of ObjectiveBayesianism, Oxford: Oxford University Press.doi:10.1093/acprof:oso/9780199228003.001.0001
Windelband, Wilhelm, 1915,Präludien. Aufsätze undReden zur Philosophie und ihrer Geschichte, fifth edition,Tübingen: Mohr Siebeck.
Winsberg, Eric, 2012, “Values and Uncertainties in thePredictions of Global Climate Models”,Kennedy Institute ofEthics Journal, 22(2): 111–137.doi:10.1353/ken.2012.0008
Wittgenstein, Ludwig, 1953 [2001],PhilosophicalInvestigations, G. Anscombe (trans.), London: Blackwell.
Worrall, John, 2002, “What Evidence inEvidence‐Based Medicine?”,Philosophy of Science,69(S3): S316–S330. doi:10.1086/341855
Wylie, Alison, 2003, “Why Standpoint Matters”, inScience and Other Cultures: Issues in Philosophies of Science andTechnology, Robert Figueroa and Sandra Harding (eds), New York,NY and London: Routledge, pp. 26–48.
Ziliak, Stephen Thomas and Deirdre N. McCloskey, 2008,TheCult of Statistical Significance: How the Standard Error Costs UsJobs, Justice and Lives, Ann Arbor, MI: University of MichiganPress.

Academic Tools

How to cite this entry.
Preview the PDF version of this entry at theFriends of the SEP Society.
Look up topics and thinkers related to this entry at the Internet Philosophy Ontology Project (InPhO).
Enhanced bibliography for this entryatPhilPapers, with links to its database.

Other Internet Resources

Norton, John, manuscript,The Material Theory of Induction, retrieved on 9 January 2020.
Objectivity, entry by Dwayne H. Mulder in theInternet Encyclopedia ofPhilosophy.

Open access to the SEP is made possible by a world-wide funding initiative.
The Encyclopedia Now Needs Your Support
Please Read How You Can Help Keep the Encyclopedia Free

Browse

About

Support SEP

Mirror Sites

View this site from another server:

USA (Main Site)Philosophy, Stanford University

Info about mirror sites

Library of Congress Catalog Data: ISSN 1095-5054

Movatterモバイル変換