Who Speaks For Science

Law, Probability and Risk (2010) 9, 2546
Advance Access publication on December 16, 2009
doi:10.1093/lpr/mgp032
Who speaks for science? A response to the National Academy of

Sciences Report on forensic science
S IMON A. C OLE
Associate Professor of Criminology, Law & Society, University of California, Irvine, CA
92697-7080, USA
[Received on 15 May 2009; revised on 15 September 2009;
accepted on 16 September 2009]
Keywords: National Academy of Science; forensic science; fingerprint; expert witnesses; general
acceptance.
1. Introduction
Of the many topics addressed by the recent National Academy of Science (NAS) Report on forensic
science, this response will focus on the forensic discipline I know best, a discipline, that enjoys great
prominence in the report: latent print analysis. In Section 2, I will place the NAS Report, as it pertains
to latent print identification, in historical context. I will argue that, by the time of the publication of
the NAS Report, a robust controversy was underway that had become starkly polarized between
scholars and scientists on the one hand and practitioners and judges on the other. In Section III, I
argue that the significance of the NAS Report lies in its status as an intervention by a mainstream
scientific institution in this ongoing controversy. As such, I suggest that the Report raises the question
of Who Speaks for Science?i.e. whether it will ultimately be judges or scientists who have the
last word on the issues raised by latent print analysis. I suggest that the resolution of this issue
implicates the broader issue of to what extent scientific institutions like the NAS have jurisdiction
over scientific matters that arise in legal settings. In Section IV, I assume that the NAS Report will
Email: scole@uci.edu
c The Author [2009]. Published by Oxford University Press. All rights reserved.

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016
This response focuses on the treatment of latent print identification by the recent National Academy
of Science (NAS) Report on forensic science. It begins by situating the Report in the historical context
of a decade of controversy over the validity of latent print identification. Stark disagreement between
the academic and judicial communities over this issue created a situation in which the question of
which of these two communities would speak for science became contested. The Reports support
of the academic position demonstrated the lack of support among non-practitioners for the claims
of extreme discrimination and accuracy advanced on behalf of latent prints. The Report in some
sense constitutes the response of institutionalized science to this issue. Nonetheless, it is still unclear
whether the Report will function, as some may have hoped, as a court of last resort on this issue
or whether the courts themselves will again arbitrate it. The response then turns to the issue of how
latent print conclusions can be reported in the wake of the NAS Report. The Report expresses clear
disapproval of the reporting framework currently mandated by latent print professional organizations,
creating a tension around the reporting of analyses. The response concludes that semantic resolutions
to this tension are undesirable compared to resolutions based on empirical data.
26
S. A. COLE
be treated as authoritative, and I ask a different version of the question Who Speaks for Science?
How, I ask, will the results of latent print analyses be reported, given that the NAS Report states
that the current mode of reporting is unacceptable? I explore some of the possible ways that latent
print conclusions may be reported, and I conclude that the reporting of conclusions is a crucial, and
underexplored, issue for forensic science.
2. Background
To situate the Report in historical context, insofar as it addressed latent print identification, the Report
marked the first intervention by a credentialed mainstream scientific institution into a debate that had
been going on for at least a decade. Latent print identification was introduced to the United States
during the first decade of the 20th century, and it was first deemed admissible evidence in a reported
decision in 1911 (People v. Jennings, 1911). Only in the last decade of the 20th century, however,
were serious questions raised about the validity and admissibility of latent print analysis. During the
1990s, a number of scholars raised questions about the validity of latent print analysis (e.g. Berger,
1994; Faigman et al., 1997; Woodworth, 1997; Cole, 1998; Saks, 1998; Starrs, 1999; Mnookin, 2001;
Stoney, 2001). In 1999, these questions resulted in a legal challenge to the admissibility of latent
print evidence (United States v. Mitchell, 2004) under the Supreme Courts Daubert and Kumho
Tire standards for expert evidence (Daubert v. Merrell Dow Pharmaceuticals, 1993; Kumho Tire
v. Carmichael, 1999). Kumho Tire mandates that the trial court assess the reliability of proffered
expert evidence. In such proceedings, the burden is clearly on the proponent of the evidence to
provide evidence of reliability (e.g., Faigman et al., 1997; Mearns, 2009). The defense, conceding
the uniqueness of all human friction ridge skin (United States v. Mitchell, 1999a), contended that
there was no evidence of reliability.
Whatever may be said about the accuracy of latent print identification, the arguments mounted
in its defense were surprisingly weak, indeed irrational. At the Daubert hearing in Mitchell, the
government contended, among other things, that the reliability of latent print identification was
demonstrated by the uniqueness of friction ridge skin and the longstanding use of the technique
in casework and criminal trials (what would later become known derisively as adversarial testing)
(United States v. Mitchell, 2004, p. 238), that the error rate of the technique could be meaningfully
parsed into methodological and human categories (p. 240) and that the methodological error
rate was zero (p. 227). Mitchell replied that one could not infer the accuracy of the technique from
the uniqueness of its target of analysis or from casework or trials, that there was no meaningful
distinction between the examiner and the method, and that the error rate, though unknown, was
certainly not zero (Epstein, 2002).
Over the ensuing decade, this debate was also joined in the scholarly literature. Most scholars
agreed with Mitchell that validation studies of latent print identification were lacking (e.g. Berger,
1994, p. 1354; Faigman et al., 1997; Haber and Haber, 2003, p. 358; Kennedy, 2003; La Morte,
2003, p. 187; Lawson, 2003, p. 65; Cole, 2004, p. 1205; Schwinghammer, 2005; Siegel et al., 2006,
p. 35; Dwyer, 2007, p. 391; Mnookin, 2008; Saks & Faigman, 2008, p. 152), that one could not
infer the accuracy of the technique from the uniqueness of its target of analysis (e.g. Robertson,
1990, p. 255; Stoney, 1997; Starrs, 1999, p. 243; Champod and Evett, 2001, p. 115; Epstein, 2002,
p. 613; Kaye, 2003b, p. 1080; Lawson, 2003, p. 9; Benedict, 2004, p. 528; Cole, 2004, p. 1199;
Moriarty, 2004, Section 12, p. 24; Zabell, 2005, p. 163; Saks & Faigman, 2008, p. 155) or from
casework or trials (e.g. Saks, 2000; Mnookin, 2001, p. 65; Faigman, 2002; Giannelli, 2002; Haber
WHO SPEAKS FOR SCIENCE?
27
and Haber, 2003, p. 343; Kaye, 2003b, p. 1081; Benedict, 2004, p. 529; Cole, 2004, p. 1210; Zabell,
2005, p. 168; Siegel et al., 2006, p. 44; Koehler, 2008, p. 1085), that latent print examiners could
not support their claims to be able to accurately individualize latent prints (e.g. Stoney, 1997, p.
72; Saks, 2000, p. 881; Mears & Day, 2003, p. 731; Cole, 2006; Mnookin, 2008; Saks & Faigman,
2008, p. 152), that there was no meaningful distinction between the examiner and the method for
calculating error rates (e.g. Mnookin, 2001, p. 60; Kaye, 2003b, p. 1083; Cole, 2005, p. 1039; Zabell,
2005, p. 177; Koehler, 2008, p. 1088; Saks & Faigman, 2008, p. 159) and that the error rate, though
unknown, was certainly not zero (e.g. Starrs, 1999, p. 243; Saks, 2000, p. 885; Mnookin, 2001, p.
59; Epstein, 2002, p. 633; Haber and Haber, 2003; La Morte, 2003, p. 184; Lawson, 2003, p. 43;
Mears & Day, 2003, p. 732; Cole, 2005; Zabell, 2005, p. 178; Siegel et al., 2006, p. 40; Cooley,
2007, p. 390; Saks & Faigman, 2008, p. 159). Some scholars and forensic practitioners, however,
agreed with the government that the reliability of latent print identification was demonstrated by
the uniqueness of friction ridge skin (Moenssens, 2003) and the longstanding use of the technique
in casework and criminal trials (Moenssens, 2002), that the error rate of the technique could be
meaningfully parsed into methodological and human categories and that the methodological error
rate was zero (testimony of Stephen Meagher and Bruce Budowle; United States v. Mitchell, 1999b).
What was the status of this debate prior to the NASs intervention? The questions raised about
latent print identification provoked divergent responses from courts and from scholars (Haber and
Haber, 2008; Mnookin, 2008). Courts deemed latent print evidence admissible with near unanimity (Faigman et al., 2007). In contrast, almost all scholars who addressed the issue concluded that
latent print examiners claims to be able to individualizethat is to determine the source of a latent
print of unknown origin to the exclusion of all other possible sources in the universe (SWGFAST,
2002)were not supported by evidence and, further, that measurements of the accuracy of latent
print analysis were lacking.
In 2005, a group of scientists and scholars (including the author) working with the New England Innocence Project filed an amicus curiae brief in an appeal of latent print admissibility to the
Supreme Judicial Court of Massachusetts (Siegel et al., 2006). This brief, signed by more than 15
scientists and scholars from a wide variety of disciplines, including members of the NAS, stated
that the reliability of latent print individualization had not been demonstrated,. It showed that the
arguments that had previously been made by various scientists and scholars were not those of an
extreme few, but rather the views of almost every non-practitioner scientist or scholar who had examined the question. However, despite having stated in an earlier case that the relevant scientific
community should be defined broadly enough to include a sufficiently broad sample of scientists so
that the possibility of disagreement exists, not so narrowly that the experts opinion will inevitably
be considered generally accepted the court ruled that the trial court had acted within its discretion in limiting the relevant scientific community to latent print examiners only (Commonwealth v.
Patterson, 2005, p. 25, quoting Canavans Case 2000).
One way of characterizing the status of the debate prior to the NASs intervention is to ask
what we might call the Frye question. Although the Report, like most evidence scholarship today,
focuses on the more recent Daubert/Kumho approach to expert evidence, the general acceptance
approach to assessing expert evidence espoused in Frye v. United States (1923) is still adhered to
in many state courts, espoused by some evidence scholars (Schwartz, 1997), and incorporated into
Daubert and Kumho themselves. The Frye approach asks whether a claim made by an expert is
generally accepted in the relevant scientific community. Among the most prominent of the many
criticisms of this approach are that this supposed test only raises further questions, including who
28
S. A. COLE
should constitute the relevant scientific community and what constitutes general acceptance. Frye
has also been criticized as crude head counting (e.g. People v. Leahy, 1994; Brim v. State, 1997).
However, other courts have defended head counting as the appropriate way to measure general
acceptance (e.g. Jones v. United States, 1988; Goeb v. Tharaldson, 2000).
The year before the release of the NAS Report, I undertook the sort of crude head count envisioned by at least some interpretations of the Frye approach. In doing so, I should note that I was
not necessarily endorsing the position that courts should make decisions about the admissibility of
evidence based on crude head counts. Rather, I was using the device of the head count to convey
a general sense of the state of opinion within the scholarly community on this issue. Although a
head count in a particular direction might not necessarily be dispositive on the issue of admissibility, it might be considered a sort of red flag. I surveyed three sources to measure the opinions
of scientist and scholars on the question of whether latent print examiners claims of individualization had been validated: sworn expert testimony, amicus curiae briefs and published literature. The
results showed that the vast majority of scholars had found evidence validating latent print examiners claims to accurate individualization lacking. A full discussion of this survey may be found
elsewhere (Cole, 2008b), but, in sum, a rough head count of non-practitioner scientists and scholars yielded 25 who had expressed the viewthrough sworn testimony, signing an amicus curiae
brief, publishing an article or a combination of those means of expressionthat latent print individualization lacked validation and three who had, through the same mechanisms, stated the contrary.
As noted elsewhere (Cole, 2008b), it possible to quibble with the numbers by eliminating some
individuals for a various reasons (e.g. disqualifying those who have testified as expert witnesses
or disqualifying legal scholars for not being scientists), but none of these alternative interpretations supports any conclusion other than that very few non-practitioner scientists or scholars had
made public statements holding that latent print individualization had been validated. This analyses
showed that, while some of the earliest scholarly arguments had been dismissed as the misinformed
opinions of a handful of critics (Moenssens, 2005) on the margins of a professional discipline
(Moenssens, 2002) with a self-serving agenda (IAI, 2007b), by 2008 the point that latent print
individualization lacked validation had been articulated by a large number of scholars from a wide
variety of disciplines who were harder to dismiss as being motivated by financial interests or having
a self-serving agenda.
This head counting exercise illustrated the starkly different results that can be generated depending upon whether a court treats as the relevant scientific community for a particular forensic technique the practitioners of the technique itself or the broader scientific community (Faigman et al.,
1997). More specifically, it showed that, although latent print practitioners themselves attested to the
reliability of their analyses, hardly any external observers, with the important exception of judges,
had been convinced of this reliability. Contrary to the representation made by the International Association for Identification (IAI) to the NAS Committee, that scientists and legal authorities . . .
generally hold that the error rate for fingerprint identification is extremely small, statistically insignificant, and not due to the methodology but instead to the inherent risk of error in any human
endeavor (IAI, 2007b) (without naming any scientists or legal authorities), it would seem that few,
if any, non-practitioners scientists or scholars hold this view.
However, the view in the courts was the opposite; nearly every court found latent print evidence
admissible, and most of those, in doing so, stated that it was valid. In their efforts to find latent
print identification reliable, courts stretched the limits of rational argument. Consider, e.g. the courts
that wrote opinions stating that the reliability of latent print identification was established by having
29
1 The author of the Havvard opinion recently became the first nominee to a Circuit Court judgeship by the Obama Administration, which has promised to develop a strategy for restoring scientific integrity to government decision-making and
base our public policies on the soundest science. In noting this mild irony, I am certainly not disputing Judge Hamiltons
fitness for the position for which he was nominated because it would not be appropriate to evaluate a judge based on a single decision. Moreover, the more relevant lesson here might well be that even good judges have had difficulty constructing
scientifically defensible arguments in support of latent print identification.
The Reports idiosyncratic reading of the Havvard case (pp. 103104) must also be noted. The Report seems at pains to
absolve both Judge Hamilton and the governments expert for what it appears to view as a poor opinion by the Seventh Circuit.
Instead, the Report blames the Seventh Circuit for allegedly overstat[ing] . . . the experts equivocal testimony, misreading
Judge Hamiltons supposedly more accurate representation of that testimony and giv[ing] fuel to the misconception that the
forensic discipline of fingerprinting is infallible.
We do not know what, precisely, the expert witness testified to in Havvard because neither the Report nor any of the scholars
who have written about the case (including the author) has obtained the transcript. In the absence of a transcript, however,
one can reasonably infer that the testimony given in Havvard was substantially similar to the testimony the same witness gave
in United States v. Mitchell, not long prior, for which a transcript has been made publicly available. In Mitchell, the witness
testified as follows:
Well, when youre dealing with a scientific methodology such as we have for ever
since Ive been trained, there are distinctionstheres two parts of errors that can
occur. One is the methodological error, and the other one is a practitioner error.
If the scientific method is followed, adhered to in your process, that the error in the
analysis and comparative process will be zero. It only becomes the subjective opinion
of the examiner involved at the evaluation phase. And that would become the error
rate of the practitioner (United States v. Mitchell, 1999b, p. 154).
Based on that experts unrebutted testimony, the District Court in Havvard wrote that [The expert] testified that the error
rate for fingerprint comparison is essentially zero. Though conceding that a small margin of error exists because of differences
in individual examiners, he opined that this risk is minimized because print identifications are typically confirmed through
peer review.
Given that the witness stated that the error rate of latent print identification was zero and the District Court credited this by
stating that it was essentially zero, it is puzzling why the Report would describe the Seventh Circuits characterization of the
error rate as essentially zero as an overstate[ment], suggest that the appellate court was responsible for the misconception
that the forensic discipline of fingerprinting is infallible or characterize the witnesss testimony as equivocal.
2 Julia Leighton is General Counsel at the Public Defender Service for the District of Columbia.
been used in court for 100 years (United States v. Havvard, 2000),1 for having withstood the test
of time (United States v. Crisp, 2003) or for having survived a regimen of implicit testing (United
States v. Mitchell, 2004). Consider the courts that declared that the error rate of latent print identification was vanishingly small (United States v. Havvard, 2000, p. 854), essentially zero (United
States v. Havvard, 2001, p. 599), negligible (United States v. Crisp, 2003, p. 269) or microscopic
(United States v. Mitchell, 2004, p. 241, n. 220), assertions made without reference to any empirical data whatsoever or to what actual numerical error rates led to these verbal characterizations.
Consider the courts that approvingly cited and relied upon the notorious, unpublished 50K study
(later criticized in the NAS Report (p. 144)) (United States v. Mitchell, 2004, p. 237; State v. Kim,
2004, p. 7). And yet, at the same time, neither these courts nor any other have ever cited any of the
many publishedsome peer-reviewedarticles critiquing this study (Wayman, 2000; Champod and
Evett, 2001; Stoney, 2001; Pankanti et al., 2002; Kaye, 2003a; Zabell, 2005). Instead, in an extraordinary feat of legal reasoning, one court ruled that latent print identification satisfied the peer review
and publication prong of Daubert and Kumho because the 50K study, though never published itself,
had been severely criticized in peer-reviewed publications (State v. Sullivan, 2005)!
Thus, latent print admissibility rulings do not appear to have been the finest hour of either legal
reasoning or scientific reasoning by judges. Why? Judge Edwards, Co-Chair of the NAS Committee, when asked this question by Julia Leighton2 at the symposium that occasioned this journal issue,
suggested (speaking for himself, not for the Committee) that judges, in perhaps too crude terms,
30
S. A. COLE
THE COURT: Except we do have at least a hundred years of case law where the courts have
found that over the passage of time there is some validity to this process called
fingerprint identification.
had been snookeredthat we just didnt get it (Edwards, 2009, video at 1:02:10). Some commentators have suggested that judges did not apply the same degree of scrutiny to evidence offered
by the government in criminal cases that they applied to evidence offered by plaintiffs in civil cases,
out of concern for allowing accused criminals to go free (Risinger, 2000; Neufeld, 2005; Cooley and
Oberfield, 2007, p. 285; Rozelle, 2007, p. 597). A subtler psychological mechanism has also been
suggested, in which judges recognize forensic evidence as, in a sense, their own creation and thus
experience cognitive resistance to the notion that their faith in it was perhaps over-hasty (Cole, 2004,
p. 1275; Risinger, 2007, p. 473).
However, one might equally well argue that judges simply have a strong intuition that latent print
identification is highly accurate and that they are reluctant to overrule this intuition in deference to
either legal or scientific formalism. Judges, of course, may well be correct in this intuition. It does
not, of course, justify countenancing the refusal to provide any data to support this intuition. Nor
does it justify the institutionalized exaggeration of the probative value of the evidence implied by the
requirement of testimonial reports of individualization, reports that overstate the probative value
of the evidence, regardless of the accuracy of the technique. Note, however, that by not generating
accuracy data and not modifying the testimonial claim of individualization, the government and
practitioners put judges in a quandary, in which they must either admit the evidence despite its
exaggerated testimonial claim or exclude it despite their intuition of its accuracyunless, i.e. they
opt for the solution of limiting, not admissibility, but the testimonial claim (e.g. State v. Pope, 2008).
If judges have been correct in their intuitions about the high accuracy of latent print identification,
then the material damage wrought by their admissibility rulings would be limited to those relatively
few individuals who were erroneously implicated by evidence that was presented to jurorswith
the courts approvalas highly accurate. However, it might also be argued that larger damages have
been wrought by the past decade of controversy over latent print identification. Even if one sets aside
the ultimate accuracy of latent print identification, one is still left with the nagging facts that courts do
not appear to have applied their own admissibility standards in an evenhanded manner; that, when
making such rulings, they did not take the trouble to support them with either legal or scientific
reasoning that would be recognizable as high quality; and that the courts disregarded or dismissed
the views of scientists and scholars on a scientific issue. These are damages wrought to the dynamic
relationship between the institutions of science and law (Jasanoff, 2005) and even, potentially, to
legitimacy of the courts.
Judicial opinions such as these had an often overlooked but crucial consequence: the judicial
opinions themselves came to proxy for the very validation studies whose absence was being debated
(Cole, 2004, p. 1275; Risinger, 2007, p. 467). When asked for evidence of the reliability of latent
print identification, prosecutors, and even courts themselves, tended to point to court opinions rather
than empirical data or studies. In one case discussed in the Report (p. 104), the court referred to
the consensus of the expert and judicial communities that the fingerprint identification technique is
reliable (United States v. Crisp, 2003, p. 269, emphasis added), thus bypassing the external scientific
community altogether in instilling legitimacy on the technique. Courts notions that they themselves
can confer validity on evidence is well illustrated by the following exchange from an admissibility
hearing (the author is The Witness in this exchange):
31
3 People v. Kelly (1976).

4 The expert witnesses who testified at the Daubert hearing in Llera Plaza were Janine Arvizu, a quality assurance auditor;
Ralph Haber, a psychologist; and Allan Bayle, Stephen Meagher and Kenneth Smith, who were latent print examiners (United
States v. Llera Plaza, 2002).
THE WITNESS: I think that scientists and scholars would not view case law as scientific data.
THE COURT: But there is a certain ideaEvery system of analysis appreciates a continued
respect for a process in any kind of forum or arena; right? If its continually to be
relied upon as a credible instrument of identification, that fact, I think, has some
bearing on any kind of contemporary assessment of its validity in a proceeding
like this; isnt that right?
THE WITNESS: Well, respectfully I think that one of the principles of science is not to have
respect for continuing institutions just because theyre continuing.
THE COURT: Well, if the courts have found that there was sufficient challenge to a process, they
would say that the evidence is not admissible as in the case of the lie detector;
isnt that right? . . . If the courts over the passage of time found that a particular
process of scientific information like lie detector polygraph machines, if they
found that that wasnt a valid process, they would say its inadmissible, they
would say its not going to be considered, its not satisfying what we call a Kelly3
or Daubert standard, which Im sure youre familiar with. We dont have that
here; do we? Fingerprint analysis hasnt been rejected by either Daubert or
Kelly or Frye, or whatever state jurisdiction youre relying upon; isnt that right?
THE WITNESS: Thats correct, but to a scientist that does not trump the absence of data. The fact
that courts have accepted it for nearly a century doesnt trump the absence of
data.
THE COURT: But youre not making the proposition that all scientists have decided as a universe of individuals to reject fingerprints as a proof of anything. Youre saying
a group of scientists, including yourself, have rejected fingerprints but not all
scientists.
THE WITNESS: No, of course I could never say that because they have not rendered an opinion.
THE COURT: Right.
THE WITNESS: But what I am saying is that if you look at what scientists have said, almost all
of them are falling on the side of this claim has not been proven. Almost none of
them are saying the opposite.
THE COURT: Well, the Llera Plaza case you identified there were two out of five experts that
testified, three said one thing and two said something different.
THE WITNESS: Thats right. And the threethe twotwo of the five of those people were scientists, and they both were testifying for the defendant.4
THE COURT: Yes.
THE WITNESS: The other witnesses were latent print examiners.
THE COURT: Yes. Who the judge essentially found to be apparently more credible.
THE WITNESS: Apparently so.
32
S. A. COLE
THE COURT: Okay (Testimony of Simon Cole, People v. Caradine, 2008).5
5 The witness in this exchange was the author of this article. As the Report notes (p. 97), it is difficult to get a clear[] picture of how trial courts handle[] . . . judicial dispositions of Daubert-type questions because most cases are decided without
published opinions. My experience consulting on admissibility motions gives me access to some anecdotal information about
what occurs on the trial court level, information that may not be available to other observers without those experiences. In the
interest of better informing readers about what may be occurring in the myriad trial courts across the United States, in this
article, where relevant, I draw on the transcripts of some of those hearings.
The NAS Report, in a crucial passage that should not be overlooked, uses the disturbing term
judicially certified for this phenomenon (p. 86). It is important to emphasize what the Report is
saying here: it is saying that judicial opinions have emerged as an alternative source of legitimacy
for forensic techniques that are unable to derive legitimacy from normal scientific channels. Thus,
one leading latent print practitioner, when asked on national television about the reliability of latent
print identification in 2003 replied, Were winning 41 times out of 41 [admissibility] challenges.
I think that says something (Fingerprints, 2003). This same practitioner recently co-authored an
article acknowledging that for many years the forensic science community has pointed to successful admissibility of its science findings, and the opportunity to cross examine expert witnesses, as
support of a techniques general acceptance and reliability. Remarkably, however, he appears to
have now disavowed this rhetorical strategy, warning philosophically we do not advocate successful
admissibility as demonstrating good science (Budowle et al., 2009, p. 799). Perhaps this turnaround
was prompted by the Rose case (see below) which rendered it no longer possible to argue that latent
print identification always won admissibility challenges. However, because common law is based on
precedent, this turnaround is disturbing. By the time forensic practitioners got around to disavowing
the tactic of treating admissibility hearing victories as de facto evidence of both general acceptance
and reliability in 2009, they had amassed a large body of legal opinions rendering many forensic
techniques, especially latent print identification, legally admissible. Future courts asked to rule on
admissibility challenges may well simply rely on precedent, despite practitioners belated concession
that admissibility hearing victories do not constitute empirical evidence of a techniques reliability.
In this period, during which what we might call the scientific perspective had won nearly unanimous acceptance among scientists and scholars but unanimous rejection in the courts, the National
Academies may have come to seem like a court of last resort for this controversy. It seemed clear
that the courts would continue to support the claims that latent print identification had satisfied the
Kumho reliability requirement and continue to construe the relevant scientific community as practitioners, rather than the broader scientific community. It also seemed that the shelf life of admissibility challenges was nearing expiration as courts were increasingly taking the position that the
issue had been sufficiently litigated and defendants were no longer entitled to admissibility hearings
(United States v. Mitchell, 2004, p. 246). This was disturbing because it seemed that the courts, as
essentially the sole or main customer[s] (Meagher, 2007) of latent print analysis, were the only
institutions capable of forcing change or improvement upon this particular forensic practice. The
National Academies had intervened effectively in prior controversies over other forensic techniques
(National Research Council, 1992, 1996, 2004) and thus seemed to have a special authority in controversies over forensic science, and so it seemed that only the National Academies could put an end
to scientifically and legally questionable claims like the zero error rate, adversarial and implicit
testing, and severe criticism constituting positive peer review.
It should, however, be noted that while the NAS Committee was conducting its study, one court
did hold the government to its burden to demonstrate the reliability of the evidence it proffered.
33
Recognizing that the governments proffer at the admissibility hearing concerning the formation
and individuality of friction ridge skin (the anatomical structure that creates finger, palm and sole
prints) was unresponsive to the issue of the reliability of latent print individualization, Judge Souder
deemed latent print evidence inadmissible (State v. Rose, 2007). This lone ruling, however, had
no precedential value for any other court. Indeed, Souders ruling in Rose, which was discussed
with apparent approval in the NAS Report (p. 105), was recently even further mooted when the
government refiled the case in a federal court, which then deemed latent print evidence admissible
and reliable without holding an evidentiary hearing (Tamber, 2009). This gave rise to the unlikely
spectacle of the government forum shopping in order to use fingerprint evidence. Whetherwithout
the NAS ReportRose would have constituted an anomaly or the harbinger of a new trend is a
counterfactual question whose answer we will never know. Similarly, whether change would have
occurred in latent print practice without the NAS Report is something we will never know.
This was the unusual context into which the NAS Report was issued in February 2009: stark disagreement between scholars and scientists on the one hand and courts and forensic practitioners on
the other. At issue was, in some sense, whothe scientific community or the courtswould have
the last word on an issue that is clearly a matter of science, but science that is primarily, if not
exclusively, used in legal settings?
The Report wholly endorsed what we might call the scientific position. It repudiated the notion
of adversarial testing (p. 42), claims of infallibility (p. 87) and the claim of a zero error rate
(p. 142). It rejected the claim that the reliability of latent print individualization could be inferred
from the uniqueness of friction ridge skin (pp. 4344, 144). It stated that no forensic assay other than
nuclear DNA profiling could support claims of individualization, an assessment that included latent
print analysis (p. 87). And, it stated that there was only limited information about the accuracy and
reliability of friction ridge analyses (p. 142).
In terms of the head counting exercise discussed above, the NAS Report was highly significant. Although it adds 17 or so names to the list of informed external observers who have adopted
the broad scientific position, more important was the endorsement of a scientific institution and not,
of course, just any scientific institution, but the NAS. Most important of all is the fact that the Report
represented the outcome of the standard NAS Report-producing process. Thus, the process of convening and vetting a committee composed from the nations foremost scientists, engineers, health
professionals, and other experts (National Academies, 2009) with no vested interest in the controversy and exposing them to the available data, literature and information yielded the conclusions
that they did. Thus, the issuing of the NAS Report rendered the already remarked upon discrepancy
between the legal and scientific communities as to the demonstrated reliability of latent print individualization even starker than it already was. It put the weight of a scientific institution, rather than
merely the accumulated weight of individual scientists and scholars, behind the scientific position.
The Report would seem to speak not only for the Committee but also for the NAS and not only for
the NAS for science itself (Hilgartner, 2000, p. 88). Moreover, it again demonstrated the difficulty
latent print examiners have had convincing any external observers, other than judges, that they have
validated their claims to be able to individualize latent prints.
Nonetheless, it raises the question: does it matter? I am not merely being facetious. By my
estimate, the NAS Committee included only one member who has probably ever analysed a latent
3. Who speaks for forensic science?
34
S. A. COLE
4. Speaking for evidence

Assuming then that the controversy does return to the courts and that the Report is treated with the
authority that it deserves, what will happen in the courtroom? There will be admissibility challenges
to be sure. But there will also be challenges to the nature of latent print testimony itself. Many
evidence scholars have taken the position that courts and scholars alike have focused too much on
the underlying validity of forensic techniques and perhaps not enough on an equally important issue:
the nature of forensic testimony (Beecher-Monas, 1999, p. 1062; Berger, 2003, p. 1140; Black,
2003; Friedman, 2003, p. 1063; Gross & Mnookin, 2003, p. 143; Nance, 2003, p. 253; Faigman,
2004, p. 258; Imwinkelried, 2004, p. 277). Forensic science, after all, is literally science that
speaksin a courtroom, to a factfinder, which is usually, in the United States, a jury of laypersons. I
have elsewhere proposed putting the forensic back into forensic science and devoting some serious
thought not merely to the dichotomous question of whether particular expert witnesses should or
should not be permitted to testify before a jury but also to the continuous question of what evidentiary
claims they should be permitted to make before that jury (Cole, 2007). The Report takes a similar
position in its Recommendation 2: that the proposed National Institute of Forensic Science should
establish standard terminology to be used in reporting on and testifying about the results of forensic
science investigations (p. 22).
Let us consider how this recommendation would play out for latent print evidence. Latent print
analysis is unusual in that, unlike many other trace evidence disciplines, it has taken a quite clear
position on the way in which testimonial conclusions are framed, at least in the United States.
6 I am thinking of Jay Siegel.
print.6 Since previous interventions by scientists and scholars had been dismissed by practitioners
as the efforts of a handful of critics, mostly academics and law professors who, beyond literature
searches, have not acquired any practical experience in comparing prints (Moenssens, 2005), should
or will the NAS Report be dismissed on the same grounds? If not, we must ask hard questions about
why scientists and scholars were not heeded until the NAS Report. We must take note of the fact that
the courts left here a record of ignoring scientists and scholars until the intervention of the NAS, a
record that led the Committee to characterize the courts as utterly ineffective in dealing with the
lack of validity and accuracy data in many forensic disciplines (p. 109). This raises discomfiting
questions about judicial dealings with sciencesince surely we cannot rely on the NAS to correct
all such situations. What other scientific matters are being decided by courts in this manner without
remediation through an NAS Report? Indeed, even the present Report had a long and tortured
pre-history before it was officially commissioned (Kennedy, 2003), and it might easily never have
seen the light of day. What would be the legal status of latent print identification and other forensic
disciplines had this Report never been commissioned?
If, on the other hand, the NAS Report is dismissed on the same grounds as previous scientific
interventions, what then? We will still have standoff between law and science, but now we will
have not just lawyers, judges, scientists and scholars involved but scientific institutions as well. The
supposed court of last resort will turn out not to have been a court of last resort at all. Further
adjudication will be necessary. And where will this adjudication occur? In the courts, as defendants
leverage the Report in challenges to forensic evidence. Thus, the court of last resort for forensic
science may again be the courts themselves.
35
WILLIAM THOMPSON8 : Theres some questions about how forensic scientists should be characterizing their findings at this point in time in light of the Report. For a long time, latent print
examiners and tool mark examiners and others who examine marks have been coming into court
and testifying that they can identify the source of the mark to the exclusion of all other sources and
with complete certainty. Do you think that kind of testimony is justified at this point in time pending
further research, and, if not, what should forensic scientists be saying about that kind of examination
at this point in time? And whats the FBI going to be saying about it?
CHRISTIAN HASSELL9 : Those absolute individualization statements are no longer a part of the
practice, especially the tool marks, the pattern based areas.
WILLIAM THOMPSON: What about latent prints?
7 Granting individualization to DNA profiling was an odd decision given that many DNA scientists do not even claim to
be able to achieve individualization, and, even for those who do, DNA individualizations must always be accompanied by the
caveat, barring monozygous twins, as SWGFAST (2009, p. 3) points out.
8 William Thompson is a Professor of Criminology, Law & Society at the University of California, Irvine.
9 Christian Hassell, an analytical chemist, had been named Director of the FBI Laboratory in June 2008.
According to both the Scientific Working Group on Friction Ridge Analysis, Science & Technology
(SWGFAST) and the IAI, only three testimonial conclusions are permitted: exclusion, inconclusive
and individualization (IAI, 1979, 1980; SWGFAST, 2003b). Individualization, therefore, is the
only permissible conclusion that implicates a defendant. Individualization has been defined as The
determination that corresponding areas of friction ridge impressions originated from the same source
to the exclusion of all others (identification) (SWGFAST, 2003a). As numerous commentators (e.g.
Stoney, 1997, p. 70; Champod and Evett, 2001, p. 113; Epstein, 2002, p. 612; Cole, 2004, p. 1196;
Zabell, 2005, p. 155; Champod, 2008, p. 117; Mnookin, 2008) and the Report (p. 142) have noted,
such categorical conclusions are problematic.
The main issue is that such conclusions require knowledge about the rarity of observed features
in the relevant population. Even if estimates of rarity are available, as, e.g. in the case of DNA
profiling, a conclusion of individualization involves a process of rounding off a small probability into
a probability of zero that is controversial even for DNA profiling (Buckleton, 2005). However, in the
case of latent prints, individualization becomes even more problematic because rarity estimates are
not derived from formal, systematically collected data, but rather intuited by the examiner (Champod
et al., 2004; IEEGFI II, 2004; Thompson & Cole, 2007).
Recognizing this, the Report clearly states that claims of individualization are not supported for
any forensic assay other than DNA profiling (p. 87).7 Thus, claims of latent print individualization are not supported. In this, the Report is not alone. At least one court that admitted latent print
evidence precluded testimony of individualization (State v. Pope, 2008). Here one wishes that the
Report had gone further and noted the bankruptcy of individualization as a concept, no matter what
the assay (Biedermann et al., 2008; Champod, 2008; Saks & Faigman, 2008; Saks & Koehler, 2008;
Champod, 2009; Cole, 2009). DNA profiling has shown that it is possible to have highly probative
evidence without rounding down the probability of the alternative hypothesis for the factfinder, and
this renders individualization at least conceptually, needless (Biedermann et al., 2008, p. 130).
What, however, does the future hold for latent print testimony? We now have a situation in which
latent print examiners are restricted by professional guidelines to giving testimony which, according
to the NAS Report, they cannot support. Among the high points of the Symposium that generated
this journal issue was the following exchange:
36
S. A. COLE
CHRISTIAN HASSELL: I have two of my examiners here in the room so I . . . Im not sure, I
cant remember what we do exactly. Like I say, Im not the expert in these areas (Forensic Science
for the 21st Century, 2009, Disk 4, Q&A section, at 0:0001:20).10
MELISSA GISCHE11 : As far as fingerprinting is concerned, we do still identify, we will effect
individualizations, to the exclusion of all others based on the science behind our examinations.
WILLIAM THOMPSON: And does the NRC Reports statement about the lack of science in this
area change your impression about how you should be
CHRISTIAN HASSELL: No, because we dont agree with the lack of science. Weve got
MICHAEL RISINGER12 : For individualization?
MICHAEL RISINGER: For perfect individualization, you dont agree with the lack of science?
CHRISTIAN HASSELL: Oh, not perfect, but
MICHAEL RISINGER: Well, thats what she said.
UNIDENTIFIED: She used the words to the exclusion of all others.
HENRY LEE13 : We do have chances of perfect individualization. Just now I give an example, . . .
in [inaudible] case I have, they leave the bumper behind with the license plate. In my opinion, thats
a perfect individualization. I have no other cases with that license plate, with the bumper, that the
car lost the bumper and license plate. Thats a perfect individualization example. Of course, it does
not address the fingerprint issue.
WILLIAM THOMPSON: Right.
[other questions asked of the panel]
JAY KOEHLER14 : I also have a question for Dr Hassell. I just want to get clarification on what Bill
Thompson was asking. Is it the FBIs position that an examiner, a fingerprint examiner, can identify
the source of a fingerprint with 100% certainty? Is it their position that they can individualize a latent
print? . . .
CHRISTIAN HASSELL: Ok, I dont mean to weasel out of this, but I am admitting my navete in
some of these areas, just being so new. Thats why I have two of the top people here with me. So
Im going to ask Melissa to answer the first part of your question because I think its very important.
However, I will say . . . the statements made earlier that the areas of fingerprint science has no
scientific foundation, that was in error and, in fact we talked about that at the break. I mean who
10 The remainder of the dialogue presented here was not recorded on the publicly available video record of the conference
proceedings. With the help of Laurie Ralston, Sandy Askland and Jay Koehler, I obtained the raw footage of the Question &
Answer section of this panel discussion. Because not everyone involved in the discussion spoke into microphone and some
people spoke simultaneously, some portions of the discussion were difficult or impossible to hear. What is presented here
represents my best effort at a faithful transcription, with some assistance from Jay Koehler, edited for flow and clarity.
11 Melissa Gische is a latent print examiner for the FBI Laboratory.
12 Michael Risinger is John J. Gibbons Professor of Law at Seton Hall University.
13 Henry Lee is the Chief Emeritus for the Connecticut Department of Public Safety, Division of Scientific Services.
14 Jay Koehler is a Professor of Law and Business at Arizona State University.
UNIDENTIFIED: Individualization?
37
made that statement and the basis for it? Thats just wrong. Theres quite a bit of work done on areas
of uniqueness, persistence, and others, so Im going to, Melissa has the . . .
JAY KOEHLER: Well, I still dont know whether . . . so the opinions are made with 100% certainty?
Source opinions, source identification? You would say Ive identified this latent to Jay Koehler with
100% certainty. Is that still what you would say?
MELISSA GISCHE: Again, it is my opinion as far as the certainty, I dont know that I would ever
attribute a percentage to that. But in my opinion, when I identify a print, I am saying that there is, in
fact, enough information present to determine the source of that print.
WILLIAM THOMPSON: To the exclusion of all others in the universe?
MELISSA GISCHE: That is correct.
MICHAEL RISINGER: You might want to think about modifying that. You wont lose that much,
and you wont be fighting a losing battle over what the science can do.
Thus, by April 2009, the NAS Report by itself had not persuaded the FBI to modify the testimonial claim of individualization, and a recent FBI publication continues to defend individualization
(Peterson et al., 2009). Similarly, SWGFASTa body that is sponsored, though not controlled
by the FBI (Grieve, 1999, p. 145) has continued to defend individualization. It issued a statement
respectfully disagree[ing] with the NASs assertion that only nuclear DNA analysis could support
claims of individualization, stating History, practice, and research have shown that fingerprints can,
with a very high degree of certainty, exclude incorrect sources and associate the correct individual
to an unknown impression (SWGFAST, 2009, p. 3). The statement confuses the issue of reaching
correct conclusions with the issue of making a scientific claim about the size of the potential donor
pool, and no data are cited to support or further specify the claimed high degree of certainty.
Whether courts will continue to accept such arguments remains to be seen. Unless latent print examiners and the litigants who employ them want to be cross-examined on the NAS Report, they may
need an alternative testimonial claim. Already a number of emerging alternatives can be discerned.
Let us examine them in turn.
15 Note that, as discussed above, the Report, as well as many scholars, reject this distinction. The FBIs conception of this
distinction has been further elucidated by Peterson et al. (2009), who state:
Because latent print examinations do not employ instrumentation that can introduce systematic or random errors, the only
general type of scientific error in the latent print discipline is human error, also commonly referred to as practitioner error. . . .
It is more accurate to say that the ACE-V methodology does not have a calculable error rate because it has no inherent error.
MELISSA GISCHE: If I can clarify, as far as the absolute certainty argument, its still the examiners opinion that there is enough information present that they can identify the source of that print
to a single area of friction ridge skin. Now, as far asI may as well cover this nowthe zero error
rate claim, Im not sure why this keeps getting so misinterpreted. Fingerprint examiners have not,
nor do we currently, claim to have a zero error rate. We know that errors have occurred. I think where
some of the confusion in the past has come is in trying to distinguish between a methodological versus a practitioner error.15 I think things have gotten all twisted up in that, but as examiners, errors
have been publicized. Any time the human is involved we know there is a potential for error, which
is why we have all the various quality assurance measures, and so on . . . its too long for me to
list here. But, as far as still individualizing, yes we still do it, based on the underlying principles of
friction ridge skin. Jay, does that answer your question?
38
4.1
S. A. COLE
Redefine individualization
SWGFASTs response to NAS Recommendation 2 states that SWGFAST uses the development of
our glossary to standardize terminology (SWGFAST, 2009, p. 6). The new SWGFAST Glossary
removes the term to the exclusion of all others and substitutes the term conclusion for determination, thus redefining individualization as The conclusion that corresponding areas of friction
ridge impressions originated from the same source (SWGFAST, 2008). Although removing the term
to the exclusion of all others may make the testimony less hyperbolic, the analyst is still attesting
to something she does not know (or, in Bayesian terms, to a posterior probability)that the two impressions derive from the same source. Redefining the statement as a conclusion, rather than a determination is perhaps a slight improvement, but still conveys the impression that the statement derives
from an analysis of data rather than being what it really is: a decision (Biedermann et al., 2008).
Identification
Another potential alternative is substituting the term identification for individualization. In a recent admissibility hearing on latent print evidence, a member of SWGFAST disavowed the notion
of individualization, conceded that it was unsustainable and offered in its stead a conclusion of
identification, defined as a decision given the relevant population . . . that the chance that someone
could have left [the latent print] is so remotely small that he is willing to dismiss it and say yes, I
believe that this latent print in my opinion was produced by that individual (Testimony of Glenn
Langenburg, State v. Hull, 2008, p. 149). There ensued some confusion about whether individualization and identification are in fact distinct concepts given that the entry for Identification in the
SWGFAST (of which the witness was a member) Glossary reads See Individualization (SWGFAST, 2002, 2003a). If they are synonymous, then nothing has changed. If they are distinct, then
the witness violated the SWGFAST and IAI guidelines which ban inclusive conclusions other than
individualization.
This confusion about whether these two terms are synonymous or distinct is common. For many
forensic practitioners, identification is used in a manner tantamount to individualization, as, e.g.
in toolmark examiners definition of identification as the likelihood another tool could have made
the mark is so remote as to be considered a practical impossibility (Association of Firearms &
Toolmark Examiners, 1998; Schwartz, 2005). However, identification has been defined by some
forensic scientists to indicate that the potential donor pool of a trace is greater than one, whereas
individualization is reserved for situations in which the donor pool has been reduced to one (Kirk,
1963; Inman and Rudin, 2001, p. 115; Thornton & Peterson, 2002, p. 8). The problem with this
notion of identification is that it is vague: has the donor pool been reduced to a class of a million
potential sources or only two? In any case, this does not appear to be the usage proposed by the
witness in Hull, which appears to be drawn more from the reasoning of Biedermann et al. (2008), in
which the relevant donor pool contains only one member but the reduction to this pool is conceived
as a decision rather than as, say, a conclusion. If this is indeed the case, the testimonys status as
decision would need to be made clear to the factfinder. The term identification would not seem to
do this job well; laypersons would seem to be apt to construe identification to mean precisely what
some forensic scientists claim is meant only by individualization, a possibility that is given added
plausibility by the fact that, as discussed above, SWGFAST itself presents ambiguous information
about whether the terms are synonymous.
4.2
39
Moreover, of course, questions might be raised about the empirical basis upon which the expert
witnesses draws conclusions about the probability of false association. Currently, there is no basis
other than the experts intuition based on her experience in casework (Thompson & Cole, 2007). This
seems inadequate, given that experts do not systematically record and compile data about the rarity
of features during casework and then retrieve that data to generate an estimate of the probability of
false association. Even if such data were recorded, there would be difficulties extrapolating from
the relatively small populations observed to the world population invoked by the decision described
above (Zabell, 2005; Saks & Koehler, 2008). Resorting to identification as a testimonial claim
allows latent print practitioners to elude what, presumably, is one of the main objectives of efforts at
forensic reform, as embodied by the NAS Report: generating rarity data and encouraging its use in
formulating defensible reports.
Opinionization
A related possibility is that latent print reports could be rehabilitated by couching them as opinions,
rather than determinations, conclusions or facts. This is the solution urged by a recent report
from the United Kingdom (Nuffield Council on Bioethics, 2007). We can also see signs of the ascendance of this approach in statements to the NAS Committee by the IAI (2007a). This solution also
seems undesirable. For one thing opinionization (Cole, 2008a) seems to function like a universal
solvent (Risinger, 2009), excusing all sins. (I was wrong? Sorry, it was only my opinion.) Like
the identification solution, opinionization allows forensic scientists to elude developing rarity data
that would allow them to more precisely convey the probative value of the evidence.
4.4
Accuracy rates
One possible alternative that has yet to show signs of acceptance in the practitioner committee is
the notion of attaching estimated accuracy rates to latent print examiners conclusions. As discussed
above, historically latent print have conveyed accuracy rates to factfinders but these were assertions
not derived from measurements: the claim that the methodological error rate is zero, and the practitioner error rate is vanishingly small, essentially zero, negligible or microscopic. Such claims
were founded upon a supposed separation of the causes of error into two categories, methodological and human or practitioner. This separation was dubious, first, because, as the Report notes,
there is no method without a practitioner (p. 143) (Giannelli, 2009); second, because the category
of methodological error was constructed in such a way that no errors would ever be assigned to
it and, thus, it would remain, eternally, definitionally fixed at zero (Peterson et al., 2009); third,
because if the methodological error rate is fixed as zero, it seems misleading to state it to the jury
as if it were some sort of meaningful finding; and, fourth, because even if there were two categories
of error the most meaningful single piece of information to convey to the consumer of the evidence
would be the total error rate (Cole, 2005).
The NAS Reports criticism of the zero error rate claim should render it extinct in the courtroom, and the IAI has now recommended that latent print examiners not make such claims in their
testimony (Garrett, 2009). While all rational thinkers should rejoice at the disappearance of the zero
error rate claim, it is important not to forget that it really was made in the first place (testimony of
Stephen Meagher and Bruce Budowle United States v. Mitchell, 1999b), and it really was accepted
by some courts (United States v. Havvard, 2000). It is also important to note that the zero error rate
was showing no signs of disappearing before the NAS intervened. I have seen affidavits asserting
4.3
40
S. A. COLE
5. Conclusion
I have focused my remarks here on latent print identification, but they raise implications that apply to
the rest of forensic science and, indeed, to the use of science in the courts more generally. The issue
is not necessarily the quality of particular techniques, but rather the irrationality of the arguments that
were mounted on their behalfthe fact that courts were, in the Reports words, utterly ineffective
at detecting that irrationality, and the fact that scholars and scientists were largely deemed irrelevant
to discussions of validity.
My concerns about the need for validation from some external scientific community, rather than
relying merely on self-validat[ion] (Black, 1988, p. 633) by a practitioner community or judicial
a zero error rate as recently as 2006 (Declaration of Erik Carpenter, United States v. Mikhel et al.,
2006, p. 3), I have heard the claim made in admissibility hearings as recently as 2008 (Letter from
Elisa J. Macken, State v. Sheehan, 2008), and I have heard it defended by practitioners even after the
publication of the NAS Report and the IAI response.
If latent print examiners can no longer tell jurors that the error rate of latent print analysis is zero,
can they tell them something else? The NAS Report states that there is only limited information
about the accuracy and reliability of friction ridge analyses (p. 142), so it does not seem like latent
print examiners can reasonably offer either quantitative or qualitative estimates of their accuracy.
That does not, of course, mean that researchers will not eventually be able to generate responsible
estimates of the accuracy of latent print examiners conclusion of common source that might be
conveyed to jurors. This is one potential solution to the dilemma of latent print testimony that was
discussed at the symposium under the label black-box validation (Risinger et al., 1998).
Another potential solution is the likelihood ratio approach being explored by Champod and his
colleagues (Egli et al., 2006; Meuwly, 2006; Neumann et al., 2006, 2007). This approach does attempt to estimate the rarity of observed features in the relevant population. Although there was some
debate at the symposium over these two approaches, it would seem that they are complementary,
and neither piece of information (accuracy or rarity) would be sufficient without the other. It will
be important to try to estimate the rarity of features, but, if analyses are to be carried out by human
examiners, it will also be important to qualify that estimation with data on actual performance rates
of comparable human examiners.
Any such efforts will need to be accompanied by educational efforts. On the one hand, one
might need to train analysts accustomed to reporting conclusions in categorical terms to report in
probabilistic terms. It will be important, however, to avoid the temptation to use computer software
to enable expert witnesses to give testimony using impressive probability figures without a thorough
understanding of those probabilities. Future analysts may require much greater education in statistics
than in the past.
Finally, it should be noted that developing a scientifically supportable testimonial claim is only
half the task, or perhaps even less than half, that lies before the justice system in order to make
appropriate use of forensic evidence. There remains the task of figuring out how to effectively convey the probative value of forensic evidence to jurors. Numerous psychological studies have shown
that jurors are innumerate and prone to fallacies in reasoning even when presented with testimonial
claims that are themselves appropriately stated (e.g. Koehler, 2001). Thus, even the development of
likelihood ratios for latent print comparisons, e.g. will not fully solve our problem because we know
that jurors have great difficulty understanding likelihood ratios.
41
certification by the courts apply to other disciplines as well. The NAS, the scientific community and
indeed everyone should be concerned by the counterfactual that this Report might never have been
commissioned. In that case, who, ultimately, would have spoken for science?
Likewise, the problem of how to report the probative value of conclusions about forensic evidence applies to other disciplines as well. Although other disciplines are less strict than latent prints
about insisting only on unsustainable claims of individualization, many of them currently make
unacceptably vague claims of match or identification. All disciplines should abandon individualization as a telos (Cole, 2009) and strive towards using accuracy and rarity to report defensible
claims about the probative value of forensic associations, fulfilling their mandate to speak, with as
much transparency (Champod and Evett, 2001) as possible, for science.
The author was a expert witness and/or consultant in several of the cases discussed in this article,
including United States v. Mitchell, United States v. Llera Plaza, People v. Carradine, State v. Rose,
State v. Hull, State v. Pope, State v. Sheehan and State v. Sullivan. The author was a signatory to the
amicus curiae brief filed in Commonwealth v. Patterson that is discussed in this article.
Funding
National Science Foundation (SES-0115305).
Acknowledgements
I am grateful to D. Michael Risinger, Jay Koehler and an anonymous referee for comments on this
paper and to Jay Koehler, Michael Saks and David Kaye for inviting me to the Symposium on
Forensic Science in the 21st century. I am grateful to Ian Tingen, Laurie Ralston, Sandy Askland
and Jay Koehler for assistance obtaining video footage of the conference. Any opinions, findings
and conclusions or recommendations expressed in this material are those of the author and do not
necessarily reflect the views of the National Science Foundation.
R EFERENCES
A SSOCIATION OF F IREARMS AND T OOL M ARK E XAMINERS (1998) Theory of Identification as It Relates to
Toolmarks. AFTE Journal, 30, 8688.
B EECHER -M ONAS , E. (1999) A Ray of Light for Judges Blinded by Science: Triers of Science and Intellectual
Due Process. Ga. L. Rev., 33, 1047.
B ENEDICT, N. (2004) Fingerprints and the Daubert Standard for Admission of Scientific Evidence: Why
Fingerprints Fail and a Proposed Remedy. Ariz. L. Rev., 46, 519549.
B ERGER , M. A. (1994) Procedural Paradigms for Applying the Daubert Test. Minn. L. Rev., 78, 13451386.
B ERGER , M. A. (2003) Expert Testimony in Criminal Proceedings: Questions Daubert Does Not Answer.
Seton Hall L. Rev., 33, 11251140.
B IEDERMANN , A., B OZZA , S. & TARONI , F. (2008) Decision Theoretic Properties of Forensic Identification:
Underlying Logic and Argumentative Implications. For. Sci. Int., 177, 120132.
B LACK , B. (1988) A Unified Theory of Scientific Evidence. Fordham L. Rev., 56, 595.
Conflict of interest disclosure
42
S. A. COLE
B LACK , B. (2003) Focus on Science, Not Checklists. Trial, 39, 26.

Brim v. State (1997) 695 So.2d 268 (Fla.).
B UCKLETON , J. (2005) Population Genetic Models. In: Buckleton, J., Triggs, C.M., Walsh, S.J. (Eds.), Forensic DNA Evidence Interpretation. CRC Press, Boca Raton, Fla., pp. 65122.
B UDOWLE , B., B OTTRELL , M. C., B UNCH , S. G., F RAM , R., H ARRISON , D., M EAGHER , S., O IEN , C. T.,
P ETERSON , P. E., S EIGER , D. P., S MITH , M. B., S MRZ , M. A., S OLTIS , G. L. & S TACEY, R. B. (2009)
A Perspective on Errors, Bias, and Interpretation in the Forensic Sciences and Direction for Continuing
Advancement. J. Forensic Sci., 54, 798809.
C HAMPOD , C. (2008) Fingerprint Examination: Towards More Transparency. Law, Probability and Risk, 7,
111118.
C HAMPOD , C. (2009) Identification and Individualization. In: Moenssens, A., Jamieson, A. (Eds.), Encyclopedia of Forensic Sciences. Wiley, London.
C HAMPOD , C. & E VETT, I. W. (2001) A Probabilistic Approach to Fingerprint Evidence. J. Forensic Identification, 51, 101122.
C HAMPOD , C., L ENNARD , C., M ARGOT, P. & S TOILOVIC , M. (2004) Fingerprints and Other Ridge Skin
Impressions. CRC Press, Boca Raton, Fla.
C OLE , S. A. (1998) Witnessing Identification: Latent Fingerprint Evidence and Expert Knowledge. Social
Studies of Science, 28, 687712.
C OLE , S. A. (2004) Grandfathering Evidence: Fingerprint Admissibility Ruling from Jennings to Llera Plaza
and Back Again. Am. Crim. L. Rev., 41, 11891276.
C OLE , S. A. (2005) More Than Zero: Accounting for Error in Latent Fingerprint Identification. J. Crim. L. &
Criminology, 95, 9851078.
C OLE , S. A. (2006) Is Fingerprint Identification Valid? Rhetorics of Reliability in Fingerprint Proponents
Discourse. Law & Poly, 28, 109135.
C OLE , S. A. (2007) Where the Rubber Meets the Road: Thinking About Expert Evidence as Expert Testimony.
Villanova Law Review, 52, 803842.
C OLE , S. A. (2008a) The Opinionization of Fingerprint Evidence. BioSocieties, 3, 105113.
C OLE , S. A. (2008b) Out of the Daubert Fire and into the Fryeing Pan? The Admissibility of Latent Print
Evidence in Frye Jurisdictions. Minn. J. L. Sci. & Tech., 9, 453541.
C OLE , S. A. (2009) Forensics without Uniqueness, Conclusions without Individualization: The New Epistemology of Forensic Identification. Law, Probability and Risk, 8, 233255.
Commonwealth v. Patterson (2005) 840 N.E.2d 12 (Mass.).
C OOLEY, C. M. (2007) Forensic Science and Capital Punishment Reform: An Intellectually Honest Assessment. Geo. Mason U. Civ. Rts. L.J., 17, 299422.
C OOLEY, C. M. & O BERFIELD , G. S. (2007) Increasing Forensic Evidences Reliability and Minimizing
Wrongful Convictions: Applying Daubert Isnt the Only Problem. Tulsa L. Rev., 43, 285380.
Daubert v. Merrell Dow Pharmaceuticals (1993) 509 U.S. 579 (U.S.).
DWYER , D. (2007) (Why) Are Civil and Criminal Expert Evidence Different? Tulsa L. Rev., 43, 381396.
E DWARDS , H. T. (2009) Solving the Problems that Plague the Forensic Science Community, Forensic Science
for the 21st Century. Sandra Day OConnor College of Law, Arizona State University, April 34, 2009:
Disk 3.
E GLI , N., C HAMPOD , C. & M ARGOT, P. (2006) Evidence Evaluation in Fingerprint Comparison and Automated Fingerprint Identification SystemsModeling with Finger Variability. For. Sci. Int., 167,
189195.
E PSTEIN , R. (2002) Fingerprints Meet Daubert: The Myth of Fingerprint Science is Revealed. So. Cal. L.
Rev., 75, 605657.
FAIGMAN , D. L. (2002) Is Science Different for Lawyers? Science, 297, 339340.
43
FAIGMAN , D. L. (2004) Expert Evidence in Flatland: The Geometry of a World Without Scientific Culture.
Seton Hall L. Rev., 33, 255267.
FAIGMAN , D. L., K AYE , D. H., S AKS , M. J. & S ANDERS , J. (1997) Modern Scientific Evidence: The Law
and Science of Expert Testimony. 1st ed. West, St. Paul, Minn.
FAIGMAN , D. L., K AYE , D. H., S AKS , M. J. & S ANDERS , J. (2007) Modern Scientific Evidence: The Law
and Science of Expert Testimony. 3rd ed. West, St. Paul, Minn.
F INGERPRINTS (2003), 60 Minutes (5 Jan.).
F ORENSIC S CIENCE FOR THE 21 ST C ENTURY (April 3-4, 2009). Institutional Structure. Sandra Day OConnor
College of Law, Arizona State University, Tempe, AZ.
F RIEDMAN , R. D. (2003) Squeezing Daubert Out of the Picture. Seton Hall L. Rev., 33, 10471070.
G ARRETT, R. (2009) Letter to All Members of the International Association for Identification, Feb. 19, available at http://www.clpex.com/Articles/TheDetail/300-399/TheDetail394.htm.
G IANNELLI , P. (2002) Fingerprints Challenged!, Criminal Justice, 17, Spring, 3335.
G IANNELLI , P. (2009) Forensic Science NAS Report, Forensic Science for the 21st Century. Sandra Day
OConnor College of Law, Arizona State University.
Goeb v. Tharaldson (2000) 615 N.W.2d 800 (Minn.).
G RIEVE , D. L. (1999) The Identification Process: TWGFAST and the Search for Science. Fingerprint Whorld,
25, 315325.
G ROSS , S. & M NOOKIN , J. L. (2003) Expert Information and Expert Evidence: A Preliminary Taxonomy.
Seton Hall L. Rev., 34, 143.
H ABER , L. & H ABER , R. N. (2003) Error Rates for Human Fingerprint Examiners. In: Ratha, N.K., Bolle, R.
(Eds.), Automatic Fingerprint Recognition Systems. Springer-Verlag, New York, pp. 339360.
H ABER , L. & H ABER , R. (2008) Scientific Validation of Fingerprint Evidence under Daubert. Law, Probability
and Risk, 7, 87109.
H ILGARTNER , S. (2000) Science on Stage: Expert Advice as Public Drama. Stanford University Press,
Stanford.
I MWINKELRIED , E. J. (2004) The Relativity of Reliability. Seton Hall L. Rev., 34, 269287.
I NMAN , K. & RUDIN , N. (2001) Principles and Practice of Criminalistics: The Profession of Forensic Science.
CRC Press, Boca Raton, Fla.
I NTERNATIONAL A SSOCIATION FOR I DENTIFICATION (IAI) (1979) Resolution VII. Identification News,
29, 1.
I NTERNATIONAL A SSOCIATION FOR I DENTIFICATION (IAI) (1980) Resolution V. Identification News, 30, 3.
I NTERNATIONAL A SSOCIATION FOR I DENTIFICATION (IAI) (2007a) IAI Position Concerning Latent Fingerprint Identification, http://www.clpex.com/Information/IAI Position Statement 11-29-07.pdf.
I NTERNATIONAL A SSOCIATION FOR I DENTIFICATION (IAI) (2007b) Letter to The National Academies of
Sciences Committee to Review the Forensic Sciences, Sept. 19, available at http://www.theiai.org/nas
letter 20070919.pdf.
I NTERPOL E UROPEAN E XPERT G ROUP ON F INGERPRINT I DENTIFICATION II (IEEGFI II) (2004) Method for
Fingerprint Identification, http://www.interpol.int/Public/Forensic/fingerprints/WorkingParties/IEEGFI2/
default.asp.
JASANOFF , S. (2005) Laws Knowledge: Science for Justice in Legal Settings. Am. J. Pub. Health, 95, S49
S58.
Jones v. United States (1988) 548 A.2d 35 (D.C.).
K AYE , D. H. (2003a) Questioning a Courtroom Proof of the Uniqueness of Fingerprints. International Statistical Review, 71, 521533.
K AYE , D. H. (2003b) The Nonscience of Fingerprinting: United States v. Llera Plaza. QLR, 21, 10731088.
K ENNEDY, D. (2003) Forensic Science: Oxymoron? Science, 302, 1625.
44
S. A. COLE
K IRK , P. L. (1963) The Ontogeny of Criminalistics. J. Crim. L., Crim., & Pol. Sci, 54, 235238.
KOEHLER , J. J. (2001) When Are People Persuaded by DNA Match Statistics? Law and Human Behavior, 25,
493513.
KOEHLER , J. J. (2008) Fingerprint Error Rates and Proficiency Tests: What They Are and Why They Matter.
Hastings L.J., 59, 10771099.
Kumho Tire v. Carmichael (1999) 526 U.S. 137 (U.S.).
L A M ORTE , T. M. (2003) Sleeping Gatekeepers: United States v. Llera Plaza and the Unreliability of Forensic
Fingerprinting Evidence under Daubert. Alb. L.J. Sci. & Tech., 14, 171214.
L AWSON , T. F. (2003) Can Fingerprints Lie? Re-weighing Fingerprint Evidence in Criminal Jury Trials. Am.
J. Crim. L., 31, 166.
M EAGHER , S. (2007) Post-decision Comment. The Weekly Detail, 285, Jan. 29, http://www.clpex.com/Articles/
TheDetail/200-299/TheDetail285.htm.
M EARNS , G. (2009) Legal Rules for Forensic Science Evidence, Forensic Science for the 21st Century. Arizona
State University.
M EARS , M. & DAY, T. M. (2003) The Challenge of Fingerprint Comparison Opinions in the Defense of a
Criminally Charged Client. Ga. St. U. L. Rev., 19, 705.
M EUWLY, D. (2006) Forensic Individualisation from Biometric Data. Science & Justice, 46, 205213.
M NOOKIN , J. L. (2001) Fingerprint Evidence in an Age of DNA Profiling. Brook. L. Rev., 67, 1370.
M NOOKIN , J. L. (2008) The Validity of Latent Fingerprint Identification: Confessions of a Fingerprinting
Moderate. Law, Probability and Risk, 7, 127141.
M OENSSENS , A. (2002) The Reliability of Fingerprint Identification: A Case Report. Forensic Evidence.com,
http://www.forensic-evidence.com/site/ID/pollak2002.html.
M OENSSENS , A. (2003) Palmprint and Handwriting I.D. Satisfy Daubert Rule. Forensic Evidence.com,
http://www.forensic-evidence.com/site/ID/ID palmprint.html.
M OENSSENS , A. (2005) Court Challenges to Friction Ridge Impression EvidenceHow Long Will They Last?
Annual Meeting of the Chesapeake Bay Division of the International Association for Identification.
M ORIARTY, J. C. (2004) Psychological and Scientific Evidence in Criminal Trials. Rel. 8 ed. Thomson West,
Eagan, Minn.
NANCE , D. A. (2003) Reliability and the Admissibility of Experts. Seton Hall L. Rev., 34, 191253.
NATIONAL ACADEMIES (2009) The National Academies Study Process.
http://www.nationalacademies.org/studyprocess/index.html.
NATIONAL R ESEARCH C OUNCIL (1992) DNA Technology in Forensic Science. National Academy Press,
Washington, DC.
NATIONAL R ESEARCH C OUNCIL (1996) The Evaluation of Forensic DNA Evidence. National Academy Press,
Washington.
NATIONAL R ESEARCH C OUNCIL (2004) Weighing Bullet Lead Evidence. The National Academies Press,
Washington.
N EUFELD , P. (2005) The (Near) Irrelevance of Daubert to Criminal Justice and Some Suggestions for Reform.
Am. J. Pub. Health, 95, S107S113.
N EUMANN , C., C HAMPOD , C., P UCH -S OLIS , R., E GLI , N., A NTHONIOZ , A. & B ROMAGE -G RIFFITHS , A.
(2007) Computation of Likelihood Ratios in Fingerprint Identification for Configurations of Any Number
of Minutiae. J. Forensic Sci., 52, 5464.
N EUMANN , C., C HAMPOD , C., P UCH -S OLIS , R., E GLI , N., A NTHONIOZ , A., M EUWLY, D. & B ROMAGE G RIFFITHS , A. (2006) Computation of Likelihood Ratios in Fingerprint Identification for Configurations
of Three Minutiae. J. Forensic Sci., 51, 112.
N UFFIELD C OUNCIL ON B IOETHICS (2007) The Forensic Use of Bioinformation: Ethical Issues. Nuffield
Council on Bioethics, London.
45
PANKANTI , S., P RABHAKAR , S. & JAIN , A. K. (2002) On the Individuality of Fingerprints. IEEE Transactions on PAMI, 24, 10101025.
People v. Caradine (2008) SCN 196416 & 200050 (Super Ct. Cal. Cty. San Francisco), on file with the author.
People v. Jennings (1911) 96 N.E. 1077 (Ill.).
People v. Kelly (1976) 549 P.2d 1240 (Cal.).
People v. Leahy (1994) 882 P.2d 321 (Cal.).
P ETERSON , P. E., D REYFUS , C. B., G ISCHE , M. R., H OLLARS , M., ROBERTS , M. A., RUTH , R. M.,
W EBSTER , H. M. & S OLTIS , G. L. (2009) Latent Prints: A Perspective on the State of the Science. Forensic Science Communications. 11, http://www.fbi.gov/hq/lab/fsc/current/review/2009 10 review01.htm.
R ISINGER , D. M. (2000) Navigating Expert Reliability: Are Criminal Standards of Certainty Being Left on
the Dock? Alb. L. Rev., 64, 99152.
R ISINGER , D. M. (2007) Goodbye to All That, or a Fools Errand, by One of the Fools: How I Stopped
Worrying about Court Responses to Handwriting Identification (and Forensic Science in General) and
Learned to Love Misinterpretations of Kumho Tire v. Carmichael. Tulsa L. Rev., 43, 447475.
R ISINGER , D. M. (2009) Legal Rules for Forensic Science Evidence, Forensic Science for the 21st Century.
Arizona State University.
R ISINGER , D. M., D ENBEAUX , M. & S AKS , M. J. (1998) Brave New Post-Daubert WorldA Reply to
Professor Moenssens. Seton Hall L. Rev., 29, 405490.
ROBERTSON , B. W. N. (1990) Fingerprints, Relevance and Admissibility. New Zealand Recent Law Review,
2, 252258.
ROZELLE , S. D. (2007) Daubert, Schmaubert: Criminal Defendants and the Short End of the Science Stick.
Tulsa L. Rev., 43, 597607.
S AKS , M. (1998) Merlin and Solomon: Lessons from the Laws Formative Encounters with Forensic Identification Science. Hastings L.J., 49, 10691141.
S AKS , M. J. (2000) Banishing Ipse Dixit: The Impact of Kumho Tire on Forensic Identification Science.
Washington and Lee Law Review, 57, 879900.
S AKS , M. J. & FAIGMAN , D. L. (2008) Failed Forensics: How Forensic Science Lost Its Way and How It
Might Yet Find It. Annu. Rev. Law Soc. Sci., 4, 149171.
S AKS , M. J. & KOEHLER , J. J. (2008) The Individualization Fallacy in Forensic Science Evidence. Vand. L.
Rev., 61, 199219.
S CHWARTZ , A. (1997) A Dogma of Empiricism Revisited: Daubert v. Merrell Dow Pharmaceuticals, Inc.
and the Need to Resurrect the Philosophical Insight of Frye v. United States. Harv. J. Law & Tech., 10,
149237.
S CHWARTZ , A. (2005) A Systemic Challenge to the Reliability and Admissibility of Firearms and Toolmark
Identification. Colum. Sci. & Tech. L. Rev., 6, 142.
S CHWINGHAMMER , K. (2005) Fingerprint Identification: How The Gold Standard Of Evidence Could Be
Worth Its Weight. Am. J. Crim. L., 32, 265289.
S CIENTIFIC W ORKING G ROUP ON F RICTION R IDGE A NALYSIS S TUDY AND T ECHNOLOGY (SWGFAST)
(2002) Friction Ridge Examination Methodology for Latent Print Examiners. ver. 1.01, Aug. 22, SWGFAST,
http://www.swgfast.org/Friction Ridge Examination Methodology for Latent Print Examiners 1.01.pdf.
(2003a) GlossaryConsolidated. ver. 1.0, Sept. 9, SWGFAST, http://www.swgfast.org/Glossary
Consolidated ver 1.pdf.
(2003b) Standards for Conclusions. ver. 1.0, Sept. 11, SWGFAST, http://www.swgfast.org/ Standards for
Conclusions ver 1 0.pdf.
46
S. A. COLE
(2008) Draft for CommentGlossary. ver. 1.0, Sept. 12, SWGFAST.
(2009) Position Statement Regarding NAS Report. SWGFAST.
S IEGEL , D. M., ACREE , M. A., B RADLEY, R., C OLE , S. A., FAIGMAN , D. L., F IENBERG , S. E.,
G IANNELLI , P., H ABER , L., H ABER , R., K ENNEDY, D., M NOOKIN , J. L., M ORENO , J. A.,
M ORIARTY, J. C., R ISINGER , D. M., VOKEY, J. R. & Z ABELL , S. L. (2006) The Reliability of Latent Print Individualization: Brief of Amici Curiae submitted on Behalf of Scientists and Scholars by The
New England Innocence Project, Commonwealth v. Patterson. Crim. L. Bull., 42, 2151.
S TARRS , J. E. (1999) Judicial Control Over Scientific Supermen: Fingerprint Experts and Others Who Exceed
the Bounds. Crim. L. Bull., 35, 234276.
State v. Hull (2008) No. 48-CR-07-2336 (Minn. D. Ct. Cty. of Mille Lacs).
State v. Kim (2004) Nos. 03-S-1196, 1197 (Hillsborough Super. Ct. N.H.), http://www.oninonin.com/fp/NH
Daubert Kim 2004.pdf.
State v. Pope (2008) 07CR62135 (Super. Ct. Bibb Cty. Geo.), on file with the author.
State v. Rose (2007) Case No. K06-0545 (Cir. Ct. Baltimore Cty. Md.).
State v. Sheehan (2008) Case No. 061908535 (D. Cty. Salt Lake Utah), on file with the author.
State v. Sullivan (2005) No. 03S 1635, et al. (N.H.).
S TONEY, D. A. (1997) Fingerprint Identification: Scientific Status. In: Faigman, D.L., Kaye, D.H., Saks, M.J.,
Sanders, J. (Eds.), Modern Scientific Evidence: The Law and Science of Expert Testimony. 1st ed. West,
St. Paul, pp. 5578.
S TONEY, D. A. (2001) Measurement of Fingerprint Individuality. In: Lee, H.C., Gaensslen, R.E. (Eds.), Advances in Fingerprint Technology. CRC Press, Boca Raton, Fla., pp. 327387.
TAMBER , C. (2009) Fingerprints Are Back in as Federal Judge Finds Them Reliable, Daily Record, Maryland.
T HOMPSON , W. C. & C OLE , S. A. (2007) Psychological Aspects of Forensic Identification Evidence. In:
Costanzo, M., Krauss, D., Pezdek, K. (Eds.), Psychological Testimony for the Courts. Erlbaum, Mahwah,
NJ.
T HORNTON , J. I. & P ETERSON , J. L. (2002) The General Assumptions and Rationale of Forensic Identification. In: Faigman, D.L., Kaye, D.H., Saks, M.J., Sanders, J. (Eds.), Science in the Law: Forensic Science
Issues. West, St. Paul, pp. 145.
United States v. Crisp (2003) 324 F.3d. 261 (4th Cir.).
United States v. Havvard (2000) 117 F.Supp.2d 848 (S.D. Ind.).
United States v. Havvard (2001) 260 F.3d 597 (7th Cir.).
United States v. Llera Plaza (2002) 188 F.Supp. 2d 549 (E. D. Pa.).
United States v. Mikhel et al. (2006) CR 02-220 (B) - NM (C.D. Cal.), on file with the author.
United States v. Mitchell (1999a) Cr. No. 96407 Memorandum of Law in Support of Mr. Mitchells Motion to
Exclude the Governments Fingerprint Evidence (E.D. Pa.).
United States v. Mitchell (1999b) Cr. No. 96407 Tr. trans. (E.D. Pa.).
United States v. Mitchell (2004) 365 F.3d 215 (3d Cir.).
WAYMAN , J. L. (2000) When Bad Science Leads to Good Law: The Disturbing Irony of the Daubert Hearing
in the Case of U.S. v. Byron C. Mitchell, Biometrics in the Human Services User Group Newsletter, 4, Jan.
W OODWORTH , F. (1997) A Printer Looks at Fingerprints. The Match!, Winter, pp. 3639.
Z ABELL , S. L. (2005) Fingerprint Evidence. Journal of Law and Policy, 13, 143170.

Who Speaks For Science

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Who Speaks For Science

Uploaded by

Copyright:

Available Formats

Law, Probability and Risk (2010) 9, 2546

Advance Access publication on December 16, 2009

Who speaks for science? A response to the National Academy of

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

WHO SPEAKS FOR SCIENCE?

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

WHO SPEAKS FOR SCIENCE?

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

WHO SPEAKS FOR SCIENCE?

3 People v. Kelly (1976).

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

THE COURT: Okay (Testimony of Simon Cole, People v. Caradine, 2008).5

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

WHO SPEAKS FOR SCIENCE?

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

3. Who speaks for forensic science?

4. Speaking for evidence

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

WHO SPEAKS FOR SCIENCE?

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

WHO SPEAKS FOR SCIENCE?

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

WHO SPEAKS FOR SCIENCE?

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

WHO SPEAKS FOR SCIENCE?

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

Conflict of interest disclosure

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

B LACK , B. (2003) Focus on Science, Not Checklists. Trial, 39, 26.

WHO SPEAKS FOR SCIENCE?

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

WHO SPEAKS FOR SCIENCE?

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

Downloaded from http://lpr.oxfordjournals.org/ by guest on December 6, 2016

You might also like