Instituting Science: Discovery or Construction of Scientific Knowledge?

Abstract

Is knowledge in the natural sciences discovered or constructed? For objectivists, scientific knowledge is discovered through investigations into a mind‐independent, natural world. For constructivists, such knowledge is produced through negotiations among members of a professional guild. I examine the clash between the two positions and propose that scientific knowledge is the concurrent outcome from investigations into a natural world and from consensus reached through negotiations of a professional guild. Specifically, I introduce the general methodological notion, instituting science, which incorporates both the discovery and the construction processes in the generation of scientific knowledge. To that end, I use a case study from the biomedical sciences to illustrate the notion. I conclude with a discussion of how this methodological notion helps to address the debate between objectivists and constructivists over the generation of scientific knowledge, and of how it compares with others who have also attempted to address the debate.

Acknowledgements

I thank especially Ron Anderson, in whose memory this article is dedicated, Susie Fisher, and Ernan McMullin, as well as James W. McAllister and two anonymous referees for International Studies in the Philosophy of Science, for comments on earlier versions of this paper and Harry Rubin for clarifying events in Dulbecco’s laboratory. A version of this paper was delivered at the 55th Annual Meeting of the New Mexico/West Texas Philosophical Society in San Antonio, Texas. Baylor University supported my research through a sabbatical award and a University Research Committee grant.

Notes

[1] Besides the terms objectivist and constructivist other dichotomous terms have been used, such as algorithmic and enculturation or rationalist and sociological (Longino 2002, 23).

[2] The following attempt to summarize the objectivist and constructivist positions is not meant to be exhaustive or to represent any one person or school. There is simply too much variation among the different positions to accomplish that task within the space allotted; rather, the purpose of this summary is to frame adequately the debate between the protagonists of both positions in order to provide sufficient background for my position.

[3] For an introduction to the science wars, see Baringer (2001) and Brown (2001). Again, it must be stressed that ‘although it is fairly easy to list the names of the most noted protagonists in the Science Wars, it is difficult to produce a characterization of the clusters of opposing views that is satisfactory to any of the participants’ (Koertge 2000, S46).

[4] Stephen Turner (2003) claims that the previous science wars have morphed into a ‘third science war’, which now involves the political agenda of the previous science wars. Although this war is an important skirmish, it is not the focus of this paper; rather, I focus on the epistemological and methodological issues surrounding the conflict between objectivist and constructivist conceptions of science and its knowledge and practice.

[5] Hans Radder (2006) has proffered a recent attempt to address the debate, in terms of connecting the world perceived and the world conceived. The first step involves explicating observational processes vis‐à‐vis ‘material realization’ and ‘conceptual interpretation’, while the next step entails demonstrating not only how concepts structure the world but also how they are abstracted from it. Thus, ideas of the world not only have their origins in the world but also shape how that world is conceived.

[6] The features are compiled from Barnes (1974), Bloor (1976), Knorr‐Cetina (1981), Pickering (1984), Collins (1985), Latour and Woolgar (1986), Kitcher (1993; 2001), Nelson (1994), Ross (1996), Gross and Levitt (1998), Goldman (1999), Hacking (1999), Kukla (2000), Brown (2001), and Longino (2002).

[7] The term ‘artefact’ is used here instead of the term ‘artifact’, since the latter term—especially in the United States—refers to a misrepresentation of nature due to experimental defect. The former term is taken to represent better the construction of facts through human agency.

[8] It is possible for standards, warrant, facts, truth, and ontology to remain local for a longer period of time or to be universal for a shorter period of time.

[9] Timothy Lenoir (1997) also introduced a notion he calls ‘instituting science’ in a book with the same title. In the book, however, he is concerned not with the methodological issues I raise here but with the institutionalization of science, i.e. the social formation of scientific disciplines.

[10] By methodological notion, I mean the process by which the practice of science is reconstructed in order to explicate that practice.

[11] The notion of instituting science applies more to the experimental sciences than to the theoretical sciences. Although there is significant overlap between these sciences, the differences between them prohibit any common methodology.

[12] See also Achinstein (2004) for a discussion of the history of the scientific method(s).

[13] Scientists advocate a scientific method, for which there are several introductory texts (Carey 2003; Gauch 2003; Valiela 2000). However, philosophers and historians of science advocate a plurality of scientific practices that depend upon a given natural science. Scientists also recognize a plurality of practices, but these practices represent specific protocols by which scientists conduct their trade within a particular discipline, which then fit ultimately into the overall method as expounded in their introductory texts.

[14] Ian Hacking (1999) proposes a means by which to resolve the tension between reality and scientific knowledge. Specifically, he distinguishes between the notion of an entity and the entity itself. For example, Hacking claims ‘quarks, the objects themselves, are not constructs, are not social, and are not historical’ (Hacking 1999, 30). Thus, the notion or idea of quarks is a construction but not quarks per se. Although Hacking’s intuition provides a valuable distinction between the objectivist and constructivist positions, it fails to resolve the tension between them.

[15] Rachel Ankeny (2001) has shown that early stages in experimental science may not involve hypothesis or theory testing but rather are ‘proto‐explanatory’ in nature, in that theories function simply as a backdrop for conducting experiments.

[16] A change in a paradigm during a scientific revolution, according to Kuhn (1970, 111–135), results in a different view of the world and to some degree scientists live in a different world after a revolution.

[17] Of course, there are alternative approaches to the practice of the natural sciences. For example, there is considerable debate over the reductionist approach to cancer research, with some researchers arguing for a holistic approach (Marcum 2005).

[18] Ruse (1999) also contends that non‐epistemic values can still operate as metaphors, even in the most robust science.

[19] Prions are known to exist but their exact nature with respect to the satisfaction of the scientific community is yet to be established.

[20] Organisms and artificial systems also function as models for conducting research (Ankeny 2000; Griesemer 2000).

[21] Since Isaac Newton a hypothesis is often considered too metaphysical or speculative, especially in terms of no empirical grounding. The use of hypothesis here also rejects the non‐empirically grounded metaphysical or speculative dimension and substitutes an empirically grounded speculative dimension, as well as a tentative dimension. In other words, a hypothesis is speculative and tentative but from an empirical basis. A hypothesis is generally proposed first, followed by construction of a theory that then becomes part of a discipline’s paradigm. However, experiments can be conducted first that then give rise to a hypothesis.

[22] In the early 1970s, the then US President Nixon declared war on cancer (Sporn 1996). Importantly, the military metaphor delimits the boundaries for what is permissible cognitively for biomedical research. In clinical science and practice, as well, metaphor plays an important philosophical role. According to Nelson Goodman, ‘to suppose that science is flatfootedly linguistic, literal, and denotational would be to overlook, for instance, the analog instruments often used, the metaphor involved in measurement when numerical scheme is applied in a new realm, and the talk in current physics and astronomy of charm and strangeness and black holes. Even if the ultimate product of science, unlike that of art, is a literal, verbal, or mathematical, denotational theory, science and art proceed in much the same way with their searching and building’ (Goodman 1978, 107).

[23] The term establish is used instead of the term justify or prove to best represent the function of warrant, since the former is not so logically constrained as the latter terms.

[24] For example, William Harvey’s warrant for his notion of the circulation of the blood in terms of cardiac output remains generally accepted today not in terms of Harvey’s quantitative data, which grossly underestimate the output, but in terms of the general conception that too much blood is pumped over a given period of time to warrant the Galenic theory of blood circulation.

[25] My position here reflects Longino’s one in addressing Collins’s programme (Longino 2002, 22–25).

[26] Obviously, there can be direct perceptual access to the natural world (Radder 2006).

[27] This constraint obviously depends upon the proper functioning of perceptual (and cognitive) faculties.

[28] ‘The whole truth,’ according to Nelson Goodman, ‘would be too much; it is too vast, variable, and clogged with trivia’ (Goodman 1978, 19). The notion of substantial truth is very similar to notion of significant truth (Kitcher 1993, 2001).

[29] Temin probably refers to the notion of provirus as a hypothesis, given its speculative and tentative nature (see note 21 for further discussion of the nature of hypothesis).

[30] Dulbecco, at this time, was working mainly on poliovirus at Pasadena’s Huntington Hospital (Rubin, pers. commun., 2005). However, he was conversant with Rubin’s and Temin’s work as evident from a lecture delivered at a cancer symposium sponsored by the American Chemical Society in September 1959 (Dulbecco 1960).

[31] Cell conversion is the process by which a phage or virus changes the phenotype of a prokaryotic or eukaryotic cell. For instance, viruses can transform a cell with a normal phenotype, e.g. regulated growth, into one with a cancerous phenotype, e.g. unregulated growth.

[32] Prophage is the latent form of a bacteriophage in which its genome is incorporated into a bacterium’s genome, without lysing or destroying the host.

[33] Lysogeny is the incorporation of the bacteriophage’s genome (prophage) or virus’s genome (provirus) into the host cell and represents the process involved in the lysogenic cycle in which the prophage or provirus can be transmitted to progeny during cell division. This cycle is in contrast to the lytic cycle, in which the bacteriophage or virus commandeers the host‐cell metabolic machinery to produce bacteriophages or viruses with cell lysis and release of the newly produced bacteriophages or viruses.

[34] Besides the institutional support of the McArdle, Temin also received funding support from granting agencies, such as the National Cancer Institute.

[35] Actinomycin D is a member of the family of antibiotics from Streptomyces, a soil bacterium, and inhibits RNA transcription from a DNA template.

[36] Nucleic hybridization is the technique in which a soluble, single‐stranded nucleic acid probe, which is often labelled, is used to detect homologous, immobilized nucleic acid sequences.

[37] Nucleic acid polymerases represent a family of enzymes that catalyze the synthesis of polymers from a polynucleotide template. For example, a DNA polymerase synthesizes DNA from a DNA template.

[38] For further discussion of construction and constraint, especially in terms of scientific rationality, see McMullin (1988).

[39] For example, the scientific community cannot be a point of origin for either scientific knowledge or reality, since such unconstrained activity—in terms of scientific knowledge or reality—runs to great a risk of being illusionary.

[40] According to Goodman, ‘Worldmaking as we know it always starts from worlds already on hand; the making is a remaking’ (Goodman 1978, 6). Although a scientist can have, obviously, unmediated perception of the natural world, such exclusive perception without eventual cognitive mediation would lead to confusion.

[41] Although Longino’s position appears to be a nondichotomizer one, it does rely on the various categorical dichotomies—in that it combines a category from each of the dichotomous positions. Given these dichotomies for the three categories there are actually eight possible combinations, of which Longino, the rationalizers, and the sociologizers represent only one combination each. Five other combinations are possible. What about each of these combinations? Are they nondichotomizer or dichotomizer ones? For further discussion of these possible combinations and a critique of Longino’s third way, see Kitcher (2002).

[42] A pluralist position does constrain the possibility of a unifying theory, as Longino recognizes: ‘Pluralists hold that the complexity of the natural world is such that a single unified picture of the world is not possible’ Longino (2002, 142). What Longino seems to mean is that the grand unification of scientific knowledge is the integration of different disciplinary approaches to the natural world (Longino 2002, 200–202).

[43] Longino claims that ‘the realist interpretation of nonmonism holds that while no single account, no single theory or model, captures the entirety of physical and biological processes operating in a single real‐world system, each satisfactory account captures some aspect. Thus, multiple accounts (theories, models) are not just permissible but necessary’ (Longino 2002, 92). There is an ambiguity here. Does the passage refer to a single natural phenomenon, like the role of nucleotides in RSV replication, for which one theory is correct or does it refer to a larger natural phenomenon that may require several theories to account for it? Longino seems to rely upon both possibilities, especially as she discusses the pluralism found in biology (Longino 2002, 176–183).

[44] The notion of instituting science does differ from the notion of worldmaking in terms of the degree of relativism. Goodman’s relativism is ‘radical’ but still ‘severe restraints are imposed’ (Goodman 1978, 94). He refers to his position as irrealism, in that there is no primary ‘world‐version’ that undergirds reality. Hence, he advocates a type of pluralism. For instituting science the restraints are imposed by the natural world, which supports its monistic view of the natural world.

[45] Under these conditions, artefacts are artifacts or misrepresentations of the natural world.

[46] In discussing Stanley Fish’s op‐ed piece in The New York Times concerning the Sokal hoax, Hacking critiques Fish for using John Searle’s distinction between epistemological objectivity and ontological subjectivity: ‘There would be no strikes without the institution of baseball, without the rules and practices of people’ (Hacking 1999, 30).