Towards a sociology of measurement: The meaning of measurement error in the case of DNA profiling

Social Studies of Science

Volumn 30, Issue 6, 2000, Pages 803-845

  Derksen, Linda ^a,b  

^a UNIVERSITY OF CALIFORNIA   (United States)

^b UNIVERSITY OF ALBERTA   (Canada)

Author keywords

Credibility; DNA fingerprinting; DNA Wwars; Expertise; Judgement; Legal; Objectivity; Subjectivity

Indexed keywords

EID: 0034345119     PISSN: 03063127     EISSN: None     Source Type: Journal    
DOI: 10.1177/030631200030006001     Document Type: Article

References (175)

1. In the United States, DNA typing was first admitted as evidence in the 1987 rape case, State of Florida v. Tommy Lee Andrews. The conviction was appealed (and upheld) in Andrews v. State of Florida, 533 So.2d 841 (Fla.App.5 Dist. 1988), 843. In July 1987, Jeffrey L. Ashton, an assistant state attorney for the state of Florida, was reading a lawyer's magazine, and saw an advertisement for paternity testing. He called the company which had placed the ad, Lifecodes Corporation of Valhalla, NY, and found that the same test could be used for forensic identification. Ashton got in touch with Tim Berry, who was prosecuting the case of Tommy Lee Andrews, who was charged with six separate counts of rape. Ashton and Berry decided to use the new type of evidence. They could find no indication 'that anyone had ever done this before, but . . . felt it was worth a try'. Ashton and Berry sent six sets of samples from the victims and the accused to Lifecodes, and the analysis came back positive on two of the six. In November of 1987, the first of the six cases went to trial in the Circuit Court in Orange County, Florida. See Jeffrey L. Ashton in his testimony before the Committee on the Judiciary (1990), Hearing before the Subcommittee on the Constitution of the Committee on the Judiciary, US Senate, 101st Congress, First Session on Genetic Testing as a Means of Criminal Investigation, 15 March 1989 (Washington, DC: US Government Printing Office, 1990), 76, 77. The evidence of DNA identity testing was given to the jury in the Andrews case by Dr Michael Baird, manager of forensic testing at Lifecodes. Baird testified that the semen samples taken from the victim matched the blood samples taken from the accused, and that the probability that the DNA samples could belong to anyone but the accused were 1 in 839,914,540.
2. Peter Neufeld, quoted in Roger Parloff, 'How Barry Scheck and Peter Neufeld Tripped Up the DNA Experts', The American Lawyer (December 1989), 1-2, 53-55, at 2. It is not unusual that scientific evidence would function in a courtroom like a 'verdict'. As noted by Roger Smith and Brian Wynne, in adversarial settings like the courtroom, it is hoped 'that the objectivity of science will provide a firm and authoritative input, giving decisions a factual basis that cannot be questioned. That the science often appears equivocal is put down to procedural problems rather than inherent properties of scientific knowledge or methods': R. Smith and B. Wynne, 'Introduction', in Smith and Wynne (eds), Expert Evidence: Interpreting Science in the Law (London & New York: Routledge, 1989), 1-23, at 1.
3. Peter Neufeld, quoted in Roger Parloff, 'How Barry Scheck and Peter Neufeld Tripped Up the DNA Experts', The American Lawyer (December 1989), 1-2, 53-55, at 2. It is not unusual that scientific evidence would function in a courtroom like a 'verdict'. As noted by Roger Smith and Brian Wynne, in adversarial settings like the courtroom, it is hoped 'that the objectivity of science will provide a firm and authoritative input, giving decisions a factual basis that cannot be questioned. That the science often appears equivocal is put down to procedural problems rather than inherent properties of scientific knowledge or methods': R. Smith and B. Wynne, 'Introduction', in Smith and Wynne (eds), Expert Evidence: Interpreting Science in the Law (London & New York: Routledge, 1989), 1-23, at 1.
4. See Michael Lynch, 'The Discursive Production of Uncertainty: The OJ Simpson "Dream Team" and the Sociology of Knowledge Machine', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 829-68, for a discussion of how, in the case of forensic DNA analysis, defense lawyers often undertook the same type of deconstructive work that papers such as this one are attempting to do, albeit for different purposes. Using the example of the OJ Simpson double murder trial, Lynch shows how prosecutors took a 'realist' position towards the technology, and defense lawyers wrote a motion which took a constructivist approach, showing the extent to which quantitative estimates of random matches were dependent on local laboratory practices and interested judgements. I think it can be argued that after the formation of the DNA Task Force in 1990, defense attorneys had more resources for deconstruction on hand, some of them obtained in expert testimony by high-ranking population geneticists. There is a burgeoning literature on the development or 'black boxing' of various techniques used in molecular biology: see Kathleen Jordan and Michael Lynch, 'The Sociology of a Genetic Engineering Technique: Ritual and Rationality in the Performance of the Plasmid Prep', in Adele E. Clarke and Joan H. Fujimura (eds), The Right Tools for the Job (Princeton, NJ: Princeton University Press, 1992), 77-114. Jordan and Lynch critically examine the notion of 'black box' in reference to another tool of molecular biology, the 'plasmid prep', which has become a well-established procedure in molecular biology. By examining variations in the use of the plasmid prep, Jordan and Lynch show that, despite the current stability of the technique and its status as an easily accessible tool for molecular biologists, it remains locally problematic and must be 'constructed' anew whenever it is adapted to new circumstances (ibid., 77). For a related study of the Polymerase Chain Reaction, or PCR, see K. Jordan and M. Lynch, 'The Mainstreaming of a Molecular Biological Tool: A Case Study of a New Technique', in Graham Button (ed.), Technology in Working Order: Studies of Work, Interaction and Technology (London & New York: Routledge, 1992), 160-80. See also K. Jordan and M. Lynch, 'The Dissemination, Standardization, and Routinization of a Molecular Biological Technique', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 773-99, and Joan Fujimura, 'The Molecular Biological Bandwagon in Cancer Research: Where Social Worlds Meet', Social Problems, Vol. 35 (1988), 261-83.
5. See Michael Lynch, 'The Discursive Production of Uncertainty: The OJ Simpson "Dream Team" and the Sociology of Knowledge Machine', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 829-68, for a discussion of how, in the case of forensic DNA analysis, defense lawyers often undertook the same type of deconstructive work that papers such as this one are attempting to do, albeit for different purposes. Using the example of the OJ Simpson double murder trial, Lynch shows how prosecutors took a 'realist' position towards the technology, and defense lawyers wrote a motion which took a constructivist approach, showing the extent to which quantitative estimates of random matches were dependent on local laboratory practices and interested judgements. I think it can be argued that after the formation of the DNA Task Force in 1990, defense attorneys had more resources for deconstruction on hand, some of them obtained in expert testimony by high-ranking population geneticists. There is a burgeoning literature on the development or 'black boxing' of various techniques used in molecular biology: see Kathleen Jordan and Michael Lynch, 'The Sociology of a Genetic Engineering Technique: Ritual and Rationality in the Performance of the Plasmid Prep', in Adele E. Clarke and Joan H. Fujimura (eds), The Right Tools for the Job (Princeton, NJ: Princeton University Press, 1992), 77-114. Jordan and Lynch critically examine the notion of 'black box' in reference to another tool of molecular biology, the 'plasmid prep', which has become a well-established procedure in molecular biology. By examining variations in the use of the plasmid prep, Jordan and Lynch show that, despite the current stability of the technique and its status as an easily accessible tool for molecular biologists, it remains locally problematic and must be 'constructed' anew whenever it is adapted to new circumstances (ibid., 77). For a related study of the Polymerase Chain Reaction, or PCR, see K. Jordan and M. Lynch, 'The Mainstreaming of a Molecular Biological Tool: A Case Study of a New Technique', in Graham Button (ed.), Technology in Working Order: Studies of Work, Interaction and Technology (London & New York: Routledge, 1992), 160-80. See also K. Jordan and M. Lynch, 'The Dissemination, Standardization, and Routinization of a Molecular Biological Technique', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 773-99, and Joan Fujimura, 'The Molecular Biological Bandwagon in Cancer Research: Where Social Worlds Meet', Social Problems, Vol. 35 (1988), 261-83.
6. See Michael Lynch, 'The Discursive Production of Uncertainty: The OJ Simpson "Dream Team" and the Sociology of Knowledge Machine', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 829-68, for a discussion of how, in the case of forensic DNA analysis, defense lawyers often undertook the same type of deconstructive work that papers such as this one are attempting to do, albeit for different purposes. Using the example of the OJ Simpson double murder trial, Lynch shows how prosecutors took a 'realist' position towards the technology, and defense lawyers wrote a motion which took a constructivist approach, showing the extent to which quantitative estimates of random matches were dependent on local laboratory practices and interested judgements. I think it can be argued that after the formation of the DNA Task Force in 1990, defense attorneys had more resources for deconstruction on hand, some of them obtained in expert testimony by high-ranking population geneticists. There is a burgeoning literature on the development or 'black boxing' of various techniques used in molecular biology: see Kathleen Jordan and Michael Lynch, 'The Sociology of a Genetic Engineering Technique: Ritual and Rationality in the Performance of the Plasmid Prep', in Adele E. Clarke and Joan H. Fujimura (eds), The Right Tools for the Job (Princeton, NJ: Princeton University Press, 1992), 77-114. Jordan and Lynch critically examine the notion of 'black box' in reference to another tool of molecular biology, the 'plasmid prep', which has become a well-established procedure in molecular biology. By examining variations in the use of the plasmid prep, Jordan and Lynch show that, despite the current stability of the technique and its status as an easily accessible tool for molecular biologists, it remains locally problematic and must be 'constructed' anew whenever it is adapted to new circumstances (ibid., 77). For a related study of the Polymerase Chain Reaction, or PCR, see K. Jordan and M. Lynch, 'The Mainstreaming of a Molecular Biological Tool: A Case Study of a New Technique', in Graham Button (ed.), Technology in Working Order: Studies of Work, Interaction and Technology (London & New York: Routledge, 1992), 160-80. See also K. Jordan and M. Lynch, 'The Dissemination, Standardization, and Routinization of a Molecular Biological Technique', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 773-99, and Joan Fujimura, 'The Molecular Biological Bandwagon in Cancer Research: Where Social Worlds Meet', Social Problems, Vol. 35 (1988), 261-83.
7. See Michael Lynch, 'The Discursive Production of Uncertainty: The OJ Simpson "Dream Team" and the Sociology of Knowledge Machine', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 829-68, for a discussion of how, in the case of forensic DNA analysis, defense lawyers often undertook the same type of deconstructive work that papers such as this one are attempting to do, albeit for different purposes. Using the example of the OJ Simpson double murder trial, Lynch shows how prosecutors took a 'realist' position towards the technology, and defense lawyers wrote a motion which took a constructivist approach, showing the extent to which quantitative estimates of random matches were dependent on local laboratory practices and interested judgements. I think it can be argued that after the formation of the DNA Task Force in 1990, defense attorneys had more resources for deconstruction on hand, some of them obtained in expert testimony by high-ranking population geneticists. There is a burgeoning literature on the development or 'black boxing' of various techniques used in molecular biology: see Kathleen Jordan and Michael Lynch, 'The Sociology of a Genetic Engineering Technique: Ritual and Rationality in the Performance of the Plasmid Prep', in Adele E. Clarke and Joan H. Fujimura (eds), The Right Tools for the Job (Princeton, NJ: Princeton University Press, 1992), 77-114. Jordan and Lynch critically examine the notion of 'black box' in reference to another tool of molecular biology, the 'plasmid prep', which has become a well-established procedure in molecular biology. By examining variations in the use of the plasmid prep, Jordan and Lynch show that, despite the current stability of the technique and its status as an easily accessible tool for molecular biologists, it remains locally problematic and must be 'constructed' anew whenever it is adapted to new circumstances (ibid., 77). For a related study of the Polymerase Chain Reaction, or PCR, see K. Jordan and M. Lynch, 'The Mainstreaming of a Molecular Biological Tool: A Case Study of a New Technique', in Graham Button (ed.), Technology in Working Order: Studies of Work, Interaction and Technology (London & New York: Routledge, 1992), 160-80. See also K. Jordan and M. Lynch, 'The Dissemination, Standardization, and Routinization of a Molecular Biological Technique', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 773-99, and Joan Fujimura, 'The Molecular Biological Bandwagon in Cancer Research: Where Social Worlds Meet', Social Problems, Vol. 35 (1988), 261-83.
8. See Michael Lynch, 'The Discursive Production of Uncertainty: The OJ Simpson "Dream Team" and the Sociology of Knowledge Machine', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 829-68, for a discussion of how, in the case of forensic DNA analysis, defense lawyers often undertook the same type of deconstructive work that papers such as this one are attempting to do, albeit for different purposes. Using the example of the OJ Simpson double murder trial, Lynch shows how prosecutors took a 'realist' position towards the technology, and defense lawyers wrote a motion which took a constructivist approach, showing the extent to which quantitative estimates of random matches were dependent on local laboratory practices and interested judgements. I think it can be argued that after the formation of the DNA Task Force in 1990, defense attorneys had more resources for deconstruction on hand, some of them obtained in expert testimony by high-ranking population geneticists. There is a burgeoning literature on the development or 'black boxing' of various techniques used in molecular biology: see Kathleen Jordan and Michael Lynch, 'The Sociology of a Genetic Engineering Technique: Ritual and Rationality in the Performance of the Plasmid Prep', in Adele E. Clarke and Joan H. Fujimura (eds), The Right Tools for the Job (Princeton, NJ: Princeton University Press, 1992), 77-114. Jordan and Lynch critically examine the notion of 'black box' in reference to another tool of molecular biology, the 'plasmid prep', which has become a well-established procedure in molecular biology. By examining variations in the use of the plasmid prep, Jordan and Lynch show that, despite the current stability of the technique and its status as an easily accessible tool for molecular biologists, it remains locally problematic and must be 'constructed' anew whenever it is adapted to new circumstances (ibid., 77). For a related study of the Polymerase Chain Reaction, or PCR, see K. Jordan and M. Lynch, 'The Mainstreaming of a Molecular Biological Tool: A Case Study of a New Technique', in Graham Button (ed.), Technology in Working Order: Studies of Work, Interaction and Technology (London & New York: Routledge, 1992), 160-80. See also K. Jordan and M. Lynch, 'The Dissemination, Standardization, and Routinization of a Molecular Biological Technique', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 773-99, and Joan Fujimura, 'The Molecular Biological Bandwagon in Cancer Research: Where Social Worlds Meet', Social Problems, Vol. 35 (1988), 261-83.
9. Barry Scheck and Peter Neufeld, 'DNA Task Force Report', The Champion (June 1991), 13-21, at 19. (The Champion is a publication of the National Association of Criminal Defense Lawyers.)
10. People v. Castro, 545 N.Y.S.2d 985 (Su 1989).
11. Sheila Jasanoff argues that the process of deconstruction of the scientific claims made by the private laboratories was set in motion by the 'legal system's normative commitment to finding two sides in every case [which] led in Castro to a confrontation among experts who finally questioned some of the methodological premises of DNA testing and testimony': Sheila Jasanoff, Science at the Bar (Cambridge, MA & London: Harvard University Press, 1995), 57.
12. United States v. John Ray Bonds, Mark Verdi and Steven Wayne Yee, 12 F.3d 540; 1993 US App. Lexis 32574; 1994 FED App. 0085P.
13. Jasanoff, op. cit. note 6, 55.
14. National Research Council, DNA Technology in Forensic Science (Washington, DC: National Academy Press, 1992).
15. Thomas Nagel, The View From Nowhere (Oxford: Clarendon Press, 1986). Historian of science Theodore Porter argues that the challenge for every scientist is that '[e]very scientific result begins its career as a view from somewhere - say some particular laboratory - and it is really the most fundamental task of every scientist to transform it as much as possible into a view from nowhere, at least nowhere in particular': Theodore Porter, 'Quantification and the Accounting Ideal in Science', Social Studies of Science, Vol. 22, No. 4 (November 1992), 633-52, at 646-47.
16. Thomas Nagel, The View From Nowhere (Oxford: Clarendon Press, 1986). Historian of science Theodore Porter argues that the challenge for every scientist is that '[e]very scientific result begins its career as a view from somewhere - say some particular laboratory - and it is really the most fundamental task of every scientist to transform it as much as possible into a view from nowhere, at least nowhere in particular': Theodore Porter, 'Quantification and the Accounting Ideal in Science', Social Studies of Science, Vol. 22, No. 4 (November 1992), 633-52, at 646-47.
17. Thomas F. Gieryn argues that a familiar notion of objectivity is that of 'objectivity as denial' - denial of all that is biographical and personal. Associated with the goal of making knowledge claims that with 'increasing representational accuracy, the means are a distinctive set of trans-situational procedures and practices often associated with the epistemic culture of science: quantification; standardization of tools, research materials, and procedures; public and critical communication; impersonal rules of logic and evidence. On this account, accurate representations of reality depend on an excision of passions, emotions, values, interests, and other subjectivities from the warranting of reality claims': T.F. Gieryn, 'Objectivity for These Times', Perspectives on Science, Vol. 2, No. 3 (Fall 1994), 324-49, at 325. Gieryn and I share the same goal, a view of objectivity that does not deny the rôle of the representing subject. In this paper I will address one of the familiar means of achieving objectivity, that of quantification, specifically the construction of 'measurement error', and show how various procedures, negotiations and decisions in the process of constructing the measure of error make the representing subject invisible. Once made invisible through the construction of a number, the representing subject disappears, and the knowledge claim put forth is a step closer to being an objective knowledge claim. What I am doing is showing the moment when one of the 'God-tricks' that give modern science its claim to objectivity takes place: for discussion of 'God-tricks', see Donna Haraway, Simians, Cyborgs and Women (New York: Routledge, 1991), 189.
18. Thomas F. Gieryn argues that a familiar notion of objectivity is that of 'objectivity as denial' - denial of all that is biographical and personal. Associated with the goal of making knowledge claims that with 'increasing representational accuracy, the means are a distinctive set of trans-situational procedures and practices often associated with the epistemic culture of science: quantification; standardization of tools, research materials, and procedures; public and critical communication; impersonal rules of logic and evidence. On this account, accurate representations of reality depend on an excision of passions, emotions, values, interests, and other subjectivities from the warranting of reality claims': T.F. Gieryn, 'Objectivity for These Times', Perspectives on Science, Vol. 2, No. 3 (Fall 1994), 324-49, at 325. Gieryn and I share the same goal, a view of objectivity that does not deny the rôle of the representing subject. In this paper I will address one of the familiar means of achieving objectivity, that of quantification, specifically the construction of 'measurement error', and show how various procedures, negotiations and decisions in the process of constructing the measure of error make the representing subject invisible. Once made invisible through the construction of a number, the representing subject disappears, and the knowledge claim put forth is a step closer to being an objective knowledge claim. What I am doing is showing the moment when one of the 'God-tricks' that give modern science its claim to objectivity takes place: for discussion of 'God-tricks', see Donna Haraway, Simians, Cyborgs and Women (New York: Routledge, 1991), 189.
19. See Evelyn Fox Keller, 'The Paradox of Scientific Subjectivity', Annals of Scholarship, Vol. 9, Nos 1-2 (1992), 135-54, for a discussion on the history of the 'subjectless' representation of the world that is equated with scientific objectivity in the 20th century.
20. The perspective on objectivity as the ridding or erasure of subjectivity is the result of much work in what has come to be called the history of objectivity. Today objectivity has a 'negative' quality, in the sense of 'an aspiration to eliminate or minimize the distortions of subjectivity, to go beyond the merely personal': Theodore M. Porter, 'Introduction' to the Symposium on 'The Social History of Objectivity', Social Studies of Science, Vol. 22, No. 4 (November 1992), 595. See also Lorraine Daston, 'Objectivity and the Escape from Perspective', ibid., 597-618.
21. The perspective on objectivity as the ridding or erasure of subjectivity is the result of much work in what has come to be called the history of objectivity. Today objectivity has a 'negative' quality, in the sense of 'an aspiration to eliminate or minimize the distortions of subjectivity, to go beyond the merely personal': Theodore M. Porter, 'Introduction' to the Symposium on 'The Social History of Objectivity', Social Studies of Science, Vol. 22, No. 4 (November 1992), 595. See also Lorraine Daston, 'Objectivity and the Escape from Perspective', ibid., 597-618.
22. Part of the problem with theorizing objectivity from the viewpoint of practices which consist of judgement, evaluation and subjectivity is that the doctrine of logical positivism, which has dominated Western philosophy of science for much of the 20th century, tied objectivity to propositions: formal or factual statements subject to clear criteria of truth and falsehood. On this view, statements of evaluation, which are normative statements, such as 'good', 'bad', 'big enough', 'small enough' -'reasonable' - are personal self-reports. They are of the realm of opinion and, as such, are not subject to truth claims; and hence are outside the realm of objectivity. Objective knowledge is knowledge which does not include personal self-reports, opinion or judgement. No objective knowledge can be had about them. The doctrine of emotivism held 'that all evaluative (especially moral) judgements were simply expressions of preference on the part of the speaker. An expression such as "this is good" meant "I approve of this; do so as well". It was neither true nor false, but simply displayed the speaker's attitude and attempted to produce a similar attitude in others': from Charles Leslie Stevenson, Ethics and Language (New Haven, CT: Yale University Press, 1945), cited in Lena Jayyusi, 'Values and Moral Judgement: Communicative Praxis as Moral Order', in Graham Button (ed.), Ethnomethodology and the Human Sciences (Cambridge: Cambridge University Press, 1991), 227-51, at 232. In contrast to this view from philosophy, I will argue that evaluative statements are an essential part of the measurement process and, as such, are epistemic acts which are crucial to the construction of objective knowledge.
23. Part of the problem with theorizing objectivity from the viewpoint of practices which consist of judgement, evaluation and subjectivity is that the doctrine of logical positivism, which has dominated Western philosophy of science for much of the 20th century, tied objectivity to propositions: formal or factual statements subject to clear criteria of truth and falsehood. On this view, statements of evaluation, which are normative statements, such as 'good', 'bad', 'big enough', 'small enough' - 'reasonable' - are personal self-reports. They are of the realm of opinion and, as such, are not subject to truth claims; and hence are outside the realm of objectivity. Objective knowledge is knowledge which does not include personal self-reports, opinion or judgement. No objective knowledge can be had about them. The doctrine of emotivism held 'that all evaluative (especially moral) judgements were simply expressions of preference on the part of the speaker. An expression such as "this is good" meant "I approve of this; do so as well". It was neither true nor false, but simply displayed the speaker's attitude and attempted to produce a similar attitude in others': from Charles Leslie Stevenson, Ethics and Language (New Haven, CT: Yale University Press, 1945), cited in Lena Jayyusi, 'Values and Moral Judgement: Communicative Praxis as Moral Order', in Graham Button (ed.), Ethnomethodology and the Human Sciences (Cambridge: Cambridge University Press, 1991), 227-51, at 232. In contrast to this view from philosophy, I will argue that evaluative statements are an essential part of the measurement process and, as such, are epistemic acts which are crucial to the construction of objective knowledge.
24. Thomas Kuhn argues that in scientific measurement there is always a discrepancy between what is predicted by theory and what is observed empirically; that scientists do not expect perfect agreement between theory and data, and indeed become suspicious if empirical measurements are too close to those predicted by theory: Thomas Kuhn, 'The Function of Measurement in the Physical Sciences', in his The Essential Tension: Selected Studies in Scientific Tradition (Chicago, IL: The University of Chicago Press, 1977), 178-224, at 182-86. For a systematic introduction to the idea that the social includes the cognitive - that it is the strength of empirical observations accepted by scientists that they are both constitutively social and judgemental and based on material inputs - see Barry Barnes, David Bloor and John Henry, Scientific Knowledge: A Sociological Analysis (London: Athlone Press; Chicago, IL: The University of Chicago Press, 1996), esp. Chapters 1-4. See also Michael Lynch, 'Method: Measurement - Ordinary and Scientific Measurement as Ethnomethodological Phenomena', in Button (ed.), op. cit. note 14, 77-108.
25. Thomas Kuhn argues that in scientific measurement there is always a discrepancy between what is predicted by theory and what is observed empirically; that scientists do not expect perfect agreement between theory and data, and indeed become suspicious if empirical measurements are too close to those predicted by theory: Thomas Kuhn, 'The Function of Measurement in the Physical Sciences', in his The Essential Tension: Selected Studies in Scientific Tradition (Chicago, IL: The University of Chicago Press, 1977), 178-224, at 182-86. For a systematic introduction to the idea that the social includes the cognitive - that it is the strength of empirical observations accepted by scientists that they are both constitutively social and judgemental and based on material inputs - see Barry Barnes, David Bloor and John Henry, Scientific Knowledge: A Sociological Analysis (London: Athlone Press; Chicago, IL: The University of Chicago Press, 1996), esp. Chapters 1-4. See also Michael Lynch, 'Method: Measurement - Ordinary and Scientific Measurement as Ethnomethodological Phenomena', in Button (ed.), op. cit. note 14, 77-108.
26. Thomas Kuhn argues that in scientific measurement there is always a discrepancy between what is predicted by theory and what is observed empirically; that scientists do not expect perfect agreement between theory and data, and indeed become suspicious if empirical measurements are too close to those predicted by theory: Thomas Kuhn, 'The Function of Measurement in the Physical Sciences', in his The Essential Tension: Selected Studies in Scientific Tradition (Chicago, IL: The University of Chicago Press, 1977), 178-224, at 182-86. For a systematic introduction to the idea that the social includes the cognitive - that it is the strength of empirical observations accepted by scientists that they are both constitutively social and judgemental and based on material inputs - see Barry Barnes, David Bloor and John Henry, Scientific Knowledge: A Sociological Analysis (London: Athlone Press; Chicago, IL: The University of Chicago Press, 1996), esp. Chapters 1-4. See also Michael Lynch, 'Method: Measurement - Ordinary and Scientific Measurement as Ethnomethodological Phenomena', in Button (ed.), op. cit. note 14, 77-108.
27. While I argue in this paper that the standard error of measurement, in any context in which it is used, serves to make invisible all the mediations through which human DNA is processed, Sheila Jasanoff makes the argument that once the DNA is in the courtroom context, it is the expert witness who serves as the 'container' or 'concealer' of 'much of the behind-the-scenes work of translation, from the investigative site through the forensic laboratory into testimony at trial'. Similarly, Jasaonff argues that the very dynamics of litigation obscure 'the complexity of the translations by which samples, artifacts, recordings or pictures become evidence': Sheila Jasanoff, 'The Eye of Everyman: Witnessing DNA in the Simpson Trial', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 713-40, quotes at 716. Jasanoff further notes that the 'metonymic genius of language that converts "DNA" into a stark signifier of truth suppresses a world of social activity' (ibid., 718). This paper addresses this world of social activity that takes place before DNA profiles reach the courtroom.
28. For a much fuller exploration of the introduction of DNA typing to the United States and the scientific controversies which ensued, see Linda Derksen, The DNA Wars: Negotiating the Meaning of Measurement (PhD dissertation, Department of Sociology, University of California, San Diego, forthcoming).
29. The pervasive myth that scientific knowledge is not man-made, constructed, articulated or crafted - but rather discovered - has its roots in a particular style of scientific reporting that began in the 17th century: see Steven Shapin, 'Pump and Circumstance: Robert Boyle's Literary Technology', Social Studies of Science, Vol. 14, No. 4 (November 1984), 481-519, at 510.
30. Michael Lynch, op. cit. note 15, makes the argument that measurement is a contingent, local achievement. He gives an example (ibid., 100-01) of how the seemingly simple act of measuring out a quantity of powdered chemical is tied to the local context in which the measurement occurs, and to a string of 'thisses' and 'thats' which cannot be specified more explicitly or concretely. The instructive demonstration [of how to measure the powder] is riddled with deictic references individuating its terms in an assemblage of haecceities: 'take your plastic thing'; 'set it on there'; 'Get a clean metal spatula'; 'take it there'; 'add it while that thing's on there'; 'like that'; 'that way you don't have to touch anything but this thing'; 'wash that thing off really well'. The assemblage is enunciated through an order of 'thisses, thats, and theres' immediately tagged to the equipment at hand. . . . The action is thus hopelessly and irremediably scientific in its every detail, but this 'science' is nothing other than a vulgar competence in a singular laboratory environment.
31. See Henry Ely Kyburg, Theory and Measurement (Cambridge: Cambridge University Press, 1984), cited in Lynch, op. cit. note 15, 77.
32. See Henry Ely Kyburg, Theory and Measurement (Cambridge: Cambridge University Press, 1984), cited in Lynch, op. cit. note 15, 77.
33. Stressing the importance of communication between disparate groups in the making of objective knowledge, Theodore Porter argues that objectivity came to be associated with quantification because numbers provide a language which facilitates communication between groups. Translating properties of the world into numbers serves as a 'technology of distance' - distancing the personal, the geographical, the intellectual and the social. The quantitative protocols and the rules of mathematics provide a form of knowledge that is 'decidedly public in character. Such knowledge is especially useful to coordinate the activities of diverse actors, and to lend credibility to forms of belief and action when personal trust is in short supply': Porter, op.cit. note 10, 640. For further discussion see Theodore Porter, 'Objectivity as Standardization: The Rhetoric of Impersonality in Measurement, Statistics, and Cost-Benefit Analysis', Annals of Scholarship, Vol. 9, Nos 1-2 (1992), 19-60.
34. Stressing the importance of communication between disparate groups in the making of objective knowledge, Theodore Porter argues that objectivity came to be associated with quantification because numbers provide a language which facilitates communication between groups. Translating properties of the world into numbers serves as a 'technology of distance' - distancing the personal, the geographical, the intellectual and the social. The quantitative protocols and the rules of mathematics provide a form of knowledge that is 'decidedly public in character. Such knowledge is especially useful to coordinate the activities of diverse actors, and to lend credibility to forms of belief and action when personal trust is in short supply': Porter, op.cit. note 10, 640. For further discussion see Theodore Porter, 'Objectivity as Standardization: The Rhetoric of Impersonality in Measurement, Statistics, and Cost-Benefit Analysis', Annals of Scholarship, Vol. 9, Nos 1-2 (1992), 19-60.
35. See also Porter, op. cit. note 10, 646-47.
36. Kuhn (op. cit. note 15) argues that measurement necessarily involves approximation. This approximation may be due to technical limits in precision of measurement, to variation between observers or, for example, to attempts to control essential variables, such as to achieve totally frictionless planes or completely perfect vacua.
37. For a discussion of the origins of this sense of measurement error, see Simon Schaffer, 'Astronomers Mark Time: Discipline and the Personal Equation', Science in Context, Vol. 2 (1988), 115-45.
38. Sheila Jasanoff (op. cit. note 6, xiv) makes the point that all scientific claims, especially those that are implicated in legal controversies, are 'highly contested, contingent on particular localized circumstances and freighted with buried presumptions about the social world in which they are deployed'.
39. 'A DNA profile consists of a set of measurements of discrete random variables, measured with error': Kathryn Roeder, 'DNA Fingerprinting: A Review of the Controversy', Statistical Science, Vol. 9, No. 2 (May 1994), 222-47, at 222.
40. In response to an anonymous reviewer's comment, the point is not that measurement error is huge and thus biases frequency estimates. Error-free measurement is not possible. However, the point is that measurement error serves to make invisible most of the social and cultural processes which go into translating properties of nature, or DNA, into numbers. By peeling back the layers of the way statistics of measurement error are constructed, we can see exactly at what moment 'personal judgement and subjectivity' vanish, and knowledge claims move one step closer to being 'objective'.
41. David Kaye, 'DNA Evidence: Probability, Population Genetics, and the Courts', Harvard Journal of Law and Technology, Vol. 101, No. 7 (Fall 1993), 101-71.
42. David H. Kaye, 'The Forensic Debut of the National Research Council's DNA Report: Population Structure, Ceiling Frequencies and the Need for Numbers', Jurimetrics Journal, Vol. 34, No. 4 (Summer 1994), 369-82, at 370.
43. Generally, when new scientific techniques are brought before the courts, they must pass what has come to be called 'the Frye standard'. The Frye standard for evaluating novel scientific evidence dates back to a case in 1923 involving the admissibility of lie detector results. In Frye v. United States (293 F. 1013 [D.C.Cir. 1923]), the court held that expert testimony which relates to novel scientific evidence must satisfy special foundational conditions which are not applicable to other types of expert testimony. Specifically, the judge declared that 'Just when a scientific principle or discovery crosses the line between the experimental and demonstrable stages is difficult to define. Somewhere in the twilight zone the evidential force of the principle must be recognized, and while the courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field to which it belongs' (ibid., 293F at 1014; emphasis in original). In the Andrews case, the judge did not follow the Frye standard of general acceptance in the scientific community, but rather ruled that the evidence was relevant, following United States v. Downing (753 F.2d 1224 [3d Cir. 1985]). The relevancy test of Downing says that where no established 'track record' for a novel scientific procedure exists, the court must look to 'other factors' which bear on its reliability. Since 1993, the Frye test has largely been superseded by the decision of the Supreme Court in Daubert v. Merrell Dow Pharmaceuticals, Inc. (509 U.S. 579, 1993), where the court ruled that if it can be proven that new scientific evidence has relevance to the case being tried, it need not meet with the general acceptance rule of Frye. This gives US courts the option of following admissibility standards specified in the Federal Rules of Evidence. This decision places more power in the hand of judges to determine what is relevant to a particular case, whether or not it has been generally accepted in the scientific community. Some states (like California) continued to adhere to the Frye standard, but most others began to follow Daubert and the Federal Rules of Evidence.
44. James E. Starrs (Professor of Law and Forensic Sciences, George Washington University), testimony before the Subcommittee on the Constitution, Committee on the Judiciary, US Senate, hearings on Genetic Testing as a Means of Criminal Investigation, 15 March 1989 (Washington, DC: US Government Printing Office, 1990), 4-40. See also Kaye, op. cit. note 28, 101.
45. James E. Starrs (Professor of Law and Forensic Sciences, George Washington University), testimony before the Subcommittee on the Constitution, Committee on the Judiciary, US Senate, hearings on Genetic Testing as a Means of Criminal Investigation, 15 March 1989 (Washington, DC: US Government Printing Office, 1990), 4-40. See also Kaye, op. cit. note 28, 101.
46. Quotes from Starrs, op. cit. note 31, 11, 12. As an example of how poorly understood the technique was by judges, Starrs (ibid., 11) said that in the Timothy Spencer trial for capital murder and rape, Judge Kendrick . . . . . . listened attentively . . . to the scrupulously precise and detailed testimony of prosecution expert Dr Michael Baird [of Lifecodes Corp.]. After Dr Baird had concluded spelling out the intricacies of each of the laboratory steps in DNA matching, Judge Kendrick, in some obvious puzzlement, asked for clarification on one point. The judge had counted only seven steps, but the expert said there were eight in all. Was not the digestion of DNA by a restriction enzyme identical to the process of separation of fragments by gel electrophoresis, the judge inquired in so many words. A muzzled gasp was almost audible in the courtroom from the observers schooled in the rudiments of DNA matching or even in just plain instrumental analysis. . . . His honor was gently informed that the processes were distinctly different, as distinctly different one might add, as bicycling is from swimming even though in a triathlon one follows the other.
47. Ivan Balazs, Michael Baird, Mindy Clyne and Ellie Meade, 'Human Population Genetic Studies of Five Hypervariable DNA Loci', American Journal of Human Genetics, Vol. 44 (1989), 182-90.
48. Starrs, op. cit. note 31, at 23.
49. Ibid., 12.
50. Ibid., 22.
51. Julie Cannon Shoop, 'Is DNA Typing Ready for Trial?', Trial (September 1990), 11-15.
52. Kaye, op. cit. note 28.
53. Parloff, op. cit. note 2, 1.
54. Although the case used in this example does not represent a successful presentation of DNA profiles by Lifecodes Corporation, see Arthur Daemmrich, 'The Evidence Does Not Speak for Itself: Expert Witnesses and the Organization of DNA-Typing Companies', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 741-72, for an examination of how the organization of private DNA-typing companies contributed to their success in presenting DNA profiles as admissible evidence in the courtroom.
55. Parloff, op. cit. note 2, 2.
56. Ibid.
57. Peter Neufeld, quoted in ibid., at 53.
58. 'Bandshifting' occurs when one or more lanes in a gel run at a faster or slower speed than the other lanes, thus physically and visually shifting the bands up or down, displaced horizontally from their potentially 'matching' bands in other lanes. For a thorough explanation of the technique of DNA profiling and some of its potential technical problems, see Keith Inman and Norah Rudin, An Introduction to Forensic DNA Analysis (New York: CRC Press, 1997), esp. 59-69.
59. Lander, quoted in Parloff, op. cit. note 2, 53.
60. Ibid., 53.
61. For a discussion of Eric Lander's rôle and history as an expert witness, see Saul Halfon, 'Collecting, Testing and Convincing: Forensic DNA Experts in the Courts', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998) 801-28, at 812-14.
62. Eric Lander, quoted in Parloff, op. cit. note 2, at 53.
63. Ibid., 54.
64. Ibid., quote from Baird at 54.
65. Eric Lander, 'DNA Fingerprinting on Trial', Nature, Vol. 339 (15 June 1989) 501-05, both quotes at 503.
66. Hardy-Weinberg equilibrium is the state achieved when a population is mating at random and no sub-populations exist. This allows the assumption of independence of alleles to be made, which should then allow for the product-rule to be used in calculating random match probabilities.
67. Lander, op. cit. note 51, 503.
68. Parloff, op. cit. note 2, both quotations at 55.
69. Lander, op. cit. note 51, 501.
70. Readers who would prefer a brief introduction to the general procedures involved in single locus DNA analysis before getting into the technicalities of measurement error should consult the Appendix (829-31).
71. Balazs et al., op. cit. note 33, at 182-83.
72. Lifecodes' first attempts to construct frequency distributions used the blood of 700 people, obtained from the New York Blood Center. In their analysis they found statistically significant differences in allele frequencies between ethnic groups: that is, by the rules of statistics, it was not just the luck of the draw that certain fragment lengths occurred more frequently in the black group than in the Caucasian group. There is only a very small chance that in a different sample of 700 people, blacks and Caucasians would come up with the same frequencies of alleles. See Michael Baird, Ivan Balazs, Allan Giusti et al., 'Allele Frequency Distribution of Two Highly Polymorphic DNA Sequences in Three Ethnic Groups and Its Application to the Determination of Paternity', American Journal of Human Genetics, Vol. 39 (1986), 489-501.
73. Both quotations from Balazs et al., op. cit. note 33, at 183.
74. An easy interpretation of a standard deviation (in this case, because it is a standard deviation of measurement errors, it can be called a 'standard error') is as a 'typical', or 'average', error or deviation from the mean value. If you don't know anything about a distribution, your best guess as to what value a member of that distribution has is the mean, or average value. The standard deviation quantifies what a typical deviation from that mean, or average value is. By the rules of statistics, for normally distributed populations, 95% of all the cases fall within two standard deviations of the mean, and 99% of the cases fall within three standard deviations of the mean. The standard error is a number that expresses the range of variation one can normally expect in measuring fragment lengths.
75. Bruce S. Weir, 'Review. Population Genetics in the Forensic DNA Debate', Proceedings of the National Academy of Sciences, Vol. 89 (December 1992), 11654-59, at 11655.
76. Donald A. Berry, I.W. Evett and R. Pinchin, 'Statistical Inference in Crime Investigations using Deoxyribonucleic Acid Profiling', Applied Statistics, Vol. 41, No. 3 (1992), 499-531, at 502.
77. Bruce Budowle, Alan M. Giusti, John S. Waye, F. Samuel Baechtel, Ronald M. Fourney, Dwight E. Adams, Lawrence A. Presley, Harold A. Deadman and Keith L. Monson, 'Fixed Bin Analysis for Statistical Evaluation of Continuous Distributions of Allele Data from VNTR Loci, for Use in Forensic Comparisons', American Journal of Human Genetics, Vol. 48 (1991), 841-55, at 844.
78. Eric S. Lander, 'Invited Editorial: Research on DNA Typing Catching up with Courtroom Application', American Journal of Human Genetics, Vol. 48 (1991), 819-23, at 820. For a discussion of variation in sizes of match windows and standard deviations, see also Kaye, op. cit. note 29.
79. Eric S. Lander, 'Invited Editorial: Research on DNA Typing Catching up with Courtroom Application', American Journal of Human Genetics, Vol. 48 (1991), 819-23, at 820. For a discussion of variation in sizes of match windows and standard deviations, see also Kaye, op. cit. note 29.
80. Neil J. Risch and Bernard Devlin, 'On the Probability of Matching DNA Fingerprints', Science, Vol. 225 (7 February 1992), 717-20, at 720.
81. Roeder, op. cit. note 26, 222.
82. See William C. Thompson and Simon Ford, 'The Meaning of a Match: Sources of Ambiguity in the Interpretation of DNA Prints', in Mark A. Farley and James J. Harrington (eds), Forensic DNA Technology (Chelsea, MI: Lewis Publishers, 1991), 93-152. See also W.C. Thompson, 'Evaluating the Admissibility of New Genetic Identification Tests: Lessons from the "DNA Wars"', Journal of Criminal Law and Criminology, Vol. 84, No. 1 (Spring 1994), 22-104.
83. See William C. Thompson and Simon Ford, 'The Meaning of a Match: Sources of Ambiguity in the Interpretation of DNA Prints', in Mark A. Farley and James J. Harrington (eds), Forensic DNA Technology (Chelsea, MI: Lewis Publishers, 1991), 93-152. See also W.C. Thompson, 'Evaluating the Admissibility of New Genetic Identification Tests: Lessons from the "DNA Wars"', Journal of Criminal Law and Criminology, Vol. 84, No. 1 (Spring 1994), 22-104.
84. See Lander, op. cit. note 51; Lander, op. cit. note 64; W.M. Shields, 'Forensic DNA Typing as Evidence in Criminal Proceedings: Some Problems and Potential Solutions', in Proceedings of the Third International Symposium on Human Identification (Madison, WI: Promega Corp., 1992), 1-50.
85. See Lander, op. cit. note 51; Lander, op. cit. note 64; W.M. Shields, 'Forensic DNA Typing as Evidence in Criminal Proceedings: Some Problems and Potential Solutions', in Proceedings of the Third International Symposium on Human Identification (Madison, WI: Promega Corp., 1992), 1-50.
86. See Lander, op. cit. note 51; Lander, op. cit. note 64; W.M. Shields, 'Forensic DNA Typing as Evidence in Criminal Proceedings: Some Problems and Potential Solutions', in Proceedings of the Third International Symposium on Human Identification (Madison, WI: Promega Corp., 1992), 1-50.
87. Roeder, op. cit. note 26, 227.
88. David Kaye notes that this variability in the size of measurement error, or matching rules, leads to potentially confusing discourse in the courtroom: 'The prosecution says to the defendant, "under our match rule, you match". The defendant replies, "That's your rule. Under a different rule, I don't match"': Kaye, op. cit. note 28, 113.
89. William C. Thompson, 'Comment' [on Roeder, op. cit. note 26], Statistical Science, Vol. 9, No. 2 (May 1994), 263-66, at 264. See also Thompson, op. cit. note 67, 40.
90. William C. Thompson, 'Comment' [on Roeder, op. cit. note 26], Statistical Science, Vol. 9, No. 2 (May 1994), 263-66, at 264. See also Thompson, op. cit. note 67, 40.
91. Roeder, op. cit. note 26, 227.
92. Thompson, op. cit. note 71, 264; William C. Thompson and Simon Ford, 'Is the Probative Value of Forensic DNA Evidence Undermined by Subjectivity in Determination of Matches?', in W. Harnagel and David H. Kaye (eds), Proceedings of the Second International Conference on Forensic Statistics (Tempe, AZ: Arizona State University Center for Law, Science and Technology, 1993), A47-A60.
93. Thompson, op. cit. note 71, 264; William C. Thompson and Simon Ford, 'Is the Probative Value of Forensic DNA Evidence Undermined by Subjectivity in Determination of Matches?', in W. Harnagel and David H. Kaye (eds), Proceedings of the Second International Conference on Forensic Statistics (Tempe, AZ: Arizona State University Center for Law, Science and Technology, 1993), A47-A60.
94. William C. Thompson and Simon Ford, 'DNA Typing: Acceptance and Weight of the New Genetic Identification Tests', Virginia Law Review, Vol. 75, No. 1 (February 1989), 45-108, at 87-88.
95. However, it should be noted that differing sizes of match windows have tremendous forensic importance. Having a large match window increases the probability of a match between two samples, but also increases the chances of a false positive match. Having a small match window makes it more difficult to declare a match, but opens up the possibility of false negatives (that is, when two fragments are declared not to be a match, but come from the same person). See Kaye, op. cit. note 28, at 113, for a discussion of court cases in which various defense counsel argued that the size of the match window was too wide, and hence that their clients' profiles should be declared as non-matching.
96. The acts of judgement involved in the forensic analysis of DNA involve evaluative statements, such as 'this is good enough', or 'this is not good enough'. In the context of creating a standard measure of error, statements such as these are performatives -they bring about a certain state of affairs in the world. The judgements are not propositional, nor are they subjective emotions or reflections of the attitude on the part of the speaker. I argue that these kinds of evaluative statements can be epistemic acts. In peeling back the layers of the onion of objectivity, evaluative, judgemental statements are part of the work of making objectivity. Far from being a source of bias or a threat to objectivity, evaluative statements are the very stuff of which objectivity is made. Objectivity is constituted in the process of communicative action: see Jürgen Habermas, The Theory of Communicative Action, Volume 1 (Boston, MA: Beacon Press, 1984). To see measurement as sociological is to make visible its normative, or evaluative components. This requires a new conception of the moral where values are not 'contaminant, regulative or determinant, but constitutive': Jayyusi, op. cit. note 14, 231 (emphasis in original).
97. The acts of judgement involved in the forensic analysis of DNA involve evaluative statements, such as 'this is good enough', or 'this is not good enough'. In the context of creating a standard measure of error, statements such as these are performatives - they bring about a certain state of affairs in the world. The judgements are not propositional, nor are they subjective emotions or reflections of the attitude on the part of the speaker. I argue that these kinds of evaluative statements can be epistemic acts. In peeling back the layers of the onion of objectivity, evaluative, judgemental statements are part of the work of making objectivity. Far from being a source of bias or a threat to objectivity, evaluative statements are the very stuff of which objectivity is made. Objectivity is constituted in the process of communicative action: see
98. The socially transmitted nature of successful scientific practice is well explored in the work of Harry Collins. Collins has developed an enculturational model of scientific knowledge transfer, whereby the transmission of scientific knowledge 'does not consist of the logical accumulation of packages of knowledge': H.M. Collins, 'The TEA Set: Tacit Knowledge and Scientific Networks', Science Studies, Vol. 4, No. 2 (April 1974), 165-86, at 174. Rather, it occurs through members coming to share a culture or form of life. Knowledge is located in a community of expert practitioners, who must interact with each other to acquire the necessary skills and knowledge: Collins, Changing Order (London: Sage, 1985), 159. The enculturational model of science holds that knowledge of a certain sort can only be transmitted tacitly, that is, by sitting with or being with someone who possesses the knowledge. Collins' larger claim is that all knowledge contains a tacit dimension, in that even transfer of knowledge by word or language requires a shared culture or form of life.
99. The socially transmitted nature of successful scientific practice is well explored in the work of Harry Collins. Collins has developed an enculturational model of scientific knowledge transfer, whereby the transmission of scientific knowledge 'does not consist of the logical accumulation of packages of knowledge': H.M. Collins, 'The TEA Set: Tacit Knowledge and Scientific Networks', Science Studies, Vol. 4, No. 2 (April 1974), 165-86, at 174. Rather, it occurs through members coming to share a culture or form of life. Knowledge is located in a community of expert practitioners, who must interact with each other to acquire the necessary skills and knowledge: Collins, Changing Order (London: Sage, 1985), 159. The enculturational model of science holds that knowledge of a certain sort can only be transmitted tacitly, that is, by sitting with or being with someone who possesses the knowledge. Collins' larger claim is that all knowledge contains a tacit dimension, in that even transfer of knowledge by word or language requires a shared culture or form of life.
100. Thompson & Ford, op. cit. note 67, at 136, 137.
101. Karin Knorr Cetina's ethnographic work has led her to argue that the epistemic culture in molecular biology is such that to produce knowledge, molecular biologists have to become 'repositories of unconscious experience' and individual scientists have to develop an embodied sense of a reasonable response to different situations: Karin Knorr Cetina, 'The Couch, the Cathedral, and the Laboratory: On the Relationship between Experiment and Laboratory in Science', in Andrew Pickering (ed.), Science as Practice and Culture (Chicago, IL: The University of Chicago Press, 1992), 113-38, at 119. Molecular biologists often have to make an educated guess as to what procedure is best in a given situation, and so the sense of what counts as a successful procedure depends heavily on an individual's experience and 'upon the prognostic knowledge which individuals must somehow synthesize from features of their previous experience, and which remains implicit, embodied, and encapsulated within the person' (ibid., 121). The sense of what counts as a successful procedure is implicit - it is a blend of the individual's experience and the narrative culture in the laboratory. Knorr Cetina calls the epistemological culture within a molecular biology laboratory, and the form of reasoning practitioners use, 'biographical', because this form of reasoning is sustained by the entire body of knowledge and experience which the researcher bulds up over a professional lifetime. 'It is a knowledge which draws upon scientists' bodies rather than their minds' (ibid., 121). Harry Collins has also demonstrated that tacit knowledge and 'body knowledge' are an essential part of the transmission of scientific knowledge and skills: on the transmission of skills, see Collins (1974), op. cit. note 77. For my purposes, the important part of Knorr Cetina's argument is that there exist different epistemic cultures between sciences, and that each science is characterized by a particular culture. This distinctive culture affects how numbers are assigned to properties of the world in practice. Thus making the numbers themselves a social product - attached to the material world through the materiality of DNA samples, but influenced by the epistemic culture within which they were produced. In molecular biology, this culture is one which leads to heavy reliance on the personal experience and subjectivity of the individual practitioner. This is not an 'error' or source of bias within the science, but a defining quality of the practice of molecular biology, and a constitutive feature of the production of objective knowledge in the field. For a more complete discussion, see Karin Knorr Cetina, Epistemic Cultures (Cambridge, MA: Harvard University Press, 1999).
102. Karin Knorr Cetina's ethnographic work has led her to argue that the epistemic culture in molecular biology is such that to produce knowledge, molecular biologists have to become 'repositories of unconscious experience' and individual scientists have to develop an embodied sense of a reasonable response to different situations: Karin Knorr Cetina, 'The Couch, the Cathedral, and the Laboratory: On the Relationship between Experiment and Laboratory in Science', in Andrew Pickering (ed.), Science as Practice and Culture (Chicago, IL: The University of Chicago Press, 1992), 113-38, at 119. Molecular biologists often have to make an educated guess as to what procedure is best in a given situation, and so the sense of what counts as a successful procedure depends heavily on an individual's experience and 'upon the prognostic knowledge which individuals must somehow synthesize from features of their previous experience, and which remains implicit, embodied, and encapsulated within the person' (ibid., 121). The sense of what counts as a successful procedure is implicit - it is a blend of the individual's experience and the narrative culture in the laboratory. Knorr Cetina calls the epistemological culture within a molecular biology laboratory, and the form of reasoning practitioners use, 'biographical', because this form of reasoning is sustained by the entire body of knowledge and experience which the researcher bulds up over a professional lifetime. 'It is a knowledge which draws upon scientists' bodies rather than their minds' (ibid., 121). Harry Collins has also demonstrated that tacit knowledge and 'body knowledge' are an essential part of the transmission of scientific knowledge and skills: on the transmission of skills, see Collins (1974), op. cit. note 77. For my purposes, the important part of Knorr Cetina's argument is that there exist different epistemic cultures between sciences, and that each science is characterized by a particular culture. This distinctive culture affects how numbers are assigned to properties of the world in practice. Thus making the numbers themselves a social product - attached to the material world through the materiality of DNA samples, but influenced by the epistemic culture within which they were produced. In molecular biology, this culture is one which leads to heavy reliance on the personal experience and subjectivity of the individual practitioner. This is not an 'error' or source of bias within the science, but a defining quality of the practice of molecular biology, and a constitutive feature of the production of objective knowledge in the field. For a more complete discussion, see Karin Knorr Cetina, Epistemic Cultures (Cambridge, MA: Harvard University Press, 1999).
103. Karin Knorr Cetina's ethnographic work has led her to argue that the epistemic culture in molecular biology is such that to produce knowledge, molecular biologists have to become 'repositories of unconscious experience' and individual scientists have to develop an embodied sense of a reasonable response to different situations: Karin Knorr Cetina, 'The Couch, the Cathedral, and the Laboratory: On the Relationship between Experiment and Laboratory in Science', in Andrew Pickering (ed.), Science as Practice and Culture (Chicago, IL: The University of Chicago Press, 1992), 113-38, at 119. Molecular biologists often have to make an educated guess as to what procedure is best in a given situation, and so the sense of what counts as a successful procedure depends heavily on an individual's experience and 'upon the prognostic knowledge which individuals must somehow synthesize from features of their previous experience, and which remains implicit, embodied, and encapsulated within the person' (ibid., 121). The sense of what counts as a successful procedure is implicit - it is a blend of the individual's experience and the narrative culture in the laboratory. Knorr Cetina calls the epistemological culture within a molecular biology laboratory, and the form of reasoning practitioners use, 'biographical', because this form of reasoning is sustained by the entire body of knowledge and experience which the researcher bulds up over a professional lifetime. 'It is a knowledge which draws upon scientists' bodies rather than their minds' (ibid., 121). Harry Collins has also demonstrated that tacit knowledge and 'body knowledge' are an essential part of the transmission of scientific knowledge and skills: on the transmission of skills, see Collins (1974), op. cit. note 77. For my purposes, the important part of Knorr Cetina's argument is that there exist different epistemic cultures between sciences, and that each science is characterized by a particular culture. This distinctive culture affects how numbers are assigned to properties of the world in practice. Thus making the numbers themselves a social product - attached to the material world through the materiality of DNA samples, but influenced by the epistemic culture within which they were produced. In molecular biology, this culture is one which leads to heavy reliance on the personal experience and subjectivity of the individual practitioner. This is not an 'error' or source of bias within the science, but a defining quality of the practice of molecular biology, and a constitutive feature of the production of objective knowledge in the field. For a more complete discussion, see Karin Knorr Cetina, Epistemic Cultures (Cambridge, MA: Harvard University Press, 1999).
104. The material in this paragraph is all drawn from Thompson & Ford, op. cit. note 67; the three quotations are at (respectively) 105-06, 110 and 138.
105. I asked Bruce Budowle, chief Research Scientist at the FBI's Forensic Science Research and Training Center at Quantico, VA, whether he thought science was done differently at the FBI compared to a corporate lab or an academic context. He replied: No, I don't think the science is different but I think the mission is different. Obviously industry has to produce a product that's going to be marketable. Academia can look at things and it doesn't have to solve anything per se, it can be just for the sake of pure science. Our particular area is application-oriented research to develop things that can be used to resolve crimes. Each one of those wants to produce things, but the driving force in what gets produced is determined by their motives and their needs. b
106. Thompson, op. cit. note 71, 264.
107. Budowle et al., op. cit. note 63.
108. For details of this work, see Baird et al., op. cit. note 58, 493. The key claim is that 'The 2.41-kbp, 2.51-kbp, 2.75-kbp, and 2.92-kbp fragments are found five times more frequently in Blacks than in Caucasians' (ibid., 493).
109. The term 'DNA fingerprinting' was the name given to the technique of identification by multi-locus DNA analysis by its inventor, Dr Alec Jeffreys (of the Lister Institute, Leicester University, UK): Alec J. Jeffreys, Victoria Wilson and Swee Lay Thein, 'Hypervariable "Minisatellite" Regions in Human DNA', Nature, Vol. 314 (7 March 1985), 67-73; Jeffreys, Wilson and Thein, 'Individual-Specific "Fingerprints" of Human DNA', ibid., Vol. 316 (4 July 1985), 76-79; and Peter Gill, Alec J. Jeffreys and David J. Werrett, 'Forensic Applications of "DNA Fingerprints"', ibid., Vol. 318 (12 December 1985), 577-79. In a linguistic stroke of genius, by naming his procedure a 'fingerprint', Jeffreys piggybacked the credibility of his new technique of identification on to the credibility of an established procedure for identification, the dermal fingerprint. In fact, DNA 'fingerprints' do not uniquely identify individuals, as it is possible that other individuals share the same pattern of alleles at the same loci. The technique is only able to establish a probable match between two sources of DNA. It is widely believed that prints from the tips of fingers are unique, with no two individuals, even identical twins, sharing identical prints, thus making them ideal for establishing identity. However, Simon Cole argues that the uniqueness of dermal fingerprints was never proven, simply asserted by all members of the community of 'fingerprint examiners', thus over time establishing an unproven assertion as 'fact'. Fingerprint examiners claimed that they had never seen any two fingerprints alike. Cole asserts that this premise went unchallenged, and through repetition it eventually became conviction; he argues that unanimity within the profession of fingerprint examiners is what established both their credibility, and the credibility of the technique. See Simon A. Cole, 'Witnessing Identification: Latent Fingerprinting Evidence and Expert Knowledge', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 687-712.
110. The term 'DNA fingerprinting' was the name given to the technique of identification by multi-locus DNA analysis by its inventor, Dr Alec Jeffreys (of the Lister Institute, Leicester University, UK): Alec J. Jeffreys, Victoria Wilson and Swee Lay Thein, 'Hypervariable "Minisatellite" Regions in Human DNA', Nature, Vol. 314 (7 March 1985), 67-73; Jeffreys, Wilson and Thein, 'Individual-Specific "Fingerprints" of Human DNA', ibid., Vol. 316 (4 July 1985), 76-79; and Peter Gill, Alec J. Jeffreys and David J. Werrett, 'Forensic Applications of "DNA Fingerprints"', ibid., Vol. 318 (12 December 1985), 577-79. In a linguistic stroke of genius, by naming his procedure a 'fingerprint', Jeffreys piggybacked the credibility of his new technique of identification on to the credibility of an established procedure for identification, the dermal fingerprint. In fact, DNA 'fingerprints' do not uniquely identify individuals, as it is possible that other individuals share the same pattern of alleles at the same loci. The technique is only able to establish a probable match between two sources of DNA. It is widely believed that prints from the tips of fingers are unique, with no two individuals, even identical twins, sharing identical prints, thus making them ideal for establishing identity. However, Simon Cole argues that the uniqueness of dermal fingerprints was never proven, simply asserted by all members of the community of 'fingerprint examiners', thus over time establishing an unproven assertion as 'fact'. Fingerprint examiners claimed that they had never seen any two fingerprints alike. Cole asserts that this premise went unchallenged, and through repetition it eventually became conviction; he argues that unanimity within the profession of fingerprint examiners is what established both their credibility, and the credibility of the technique. See Simon A. Cole, 'Witnessing Identification: Latent Fingerprinting Evidence and Expert Knowledge', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 687-712.
111. The term 'DNA fingerprinting' was the name given to the technique of identification by multi-locus DNA analysis by its inventor, Dr Alec Jeffreys (of the Lister Institute, Leicester University, UK): Alec J. Jeffreys, Victoria Wilson and Swee Lay Thein, 'Hypervariable "Minisatellite" Regions in Human DNA', Nature, Vol. 314 (7 March 1985), 67-73; Jeffreys, Wilson and Thein, 'Individual-Specific "Fingerprints" of Human DNA', ibid., Vol. 316 (4 July 1985), 76-79; and Peter Gill, Alec J. Jeffreys and David J. Werrett, 'Forensic Applications of "DNA Fingerprints"', ibid., Vol. 318 (12 December 1985), 577-79. In a linguistic stroke of genius, by naming his procedure a 'fingerprint', Jeffreys piggybacked the credibility of his new technique of identification on to the credibility of an established procedure for identification, the dermal fingerprint. In fact, DNA 'fingerprints' do not uniquely identify individuals, as it is possible that other individuals share the same pattern of alleles at the same loci. The technique is only able to establish a probable match between two sources of DNA. It is widely believed that prints from the tips of fingers are unique, with no two individuals, even identical twins, sharing identical prints, thus making them ideal for establishing identity. However, Simon Cole argues that the uniqueness of dermal fingerprints was never proven, simply asserted by all members of the community of 'fingerprint examiners', thus over time establishing an unproven assertion as 'fact'. Fingerprint examiners claimed that they had never seen any two fingerprints alike. Cole asserts that this premise went unchallenged, and through repetition it eventually became conviction; he argues that unanimity within the profession of fingerprint examiners is what established both their credibility, and the credibility of the technique. See Simon A. Cole, 'Witnessing Identification: Latent Fingerprinting Evidence and Expert Knowledge', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 687-712.
112. The term 'DNA fingerprinting' was the name given to the technique of identification by multi-locus DNA analysis by its inventor, Dr Alec Jeffreys (of the Lister Institute, Leicester University, UK): Alec J. Jeffreys, Victoria Wilson and Swee Lay Thein, 'Hypervariable "Minisatellite" Regions in Human DNA', Nature, Vol. 314 (7 March 1985), 67-73; Jeffreys, Wilson and Thein, 'Individual-Specific "Fingerprints" of Human DNA', ibid., Vol. 316 (4 July 1985), 76-79; and Peter Gill, Alec J. Jeffreys and David J. Werrett, 'Forensic Applications of "DNA Fingerprints"', ibid., Vol. 318 (12 December 1985), 577-79. In a linguistic stroke of genius, by naming his procedure a 'fingerprint', Jeffreys piggybacked the credibility of his new technique of identification on to the credibility of an established procedure for identification, the dermal fingerprint. In fact, DNA 'fingerprints' do not uniquely identify individuals, as it is possible that other individuals share the same pattern of alleles at the same loci. The technique is only able to establish a probable match between two sources of DNA. It is widely believed that prints from the tips of fingers are unique, with no two individuals, even identical twins, sharing identical prints, thus making them ideal for establishing identity. However, Simon Cole argues that the uniqueness of dermal fingerprints was never proven, simply asserted by all members of the community of 'fingerprint examiners', thus over time establishing an unproven assertion as 'fact'. Fingerprint examiners claimed that they had never seen any two fingerprints alike. Cole asserts that this premise went unchallenged, and through repetition it eventually became conviction; he argues that unanimity within the profession of fingerprint examiners is what established both their credibility, and the credibility of the technique. See Simon A. Cole, 'Witnessing Identification: Latent Fingerprinting Evidence and Expert Knowledge', Social Studies of Science, Vol. 28, Nos 5-6 (October-December 1998), 687-712.
113. Timothy Wilson Spencer v. Commonwealth of Virginia, 384 S.E.2d 785 (VA 1989), 791. '[T]he three probes used in this case were pAC 255/locus D2S44, pAC 256/locus D17S79, and pAC 225/locus DXYS14. Probe pAC 255/locus D2S44 was tested on a data base comprised of 274 unrelated American blacks. Probe pAC 256/locus D17S79 was tested on 201 unrelated black individuals. The black data base for probe pAC 225/locus DXYS14 consisted of 220 unrelated individuals. These tests showed that the statistical likelihood of each probe's matching DNA fragment appearing in the DNA of unrelated black individuals is 1/657, 1/1292, and 1/159, respectively. The overall likelihood of finding all three fragments in the same sample is determined by multiplying 1/657 × 1/1292 × 1/159, resulting in an overall likelihood of 1/135,000,000' (ibid., 782, note 8).
114. Leslie Roberts, 'Fight Erupts Over DNA Fingerprinting', Science, Vol. 254 (20 December 1991), 1721-23, at 1721: '[P]roponents of "DNA fingerprinting" have claimed that the probability of two DNA samples matching by chance is minuscule -citing figures like 1:500,000 or 1:738,000,000,000,000'.
115. Jean L. Marx, 'DNA Fingerprinting Takes the Witness Stand', Science, Vol. 240 (17 June 1988), 1616-19, at 1617.
116. Colin Norman, 'Maine Case Deals Blow to DNA Fingerprinting', Science, Vol. 246 (22 December 1989), 1556-58. For arguments that the calculation of very small probabilities were calculated in a 'conservative' (that is, pro-defendant) manner, see Bruce Budowle, Ranajit Chakraborty, Allan M. Giusti, Arthur J. Eisenberg and Robert C. Allen, 'Analysis of the VNTR Locus DIS80 by the PCR Followed by High-Resolution PAGE', American Journal of Human Genetics, Vol. 48 (1991), 137-44; Bernard Devlin, Neil Risch and Kathryn Roeder, 'Estimation of Allele Frequencies for VNTR Loci', ibid., 662-76; Budowle et al., op. cit. note 63.
117. Colin Norman, 'Maine Case Deals Blow to DNA Fingerprinting', Science, Vol. 246 (22 December 1989), 1556-58. For arguments that the calculation of very small probabilities were calculated in a 'conservative' (that is, pro-defendant) manner, see Bruce Budowle, Ranajit Chakraborty, Allan M. Giusti, Arthur J. Eisenberg and Robert C. Allen, 'Analysis of the VNTR Locus DIS80 by the PCR Followed by High-Resolution PAGE', American Journal of Human Genetics, Vol. 48 (1991), 137-44; Bernard Devlin, Neil Risch and Kathryn Roeder, 'Estimation of Allele Frequencies for VNTR Loci', ibid., 662-76; Budowle et al., op. cit. note 63.
118. Colin Norman, 'Maine Case Deals Blow to DNA Fingerprinting', Science, Vol. 246 (22 December 1989), 1556-58. For arguments that the calculation of very small probabilities were calculated in a 'conservative' (that is, pro-defendant) manner, see Bruce Budowle, Ranajit Chakraborty, Allan M. Giusti, Arthur J. Eisenberg and Robert C. Allen, 'Analysis of the VNTR Locus DIS80 by the PCR Followed by High-Resolution PAGE', American Journal of Human Genetics, Vol. 48 (1991), 137-44; Bernard Devlin, Neil Risch and Kathryn Roeder, 'Estimation of Allele Frequencies for VNTR Loci', ibid., 662-76; Budowle et al., op. cit. note 63.
119. Colin Norman, 'Maine Case Deals Blow to DNA Fingerprinting', Science, Vol. 246 (22 December 1989), 1556-58. For arguments that the calculation of very small probabilities were calculated in a 'conservative' (that is, pro-defendant) manner, see Bruce Budowle, Ranajit Chakraborty, Allan M. Giusti, Arthur J. Eisenberg and Robert C. Allen, 'Analysis of the VNTR Locus DIS80 by the PCR Followed by High-Resolution PAGE', American Journal of Human Genetics, Vol. 48 (1991), 137-44; Bernard Devlin, Neil Risch and Kathryn Roeder, 'Estimation of Allele Frequencies for VNTR Loci', ibid., 662-76; Budowle et al., op. cit. note 63.
120. The adversarial process in general, and the Yee case specifically, had a direct effect on the academic publications of those who had been appearing as expert witnesses. As a result of meeting their peers in the adversarial setting of the courtroom and debating scientific issues, the scientists who were called as expert witnesses for both the prosecution and the defense began to publish academic articles which reflected their opinion of the technique. The first of these articles was by Eric Lander, who set his views to print in the front-line journal Nature in 1989 (op. cit. note 51). Lander pointed out the paucity of scientific research on both the theoretical and technical issues grounding claims made in DNA typing, and argued that the correct place to work out the controversies over techniques was in the academic literature, not in the courtroom. He argued that there were severe problems with the interpretation of DNA-typing results, due to a lack of consensus and standards in the scientific community. For Lander, what had happened was simple. The perceived power of DNA typing to uniquely identify individuals was so great that it had been rushed into court by private laboratories before the scientific research necessary to ground the technique had been conducted, and before the necessary standards had been set. Lander said that the courts were premature in ruling that DNA typing was 'generally accepted in the scientific community', stating that 'the scientific community has not yet agreed on the standards that ensure the reliability of the evidence' (ibid., 501). As well, responding to challenges by defense attorneys and judges in some cases that the assumptions grounding their techniques had neither been published nor peer reviewed, the FBI began to publish numerous articles in academic, peer-reviewed journals. Speaking to the issue of the effects of population substructure on the probability calculations, in State v. Jakobetz, both Dr Kidd (Yale) and Dr Budowle (FBI) conceded 'that some substructure exists, but [they] contend that the frequency differences for VNTRs between subgroups are insubstantial and thus are more than offset by the FBI's conservative fixed bin procedures': 747 F. Sup250, 1990 US Dist. Lexis 12714, 260. The judge found it 'of some concern that neither Dr Kidd nor Dr Budowle cite any published studies to support these conclusions', but found their testimony credible and convincing anyway. See also United States v. Yee et al., op. cit. note 7.
121. United States v. Steven Wayne Yee et al., 134 F.R.D. 161; 1990 US Dist. Lexis 15908. Yee, Verdi and Bonds were members of the Hell's Angels motorcycle gang, accused of gunning down music store manager David Hartlaub as he stopped after work on 27 February 1988 to make a night deposit. The motive was not robbery - the money was left on the seat of the van. Apparently, the murder was a case of mistaken identity. Mr Hartlaub drove a yellow van identical to that of a local member of the Outlaws, a rival motorcycle gang. Yee, Verdi and Bonds had apparently planned to 'hit' this member to retaliate for the shooting of a Hell's Angels member the previous year in Joliet, IL. The police found the van used in the murder abandoned behind a nearby hotel. There was a MAC-11 9-mm semi-automatic pistol in the van which was spattered with blood. The carpet in the van was also splattered with blood. Blood tests showed that the blood did not belong to Mr Hartlaub, but matched that of John Ray Bonds, who had apparently been wounded by a ricochet from the gun.
122. Richard Lewontin and Daniel Hard, 'Population Genetics in Forensic DNA Typing', Science, Vol. 254 (20 December 1991), 1745-50.
123. However, as noted by Lynch (in 'Discursive Production of Uncertainty', op. cit. note 3), over the course of the DNA Wars, defense lawyers became experts at destabilizing the prosecution's claims by showing the contingencies involved in the production of knowledge claims, which include allele frequency distributions.
124. It is important to note that doubts about DNA technology did not appear by accident, they were created by a team of dedicated defense lawyers determined to find a way to make DNA evidence at least a little wobbly. As noted above, in March 1990, the National Association of Criminal Defense Lawyers (NACDL) held its First National DNA Symposium and created a 'DNA Task Force' to aid defense lawyers in attacking the technique. Barry Scheck and Peter Neufeld were its chairs. These challenges to DNA evidence in the courtroom were organized and spearheaded by a highly dedicated, select group of people. They sought out, created and nurtured a community of scientists who were opposed to some aspects of DNA typing. In their written accounts, they use the term 'engineered' to describe their efforts to attack DNA evidence in the courtroom. Scheck and Neufeld (op. cit. note 4, 17) reported that the cost and the complexity of this type of litigation were enormous, and the brunt of the burden was being borne by public defenders and assigned counsel. In November 1990, the NACDL held its Second National DNA Defense Symposium in Los Angeles. The DNA Task Force was organized to the level of having regional contact people. At the Los Angeles Symposium, they brought in experts to teach the defense counsel how to challenge DNA evidence, including the 'Irvine Mafia' -William Thompson, Simon Ford and Laurence Mueller - so-called because they are all faculty members at UC Irvine.
125. In addition to two FBI employees, Dwight Adams (examiner) and Bruce Budowle (who was principally responsible for developing the FBI's DNA protocols and describes himself as the FBI's 'pointman' for DNA technology), the government called Dr Patrick Conneally (a Distinguished Professor of Medicine at the Indiana University School of Medicine), Dr Stephen Daiger (a Professor at the Graduate School of Biomedical Sciences at the University of Texas Health Science Center) and Dr C. Thomas Caskey (holder of the Henry and Emma Meyer Chair in Molecular Genetics at the Baylor College of Medicine). Dr Caskey had been the Chair of the Advisory Panel to the Office of Technology Assessment's 1990 Report on genetic testing, of which Dr Eric Lander, the court's witness, was also a member. The government also called Dr Kenneth Kidd (Professor of Human Genetics, Psychiatry and Biology, Yale University School of Medicine). Defense lawyers Sheck and Neufeld called Dr Peter D'Eustachio (Associate Professor, Department of Biochemistry, New York University Medical Center); Dr Paul J. Hagerman (Associate Professor of Biochemistry, Biophysics and Genetics, University of Colorado Health Sciences Center); Dr Richard C. Lewontin (Alexander Agassiz Professor of Zoology and Professor of Population Sciences, Harvard University); Dr T. Conrad Gilliam (Assistant Professor of Neurogenetics, Department of Genetics and Development, College of Physicians and Surgeons, Columbia University); and Dr Daniel L. Hartl (who was then James S. McDonnell Professor and Head, Department of Genetics, Washington University School of Medicine).
126. United States v. Bonds et al., op. cit. note 7, 174.
127. United States v. Yee et al., op. cit. note 91, 550.
128. The judge's report gives no reason for the re-calculation of the probability, simply stating that: On 7 April 1989, the FBI's DNA laboratory submitted a report stating that there was a 'match' of DNA profiles from a bloodstain found in victim David Hartlaub's car and the DNA profiles derived from the blood of defendant Bonds. The FBI then calculated a probability of 1 in 270,000 that an unrelated individual selected randomly from the Caucasian population would have a DNA profile matching that of Bonds. In May of 1990, the FBI revised its probability figure to 1 in 35,000. This revised probability estimate was presented to the jury in this case. United States of America v. John Ray Bonds, Mark Verdi and Steven Wayne Yee, 18 F.3d 1327; 1994 US App. Lexis 4594; 1994 FED App. 0085P (6th Cir.), quote at 16.
129. See transcripts on file at courthouse in Toledo, OH, for United States v. Bonds et al., op. cit. note 7.
130. Lander was called as the court's witness in the six-week Frye hearing held in the first case brought against Bonds, Verdi and Yee: see United States v. Yee, 129 F.R.D. 629, 631 (N.D. Ohio 1990). The judge's report of the 1993 appeal (United States v. Bond et al., op. cit. note 7, 17) reads: The magistrate judge [in the 1990 original case] then conducted a six-week Frye hearing to determine whether the proposed experts' trial testimony about the DNA evidence was based on principles generally accepted in the scientific community. During the hearing, the Government called six expert witnesses, the defendants called five expert witnesses, and the court called Dr Eric Lander as the court's witness. More than 200 exhibits were introduced relating to the FBI's methods of testing. Sheila Jasanoff (op. cit. note 16, 732) notes that in both the USA and Britain, the practice of appointing an impartial witness to act on the court's behalf is becoming more and more common. She believes that: As a 'neutral' third eye in the traditional two-party format of litigation, the court-appointed expert might occupy a quasi-judicial position and would form, together with the party experts, a smaller 'facts' triad (a de facto Science Court) within the larger 'justice' triad comprised by the judge and the legal advocates. Jasanoff argues that because these witnesses come to court bearing an imprimatur of impartiality, their credibility is heightened.
131. On the issue of population substructuring, no random samples were ever drawn to ascertain the nature of substructuring among the American population: United States v. Yee et al., op. cit. note 91, 183.
132. The bracketed numbers in the following section refer to page numbers in the trial record: United States v. Bonds et al., op. cit. note 7. 'Substructure' means that a population is made up of several separate, distinct populations which within themselves may have different allele frequencies than when they are all lumped together into one large population. On the issue of random sampling and the representativeness of samples, the second NRC committee members agreed among themselves that 'samples of convenience' were for all intents and purposes the same as 'random samples', and the issues of representativeness of samples fell by the wayside: US National Research Council, The Evaluation of Forensic DNA Evidence (Washington, DC: National Academy Press, 1996), 126-62. This shows further that closure in the DNA Wars was brought about by social negotiation, not by scientific research (for a more detailed explanation of this aspect of the controversy, see Derksen, op. cit. note 17).
133. The bracketed numbers in the following section refer to page numbers in the trial record: United States v. Bonds et al., op. cit. note 7. 'Substructure' means that a population is made up of several separate, distinct populations which within themselves may have different allele frequencies than when they are all lumped together into one large population. On the issue of random sampling and the representativeness of samples, the second NRC committee members agreed among themselves that 'samples of convenience' were for all intents and purposes the same as 'random samples', and the issues of representativeness of samples fell by the wayside: US National Research Council, The Evaluation of Forensic DNA Evidence (Washington, DC: National Academy Press, 1996), 126-62. This shows further that closure in the DNA Wars was brought about by social negotiation, not by scientific research (for a more detailed explanation of this aspect of the controversy, see Derksen, op. cit. note 17).
134. United States v. Yee et al., op. cit. note 91, 182. The problem is not as simple as Lander makes it sound. When I interviewed Dr Daniel Hartl, and asked him whether this problem would have come up for the population genetics community without the demands for the knowledge arising from the justice system. He responded: No. That's why it was a challenge to population genetics. Nobody knew how to think about this kind of problem, and when in the earliest stages population geneticists made statements that would lead you to think this was a trivial problem, that was not correct. It's a very difficult problem, how you do this, and even the second Crow committee set up a number of scenarios that they didn't know how to do the calculations, still didn't know how to do the calculation. So this is not a trivial problem in population genetics. c
135. United States v. Bonds et al., op. cit. note 7, at 184.
136. This was eventually published as Lewontin & Hartl, op. cit. note 92.
137. For an independent account of this exchange, see the inset in Roberts, op. cit. note 87, 1722.
138. For another account, see Roberts, op. cit. note 87, 1722: 'Kidd and Caskey felt so strongly that they cornered one of Science's editors . . . urging her to pass on their concerns to Koshland'.
139. Chris Anderson, 'DNA Fingerprinting Discord', Nature, Vol. 354 (19/26 December 1991), 500; see also Roberts, op. cit. note 87, 1722.
140. Chris Anderson, 'DNA Fingerprinting Discord', Nature, Vol. 354 (19/26 December 1991), 500; see also Roberts, op. cit. note 87, 1722.
141. The article was published as Ranajit Chakraborty and Kenneth Kidd, 'The Utility of DNA Typing in Forensic Work', Science, Vol. 254 (20 December 1991), 1735-39.
142. Roberts, op. cit. note 87, 1721.
143. NRC, op. cit. note 9.
144. For examples of criticism, see Joel Cohen, 'The Ceiling Principle is not Always Conservative in Assigning Genotype Frequencies for Forensic DNA Testing' (Letter), American Journal of Human Genetics, Vol. 51 (November 1992), 1165-68. See also Bruce S. Weir, 'Forensic Population Genetics and the National Research Council (NRC)' (Letter), ibid., Vol. 52 (February 1993), 437-40; Bernard Devlin, Neil Risch and Kathryn Roeder, 'Comments on the Statistical Aspects of the NRC's Report on DNA Typing', Journal of Forensic Sciences, Vol. 39, No. 1 (January 1994), 28-40.
145. For examples of criticism, see Joel Cohen, 'The Ceiling Principle is not Always Conservative in Assigning Genotype Frequencies for Forensic DNA Testing' (Letter), American Journal of Human Genetics, Vol. 51 (November 1992), 1165-68. See also Bruce S. Weir, 'Forensic Population Genetics and the National Research Council (NRC)' (Letter), ibid., Vol. 52 (February 1993), 437-40; Bernard Devlin, Neil Risch and Kathryn Roeder, 'Comments on the Statistical Aspects of the NRC's Report on DNA Typing', Journal of Forensic Sciences, Vol. 39, No. 1 (January 1994), 28-40.
146. For examples of criticism, see Joel Cohen, 'The Ceiling Principle is not Always Conservative in Assigning Genotype Frequencies for Forensic DNA Testing' (Letter), American Journal of Human Genetics, Vol. 51 (November 1992), 1165-68. See also Bruce S. Weir, 'Forensic Population Genetics and the National Research Council (NRC)' (Letter), ibid., Vol. 52 (February 1993), 437-40; Bernard Devlin, Neil Risch and Kathryn Roeder, 'Comments on the Statistical Aspects of the NRC's Report on DNA Typing', Journal of Forensic Sciences, Vol. 39, No. 1 (January 1994), 28-40.
147. The 'ceiling principle' was a 'conservative' procedure for calculating random match probabilities proposed by the first NRC Committee to meet the needs of the courts. After interviewing several members of that committee, it is not clear whose idea the ceiling principle was. However, the principle sets a lower bound on the frequency of alleles of 5% or 10%. This means that probabilities calculated on distributions with rare alleles with much lower frequencies would have the frequency, for the purpose of the calculation, raised to 5%, and the random match probability calculated from there. This procedure is conservative in that it acts in favour of the defendant, because it increases the estimate of the probability of a random match, thus making it seem more likely that someone other than the defendant could match the crime scene evidence. See NRC, op. cit. note 9, esp. 82-85.
148. NRC, op. cit. note 102, v-vi.
149. Eric S. Lander and Bruce Budowle, 'DNA Fingerprinting Dispute Laid to Rest', Nature, Vol. 371 (27 October 1994), 735-38.
150. Kaye also told me that he is 'persuaded that the publication of the [1st NRC] report accelerated a trend to say that there was lack of agreement in the scientific community about what was known as the product-rule for estimating the probability of a random match. So it reinforced doubts. I think it underlined the fact that there was a controversy'. d
151. Richard C. Lewontin, 'Forensic DNATyping Dispute', Nature, Vol. 372 (1 December 1994), 398; Daniel Hartl, 'Forensic DNATyping Dispute', ibid., 398-99.
152. Richard C. Lewontin, 'Forensic DNATyping Dispute', Nature, Vol. 372 (1 December 1994), 398; Daniel Hartl, 'Forensic DNATyping Dispute', ibid., 398-99.
153. For example: 'We share the view of many experts who have criticized it [the ceiling principle] on practical and statistical grounds and who see no scientific justification for its use' (NRC, op. cit. note 102, at 157).
154. How a committee composed of highly esteemed scientists hand-picked by the National Academy of Sciences could come up with a conclusion that had 'no scientific grounding whatsoever' is an interesting issue: for further exploration of it, see Derksen, op. cit. note 17.
155. Roberts, op. cit. note 87, 1721.
156. Chakraborty & Kidd, op. cit. note 109, at 1735; Bernard Devlin, Neil Risch and Kathryn Roeder, '"Response" to "Forensic DNA Tests and Hardy-Weinberg Equilibrium"', Science, Vol. 253 (30 August 1991), 1039-41, at 1039.
157. Chakraborty & Kidd, op. cit. note 109, at 1735; Bernard Devlin, Neil Risch and Kathryn Roeder, '"Response" to "Forensic DNA Tests and Hardy-Weinberg Equilibrium"', Science, Vol. 253 (30 August 1991), 1039-41, at 1039.
158. Lander, op. cit. note 51; Philip Green, 'Population Genetic Issues in DNA Fingerprinting', American Journal of Human Genetics, Vol. 50 (1992), 440-41; Lewontin & Hartl, op. cit. note 92.
159. Lander, op. cit. note 51; Philip Green, 'Population Genetic Issues in DNA Fingerprinting', American Journal of Human Genetics, Vol. 50 (1992), 440-41; Lewontin & Hartl, op. cit. note 92.
160. Lander, op. cit. note 51; Philip Green, 'Population Genetic Issues in DNA Fingerprinting', American Journal of Human Genetics, Vol. 50 (1992), 440-41; Lewontin & Hartl, op. cit. note 92.
161. See Budowle et al., op. cit. note 63, 841; Devlin, Risch & Roeder, op. cit. note 121; Bernard Devlin, Neil Risch and Kathryn Roeder, 'No Excess of Homozygosity at Loci Used for DNA Fingerprinting', Science, Vol. 249 (21 September 1990), 1416-20.
162. See Budowle et al., op. cit. note 63, 841; Devlin, Risch & Roeder, op. cit. note 121; Bernard Devlin, Neil Risch and Kathryn Roeder, 'No Excess of Homozygosity at Loci Used for DNA Fingerprinting', Science, Vol. 249 (21 September 1990), 1416-20.
163. See Budowle et al., op. cit. note 63, 841; Devlin, Risch & Roeder, op. cit. note 121; Bernard Devlin, Neil Risch and Kathryn Roeder, 'No Excess of Homozygosity at Loci Used for DNA Fingerprinting', Science, Vol. 249 (21 September 1990), 1416-20.
164. For a discussion of how laboratory 'techniques' like PCR or DNA fingerprinting can be seen as social, see Jordan & Lynch (1998), op. cit. note 3, 778. Here, Jordan and Lynch argue that techniques like PCR are social in at least two senses. First: [Techniques] are artefacts 'in' one or another social context. From this point of view, such artefacts are like raisins in a pudding, with an integrity separate from the organizational context in which they are situated . . . the relevant way to reveal the culture of an artefact is to specify how the context determines, shapes, affects, circumscribes, sets conditions for the use of, provides resistance to or sustains the technical object in question. The material design of the artefact may vary in certain respects, but such variations can be indexed to group interests, organizational tasks and objectives, and financial resources. Techniques are also social in a second sense, in that their existence and production are not guaranteed by instructions or definitions, but 'are accomplished through laborious struggles and negotiations'. See also the other work by Jordan and Lynch cited in note 3.
165. For a discussion of how laboratory 'techniques' like PCR or DNA fingerprinting can be seen as social, see Jordan & Lynch (1998), op. cit. note 3, 778. Here, Jordan and Lynch argue that techniques like PCR are social in at least two senses. First: [Techniques] are artefacts 'in' one or another social context. From this point of view, such artefacts are like raisins in a pudding, with an integrity separate from the organizational context in which they are situated . . . the relevant way to reveal the culture of an artefact is to specify how the context determines, shapes, affects, circumscribes, sets conditions for the use of, provides resistance to or sustains the technical object in question. The material design of the artefact may vary in certain respects, but such variations can be indexed to group interests, organizational tasks and objectives, and financial resources. Techniques are also social in a second sense, in that their existence and production are not guaranteed by instructions or definitions, but 'are accomplished through laborious struggles and negotiations'. See also the other work by Jordan and Lynch cited in note 3.
166. All three quotations from Roberts, op. cit. note 87, 1723.
167. See Knorr Cetina (1999), op. cit. note 79, 63. For the original article on liminal spaces, see Victor Turner, 'Betwixt and Between: The Liminal Period in "Rites de Passage"', in his The Forest People of Symbols (Ithaca, NY & London: Cornell University Press, 1976), 59-92.
168. See Knorr Cetina (1999), op. cit. note 79, 63. For the original article on liminal spaces, see Victor Turner, 'Betwixt and Between: The Liminal Period in "Rites de Passage"', in his The Forest People of Symbols (Ithaca, NY & London: Cornell University Press, 1976), 59-92.
169. Kathleen Jordan, Sociological Investigations into the Mainstreaming of the Polymerase Chain Reaction (unpublished PhD dissertation, Department of Sociology, Boston University, 1997).
170. See Ted Porter, Trust in Numbers - The Pursuit of Objectivity in Science and Public Life (Princeton, NJ: Princeton University Press, 1995), 224-31.
171. 'Disciplinary objectivity' is that sense of objectivity which does not assume a 'wholesale convergence and instead takes consensus among the members of particular research communities as its standard of objectivity': Allan Megill, 'Introduction: Four Senses of Objectivity', Annals of Scholarship, Vol. 8, Nos 3-4 (1991), 301-21, at 301.
172. For more on social worlds theory, see Adele E. Clarke, 'Social Worlds/Arenas Theory as Organizational Theory', in David Maines (ed.), Social Organization and Social Process: Essays in Honor ofAnselm L. Strauss (Hawthorne, NY: Aldine de Gruyter, 1991), 119-58; and Clarke, 'A Social Worlds Research Adventure', in Susan E. Cozzens and Thomas F. Gieryn (eds), Theories of Science in Society (Bloomington: Indiana University Press, 1990), 15-42.
173. For more on social worlds theory, see Adele E. Clarke, 'Social Worlds/Arenas Theory as Organizational Theory', in David Maines (ed.), Social Organization and Social Process: Essays in Honor ofAnselm L. Strauss (Hawthorne, NY: Aldine de Gruyter, 1991), 119-58; and Clarke, 'A Social Worlds Research Adventure', in Susan E. Cozzens and Thomas F. Gieryn (eds), Theories of Science in Society (Bloomington: Indiana University Press, 1990), 15-42.
174. This can be seen as the 'haecceity' of objectivity (where 'haecceity' means 'just thisness'). In the end, objectivity rests upon the accountability of measurement practices and standards. For discussion of haecceity, see Harold Garfinkel and D. Lawrence Weider, 'Two Incommensurable, Asymmetrically Alternate Technologies of Social Analysis', in Graham Watson and Robert M. Seiler (eds), Text in Context: Contributions to Ethnomethodology (Newbury Park, CA: Sage Publications, 1992), 175-206, esp. 203, footnote 2.
175. See Joan Fujimura, Crafting Science: A Sociohistory of the Quest for the Genetics of Cancer (Cambridge, MA & London: Harvard University Press, 1996), 80-82, for a discussion of how restriction enzymes became easily accessible 'tools' to be used in molecular biological work.