Public Interest Law ReporterVolume 11Issue 2 Summer 2006 Symposium Issue Article 6

2006

The Street, the Lab, the Courtroom, the MeetingRoomJames M. DoyleCenter of Modern Forensic Practice, John Jay College of Criminal Justice, City University of New York

Steven Penrod Ph.D.Center of Modern Forensic Practice, John Jay College of Criminal Justice, City University of New York

Margaret Bull Kovera Ph.D.Center of Modern Forensic Practice, John Jay College of Criminal Justice, City University of New York

Jennifer Dysart Ph.D.Center of Modern Forensic Practice, John Jay College of Criminal Justice, City University of New York

Follow this and additional works at: http://lawecommons.luc.edu/pilr

Part of the Courts Commons, Criminal Law Commons, and the State and Local GovernmentLaw Commons

This Article is brought to you for free and open access by LAW eCommons. It has been accepted for inclusion in Public Interest Law Reporter by anauthorized administrator of LAW eCommons. For more information, please contact law-library@luc.edu.

Recommended CitationJames M. Doyle, , Steven PenrodPh.D., , Margaret B. KoveraPh.D., & Jennifer DysartPh.D., The Street, the Lab, the Courtroom, theMeeting Room, 11 Pub. Interest L. Rptr. 13 (2006).Available at: http://lawecommons.luc.edu/pilr/vol11/iss2/6

SYMPOSIUM ISSUEThe Street, The Lab, The Courtroom,The Meeting Room

James M. Doyle', Steven Penrod, Ph.D.,Margaret Bull Kovera, Ph.D. and JenniferDysart, Ph.D.

The Mecklenburg Report 2 documenting the resultsof the Illinois Pilot Program on Sequential Double-Blind Identification Procedures will frustrate the mostviolent partisans on both sides of the debate over thefuture of eyewitness investigations.

Sadly, the Mecklenburg Report will also disappointa broad audience of practitioners in the middle thathoped for guidance-for something to do (or avoiddoing) to minimize the number of imprisoned innocentsand untouched criminals that the DNA exonerationcases warn us eyewitness memory can produce. TheMecklenburg Report represents a taxing effort, itsauthor and the participating officers and departmentsdeserve our thanks, but the Report does not succeedin combining the perspectives ofthe street investigator,the laboratory scientist and the courtroom litigator intoa working synthesis. In the Report's aftermath, it isclearer than ever that all three perspectives, theirpotentials and their limitations must be recognizedbefore there can be a basis for action informed byscience.

If the Mecklenburg Report convinces the criminaljustice system's practitioners-investigators,prosecutors, defenders and judges-that they cannotwait around for legislatures to act, but must getthemselves to the table together, engage the scientists,and work to find answers, then it can be a positivecontribution. But until that happens, the MecklenburgReport will leave us not far from where we were whenthe National Institute ofJustice issued its path-breakingEyewitness Evidence: A Guide for LawEnforcement3 ("the Guide") in 1999. We still have asubstantial body of laboratory results arguing for the

procedures that the Guide identified as good (pre-line-up instructions), or good but not preferred (sequentiallineup display), or simply potential (double-blind)options. We now have a number of satisfied

jurisdictions around the country (including New Jersey,

Boston and Minneapolis) that have instituted double-

13 | Public Interest Law Reporter

blind, sequential routines with apparent success. Even

so, the Guide did place the burden of proving the

superiority of those procedures in operation on their

advocates, and the Mecklenburg Report's numbers

certainly do not lighten that burden. In fact, at least on

the surface, the Report's numbers seem to tend in the

opposite direction: against innovation. But if we look

beneath the surface, we find that even ifthe numerical

results noted in the Mecklenburg Report's field studyhad been reversed the Report still would not have

proved the superiority of double-blind sequential

procedures. The Mecklenburg Report reveals a studythat simply was not set up to test under scientific controleither double-blind, or sequential procedures. Nor

did it test scientifically any differences between blindand not-blind simultaneous procedures.

These gaps are doubly unfortunate because five

years from now we will not be handling eyewitnessidentifications in the same way that we handle themnow. Our arrays of dog-eared mug shots and hastilyimprovised station house lineups are certain to be

supplemented-and are almost certainly doomed-by a digital revolution that promises us quick, cheap,convenient and comprehensively documentedidentification procedures. We will have-somedepartments now have-photo lineup capability onlaptops in squad cars. The capacity to present photo-arrays on Palm Pilots has already been studied in thelabs. There soon will be many alternatives to draggingvictims to the precinct house in the middle of the nightand hiring line-up fillers from the homeless shelters inorder to test witness memory.

But to say we will have new equipment doesn't tellus what we should do with it. What should we showthe witnesses on our laptops? "Sequential" displays?How should we show it? With "double-blind"techniques? What is the best procedure for the future?The Mecklenburg Report does not really answer thesequestions; in fact, the study it recounts does not reallyask them.

In hindsight it is clear that the Mecklenburg Reportreveals a crippling misunderstanding at the heart ofthefield study it describes.

The Illinois Legislature issued a directive to pursuea specific goal: compare a traditional technique ofeyewitness evidence gathering (the "simultaneous"

(Doyle, continued on page 14)

Summer 2006

1

Doyle et al.: The Street, the Lab, the Courtroom, the Meeting Room

Published by LAW eCommons, 2006

SYMPOSIUM ISSUE(Doyle, continued from page 13)

display of suspect and fillers by an officer who is awareofwhich lineup member is the suspect) with a proposedimprovement (the "sequential" display of suspect andfillers by a "double-blind" administrator). But theMecklenburg Report is pervaded by an unexplaineddetermination to treat the Legislature's statement of agoal as if it dictated a method. The failure toacknowledge the distinction between goal and methodaffected not only the Report on the study, but the designof the study itself, and it imposed serious handicaps.The muddle of method and goal explains why, in thewords of United States Attorney Patrick Fitzgerald,the study "raises more questions than it answers."-

In faimess to the Illinois Legislature, it did what itcould to signal that the studies of eyewitnessprocedures that should be conducted were not thecrude "traditional v. double-blind sequential" test theMecklenburg Report describes. The Legislature soughtan empirical answer to an empirical question by theuse of study instruments: "[D]esigned to elicitinformation for comparative evaluation purposes, and... consistent with objective scientific researchmethodology."' An appropriate objective scientificresearch methodology exists, but the MecklenburgReport shows plainly that the study it discusses stoppedshort of applying that methodology.

Four psychologists are mentioned prominently inthe Mecklenburg Report. Two, (Dr. Nancy Steblayand Dr. Gary Wells) are bitterly critical of the reportand allege that their participation is exaggerated bythe Report's author; two (Dr. Ebbe Ebbesen and Dr.Roy Malpass) consider the Mecklenburg Report onthe field study to be a valuable document. But allfour of these scientists disclaim any responsibility fordesigning the study. In fact, the study, like the Report,is the product of a single hard-working lawyer for theChicago Police without formal training in social sciencemethods. The differences between legal and socialscience practice show. For example, the Reportdescribes as "random" assignment methods which inthe legal world might be accepted as meriting the term,but which no social scientist would recognize as truerandom assignment. The result of this absence of ascience-based design is that the Mecklenburg Reportforfeits lessons that a truly scientific approach might

have taught. Because the requirements of scientificmethodology were not imposed in the design of thestudy, we now know much less than we could.

To begin with, a comparison of the new "double-blind sequential" photo arrays with traditional"simultaneous, not-blind" lineups, which merely laysthe two side-by-side, could never have beeninformative in any scientific way. To properly assessthe "sequential" photo-arrays against "simultaneous"procedures either both "simultaneous" and "sequential"would have to be "not-blind," or both would have tobe double-blind-only then could we gauge whichfactor was creating the effect we see. To properlyweigh the impact of "double-blind" procedures bothsimultaneous and sequential procedures would haveto be run in "double-blind" and"not-blind" conditions

The Mecklenburg Report speaks formany when it suggests that we expandour inquiries and address further ques-tions. But, real improvements in justicesystem processes based on science willonly occur if cops, prosecuors and de-fenders take responsibility for framingthe right questions informed from thebeginning by scientific advice.

before the impact of "blindness" on investigations couldbe assessed. In other words, a fatal "confound" isbuilt into the design ofthe Report, making it impossibleas a matter ofmethod, to retrieve authoritative answersto the question the Legislature posed.

Besides, treating the question posed by theLegislature as a methodological directive while ignoringthe Legislature's wish that "scientific methodologies"govern the study hopelessly entangled the operationalissues of what is feasible on the street or in the precinctwith the reliability research issue of whether the newprocedures are worth doing in the first place.

For operational purposes it was natural for theReport to use "suspect hits" as a proxy for "correctidentifications." A radical decline in "suspect hits" in

(Doyle, continued on page 15)

Summer 2006 Public interest Law Reporter 114Summer 2006 Public Interest Law Reporter 114

2

Public Interest Law Reporter, Vol. 11, Iss. 2 [2006], Art. 6

http://lawecommons.luc.edu/pilr/vol11/iss2/6

SYMPOSIUM ISSUE(Doyle, continued from page 14)

double-blind sequential procedures would indicate atleast in a rough and ready way a very awkwarddisconnect between that particular identificationtechnique and police field operations. But it is importantto confine the "suspect hit" criteria to its operationalsignificance. Obviously, ifwe were satisfied with everyprocedure that yields a 100 percent rate of "suspecthits" we would never have undertaken the enterprisein the first place; we would have simply agreed toregard the dozens of DNA exonerations (every one ofthem based on a "suspect hit" which seemed"corroborated") and the dozens of active criminals whoescaped justice in those cases as an inevitable cost ofdoing business. "Suspect hits" can tell us importantthings about operations, but-even in the HennepinCounty field study, where the results were radicallydifferent from those recited in two of the threeMecklenburg Reportjurisdictions -they tell us verylittle about the reliability questions at the heart of theissue of procedural superiority.

Lab methods have their own limitations, and thereis a danger ofunintentionally imposing those limitationsif we undertake "lab-like" studies in the field.

The specific limitation that concerns us here is notthe worry that in the real world crime situation humanmemory operates in a qualitatively different way. Thereis no evidence for that fear. In studies pre-dating theReport, the rate of"filler ID's" in the lab and in the realworld seemed to match up fairly closely. In theoccasional hyper-realistic laboratory study, such as Dr.Charles Morgan's controlled study of special forcestroops who were asked to identify their interrogatorsafter a high stress interrogation (more than half identifieda "filler" in conventional simultaneous arrays) the results,again, are consistent with both the more conventionallab setting experiments and with the scattered fieldresults from the United States and the UnitedKingdom.' But even while we acknowledge that thelab studies and the field studies are examining the sameprocesses of human memory, we have to rememberthat they do so in different contexts.

The Mecklenburg Report's most intriguing resultsare its account of a "zero" rate of filler identificationsin two ofthree jurisdictions. These results are uniqueamong existing studies in the lab and in the field. How

did this happen? What does it mean? Unfortunately,the design of the study and the Mecklenburg Report'srecounting of it leave open a quite simple and obviousexplanation: the failure to account for a fundamentaldifference between lab life and street life.

When the lab scientists study the efficacy of anidentification technique, the single "simultaneous" or"sequential" test they scrutinize is almost always thewitness'sfirst attempt at the identification ofa stranger-perpetrator. This places the focus on the most influential(and therefore dangerous) encounter, but it does notautomatically duplicate typical real-life practice; in reallife a witness's live line-up performance is only oneepisode in the witness's career in the criminal justicesystem. In real life-and in the experience of thewitnesses depicted in the Mecklenburg Report-a"live" lineup experience can be (in most places, usuallyis) preceded by a show-up, a "drive-by", or by aphoto-array. To treat the rate of suspect identificationsattained in first attempts in laboratories and thirdattempts in the field (by witnesses who were, in effect,pre-tested on a show-up and a photo-array) asequivalent doesn'tjust compare apples to oranges; itcompares apples to automobiles. It isn't particularlysurprising if third attempts by pre-tested witnesses (i.e.,after two successful attempts and the dismissal of allof the unsuccessful witnesses) to identify a suspect inthe field lead to fewer "filler ID's."

We don't know from the Mecklenburg Report thatthis happened, but unfortunately we can't know that itdidn 't happen, because the witnesses' history in theinvestigation is not recorded or reported. TheMecklenburg Report treats the field results as if theywere the lab results, but the study under examinationdid not follow the scientific tradition of recordingexperimental data, and so it failed to capture data thatthe lab would have noted as a matter of routineexperimental design. How many of these eyewitnessesidentified fillers in initial field procedures? How manyof these were filtered out of the process before thesubsequent, reported lineups? How many witnessesin the subsequent lineups were performing aconfirming recognition task following a successfulsuspect identification in a show-up or an array? Eitherof these features is at least as likely to have affectedthe suspect/filler identification rates as might wholesale

(Doyle, continued on page 16)

Summer 200615 Public Interest Law Reporter3

Doyle et al.: The Street, the Lab, the Courtroom, the Meeting Room

Published by LAW eCommons, 2006

(Doyle, continued from page 15)

police "tipping" ofwitnesses by police. (Although theMecklenburg Report persistently invokes the strawman of sinister allegations ofpolice misconduct, in factno one claims either that intentional police misconductis the problem in the DNA eyewitness exonerationcases, or that if it were the problem, procedural reformswould be a silver-bullet solution to all intentional"framing" of suspects.) If people are cheating, theywill continue to cheat, whatever procedures areadopted. But, as things now stand we are providedwith no authoritative refutation of dark speculationsabout of police "tipping" because a wide range ofdata points-for example, the number of "lowconfidence" filler identifications and the number offailures to identify-were not captured in the study'sdesign.

Operations and reliability are muddled in a differentway when the "double-blind" technique is at issue. If"double-blind" procedures add something to accuracy,then implementing double-blind procedures-asjurisdictions in Wisconsin, Massachusetts, New Jerseyand elsewhere have done-becomes a question ofpolice ingenuity, commitment and leadership insurmounting operational challenges. But, if we askthe police their opinion of the "double-blind" approachbefore they are persuaded that it can contribute toaccuracy, the police can't be blamed for acceptinginconvenience and unfamiliarity as sufficient answer.Besides, when the police are not invited to participatein the design ofthe specific local double-blind sequentialtechnique but are simply presented with a "take it orleave it" version in informal oral training an opportunity

The Mecklenburg Report represents ataxing effort, its author and the partici-pating officers and departments deserveour thanks, but the Report does not suc-ceed in combining the perspectives ofthe street investigator, the laboratoryscientist and the courtroom litigator intoa working synthesis.

SYMPOSIUM ISSUEto confuse the performance is created and anopportunity to exploit police expertise is lost.

The Mecklenburg Report speaks for many when itsuggests that we expand our inquiries and addressfurther questions. But, real improvement injusticesystem processes based on science will only occur ifthe cops, prosecutors and defenders take responsibilityfor framing the right questions informed from thebeginning by scientific advice. The system'spractitioners not only have to take responsibility forintegrating science into practice; they have to takeresponsibility for doing it together.

This will cause some discomfort. The MecklenburgReport documents a field study that followed the morenormal course ofreform efforts within the system: Oneactor or another is charged with (or pro-activelyassumes) responsibility for mobilizing one scientificadvance or another, chooses its own scientists, closelyholds the information developed and makes (orforegoes) reforms. This is not the only way.

Behind their adversarial routines, all criminal justicepractitioners share a common enemy-the innocentdefendant. No one wants the innocent in the system.The police do not want to waste their time on theinnocent while the guilty go free to prey on new victims;the prosecutors realize that highly publicizedexonerations in the cases they should have lost willlater cost them the cases they should win. Maybeyoung defense lawyers go to law school with dreamsof defending the innocent, but experienced defenselawyers see defending an innocent-particularly in aneyewitness case-as a nightmare. Double-blind,sequential lineup procedures-ifthey work to keepthe innocent out of the system-are to everyone'sadvantage, and they should get a genuine scientific testfor that reason alone.

It is also worth remembering that the question ofeyewitness identification reform is not an all-or-nothingmatter. Sophisticated police departments might, aftertesting, decide that some crime situations (for example,where there is a substantial amount of corroboratinginformation) call for traditional methods ofidentification, while other, shakier, cases call for themore cautious, conservative double-blind sequentialapproach. "Double-blind" administration on its own(even if "simultaneous") also serves important law

(Doyle, continued on page 31)

Summer 2006 Public Interest Law Reporter 16Public Interest Law Reporter 16Summer 2006

4

Public Interest Law Reporter, Vol. 11, Iss. 2 [2006], Art. 6

http://lawecommons.luc.edu/pilr/vol11/iss2/6

SYMPOSIUM ISSUEGetting to Truth Before It Falls intothe Hands of the Lawyers: PursuingAccuracy in Criminal Cases

James B. Zagel'

Societies have always wrestled with the overallquestion of the reliability of witnesses and, even now,when the legal rules are mostly settled, we still worryabout perjury, mistakes, delusions and the integrity ofmemory. At issue today is the relatively small subset ofthe witness problem-just that one moment when thewitness points to one person and says that is the per-son whose conduct I have described. For most wit-nesses to crime, the phrase "That's the man" is short-est part of the story they offer. Concern about its ac-curacy has been with us for centuries.

If solutions were easily found, this would not bean age old concern. Be wary of those who, withgreat confidence, offer the miracle cure to a problemwe all recognize. The results of the Illinois double-blind eyewitness pilot program offer a vivid exampleof why what some think obvious is often not so.There is another point here; failure teaches as muchor more than does success and we ought not to

turn our back on any enterprise that seeks to make

our investigations and adjudications better. The greatvalue of tests, like the one we discuss herein, is theykeep us from a terrible kind of optimism that, once

disappointed, can lead us to abandoning the search

for something better.

What Is It We Are Trying To Repair andWhy?

Our world of arrest, prosecution and defense haschanged. The idea that truth arises out of trial in anadversary system is still with us but mostly in theorynot practice. Plea bargains are the dominant mode.

My colleague, Judge Lynch in New York, hasaccurately described the process this way:

"[T]he prosecutor ... is the central adjudicator

of facts ... arbiter of ... legal issues and of the

appropriate sentence to be imposed. Potential

defenses are presented by the defendant ... to a

prosecutor, who assesses their factual accuracy and

then decides the charge of which the defendant

should be adjudged guilty..."'This is a far cry from what the Supreme Court

envisioned when it began to emphasize 75 yearsago that defense counsel at trial was essential togetting at the truth.' And the place, they thought,where truth was to be found was trial where it wasjudge and jury, after hearing prosecution and

defense, not prosecutor, who decided the outcome.This reliance on defense counsel to help us get to

the truth was a key element in the first cases in whichthe Supreme Court sought to bring constitutional

regulation to eyewitness identification.'But defenders are not duty bound to see that the

truth comes out. Ifthe client is guilty they are obliged

It is critical to establish an identification before the eyewit-ness is subject to outside suggestion and a fhise identifica-tion is made. A false identification often has ireversableconsequences.

to use all legally permissible means to see that thetruth does not come out. This became particularlyclear when we thought about what a lawyer shoulddo at a lineup. Suppose the client tells his lawyer"Yeah, I stole the stuff but I'm sure no one saw meinside." Then the lawyer sees his client in a proposedline-up of seven, six of whom are Hispanic, and hisclient is the only blond white man in the group. Does

(Zagel, continued on page 18)

Summer 200617 1 Public Interest Law Reporter5

Doyle et al.: The Street, the Lab, the Courtroom, the Meeting Room

Published by LAW eCommons, 2006

(Zagel, continued from page 17)

counsel want a fairer or more accurate lineup, onemore likely to elicit the truth? If he gets one and theidentification is made, he has deprived his client ofa good argument at trial. What if the police ask thelawyer for suggestions to improve the lineup andagree that they will follow these suggestions? Doesthe lawyer improve the process to the detriment ofhis client? The dilemma here is stark because, unlikeinterrogation, the lawyer can not simply advise hisclient not to participate in the lineup. Theidentification procedure is going to happen. Thelawyer is not authorized to decide simply that it isright to have a fair identification parade; the lawyeris only authorized to seek the kind of parade that isgood for the client.

Many eventually accepted this state of affairswhere, in a trial, getting to the truth was not thesingle overriding value.' I think they did so for tworeasons. First, the thought was that, in nearly allcases, the truth came out anyway. Second, therewere important social values found in proceduralfairness and in giving the defendant a meaningful rolein his or her defense. The price of an occasionalcriminal going free was thought to be worth payingto achieve these good things. This tradeoff hasalways been controversial. It might not survive apublic referendum.

The tradeoff also rests on premises thatprofessionals find hard to accept. It is not easy tofind scholars (though not so hard to find judges)who actually believed that trial was really a goodway to get at the truth in hard cases; the scholarlydefense of the system was based upon its serviceto other democratic values. And even where theadversarial system could work, it was dependenton having a skilled, adequately funded advocate onboth sides of the case. This last condition was oftenunmet.

In the decades that followed the criminalprocedure upheavals of the 1960s, there was alukewarm to cool acceptance of the way criminalcases were handled: lukewarm to cool because wewere in the midst of a rise in crime that lasted fordecades and, only relatively recently, subsided;accepted because there was much in popular media,

Summer 2006 Public Interest Law Reporter 118

Summer 2006

SYMPOSIUM ISSUEshows like NYPD Blue and Law and Order, thatportrayed a system that got the right result. The rightresult is the common result, but it is not because ofour system of trials. It is because, in most cases,there is no serious question of guilt. The evidence isusually more than good enough and, if it isn't, theprosecutors frequently won't take a chance on thecase. Perfection is unattainable, but getting it rightin the largest percentage of cases is not good enougheven if that percentage is in the high nineties. Theconsequences of error are too grave.

But still we toddled along with what we had. Thevolume of criminal cases, which was associated withthe rise in crime and the relatively poorly fundeddefense services, led to fewer trials and what JudgeLynch called an administrative system of criminaljustice. We would be that way today but for recentevents.

The public, as opposed to the defenders of theworld, was generally worried only about the guiltygoing free. New science and today's news madethem worry, at least a little, about the innocent beingfound guilty.

For this reason the importance of finding the truthabout guilt is valued more highly today than it hasbeen in many years. We might be in the midst of atectonic shift in perspective about crime,investigation and the accused.

Some of this might seem strange to say to thepublic. Haven't we always thought that the end ofthe criminal justice system was to find out the truth?Most people did, but no one who labors in this fieldbelieves that truth is always revealed or acted upon.There are unjustified convictions and unjustifiedacquittals. Even under the better practices that wewill someday have, we will never reach perfectionbecause the truth is elusive, often beyond the abilityof humans to discover. But we won't stopprosecuting. Crime has a devastating effect on itsvictims and a large effect on the society in whichwe live. We have never decided to leave the guiltyor the innocent to the judgment of heaven. We oughtthen to do the best we can to lock up all the guiltyand free all the innocent, knowing that we willsometimes fail.

(Zagel, continued on page 19)

Public Interest Law Reporter 1 18

6

Public Interest Law Reporter, Vol. 11, Iss. 2 [2006], Art. 6

http://lawecommons.luc.edu/pilr/vol11/iss2/6

SYMPOSIUM ISSUE(Zagel, continued from page 18)

What Has Led Us to Try Now?The short answer to this question is that there is no

sudden change, just a gaudy tipping point. The shift toconcern for truth has been building for a while.

Start with Miranda v. Arizona6 and the de-em-phasis on admissions of guilt. Miranda can now beread as an implied endorsement of the reliability ofeyewitnesses and, perhaps, of the very forensic evi-dence that is now under attack, say, bite marks, andeven of reliable evidence excludible under the FourthAmendment. The Court understood it was holding thateven voluntary confessions, whose truth value was un-questioned, were to be excluded from evidence. Thisde-emphasis of truth and re-emphasis on proceduralprotections did not last very long. The Supreme Courtlimited the scope of Miranda.' The Court (in an opin-ion written by Justice Thurgood Marshall) also de-cided to permit the police to use deception to induceconfessions' and narrowed the scope of the exclu-sionary rules based on the Fourth Amendment.9

'A-

The idea of science in law enforcement is relatively new.Identification by fingerprinting, br example, was not broadlyaccepted in the United States until the 1920s.

Despite this renewed endorsement of the value ofconfessions, there is no doubt that the Supreme Court,and every sane person for that matter, would preferthat guilt be determined by incontestable evidence likethat found three times every week in New York, Mi-ami and Las Vegas by an infallible corps of Crime SceneInvestigators. It is science that created that gaudy tip-ping point.

The entry of science into the courtroom started ata very slow pace. Most of the earliest expert testi-mony from doctors and alienists was admitted into

19 | Public Interest Law Reporter

evidence because the law permitted anybody to offeropinion evidence.

Fingerprints are a good example of the early evi-dence of experts. Fingerprints were used for identifi-cation in India in the 1850s.1o Written work appearedin 1881 and Galton's book was published in Englandin 1892.11 By 1910, fingerprinting itself was in fairlywide police use.12 Despite this, it was not until 1911that a reviewing court approved its use, but broadacceptance did not come until the 1920s, and it wasnot until the 1940s that courts said the prosecutionwould no longer have to prove that no two finger-prints are alike.13 The course of admission of otherforms of identification evidence was similar. It tooktime to get the courts to approve comparative micro-graphy, microanalysis, questioned documents. Mostof the first scientific evidence dealt with traces andmarks which the jurors themselves could perceive-friction ridges, striations on bullets and so forth. As thetwentieth century went on, and science itselfbegan todeal with things not directly observable, the law beganto take in serology, general chemistry and neutron ac-tivation analysis. In all these cases, though, one reasonthe courts moved slowly was the resistance of de-fense counsel to the admission of such evidence be-cause it rarely served any purpose other than to in-criminate their clients.

DNA evidence was accepted with amazing speedprecisely because it could exonerate as well as incrimi-nate. There were very few to fight tooth and nail againstits admissibility because the prosecutor or defenderwho objected vigorously to DNA evidence knew that,in the next case, they might be offering that same evi-dence. DNA, too, came to the courts at a time whenstandards of what constituted reliable and valid sci-ence had become clearer. DNA analysis had the ad-vantage of service as a tool in many sciences, not merelycriminalistics. The broader use ofDNA analysis meantthe discipline had been critically reviewed by manymore scientists than, say, fingerprints.

It is true that the advent of closerjudicial scrutinyof expert witnesses in recent years14 has called intoquestion much ofthe science that is offered in the court-room, but the outcome of disputes about questioneddocument examination, serology, fiber analysis,

(Zagel, continued on page 20)

Summer 2006

7

Doyle et al.: The Street, the Lab, the Courtroom, the Meeting Room

Published by LAW eCommons, 2006

SYMPOSIUM ISSUE(Zagel, continued from page 19)

(maybe, even fingerprints) are not for today's discus-sion. DNA survived its first tests and it proved thatthere were a small, but nonetheless very meaningful,number ofwrongful convictions. DNA science teachesthere is, in a certain class of cases, a class of evidencethat, if properly handled, is conclusive. Of course,properly handled is not inevitable, but it put forth animplicit message. We can get to the truth better thanwe had gotten to it before. And this is why pilot pro-grams are undertaken: not because eyewitnesses canbecome the legal equivalent ofrecombinant strands ofDNA, but because DNA showed that we can do bet-ter. Obviously only if we try to do better.

There will be attention paid to the causes ofwrongfulconvictions, now the predominant element is bad eye-witness identifications. So too, we will look at errorrates in particular kinds of cases." These studies mighthelp us in our scrutiny of past errors.

Professional investigators of crime have known forhundreds of years that evidence can lead to the wrongperson, that some eyewitness identifications are worth-less, that some confessions are worthless and that someforensic analyses are worthless. From the perspectiveof the police and prosecutors, the solution was eithernot to charge in those cases or to drop the charges ifalready brought. The defense often proposed this so-lution. The problem was handled in house and therewas always the final safeguard of the trial.

What DNA told the public is that a trial does notprotect adequately against these errors if the prosecu-tor decides to go forward with the case. DNA put anenormous dent into the idea that the adversary systemis the best way to protect against false convictions.The belief that even good faith errors made earlier inthe process will be detected and repaired as the casemoved through our adversary system has lost some ofits hold on our society. While the adversary systemmight serve many social values apart from its detect-ing the truth of accusations, all these values collec-tively seem no longer to outweigh the risk of that thejudgments it produces may be untrustworthy. It is notthat these values are to be disregarded, it is the degreeto which they are fostered that is questioned. When acrime victim sees the perpetrator unjustly acquitted, itis small consolation to tell them that it is better that ten

guilty men go free than one innocent man be convicted,but the society as a whole accepts, or at least under-stands, this policy. Now the message is, not only dowe let those ten guilty go free by the way, we alsosend quite a few innocent men to prison too. Thatmessage is not well received.

So the turn now is to making things right before thelawyers in the adversary system get their hands on it.

What Can We Expect from Science?Science proceeds by evaluating ideas, theories,

guesses, conjectures, hopes and dreams. It does thisby experiments of all sorts and observations.

The idea that there is science in law enforcement isnot so old. Early criminology was a form of moral phi-losophy as in Cesare Beccaria.'6 One of its first scien-tists was an anthropologist Cesare Lombroso,17 whodied less than one hundred years ago. Many of histheses seem laughable today but he used the inductivemethod of science as well as he could. Our under-standing of criminology is still very much in flux. Weseem to know that more police officers and morepeople in prisons are good at reducing crime rates.Some ofthe standard explanations about strong econo-mies, too many people under thirty, order maintenancepolicing, strong gun laws, capital punishment, chang-ing drug habits and markets all remain unproved." Butwe are not talking about criminology here, it iscriminalistics or police science which largely concernsitself with helping to find out, by examination ofphysi-cal objects, who did what, when and how.

In recent times, I have noted, some accepted po-lice science has come under question. This is notunique to police science. The scientific enterprise isfilled with failure and mistakes. There is a well-knownmaxim offered to some first-year medical studentswhich runs this way: "half ofwhat we teach you will bewrong, we just don't know which half."

Mistakes in science are not limited to earlier centu-ries like the phlogistan theory of fire. The theory is alaughing stock today but it was clever in its time. Theinventor believed there was a combustible substance-phlogistan-consumed by combustion which requiredair. After the phlogistan was gone the residue weighedless than the original product as is demonstrated bythe case of ashes which weigh less than the burnt log.

(Zagel, continued on page 32)

Summer 2006 Public Interest Law Reporter 20

Summer 2006 Public Interest Lawv Reporter' 20

8

Public Interest Law Reporter, Vol. 11, Iss. 2 [2006], Art. 6

http://lawecommons.luc.edu/pilr/vol11/iss2/6