Brain Fingerprinting - Harrington Testimony Analysis

Brain Fingerprinting Laboratories

Brief Analysis of Testimony by Dr. Lawrence A. Farwell, Dr. William Iacono, and Dr. Emanuel Donchin Regarding Brain Fingerprinting, the Daubert and Frye Standards, and Related Issues Including Relevant Portions of the Transcript of the Harrington vs. State hearing on Nov. 14, 2000

Lawrence A. Farwell, PhD

November 14, 2000

Harrington vs. State of Iowa

Pottawattamie County District Court

"Transcript" refers to the transcript of the November 14, 2000 Harrington vs. State hearing. The portions of the transcript referred to in this analysis are reproduced below. "Report" refers to Exhibit 1, Dr. Farwell's report on the Harrington Brain Fingerprinting test, HarringtonTechReport005Exhibit1.doc, "Supplement" refers to Exhibit 2, the supplement to that report, HarringtonSupplement003figsExhibit2.doc. Both were received as exhibits in the hearing.

A. Dr. Lawrence Farwell

1. The science of Brain Fingerprinting/P300 has been tested.

Dr. Farwell

Dr. Farwell stated that the science has been tested.
Report

p. 7, 11-13

Appendix 3 (Dr. Farwell CV)

Appendix 6 (Journal of Forensic Sciences publication)

Appendix 7 (US Patent)

Supplement

p. 20-29

Transcript

(References to transcript are of the form "page, line - page, line.")

6, 4 - 15, 23

28,8 - 29, 25

45,14 - 49, 3

2. The science of Brain Fingerprinting/P300 has been peer reviewed and published.

Dr. Farwell

Dr. Farwell stated that the science has been peer reviewed and published.

Report

p. 7, 11-13

Appendix 3 (Dr. Farwell CV)

Appendix 6 (Journal of Forensic Sciences publication)

Appendix 7 (US Patent)

Supplement

p. 19-29

Transcript

6, 4 -15, 23

19, 14 - 19, 18

28,8 - 29, 25

45,14 - 49, 3

3. The science of Brain Fingerprinting/P300 is accurate.

Dr. Farwell

Dr. Farwell stated that the science is accurate.

Report

p. 7, 11-13

Appendix 3 (Dr. Farwell CV)

Appendix 6 (Journal of Forensic Sciences publication)

Appendix 7 (US Patent)

Supplement

p. 19-29

Transcript

6, 4 -15, 23

19, 14 - 19, 18

22, 19 -23, 24

28, 8 - 29, 25

45,14 - 49, 3

56, 17 - 57, 2

57, 21 -57, 25

4. The science of Brain Fingerprinting/P300 is generally accepted in the scientific community.

Dr. Farwell

Dr. Farwell stated that the science is generally accepted in the scientific community.

Report

p. 7, 11-13

Appendix 3 (Dr. Farwell CV)

Appendix 6 (Journal of Forensic Sciences publication)

Appendix 7 (US Patent)

Supplement

p. 20-29

Transcript

6, 4 -15, 23

19, 14 - 19, 18

28, 8 - 29, 25

45,14 - 49, 3

Dr. Farwell also stated the following:

The Brain Fingerprinting tests showed that the record of the evening of the crime stored in Harrington's brain did not match the crime scene and did match his alibi.

Report

p. 17 - 20

Supplement

16 - 18

Transcript

35, 12 - 44, 8

( 44, 1 - 44, 8 contains a very brief summary statement regarding the crime.)

45, 14 - 48, 7

49, 9 - 52, 23

The Brain Fingerprinting results obtained provide relevant information bearing on the question of whether Harrington committed the crime.

Report

p. 15 - 17

Supplement

p. 13 - 16

p. 31 - 35

Transcript

40, 4 - 44, 7

Brain Fingerprinting has been tested extensively on information stored in the brain regarding real-life situations, as well as information learned in the course of laboratory simulations, and the results of these tests on information regarding real-life events have been accurate, peer reviewed, published, and accepted in the scientific community.

Report

p. 11 - 12

Supplement

p. 28 - 31

Transcript

7, 13 - 9, 12

10, 3 - 10, 20

11, 25 -14, 8

This science was unavailable at the time of the original trial.

Transcript

31, 19 - 32, 2

Brain Fingerprinting does not test guilt or innocence, but only whether certain information is stored in the brain.

Supplement

p. 31 - 32

Transcript

26, 9 - 26, 23

56, 6 - 13

98, 24 - 99, 3

141, 11 - 14

Brain Fingerprinting is to be used in conjunction with other methods of determining the facts of the case.

Transcript

26, 24 - 27, 11

141, 15 - 22

B. Dr. William Iacono

1. The science of Brain Fingerprinting/P300 has been tested.

Dr. Iacono

Dr. Iacono stated that the science has been tested.

Transcript

152, 7 - 152, 10

156, 11 - 157, 2

157, 23 - 158, 10

160, 2 - 161, 18

165, 21 - 166, 9

168, 13 -168, 25

170, 21 - 174, 25

179, 11 - 179, 21; 180, 5 - 180, 9

183, 14 - 183, 22

2. The science of Brain Fingerprinting/P300 has been peer reviewed and published.

Dr. Iacono

Dr. Iacono stated that the science has been peer reviewed and published.

Transcript

156, 11 - 157, 2

157, 23 - 158, 10

160, 2 - 161, 18

165, 21 - 166, 9

168, 13 -168, 25

170, 21 - 174, 25

179, 11 - 179, 21; 180, 5 - 180, 9

183, 14 - 183, 22

3. The science of Brain Fingerprinting/P300 is accurate.

Dr. Iacono

Dr. Iacono stated that the science is accurate.

Transcript

168, 18 - 170, 20

170, 21 - 174, 25

179, 11 - 179, 21; 180, 5 - 180, 9

184, 19 - 185, 2

4. The science of Brain Fingerprinting/P300 is generally accepted in the scientific community.

Dr. Iacono

Dr. Iacono stated that the science is generally accepted in the scientific community.

Transcript

156, 11 - 157, 2

157, 23 - 158, 10

160, 2 - 161, 18

165, 21 - 166, 9

168, 13 -168, 17 Here is a direct quote on the point of general acceptance from p. 168:

13 Q So in your view, the technique described by Dr. Farwell,

14 which is in your view an application of the guilty knowledge

15 test, does this have a high degree of acceptance in the

16 scientific community?

17 A In my view, it does, yes.

170, 21 - 174, 25

179, 11 - 179, 21; 180, 5 - 180, 9

183, 14 - 183, 22

Dr. Iacono further stated the following:
Dr. Farwell is an expert.

Transcript

188, 7 - 188, 11

This science was unavailable at the time of the original trial.

Transcript

187, 15 - 187, 20

The results of Farwell's Brain Fingerprinting test provide relevant information to the question of whether Harrington committed the crime.

Transcript

185, 24 - 187, 14

C. Dr. Emanuel Donchin

1, 2, and 4. The science of Brain Fingerprinting/P300 has been tested, peer reviewed, and published, and is generally accepted in the scientific community.

Dr. Donchin

Dr. Donchin stated that the science has been tested, peer reviewed and published, and is generally accepted in the scientific community:

a) The Farwell and Donchin 1991 study, including experiment 1 on a laboratory mock crime and experiment 2 on real-life events, constitutes a test of the science. Dr. Donchin stated that it has been peer reviewed and published, and is well accepted in the scientific community:

210, 21 - 211, 8

b) Dr. Donchin stated that Dr. Farwell used essentially the same procedures in the Harrington case as in the Farwell and Donchin study (Dr. Farwell said the same thing):

220, 2 - 220, 14

3. The science of Brain Fingerprinting/P300 is accurate.

Dr. Donchin

Dr. Donchin did not address the issue of accuracy directly in his testimony. He did, however, testify that he had published the Farwell and Donchin paper with Dr. Farwell, and did not dispute Dr. Farwell's testimony regarding the accuracy of the results reported therein.

Dr. Donchin also stated the following:

Dr. Donchin could not agree or disagree with the conclusions reached by Dr. Farwell, because he had not prepared to the extent Dr. Farwell had regarding this particular test, and did not know enough about the specific case.

203, 1 - 203, 4

209, 22 - 25

Dr. Donchin agreed with Dr. Farwell and Dr. Iacono that if there is no P300 in response to the probes, the subject does not recognize them. If there is a P300 in response to the probes, the subject recognizes them.

216, 23 - 217, 1

Dr. Donchin agreed with Dr. Farwell and Dr. Iacono 1) that the science of Brain Fingerprinting can determine whether or not someone recognizes the probes, and 2) that the answer to legal questions of guilt or innocence goes beyond the science.

217, 24 - 218, 3

Dr. Donchin stated that in the Farwell and Donchin study there was skill or art involved in applying the science (in selecting irrelevants in experiment 1 and probes in experiment 2), and that this art component of the scientific process had been accepted by the scientific community. He stated that the art or skill component was "just like fingerprints." Dr. Donchin stated that he could not criticize the probes Dr. Farwell had used in the Harrington case because he did not know enough about the case.

209, 4 - 209, 6

222, 19 - 223, 22

Dr. Donchin stated that Dr. Farwell had a history of skillfully developing the probes, as evidenced in the peer-reviewed and published Farwell and Donchin 1991 study.

225, 24 - 226, 3

1 IN THE DISTRICT COURT OF IOWA IN AND FOR POTTAWATTAMIE COUNTY

2 -------------------------------------------------------------

3 TERRY J. HARRINGTON, :

4 Plaintiff, : No. PCCV073247

5 -vs- :

6 STATE OF IOWA, : TRANSCRIPT OF PROCEEDINGS

7 Defendant. :

8 -------------------------------------------------------------

9 Before Hon. Timothy O'Grady, Judge,

at Council Bluffs, Iowa, held on

10 November 14, 2000, and November 15, 2000.

11 APPEARANCES:

12 Ms. Mary Kennedy and

Mr. Tom Frerichs

13 Attorneys at Law

209 West Fifth Street

14 Waterloo, Iowa

15 Appearing for Plaintiff.

16 Mr. Richard Crowl

Pottawattamie County Attorney

17 227 South Sixth Street

Council Bluffs, Iowa

Appearing for Defendant.

- - -

Page 1

1 TABLE OF CONTENTS

2 WITNESSES: DIRECT CROSS REDIRECT

3 Lawrence A. Farwell 3 54 139

William Iacono 148 189 199

4 Emanuel Donchin 203 210 ---

5 EXHIBITS: OFFERED RECEIVED

6 Ex 1 - Dr. Farwell Report 28 148

Ex 2 - Supplemental Report 35

7 Ex 3 - Dr. Iacono CV 148 148

Ex 4 - FBI Memo 120 ---

8 Ex 5 - Washington Post Article

Ex 6 - Des Moines Register Article

9 Ex 7 - Farwell Transcript (3-10-95)

Ex 8 -

10 Ex 9 - Dr. Donchin CV 203 203

Page 2

1 THE COURT: This case is styled Terry J. Harrington

2 versus the State of Iowa. Today we have a hearing -- the

3 first part of the trial, actually, on the Post Conviction

4 Petition, and also on the State's Motion to Dismiss and

5 Motion in Limine. I think we have an understanding that we

6 will proceed with the scientific evidence over the next day

7 or so.

8 Mr. Crowl, are you ready?

9 MR. CROWL: Yes, Your Honor.

10 THE COURT: Miss Kennedy, Mr. Frerichs, are you

11 ready?

12 MS. KENNEDY: Yes, Your Honor, we are.

13 THE COURT: Your first witness, please.

14 MS. KENNEDY: I call Lawrence Farwell.

15 THE COURT: Dr. Farwell, if you would step forward,

16 please.

17 LAWRENCE A. FARWELL

18 being produced as a witness on behalf of the plaintiff,

19 and having been first duly sworn by the Court, was examined

20 and testified as follows:

21 DIRECT EXAMINATION

22 By Ms. Kennedy:

23 Q Would you state your name, please, and spell it for the

24 record.

25 A Lawrence Ashley Farwell; L-a-w-r-e-n-c-e, A-s-h-l-e-y,

Page 3

1 F-a-r-w-e-l-l.

2 Q And Dr. Farwell, what is your business address?

3 A 107 South Maple Street, Fairfield, Iowa.

4 Q Would you please tell the Court about your educational

5 background, briefly.

6 A I have an undergraduate degree from Harvard University.

7 I have a Master's and a Ph.D. in Biological Psychology from

8 the University of Illinois.

9 Q Your current position?

10 A I'm Director and Chief Scientist of the Human Brain

11 Research Laboratory.

12 Q What other professional employment have you had prior to

13 your current occupation?

14 A Since receiving my Ph.D. in 1992, our actually beginning

15 before I had my Ph.D., I was consultant to the Central

16 Intelligence Agency on brain research and its use in

17 detecting concealed information. I was also, for a couple of

18 years -- It's listed in my C.V., but I was a member of the

19 faculty of Harvard University. I was a research associate in

20 the Department of Psychiatry of the Harvard Medical School.

21 Q Do you consider yourself a scientist?

22 A That's correct.

23 Q And do you specialize in any particular field?

24 A Yes. I specialize in cognitive psychophysiology. That

25 is what my training is in and that's the field that I have

Page 4

1 applied in this case.

2 Q Would you define psychophysiology, please; briefly, if

3 you can?

4 A I can. Psychophysiology is the measurement of

5 physiological responses with the aim of drawing conclusions

6 regarding mental or psychological processes. Cognitive

7 psychophysiology, specifically, deals with information

8 processing in the brain, detecting that through physical

9 measures.

10 Q Okay. And your current -- in your current position, as

11 psychophysiologist, are you focusing on a certain line of

12 work or a certain area of specialization?

13 A Yes. And I have been since I first started working for

14 the CIA in 1991. And that is specifically what has come to

15 be known as brain fingerprinting. There has been some

16 evolution of technology, but over the years, it's been

17 essentially the same practice, and that is detecting

18 concealed information in the human brain in order determine

19 to whether an individual has or doesn't have particular

20 information in his brain, often with forensic applications.

21 Q In this brain fingerprinting area, have you done

22 research, published papers, anything like that?

23 A Yes, I have.

24 Q What papers have you published?

25 A I have published -- Well, first of all, the details are

Page 5

1 in my C.V., but in terms of the major papers that I have

2 published, the ones perhaps most pertinent here, there was a

3 Farwell and Donchin study in 1991 -- I'm also -- that

4 accepted for publication. It will come out in January, a

5 paper in Journal of Forensic Sciences, which I'm publishing

6 with Sharon Smith of the FBI on research that we did there.

7 Q These papers focus specifically on brain fingerprinting?

8 A That's correct. The terminology was a little bit

9 different in the first paper, but the science was the same.

10 Q Who developed or who invented brain fingerprinting as

11 you are working with it today?

12 A I did. Now, that's -- That is really too brief a

13 description to say that I invented it, because it's based on

14 science which has been developed by my peers, and colleagues

15 over a period of decades.

16 Q If you then invented the brain fingerprinting, you are

17 saying that although you invented it, it's based on science,

18 accepted science from other scientists?

19 A Right. Basically, what I measure is event related brain

20 potentials. Those are specific patterns of brain activity

21 that indicates that a certain cognitive activity has taken

22 place. By measuring those event related brain potentials, we

23 can tell what information a person recognizes or doesn't

24 recognize. Now, the specific brain wave component that I

25 began with when I first -- or I should say brain wave pattern

Page 6

1 that I began with when I started this, first developed this

2 invention, was the P300. That was discovered in 1965 by

3 Sutton, et al.

4 Q Let me stop you right there. The P300, then, is what?

5 Explain that, please.

6 A It is an event related brain potential, and what an

7 event related brain potential is, is a specific pattern of

8 brain wave activity that is related to an event. That's why

9 we call it event related. The event that a P300 is related

10 to is the event of taking note of something significant

11 called context updating. This has been published in hundreds

12 of studies in many different labs throughout the world.

13 Q These studies were all centered around the P300;

14 correct?

15 A Correct.

16 Q And you used the P300 in your brain fingerprinting test?

17 A That's correct. Initially I started with the P300, and

18 what happened over time, is that I discovered that there was

19 a little bit more going on after the P300 was over. That

20 happened, by the way, throughout the course of development of

21 cognitive psychophysiology and event related potential.

22 Initially, papers were looking at short potentials that only

23 had to do with sensory processing, and started to look at

24 things that took place a little longer after the stimulus,

25 which had to do with cognitive processing. What I have done

Page 7

1 is extended out beyond the P300 and looked at the pattern

2 that takes place a little farther out in time. However, in

3 order to conduct this brain fingerprinting, and to get the

4 results that we have gotten with Terry Harrington, all I

5 really need is the P300. I can add frosting to that cake, if

6 I want to, and go into more detail, look at the brain waves a

7 little bit farther after the stimulus, but I don't need to do

8 that in order to get the results. So fundamentally, this is

9 based on the P300. The original invention was as well.

10 Q Aside from the papers that you briefly discussed, have

11 you done any research on test subjects or anything like that

12 on your brain fingerprinting?

13 A I have conducted about 170 tests, and of those,

14 approximately 100 were on real life information; not

15 information that was acquired in the laboratory, but actually

16 regarding real life events.

17 Q What was one of those studies, please? Could you

18 describe one of those studies?

19 A Yes. The first study that I did, which actually later

20 became a part of that Farwell and Donchin 1991 paper, was a

21 study of detecting information about minor crimes or socially

22 undesirable acts in the lives of college students.

23 Basically, what we did is we got four college students, we

24 went with their permission, to their roommates, and asked

25 them about some situation that they had done that was either

Page 8

1 a crime or some socially undesirable act. We didn't tell the

2 person we were testing what that situation was that we were

3 looking for. And we developed a set of words or phrases,

4 some of which were relevant to that particular event, and

5 others weren't. I can describe in more detail, but in any

6 case, some of these phrases or words were relevant to that,

7 and others weren't. We presented those on a computer screen

8 mixed in with other irrelevant items, and we could tell by

9 the brain response by this P300 which ones the subjects

10 recognized. So what happened is that we found that we were

11 able to detect that these four undergraduates had

12 participated in these four specific real life events. We

13 also ran four tests on these same subjects on events that

14 they hadn't participated in. So each subject, we tested on

15 two things; one was an event that he or she had participated

16 in, the other was a event that he or she had not participated

17 in. In the cases where they participated in the event, we

18 got a P300 in response to the items that were flashed on the

19 screen that were relevant to the event. Where they hadn't

20 participated, they didn't get a P300, and this is what you

21 predict from the literature, because the P300 indicates that

22 you take note of something.

23 Q Let me stop you there and ask you to slow down a little

24 bit for the court reporter.

25 A All right.

Page 9

1 Q He looks a little nervous to me.

2 A If I inadvertently lapse into technical jargon, please

3 stop me and I will say it in plain English.

4 Q Speaking of plain English, what you just described for

5 the Court, the test that you ran on the college student, was

6 in conjunction with the paper that you talked about earlier,

7 the 1991 paper?

8 A That's correct. Actually, it was originally published

9 in 1986. The original study on four students was published,

10 or eight tests on students, we published in 1986. Another

11 publication on essentially the same content was one of the

12 two experiments in the 1991 paper.

13 Q And the paper was done with whom?

14 A Dr. Emanuel Donchin, my advisor in my Ph.D. program at

15 the University of Illinois.

16 Q When you worked with Dr. Donchin, what was his

17 specialty? What did you work with him about, specifically?

18 A About the P300. When I first had this idea of

19 developing this test, it was on the basis of a lot that I had

20 learned from Dr. Donchin, because he was my advisor.

21 Actually, he wasn't my advisor at that time. Someone else

22 was my advisor, but he was the head of the lab and probably

23 the world's leading expert on 300. When I first came up with

24 this idea, after I had seen it work, I took it to

25 Dr. Donchin. He said, "Great, let's go with this." And then

Page 10

1 we did further research together on it. So in answer in your

2 question, his specialty is the 300.

3 Q To cut to the heart of the study again, I'm asking you

4 to simplify it way down. You asked college students

5 questions about events that you had researched; correct,

6 events in their life?

7 A Well, actually it's not quite correct that I asked them

8 questions. What we did is to present words or pictures on a

9 screen, flash them on a screen, computer screen, and some of

10 these words or pictures were relevant to the events in

11 question, some weren't. It's a little more complicated than

12 that, but that is the basics of it.

13 Q And the words and pictures, whatever, that you flashed

14 on the screen, what did that tell you about those college

15 students?

16 A What it told us, if they got a P300, in response to the

17 items that were specific to that crime or that event, then we

18 knew they recognized those items. That's what a P300 means.

19 It means you recognize something as significant. You take

20 note of it. So that's what that indicated.

21 Q Okay. Other than that particular test, did you have any

22 other tests or do any other research in the field?

23 A Yes. In that same experiment, we did a test on 40

24 subjects. I should say 20 subjects in each of two

25 conditions. One condition was they had participated in an

Page 11

1 event. The other condition was they hadn't participated in

2 an event. This was a laboratory event where it was actually

3 a laboratory mock crime experiment, and Dr. Donchin got

4 funding from the CIA to the tune of about $80,000 to do that

5 particular research. We did it together. I actually ran the

6 subjects -- I wrote the software to acquire the data, and

7 acquired the data. I wrote the software to analyze the data,

8 and I analyzed the data. And the results were that we were

9 able to tell which event the individual had participated in,

10 because when they recognized the details about those events,

11 on the computer screen they got a P300. We defined that

12 mathematically, we measured it, we came up with a

13 determination and a statistical confidence for that

14 determination. It's not just that we look at lines on a

15 graph. We mathematically analyze and the data analysis

16 algorithm says either yes, the information is there, or no,

17 it isn't. Or it can come up saying we just don't have enough

18 information to make a clear determination, and that's an

19 indeterminate response. And that happened in 12 and a half

20 percent of the cases in that Farwell and Donchin 1991 study.

21 We never got a wrong answer before or since, but in that

22 case, we did have 12 and a half percent we couldn't come up

23 with a clear answer one way or the other. So that's simply

24 listed as an indeterminate.

25 Q Any other studies that -- high profile studies or

Page 12

1 anything else that you have done with the P300 or with your

2 brain fingerprinting?

3 A Yes. There is another study that I am publishing in the

4 Journal of Forensic Sciences with Sharon Smith of the FBI.

5 She is an instructor at the FBI Academy in Quantico. This

6 was very similar to the study that we did initially on the

7 students. Again, we had pairs of individuals, and in each

8 pair, one of the individuals had committed some kind of an

9 activity. These weren't crimes. They were just real life

10 activities that they had engaged in, and one of them had not.

11 In each case as before, we had one person who we were testing

12 for knowledge of an event, and then we got that information

13 from somebody else, so we recruit a subject and somebody who

14 knew that subject, and we had permission then to find out

15 about that person's past from the other person. The person

16 who knew the subject would tell us, "Okay, here is a

17 particular event that the person participated in some years

18 ago, and here are the details about that event." We used

19 those details, and what we found was that in every case, when

20 we flashed those particular details mixed in with the

21 irrelevant items on a screen, the individual would recognize

22 them, they would emit a P300 and we could tell with a high

23 statistical confidence which event people participated in,

24 and which ones they hadn't.

25 Q To break it down, what would you consider the scientific

Page 13

1 basis for brain fingerprinting? Is it the P300?

2 A Yes, it is. Now, I have gone beyond that, and again, in

3 addition to the studies I just mentioned, I did three studies

4 under contract for the CIA in which we did the same kind of

5 research. I did another study with the FBI in which we did

6 the same kind of research in the FBI study. We were

7 detecting who was an FBI agent and who wasn't by detecting

8 information in the brain only an FBI agent would recognize.

9 The CIA studies, one was in collaboration with the Navy, and

10 we were detecting who had information about Naval medicine,

11 who had that particular expertise, the application of those

12 experiments. If we can detect an FBI agent or Naval medical

13 expert, we could detect a KGB agent, or what have you. There

14 was another study which was a mock crime which we did for the

15 CIA, my colleagues and I, and then there was another study

16 for the CIA which had, instead of smaller number of subjects,

17 it had 30 subjects, and again, we detected real life events

18 in the lives of people which happened to include a couple of

19 crimes, but specifically what we were looking for is can we

20 tell whether a person has participated in a particular event

21 or not, and what we found again, is that we could in all of

22 those CIA studies, we had 100 percent accurate results. We

23 got the correct answer. We got a definite answer in every

24 case, and the answers in every case were correct, and this is

25 not from my looking at the wave forms. This is from the

Page 14

1 mathematical analysis which says either information present,

2 or information absent, and it gives us a statistical

3 confidence for that individual determination.

4 So yes, the scientific basis is the P300, although I

5 have added to that. But to repeat, when I tested Harrington,

6 I conducted data analysis, including only the P300, the same

7 thing that I had done in the Farwell and Donchin studies,

8 same thing that I had done in previous studies.

9 Q Well, we can get to the Harrington study later. You

10 mentioned -- You used the phrase information present,

11 information absent?

12 A Right.

13 Q Again, is that another way of describing your brain

14 fingerprinting test --

15 A Sure.

16 Q -- for us non scientists?

17 A Sure. That's a way of describing the result of the

18 test.

19 Q Okay. Now, probably you have had lots of people that

20 want to draw an analogy between your brain fingerprinting

21 test and a polygraph. Is what you are doing telling whether

22 or not somebody is lying?

23 A No, it is not; has nothing to do with whether they are

24 lying or not. In fact, you get the exact same results with

25 brain fingerprinting whether the person is lying or telling

Page 15

1 the truth.

2 Q How is that?

3 A We are just detecting whether the information is there

4 in their brain. It's as if -- Say the DNA, they have DNA

5 from a crime scene and on the person of the suspect, or you

6 have fingerprints at the crime scene, you have fingerprints

7 on the fingers of the suspect. If those match, it doesn't

8 matter what he says about it. What we are doing is detecting

9 a match, or no match, between information stored in the brain

10 and information that we get from the crime scene, or relevant

11 to the crime. If it matches, it matches. If it doesn't, it

12 doesn't; has nothing to do with what the person says.

13 Q How much subjectivity is involved in looking at the

14 results, interpreting the results of the brain fingerprinting

15 test?

16 A None at all. You could push the button. I could tell

17 you which button to push. When it says analyze the data, you

18 would push the button, the system would analyze the data and

19 come up with a mathematical determination. Now, you can see

20 the results in the plots, but that's not the basis on which

21 we make the determination. We make it mathematically, and we

22 use standard mathematical techniques for doing that that are

23 well published and accepted in the statistical literature.

24 Q Your system of brain fingerprinting, you described

25 earlier in your testimony how you flash pictures or phrases

Page 16

1 or words?

2 A Correct.

3 Q What are the differences between these words or phrases

4 that you use?

5 A All right. We use three types of stimuli or words or

6 pictures. Usually we use words. One type of stimulus is

7 called a target. These are items that we know the person

8 knows, we are sure the person knows, because we have talked

9 to them about them. We give them a list of them. In many

10 cases, these are details, for example, about a crime that we

11 know he knows, because he has been told them in a trial or

12 because of information that he has some independent access

13 to. We are sure he knows them. We give him a list of them,

14 and we say, "When one of these comes up on the screen, you

15 push a special button." We are very certain he is going to

16 recognize and take note of those. And those are going to

17 emit a P300. You recognize them, he takes note of them. We

18 know we will get a P300 from those, also a MERMER extending

19 out beyond, extending out into a longer time window in

20 looking at the brain wave data from a P300, which will be

21 clearer when I have showed the plots. So we have that

22 stimulus type called targets where we know he is going to

23 recognize them, we know we will get a P300 and a MERMER. A

24 second type of stimuli are irrelevants. They have nothing to

25 do with the crime, they have nothing to do with the person in

Page 17

1 particular. We know he is not going to recognize those as

2 relevant, and we know we are not going to get that P300 or

3 MERMER response. So those are the two types, sort of set a

4 standard, here is what his brain wave looks like when he

5 recognizes the stimulus, here is what it looks like when it's

6 irrelevant to him, or here is what it looks like when he has

7 that information in his brain, here is what it looks like

8 when he doesn't have that response.

9 Then we present, mixed in with these others, a third

10 type of stimulus. We don't tell him which ones these are.

11 These are called probes, and these are stimuli that are

12 relevant to the situation we are investigating, say the

13 crime, but that he would have no way of knowing, unless he

14 did it. So if he shows a P300, and a MERMER in response to

15 these probes, that provides evidence that he actually has

16 that information about the crime stored in his brain, because

17 those are significant to him. If he doesn't have the

18 information stored in his brain, then he won't even know

19 which ones are the probes to him. They will look like

20 irrelevants, and the response will look like the irrelevants

21 response. It will not have a P300 in them. It won't have a

22 MERMER.

23 Q You keep mentioning MERMER. Would you explain briefly

24 what the MERMERS are as compared to the P300?

25 A Yes. The P300 is a positive response. By positive, I

Page 18

1 mean electrically positive component, maximal at the parietal

2 midline area of the brain, the top of the back of the head.

3 And it indicates that a person is taking note of the

4 stimulus. Now, what happened during my research initially in

5 the FBI study where we were detecting FBI agents, I noticed

6 after this positive response, there was a negative deflection

7 that seemed to take place consistently, whenever that

8 positive response happened. It's been found in other labs as

9 well, at least in one of the labs as well at that time, and

10 more since, I believe. And I started to think, well, maybe

11 there is more to this than we thought. So the positive

12 deflection of the P300 followed by a negative deflection --

13 By a negative, I mean negative voltage on the scalp,

14 constitutes the MERMER. I also am convinced that there are

15 phasic changes, short term changes in the frequency of

16 signals that take place that don't show up in the signal

17 averaging that we use to get event related brain potentials,

18 and these are interesting scientific ideas, and I have some

19 data on that, but basically the MERMER is frosting on the

20 cake, of which the P300 is the cake.

21 Q Let me stop you here. Could you do these same tests

22 using just the P300 and not extending it out to the MERMER?

23 A Yes, I could, and I have.

24 Q When did you -- If you can remember, when did you start

25 actually using the MERMER in addition to using the P300 for

Page 19

1 your tests?

2 A I started doing that in 1993 with the FBI study. But

3 the CIA studies specified the P300, so the results are

4 reported to the CIA. All three of those studies were only in

5 terms of the P300. For my own edification, I also analyzed

6 the MERMER. So I analyzed both those studies both ways. In

7 every case, what I got is I got the same result with the P300

8 or with the full MERMER, but in the MERMER, the statistical

9 confidence would increase, because -- in most cases, because

10 I'm just including more data.

11 Q Now, the MERMER is something you developed yourself; is

12 that correct?

13 A That's correct.

14 Q As you mentioned earlier, the P300 has been subject to

15 peer review for years, and basically accepted by the

16 scientists in the field; is that right?

17 A P300 has been accepted for decades by all the scientists

18 in the field, that's correct.

19 Q How about the MERMER?

20 A The MERMER has been published by myself, or is being

21 published by myself and Sharon Smith at the FBI in this

22 article in the Journal of Forensic Sciences in January of

23 this coming year. However, it is a very new thing and

24 doesn't have anything remotely approaching the level of

25 scientific acceptance the P300 has.

Page 20

1 Q I'm going to back up up just a little bit. You talked

2 about the three types of stimulus that you use on your

3 subjects, and you equated it to crime scenes and so forth,

4 but all your studies used the same three types of stimuli;

5 correct?

6 A All my studies use the same three type of stimuli. All

7 of them detected information stored in the brain. Now, this

8 could be information about a real life event, as it was in

9 about 100 of the 170 cases I used; it could be information

10 about a laboratory event, as it was in about 70 of those

11 cases; it could be information about a crime. Information in

12 the brain is information in the brain.

13 Q Now, let me ask you some questions that occurred to me

14 as a lay person, and not a scientist. You tailor the

15 stimulus to what your subject remembers or doesn't remember;

16 correct, or what is stored in the brain, or they tell you

17 that this is what I know, and this is what I don't know;

18 correct?

19 A Well, that is correct. However, before we even get to

20 that point, before I may even ever meet the subject, I

21 ascertain independently what they know and don't know insofar

22 as possible.

23 Q Okay. Let's take, example, if you are working on a

24 criminal case.

25 A Right.

Page 21

1 Q You would have to discuss with your subject at some

2 point in time what this person knew about the crime; correct?

3 A Correct.

4 Q Now, a question that comes to me, and I'm sure comes to

5 everybody, is what if the person doesn't remember. What if

6 they have read it or know it, but don't remember it, will

7 your test still show the same results?

8 A I don't really understand the question. Are you talking

9 about the probe stimuli? Are you talking about the targets?

10 Q I'm talking about the target stimuli; for instance,

11 something that happened 10 years ago, and the person doesn't

12 remember to tell you they don't remember the information to

13 give you or to tell you?

14 A Well, in the target stimuli, what we do is we take

15 information that we are sure they know, and we make sure they

16 know it, because we give them a list of the targets, and we

17 say, whenever one of these comes up, you need to push a

18 special button. You are sure you can recognize? Yeah, I'm

19 sure I can recognize. When they come up, push a special

20 button. So we know they are going to recognize the targets.

21 We know they will get an appropriate brain response, a P300,

22 to the targets.

23 Q What about the probes? The probes are what they claim

24 they don't know, the subject claims?

25 A That's correct. What we do when we are structuring the

Page 22

1 probes, is we examine the information we have about the

2 situation, say about the crime, and we come up with things

3 that the subject has never been told, say, in courtroom

4 proceedings or in interrogations that we know about, but

5 would know from having committed the crime, and then we say

6 to the subject -- we sit them down and say, all right, you

7 say you didn't commit the crime. Then if you didn't commit

8 the crime, then you wouldn't know this particular item about

9 the crime. Would you have ever been told this for any other

10 reason? No, I don't know that. And we go through a list of

11 them. We say, would you recognize this or would you have the

12 answer to this question? Would you know this information

13 about a crime? They say no, I wasn't there, so I don't know

14 that.

15 So we verify with the subject that he is claiming not to

16 know any of those items that become probe stimuli before the

17 test. Then what the test tells us is in fact whether that

18 information is stored in his brain or not.

19 Q Would there be any way for somebody to beat the test?

20 A No. Well, put it this way. Let me give an example of

21 where this -- what we are measuring. I think that will

22 clarify. Let's say that that that door opens and an elephant

23 comes into the room. Now, everyone in the room who has

24 eyesight is going to notice that. We now have a picture of

25 what is going on in the room stored in our brains, there is a

Page 23

1 scheme, there is a map of what is going on here of what the

2 physical environment are, what -- the rules, what we can

3 expect stored in our brains. If an elephant comes in the

4 room, we have to change that picture to include some new

5 information, namely an elephant. We have to update our

6 internal picture of where we are and what's going on. It's a

7 basic survival function. We have to do it to survive in a

8 changing environment. So when the elephant comes in the room

9 through the door, we are all going to notice it, and say,

10 uh-huh, yes, okay. Now we have an elephant in the room.

11 After we have noticed it, we can start thinking okay, now

12 what am I going to try to do about it. Maybe I will feed the

13 elephant. Maybe I will exit through the back door. When

14 that brain's response recognizing something significant

15 happened takes place, that's when we pick up the P300, that's

16 when we pick up the information, that's when we get the

17 result. So we get our result when they take note of the

18 elephant, not when they decide what to do about it.

19 When that probe stimulus comes up that is relevant to

20 the crime, the first thing he thinks is uh-huh, yeah, I

21 recognize that. That was the place where the victim was when

22 he was shot, or whatever it is he recognizes, that that is

23 the point at which we pick up our information. Anything else

24 would happen later.

25 Q A couple more things, Dr. Farwell. Are there any

Page 24

1 situations where the brain fingerprinting test would not work

2 on a subject?

3 A I wouldn't say there aren't any situations where it

4 wouldn't work in the sense of not working accurately. But

5 there are situations where we simply can't apply it. It's

6 just like, say, DNA evidence. If there is no DNA at the

7 scene of the crime, we will not be able to try to match the

8 DNA with the suspect at the scene of the crime. If we can't

9 discover any information from the crime that the subject

10 would know, if and only if he's committed the crime. Then we

11 can't use it. Example; say someone disappears, we don't know

12 where, why, we don't know what happened. We don't even know

13 whether a crime has been committed or not. But -- So if we

14 are going to test someone and try and determine whether they

15 committed a crime, we wouldn't know what to test them on.

16 Another difficult situation would be somebody who has already

17 been convicted of a crime, if they know everything that we

18 can find out about the crime, then we couldn't use the test,

19 because we wouldn't have any probe stimuli. What we have to

20 do in a situation where someone has already been convicted,

21 they know most of the things that we could have easily used

22 as probe stimuli if we had gotten into the case early, so we

23 have to be a little -- do a little more digging, find out

24 things that they don't know or claim not to know, but would

25 know if they had committed the crime. So yes, there are

Page 25

1 cases when it can't be applied and that's simply when we

2 don't have enough information or don't have sufficient

3 information that the person would know if he had committed

4 the crime, but that he wouldn't know if he hadn't committed

5 the crime.

6 Q Now, we are going to discuss the test you did on Terry

7 Harrington, but before that, have you done any other tests in

8 the field on anybody who has been convicted of a crime or any

9 defendants?

10 A Yes. All right. Yes. I did one test relevant to an

11 internal investigation in the Alexandria Police Department

12 and that did not go to trial. What I found was that

13 according to the brain fingerprinting, the individual had not

14 committed the crime in question, and it appears from

15 subsequent evidence that that was correct. Another case in

16 which I used it on a defendant was about a year ago, year and

17 a half ago, J. B. Grinder, in the case of the murder of Julie

18 Helton. This was in Macon, Missouri. The sheriff there

19 asked me to come down and see if I could shed some light on

20 the case, because they heard a lot of different stories from

21 this defendant, from other defendants, from other alleged

22 witnesses, everybody was contradicting each other, and this

23 particular defendant contradicted himself in many different

24 stories he told; some of which involved his participation in

25 the crime, others which did not. So I ran a test on James

Page 26

1 Grinder, and brain fingerprinting clearly showed that he had

2 the details of the murder of Julie Helton which had taken

3 place 15 years previously, stored in his brain. Shortly

4 thereafter, he had an opportunity to plead guilty in exchange

5 for life in prison. This was in Missouri where he got life

6 in prison, rather than the death sentence, and he took that

7 option. He later confessed to several other murders of young

8 women.

9 Q Dr. Farwell, can your brain fingerprinting test solve

10 crimes?

11 A The purpose of brain fingerprinting is not to solve a

12 crime. It doesn't tell us who is innocent and who is guilty.

13 What it does is provides us information about what

14 information is stored in a person's brain and what

15 information isn't. If somebody has fingerprints at the crime

16 scene and that matches the fingerprints on the fingers, that

17 gives us information that may be relevant to determining what

18 took place in the crime. But this doesn't tell you who is

19 innocent and who is guilty. It doesn't solve the crime, per

20 se. What it does is provides information about what the

21 person has stored in his brain. If that information is

22 useful for understanding what happened at the crime, then so

23 be it.

24 Q Does the brain fingerprinting test -- is the function to

25 be used in conjunction with other parts of a criminal

Page 27

1 investigation, for instance?

2 A Yes. Like any other test, the purpose of brain

3 fingerprinting is not to make other sources of evidence

4 irrelevant, but rather to contribute to the amount of

5 knowledge that we have that can shed some light on what the

6 truth is in the situation. Of course, it needs to be

7 evaluated in light of all of the other evidence that is

8 available.

9 Q Is the brain fingerprinting test meant to take the place

10 of a fact finder in any case, in a criminal case?

11 A Not at all.

12 Q Dr. Farwell, I'm going to show you a report I have

13 marked Exhibit 1. Would you take a look at that? Is that

14 your report?

15 A Yes, it is.

16 Q And when and about whom did you write that report?

17 A I wrote it May 21, 2000. I wrote it about Terry

18 Harrington.

19 MS. KENNEDY: Your Honor, I would like to offer

20 that as Exhibit 1.

21 MR. CROWL: State objects to Exhibit 1, under the

22 Rules of 402, 403 and 702, and would like to have a standing

23 objection, Your Honor, concerning this exhibit throughout the

24 trial.

25 THE COURT: I understand that your objection goes

Page 28

1 to the heart of the issues that are presented. I will

2 reserve ruling on Exhibit 1 and allow you to have a standing

3 objection.

4 MR. CROWL: Thank you, Your Honor.

5 MS. KENNEDY: Thank you, Your Honor.

6 THE COURT: But we may proceed to use Exhibit 1 in

7 further examination.

8 Q Before we go over the test, or the report, Dr. Farwell,

9 I want to just summarize your previous testimony in that your

10 technology, the brain fingerprinting, are you saying that

11 that has been thoroughly tested by scientists such as

12 yourself?

13 A That's correct. The science, the fundamental science of

14 cognitive psychophysiology event related potentials,

15 specifically the P300, has been tested in many laboratories

16 all around the world and has been very well established

17 science for decades. Dr. Donchin and I published, Sharon

18 Smith of the FBI and I have published, specifically on this

19 technique of using the P300 and event related brain

20 potentials in detecting information in the brain. Other

21 laboratories have also published on this same kind of

22 technology; namely, Bill Iacono.

23 Q Would you say that the science on which your brain

24 fingerprinting test is based is generally accepted in --

25 with -- among scientists such as yourself that are

Page 29

1 psychophysiologists?

2 A Yes. I would say it's -- particularly cognitive

3 psychophysiologists who know the field, the science on which

4 this is based, event related brain potential and the P300

5 components of the brain related potentials, would be probably

6 the most well researched and published phenomenon in all of

7 psychophysiology.

8 Q And we didn't discuss an error rate. What is the error

9 rate in the test that you have conducted with your brain

10 fingerprinting tests?

11 A I have conducted about 70 tests on information that was

12 acquired in a laboratory, about 100 tests on information from

13 real life events. In every case in which a determination has

14 been made, it was a correct determination. There have been

15 no mistakes made. It either gave us information present, or

16 information absent. Those were always correct. In I believe

17 six of those cases, there was an indeterminate result, which

18 means that the data analysis algorithm, simply says we don't

19 have enough information to decide one way or the other. So

20 it doesn't make a determination. That is not an error. It's

21 just a matter of we don't have enough information to make a

22 determination. So all of the determinations that have been

23 made, 100 percent of them have been accurate, and we have

24 been able to make a determination in over 95 percent of the

25 cases.

Page 30

1 Q As you use the brain fingerprinting test more and more,

2 can you foresee that you will fall short of the 100 percent

3 accuracy rate?

4 A There is never -- There is nothing in science that is

5 100 percent and I wouldn't stand up and say brain

6 fingerprinting is an absolute, and it makes everything else

7 irrelevant, because it's so perfect. In science, there is

8 always an uncertainty factor, and that actually is included

9 in our individual determinations. We say we have this

10 determination and we have, say, a 99 percent confidence in

11 that statistically, or a 99.99 percent confidence or what

12 have you. There is always an uncertainty factor and I would

13 not claim that this is something perfect that could never

14 make a mistake.

15 Q Thank you. Moving on to your report. This is a report

16 about Terry Harrington that you conducted the brain

17 fingerprinting test on; correct?

18 A Correct.

19 Q How did you get involved with Terry Harrington?

20 A I got a telephone call from Ann Danaher, who I believe

21 had read about me in the newspaper or television, and she

22 indicated that she knew somebody who was in prison who

23 claimed to be innocent, and she had reason to believe that he

24 might be innocent, and he had indicated to her that he would

25 consider taking such a test. She then -- I said that I would

Page 31

1 be willing to consider giving him such a test, if it turned

2 out to be possible, and she put the two of us in touch, and

3 Harrington told me that he was innocent, had always

4 maintained he was innocent, and would be eager to participate

5 in anything that might help us to determine the truth about

6 that.

7 Q Have you ever been approached by people before, asking

8 you to administer a brain fingerprinting test?

9 A Are you talking about criminal defendants?

10 Q Individuals. Pardon me.

11 A Defendants?

12 Q Individuals, yes.

13 A Yes. I have been approached. You mean before today, or

14 before Harrington?

15 Q Before Mr. Harrington contacted you.

16 A Harrington was actually the first case where somebody

17 who was a criminal defendant contacted me and asked me to run

18 a test on them.

19 Q Prior to your deciding to administer the test to

20 Mr. Harrington, how long before that point would you say you

21 were actually ready, or your technology was ready to

22 administer that test?

23 A Well, I was ready in August of 1999, when I administered

24 the test to James B. Grinder, but that was really the first

25 point at which -- or maybe a period of weeks or months before

Page 32

1 that, was really the first point at which this would have

2 been available to have been applied out in an actual case.

3 Q What specific things did you do prior to -- or did you

4 have to do prior to administering this test or getting ready

5 to administer the test to Mr. Harrington?

6 A Do you mean scientifically, or administratively?

7 Q Administratively. Did you do any research, things like

8 that, as far as his case?

9 A Yes.

10 Q What were those things?

11 A All right. I read the trial transcripts, at least the

12 relevant portions of the trial transcripts. I read relevant

13 portions of the police reports. I visited the crime scene,

14 and I talked to one of the alibi witnesses, because this was

15 for information not about the crime, but about the alibi. So

16 the task that I had there was to come up with information

17 that Harrington would have no way of knowing unless he had

18 committed the crime. So I couldn't use what was in the trial

19 transcripts directly, because he knew all that from the

20 trial, whether he committed the crime or not. So what I had

21 to do then was combine police reports, trial testimony and a

22 visit to the crime scene to find out things that he would

23 have had to have encountered in committing the crime, but

24 that never came out explicitly as such in the testimony

25 itself at the trial. And I was able to do that.

Page 33

1 Q And by doing these things, reading the transcripts,

2 visiting the crime scene, these are the ways you came up with

3 the words and phrases necessary to administer your test?

4 A That's correct.

5 Q What if -- Strike that.

6 The booklet, that's the report, does this include the

7 test results from the tests that you gave Terry Harrington?

8 Does this include the results?

9 A It includes the preliminary results that I had at the

10 time. Now, recall, I'm not being paid for this and I

11 volunteered my time. I paid my own expenses. I have finite

12 resources to devote to this, so when I wrote this report

13 originally, I wrote a brief report. I included the results I

14 had up to that time. I have included -- I have analyzed the

15 data further since that time and I have included those more

16 complete results in the supplement I wrote to the report.

17 MS. KENNEDY: May I approach the witness?

18 THE COURT: You may.

19 Q Dr. Farwell, I'm going to show you what's marked

20 Exhibit 2.

21 A Yes.

22 Q Is that the supplemental report that you just spoke

23 about?

24 A Yes, it is.

25 MS. KENNEDY: Your Honor, I would ask the Court to

Page 34

1 accept Exhibit 2.

2 MR. CROWL: State objects under Rules 402, 403,

3 702. Also, Your Honor, in this particular report, the format

4 is almost -- it's a question and answer format, almost as if

5 it takes the place of his testimony, and I'm going to also

6 object on the basis of hearsay and would ask for a standing

7 objection as well.

8 THE COURT: Certainly an expert can use hearsay in

9 forming his opinions. I will reserve ruling on the relevancy

10 objections, but we can proceed with examination of Exhibit 2.

11 Go ahead, Miss Kennedy.

12 Q Dr. Farwell, would you say then that the original report

13 and the supplement would be a culmination of the results of

14 the tests on Terry Harrington?

15 A Yes. The original report and the supplement are my

16 professional account of the results of that specific test.

17 Q Dr. Farwell, you have got some equipment set up here,

18 and perhaps the best way to show the results would be to show

19 it on the screen.

20 MS. KENNEDY: Your Honor, it is 10:30, if this

21 might be a good time to take a break?

22 THE COURT: Let's go ahead and do that. We will

23 take about 10 minutes, and you can make sure all your gear is

24 set up and ready to go.

25 (Whereupon, a recess was taken, after which the following

Page 35

1 proceedings were had and taken, to-wit:)

2 THE COURT: Dr. Farwell. Remember, sir, you are

3 still under oath.

4 THE WITNESS: Yes.

5 THE COURT: Miss Kennedy.

6 Q Dr. Farwell, you have the projector set up and you have

7 a screen up here. Do you want to tell us what this is going

8 to show the Court?

9 THE COURT: If you need to get up and show us, go

10 ahead, doctor.

11 THE WITNESS: Thank you.

12 A You will recall, that we talked about three different

13 types of stimuli. The first type of stimuli we talked about

14 are called targets, target stimuli. These are words or

15 phrases that contain information that we know the subject

16 knows. In this case, they were details about the crime that

17 we were very sure that Harrington knew, I discussed them with

18 him prior to the test, and he confirmed that he knew them. I

19 gave him a list of them, and I made sure he was going to

20 recognize that list. He said he was sure he could recognize

21 them. And we asked him to push a particular button on a

22 mouse in response to those targets, and another button for

23 everything else. So the bottom line on targets is we know he

24 will recognize them. When a person recognizes something, the

25 brain engages in a particular information processing

Page 36

1 activity, which is known as context updating and that results

2 in a P300. An analogy for this is when your computer

3 accesses the hard drive, there is a little buzzing sound and

4 you know, you can tell from outside of the computer. You may

5 not know what they are putting on the hard drive or taking

6 out, but you can tell from the outside of the computer it's

7 accessing the hard drive when we engage in the information

8 processing activity of context updating, of taking a note

9 of something. For example, an elephant comes into the room,

10 and we make note of that. Well, the brain does something and

11 the neurons in the brain fire electrically in a particular

12 pattern and we get this very well documented research

13 phenomenon known as a P300. Initially, the P stands for

14 positive voltage at the scalp and 300 initially stood for 300

15 milliseconds or three-tenths of a second after the stimulus.

16 If you have a very simple stimulus like a beep in the air or

17 something that has particular significance, you will be able

18 to evaluate it quickly enough to get a P300 after 300

19 milliseconds. If you have a stimulus that takes a while to

20 figure out what it is, like several words on a screen that

21 you have to read, that same component will take place a

22 little bit later. So the terminology gets a little

23 confusing, because it's still called a P300, but may have a

24 peak at 800 milliseconds or eight-tenths of a second. That

25 is common in the literature.

Page 37

1 Q So Dr. Farwell, this particular screen shows what from

2 Terry Harrington's test?

3 A We flashed these target stimuli mixed in with the other

4 types of stimuli on the screen. This is an average of his

5 responses, brain responses to the target stimuli, and this is

6 in the time window that we are interested in, the time window

7 where the relevant brain responses take place early on. Some

8 other things take place that have to do with just sensory

9 processing of stimulus. We don't care about that. We care

10 about how he cognitively processes what this stimulus means

11 to him. So this is from 600 to 1,600 milliseconds or point

12 six to one point six seconds. That is the response at this

13 area of the brain, the mid line parietal area of the brain.

14 And I'm pointing to the top back of my head, which is where

15 this P300 is maximum. You see that bump on the left, the

16 positive deflection on the left on the Y-axis, this is

17 voltage, and the top of that is going to be around 20

18 microvolts across the X-axis. This is time. We are going

19 from about point six seconds to about one point six seconds.

20 This is the time range of interest. That bump there on the

21 left, that is a P300. Now, I noticed in my research that

22 after the P300, there is a negative deflection. The voltage

23 goes positive, then relative to that it goes negative at the

24 parietal area, and also in the central and frontal areas of

25 the scalp. So that mountain up there, if you can see in the

Page 38

1 upper left, that positive peak, that is a P300. That plus

2 the negative deflection off to the right is a MERMER. We

3 know Harrington is going to recognize the target stimuli, he

4 does recognize them, we get a MERMER. We get a P300. I will

5 talk in terms of P300, because again, the MERMER is just

6 frosting on the cake. The second type of stimulus we present

7 are irrelevants, and here is the response to the irrelevants.

8 It doesn't contain that positive peak. It doesn't contain a

9 P300, as you can see.

10 Q Dr. Farwell, just a minute. You want to refresh our

11 memory on what you are speaking about when you are saying the

12 irrelevants?

13 A Yes. The irrelevants are stimuli, words, phrases, that

14 are not relevant to the subject. They have nothing to do

15 with the crime, they have nothing to do with the subject.

16 They are structured so that they are equally plausible to a

17 subject who doesn't know anything about the crime. For

18 example, say a target was the murder weapon. An irrelevant

19 might be some other weapon that could have been the murder

20 weapon, but it wasn't. It's not relevant in this context,

21 it's not going to mean anything to the suspect. So when he

22 recognizes the targets, the red line shows us, we get that

23 P300. When he doesn't recognize the irrelevants, the green

24 line shows us that he doesn't get that P300 and doesn't get a

25 MERMER. The critical scientific question here is is he going

Page 39

1 to recognize the probe stimuli as significant? And the

2 probes, you recall, are the ones that are relevant to the

3 crime, and he will know if and only if he committed the

4 crime. These are items that he is going to have to encounter

5 in order to commit the crime, but he doesn't have any other

6 way of knowing them, such as having heard about them at the

7 trial or in interrogations or what have you. So the question

8 is is he going to recognize the targets -- excuse me -- is he

9 going to recognize the probes, that are relevant to the

10 crime, in which case he will have a response that looks like

11 the response to the targets. It will contain a P300. Or is

12 he not, in which case the probes will have a response that

13 looks like the irrelevants. What that will tell us is, is

14 this critical information about the crime stored in that

15 particular brain or not, and to repeat myself, he recognizes

16 the targets. We know he does. If he recognizes the probes,

17 if he has that information, then the probes are going to

18 contain a P300 and they are going to look like the targets.

19 If not, if the probes are irrelevant to him because he

20 doesn't have the critical information about the crime, then

21 what's going to happen is the response to the probes are

22 going to look like the response to the irrelevants.

23 Q Dr. Farwell, would you consider those then something

24 such as a base line to go on?

25 A Exactly, exactly. The irrelevants provide a base line

Page 40

1 for something he doesn't know. The targets provide a base

2 line for something he does know. Then we say which does this

3 look like? Which kind do the probe responses look like?

4 Q Perhaps you could tell the Court some of the -- if you

5 can recall, some of the probe stimuli that you actually used?

6 A Yes. What I did is -- I will tell you how I used them

7 and got them. I went out to the crime scene with Ann

8 Danaher, as a matter of fact, and with the knowledge of what

9 had taken place in the court and in some of the

10 interrogations, so I knew a lot about the lay of the land and

11 I had also crime scene photos. What I did was to figure out

12 things about the crime scene that I could discern by

13 combining information from those various sources with

14 actually being at the crime scene, but that never came out

15 explicitly in the trial. For example, Kevin Hughes, the

16 primary prosecution witness, testified that he saw Harrington

17 and another co-defendant go onto the -- disappear around the

18 corner onto the car lot where they were intending to steal a

19 car. Well, the crime didn't take place, the murder didn't

20 take place on that car lot. It took place about a half a

21 block away. The suspect, the person who actually committed

22 the crime -- I should say the perpetrator, would have had to

23 go from the car lot over to across the street to where the

24 railroad tracks were. So I asked Harrington, "Do you know

25 where the perpetrators had to go to get to the place the car

Page 41

1 was going to be stolen to where the crime took place?" He

2 said, "No, I don't know, because I wasn't there." I said,

3 "Well, would you know if they had had to cross a street or go

4 over a bridge, under an underpass?" He said, "No, I don't

5 know. I wasn't there." Then I said, "All right. You know

6 that the victim was shot by the railroad tracks, but where

7 was he actually? What was he standing in front of when he

8 was shot?" And something clearly the perpetrator would know.

9 "What was behind the victim at the time that he was shot?" I

10 was able to discern from crime scene photos that in fact

11 behind the victim when he was shot -- and also from the

12 ballistic evidence, putting those two things together, and

13 going out onto the crime scene and looking at it. And I

14 might add here, fortunately, the crime scene, in terms of the

15 major lay of the land where the buildings are and railroad

16 tracks and street and so on, has not changed much since the

17 crime.

18 Q How did you know that?

19 A I knew that because I had crime scene aerial photos that

20 showed the various relative locations of those various items.

21 So I was able to discern what was behind the victim at the

22 time he was shot. And I asked Harrington if he knew that,

23 and he said no, he didn't know. Then after the murder was

24 committed, Hughes testified that he saw Harrington and

25 another co-defendant running around behind a particular

Page 42

1 building which was identified in the aerial photos. And I

2 asked Harrington, "Were you ever behind that building?" He

3 said, "No." I said, "Do you know what was behind that

4 building?" He said, "No."

5 I was able to discern, although it never came out

6 explicitly, stated as such in the trial, it was clear from

7 going to the scene and looking at the photos, that they would

8 have had to run through waist high weeds and grass behind

9 that building to get to where Hughes said they went. So I

10 asked Harrington, "Do you know what's behind that building?

11 Do you know if the suspects would have had to -- I should say

12 the perpetrators would have had to run through waist high

13 weeds and grass or just across the parking lot, across the

14 black top and concrete, across gravel?" He said, "I don't

15 know, because I wasn't back there." In order to get to the

16 getaway car, according to the route that Hughes testified,

17 they would have had to negotiate a deep drainage ditch, maybe

18 five feet deep, and I asked Harrington -- there was an

19 obstacle that the perpetrators had to negotiate to get to the

20 getaway car -- "Do you know what that was?" He said, "No."

21 I said, "Do you know if it was a drainage ditch?" Or I

22 believe it was a wire fence or cement wall. He said, "No, I

23 don't know, because I wasn't there." Then I asked him,

24 "Well, once you got to the -- once the perpetrator got to the

25 getaway car, what was the car parked next to?" He also

Page 43

1 claims not to know that.

2 Hughes testified that Harrington drove the car and he

3 specified the streets, but he didn't specify where they would

4 have had to go directly to get to that street. I said to

5 Harrington, "Do you know what direction the car was driven

6 immediately at the beginning of the getaway when the car was

7 started?" And he said, "No."

8 So here is a lot of detail about the lay of the land,

9 about what happened at the time of the crime that I could

10 figure out from the crime scene photos and the testimony that

11 Harrington claimed not to know. I put those details into

12 probe stimuli. For example, weeds and grass was one of the

13 probe stimuli, and --

14 Q The probe stimuli, again, were -- Remind the Court what

15 the probe stimuli --

16 A The probe stimuli were the ones that were relevant to

17 the crime that he would have had to have experienced to

18 commit the crime that he had -- Harrington had no other way

19 of knowing, and that he claimed not to know.

20 Q And those are not shown on the screen?

21 A They are not shown yet. The question is did he

22 recognize the probes. If so, we can expect there to be a

23 P300 like that red line up there. If not, we can expect the

24 probe response to be like that green line of the irrelevants

25 which he did not recognize.

Page 44

1 Here is what the average probe response looked like. As

2 you can clearly see, the probe responses looked like the

3 irrelevant responses. The scientific conclusion we draw from

4 that is that the information -- that specific information

5 about the crime was not stored in Harrington's brain. You

6 clearly do not see a P300 to those probe responses. You

7 don't see a MERMER.

8 Q Doctor Harrington -- I'm sorry. Dr. Farwell, these

9 plots, did you say that they were an average of the

10 responses?

11 A Right. That's a standard procedure in event related

12 brain potentials. The brain is doing other things other than

13 the things we are interested in at the same time, so in order

14 to get clear responses, what we do is we present the stimuli

15 a number of times and we average the responses. That's been

16 the standard procedure for half a century.

17 Q And the results -- Well, first of all, let's back up. I

18 don't think you described to the Court exactly how the test

19 was administered?

20 A All right. Harrington -- I said the things that I have

21 just said, and I gave him a few other instructions about what

22 was going to be happening, press the left button for targets,

23 the right button for everything else. I administered a head

24 band equipped with sensors which goes around the head, and

25 has a strap that goes across to measure the standard sites

Page 45

1 for measuring the P300. And that head band with sensors

2 picked up the electrical signals from the brain with standard

3 electroencephalographic electrodes. That's the -- Those are

4 all standard kind of techniques. The signals were amplified

5 by, again, EEG amplifiers which are standard for that kind of

6 research. They were digitized, which means we convert those

7 voltages into numbers, and those numbers were fed into a

8 computer to be analyzed.

9 So during the test, Harrington sat in front of a

10 computer monitor and he observed the stimuli of these three

11 types on the monitor, he pushed buttons in response to those

12 three types of stimuli. And his brain responded in the way

13 that is shown on the screen.

14 Q How certain, Dr. Farwell, are the results that you

15 obtained here?

16 A Well, we analyzed the data, not just by looking at the

17 data, but we conduct a signal analysis using a mathematical

18 algorithm known as bootstrapping. We do bootstrapping on the

19 correlations, which is a standard statistical -- a well

20 published and well researched and well accepted statistical

21 technique for comparing data of different types, especially

22 when that data has an unusual or unknown distribution, which

23 correlations of brain waves tend to have. And this

24 bootstrapping algorithm provides us with a determination,

25 either does the red line look like the green line -- I'm

Page 46

1 sorry. Does the blue line representing the probes look like

2 the green lines representing the irrelevants, or do the probe

3 responses look like the red line, which represents the

4 targets. In other words, are we getting the kind of response

5 to these probe stimuli that this brain emits when it

6 recognizes something, or are we getting the kind of response

7 to these probe stimuli that are relevant to the crime that

8 looks like what that brain emits when it doesn't recognize

9 something. We make that determination mathematically. We

10 come up with a determination one way or the other, and a

11 statistical confidence. Analyzing the P300, I came up with

12 the determination of information absent, means he didn't

13 recognize the probe stimuli with the statistical confidence

14 of 99 percent. I'm going to come up and point to you what

15 I'm talking about. This mountain here is a P300, analyzing

16 in that time range 600 to 1000 milliseconds, we came up with

17 that result. Now, when you include these other data, which

18 as I mentioned constitute the MERMER, which is something

19 which has been less thoroughly established in the scientific

20 community, and less published and less researched, although

21 it has been published by Sharon Smith and myself in the

22 Journal of Forensic Sciences, when you include the MERMER,

23 it's frosting on the cake. You get exactly the same

24 determination; information absent. He doesn't know these,

25 but we get a higher statistical confidence, 99.99 percent,

Page 47

1 instead of just 99 percent. So I analyzed the data twice,

2 and in the supplement to the report, I include that complete

3 analysis.

4 Q Dr. Farwell, you mentioned bootstrapping, and I think

5 you quickly went over it, but would you explain it a little

6 further, please?

7 A Yes. Bootstrapping is a statistical technique that is

8 standard in the literature. It's been published by Wasserman

9 and Bockenholt and by others. It is standard in the

10 literature. And what that statistical technique does is

11 allows us to come up with a statistical confidence as to

12 whether mathematically the data that go into this look like

13 data that go into this line or this line. It allows us to

14 mathematically determine whether the probe responses are

15 similar to the targets containing a P300 or whether the probe

16 responses are similar to the irrelevants and they don't

17 contain a P300. The output of that statistical procedure is

18 to give you a determination one way or other, and a

19 statistical confidence for that. So it's a standard

20 mathematical procedure.

21 Q Dr. Farwell, by doing the bootstrapping and doing the

22 mathematical and statistical work, does that take the

23 subjectivity out of interpreting these results?

24 A Yes, it does. Now, I can see it and everybody in this

25 courtroom can see the pattern there, but I don't make my

Page 48

1 decision based on that. I make the decision and the computer

2 makes its decision based on a mathematical analysis of the

3 data that we are just illustrating here. I can see this and

4 I can say, yeah, there is a P300 to the targets, there isn't

5 one to the probes, but that does not go into my report. What

6 goes in the report is the mathematical determination of that

7 same thing. So yes, it does remove the subjectivity.

8 Q And just for the record, this actual screen that you

9 just completed is contained in the supplement we marked

10 Exhibit 2; is that correct?

11 A That's correct, and contained in the initial report is

12 not only this small segment of the wave forms, but the

13 earlier part which is not relevant to our discussion, and the

14 later part which is not relevant to our discussion.

15 Q How can you be sure, other than what you just told us,

16 of the accuracy of Harrington's test?

17 A Well, this phenomenon of P300 is something that has been

18 very voluminous in research and published over quite a number

19 of decades, and it is well known in the scientific community

20 we can accurately detect the presence or absence of a P300.

21 The statistical procedures we used are well known and well

22 published. The technical procedures that I used are included

23 in the supplement to the report, and there is a separate page

24 that talks about the technical procedures, and I followed all

25 the standard things competent scientists in this field do

Page 49

1 with respect to the amplifiers and electrodes and the filter

2 settings on the amplifiers and digital filters we used and so

3 on.

4 Is that responsive to your question? I'm not quite sure

5 what you are asking for.

6 Q Yes. This screen that's showing right now is the

7 responses to crime scene information?

8 A Right.

9 Q And I believe early in your testimony, you mentioned

10 that you also did the same comparison to Terry Harrington's

11 alibi; is that correct?

12 A Right. What we can see here is that Harrington's brain

13 doesn't contain the information about the crime, but what if

14 he just doesn't remember what happened on that night? So if

15 we had done this test a week after the crime, that might be

16 not such a relevant question, and we have used very salient

17 features about the crime. There is extensive literature on

18 what is memorable, what constitutes a memorable event, and we

19 have used memorable events here. For example, something that

20 has emotional content, something that is unusual, something

21 that is unique, something that is personal to the individual,

22 something that involves actions. We used those parameters,

23 structured stimuli likely to be memorable to the subject when

24 we developed the probes, but the question can still be asked,

25 what if he just doesn't remember what happened on that

Page 50

1 evening. What we did is gather information about the events

2 of the evening as told by his alibi witnesses who claimed

3 that he was not in Council Bluffs, he was over in Omaha at a

4 concert, a specific concert where he talked to specific

5 people, and specific events happened, both before, during and

6 after that concert, and at the same time as the crime was

7 taking place. So the question that this next test is going

8 to answer is well, okay, does Harrington's brain contain a

9 record of the events of that evening as told by his alibi

10 witnesses. It doesn't contain a record of the crime. Does

11 it contain a record of the same evening, a different set of

12 events. This is exactly the same thing that we did in the

13 Farwell and Donchin study on real life events, and laboratory

14 events, and in many of the other studies, and one of the CIA

15 studies where we had 30 subjects, took an event a person had

16 participated in, an event they hadn't, and tested on both,

17 and we could tell by the brain responses which event they

18 participated in and which one they hadn't.

19 Q As you show us the screen for the alibi information, you

20 could tell us, did you use the same three types of stimuli in

21 the alibi information as you did in the crime scene

22 information?

23 A Yes, I did, except of course, they are relevant to the

24 alibi, not to the crime. There, once again, this is

25 Harrington's brain wave responses to alibi information,

Page 51

1 target stimuli. These are things we know he knows about, the

2 alibi. They are details about the alibi we have discussed

3 with him. We have given him a list of them. He pushes a

4 special button when they arise, we will get a P300 as before.

5 We also get the positivity followed by negativity. In this

6 particular time range, you see the pattern is very similar to

7 what you saw in the other brain response. We also used as

8 before, irrelevant stimuli. You can once again see we are

9 getting a clear difference between the targets and

10 irrelevants. The irrelevants do not show this marked

11 pattern; the target responses do. And to repeat myself, the

12 irrelevants are the stimuli not relevant to Harrington, not

13 relevant to the alibi, not relevant to the crime. So the

14 question now is going to become as before, we have stimuli we

15 know he is going to recognize, information we know is going

16 to be stored in that brain, represented by the targets. We

17 have information we know is irrelevant, represented by the

18 irrelevants. The question is does Harrington's brain store a

19 record of the events of that evening that constitute his

20 alibi. Details about the alibi are the details about the

21 alibi stored in the brain. Details about the crime aren't.

22 Are details of the alibi? If they are, we will get a

23 response that looks like that red line, looks like the target

24 response, because those things he knows. If the information

25 about the alibi is not stored in his brain, he doesn't have a

Page 52

1 clear record of the events of that evening stored in his

2 brain, we are going to get a response that looks like the

3 irrelevants.

4 Here is what the response looked like. As you can see,

5 this response to the probes that were relevant to his alibi

6 clearly contains a P300, clearly contains a MERMER, clearly

7 looks very, very similar to this response to the targets. In

8 other words, this is what brain response you get when the

9 person recognizes the probe stimuli, when he recognizes --

10 This is the response you get when that information is stored

11 in the brain.

12 Let me go back and forth between these two responses.

13 You see clearly he is getting a P300 to the probes that are

14 relevant to the alibi, but when we look at the probes that

15 are relevant to the crime, it's not there. There is no P300

16 to the probes relevant to the crime. In both cases, there is

17 this P300 and MERMER represented by the red line to the

18 target stimuli that we know he knows, but the stimuli about

19 the crime lacked that response. Stimuli about the alibi

20 contained that response. And the conclusion that we draw

21 here is that there is a record of that particular evening,

22 events of that evening stored in Harrington's brain. That

23 record matches the alibi; doesn't match the crime.

24 Q And Dr. Farwell, again, the screen that we are showing

25 here that showed the responses to the alibi information,

Page 53

1 that's also contained on Appendix II of the supplemental

2 report; correct?

3 A That's correct. Both of these are contained in the

4 Appendix II. Both the crime and the alibi information plots

5 of the brain waves are contained in Appendix II of the

6 supplement to the report.

7 Q And in addition to the two screens which make everything

8 look pretty simple, the rest of the information about the

9 test is contained in the report and the supplement?

10 A That's correct. Those are my professional account of

11 the scientific results of this particular test.

12 MS. KENNEDY: Okay. I don't have any further

13 questions for you. Thank you.

Page 147

1 WILLIAM IACONO

2 being produced as a witness on behalf of the plaintiff,

3 and having been first duly sworn by the Court, was examined

4 and testified as follows:

5 DIRECT EXAMINATION

6 By Ms. Kennedy:

7 Q I believe you already stated and spelled your name for

8 the record; is that correct?

9 A Yes. At least I spelled it. I don't know if I stated

10 it.

11 Q Dr. Iacono, I handed you what is marked Exhibit 3. Do

12 you recognize that document?

13 A It's a reproduction of my curriculum vitae.

14 MS. KENNEDY: Your Honor, I would ask that

15 Exhibit 3 be admitted.

16 THE COURT: Any objection?

17 MR. CROWL: No objection.

18 THE COURT: Exhibit 3 is received.

19 Q Dr. Iacono, what is your educational background, and you

20 can be brief.

21 A I have a Bachelor of Science in Psychology from

22 Carnegie-Mellon Mellon University and a Ph.D. in Psychology

23 from the University of Minnesota.

24 Q What has been your employment history since obtaining

25 your doctorate?

Page 148

1 A I have faculty appointments at the University of British

2 Columbia and currently at the University of Minnesota?

3 Q And what positions do you currently hold at the

4 University of Minnesota?

5 A I am currently a Professor of Psychology and a Professor

6 of Neuroscience. I'm also the Director of the Clinical

7 Science Training Program at the university.

8 Q What do you specialize in in the areas of psychology and

9 neuroscience?

10 A I specialize in clinical psychology and psychophysiology

11 and clinical applications of psychophysiology such as using

12 psychophysiological measures to assist in clinical

13 assessment.

14 Q What -- If you could describe, what do

15 psychophysiologists do, exactly?

16 A Psychophysiologists measure -- They do a lot of things,

17 but one of the main things they do is measure human bodily

18 responses, like palm or sweating and heart rate and blood

19 pressure and brain wave signals to make inferences about

20 what's going on in people's minds, in their brains.

21 Q Concerning your work since obtaining your doctorate,

22 have you received any special recognition from any scientific

23 organizations for your work?

24 A I have. I have received distinguishing scientific

25 contributional words from the Society for Psychophysiological

Page 149

1 Research, and from the American Psychological Association.

2 These were for early career contributions. I have been

3 elected to the status of Fellow in a number of scientific

4 organizations, including the American Association for the

5 Advancement of Science and the American Psychological

6 Association, and the American Psychological Society. I

7 served as president of the Society for Psychophysiological

8 Research. That's a elected office. And I have also received

9 an endowed share from the University of Minnesota, which is a

10 distinguished McKnight Professorship. That is another title

11 I hold at the university.

12 Q What does the title of Fellow actually mean?

13 A Well, it's typically a title given to people who are

14 recognized for having at least some prominence in the field

15 and in the area of their specialization, which is defined in

16 part by the scientific society that confers that honor.

17 Q Again, how many organizations do you hold the title of

18 fellow?

19 A I'm not exactly sure, but three or four. The American

20 Psychological Association is composed of divisions; many,

21 perhaps as many as as 50, and I have held that title in a

22 couple of those divisions, but subsequently, I'm no longer a

23 member of the American Psychological Association, even though

24 I had -- I was awarded that honor. But there may be another

25 society; I think the Association for the Advancement of

Page 150

1 Preventative Psychology, or something like that. I have

2 forgotten. It's also -- I'm no longer a member of that

3 either, because I think the society has gone defunct, but

4 several.

5 Q Have you published any scientific papers?

6 A Yes.

7 Q How many?

8 A Over 175.

9 Q And have you presented any papers or given any addresses

10 at scientific meetings of prominent organizations?

11 A Yes.

12 MR. CROWL: Your Honor, the State will stipulate to

13 his qualifications as an expert.

14 THE COURT: Thank you. I notice also, Exhibit 3,

15 this curriculum vitae is admitted, if you want to proceed,

16 Miss Kennedy.

17 Q Have you been qualified as an expert in any Iowa court?

18 A Yes.

19 Q Would you tell us about that case, please?

20 A I don't remember the details, but I think it was a

21 murder case from about 10 years ago where an individual had

22 taken a polygraph examination and had made some admissions in

23 the context of taking the polygraph examination, so I was

24 called as an expert to consider the significance of the test

25 and the admissions that were made.

Page 151

1 Q Has the Supreme Court cited your work?

2 A Yes, it has. I had a chapter in a book called Modern

3 Scientific Evidence, and in the Supreme Court Scheffer

4 decision of 1998, that chapter was cited by the court in

5 rendering their decision.

6 Q Has your work in forensic psychophysiology focused on

7 measures to assess recognition memory?

8 A I'm sorry. Could you repeat the question?

9 Q Has your work in forensic psychophysiology focused on

10 how to use psychophysiological measures to assess recognition

11 memory?

12 A That has been a focus, but I wouldn't say that's the

13 primary focus.

14 Q What would you consider the primary focus?

15 A The primary focus has been just generally using

16 psychophysiological procedures to, in forensic settings

17 mostly in the use of what we might call conventional

18 polygraph tests.

19 Q And would you explain further your work regarding

20 conventional polygraph tests?

21 A Well, I have done research studies on how they work, on

22 some of the factors that affect their outcome. I have been a

23 consultant to various government agencies, like the CIA and

24 the United States Congress Office of Technology Assessment,

25 and President Clinton's Joint Security Commission, and the

Page 152

1 Department of Defense Polygraph Institute, on government

2 practices and policies related to the use of polygraph tests.

3 I have written numerous review articles and evaluations of

4 the scientific literature as they relate to the application

5 of these different types of forensic psychophysiological

6 procedures.

7 Q Has any of your work focused on how to use electrical

8 signals recorded from humans to get into concealed memories

9 regarding the perpetrator of a crime?

10 A Yes.

11 Q Would you tell us about that, please?

12 A This is an application that in the literature is

13 referred to as the guilty knowledge test. This is a type of

14 test that is used to determine whether or not a person has

15 knowledge of a crime that only the perpetrator of the crime

16 would be expected to have. This procedure is based typically

17 on the recording of measures of autonomic nervous system

18 responsivity. In particular, it's been focused on measure of

19 palm sweating that is known as the GSR, or galvanic skin

20 response. The guilty knowledge test was discovered many

21 years ago. In fact, it was first reported in scientific

22 literature back around 1960. Since then, there have been a

23 number of studies published that use this particular

24 technique, including several that I have studied. The

25 premise behind the procedure is that you can ask an

Page 153

1 individual a number of questions in a multiple choice format,

2 but instead of the person being asked the question choosing

3 the alternative that would be associated with guilty

4 knowledge, you measure physiological responses to the

5 multiple choice alternatives, and the multiple choice

6 alternative that elicits the largest physiological response,

7 is then examined to see if that's linked to the crime

8 relevant knowledge that the test is assessing. Typically,

9 these tests might, say, have five to 10 questions on them and

10 each one would have four or five multiple choice

11 alternatives. An example might be if you robbed a bank, then

12 you would know what was written on the holdup note, and then

13 an individual might be presented with options like, don't

14 sound the alarm, I have a gun, don't move, put all your money

15 in this bag, etcetera. Now, if you robbed the bank, then you

16 would know which one of those options you actually wrote on

17 the holdup note, but if you didn't rob the bank, you wouldn't

18 know. Your body gives a recognition memory response in

19 effect to the alternative that reflects your guilty

20 knowledge, if in fact you are guilty. So you would know what

21 you wrote on the holdup note, for instance, and your body

22 would produce the largest response to that particular item

23 typically.

24 Then you would follow that up with a second item which

25 might be if you robbed this bank, then you would remember

Page 154

1 what the teller that you held up looked like and then you

2 might hold up five pictures of people; one of them was the

3 actual teller, and the other four of whom were just people

4 chosen at random, and again, if you had a recognition memory

5 response, then your galvanic skin response might be largest

6 to that particular picture. And by having a whole series of

7 these types of questions, you can record the galvanic skin

8 response to the multiple choice alternatives in each one, and

9 then determine the number of times that recognition memory is

10 shown to the guilty knowledge alternatives.

11 That's how the guilty knowledge technique has been

12 developed. That's what it constitutes.

13 Q And the guilty knowledge test, you said it was developed

14 in approximately 1960. Has it changed any since that time?

15 A The most significant change, I believe, has been the

16 work that's been done in the last 10 years or so, which has

17 been to replace the recording of the galvanic skin response

18 with the recording of the brain event related potential, but

19 basically, the principal technique has been largely unchanged

20 over most of this time.

21 Q Dr. Iacono, the articles listed on your vitae that would

22 mention assessment of memory information, would that also --

23 would that be referring to the phrase guilty knowledge?

24 A Some of the articles would specifically use the phrase

25 guilty knowledge, and those would be ones that would be

Page 155

1 concerned with this technique.

2 Q You mentioned you are Director of the Clinical Science

3 Training Program at the University of Minnesota. Are you

4 also a clinical psychologist?

5 A Yes.

6 Q What do you do as a clinical psychologist?

7 A Too much. But related to this case, I do research with

8 people who have different forms of psychopathology and use

9 psychophysiological techniques to try and understand their

10 psychopathology. I use psychophysiological techniques, for

11 instance, to identify genetic risks to help me understand the

12 mental processes of people with psychopathology, and as I

13 mentioned, I also am interested in clinical assessment. So

14 some of the research I have been involved with has been

15 application of psychophysiological procedures like those we

16 have been talking about to study memory, including memory in

17 people with psychopathology.

18 Q When you say people with -- I'm sorry. I will move on.

19 I lost my train of thought.

20 Do you also study -- As a clinical psychologist or any

21 other work, do you study memory?

22 A I study the assessment of memory. I wouldn't consider

23 myself a person who is expert in memory per se, but as

24 someone who has been interested in trying to use

25 psychophysiological procedures to see if we can determine if

Page 156

1 a person has a memory, certain memory, I have done that type

2 of work.

3 Q Do you study amnesia?

4 A I have done some research on amnesia.

5 Q Are there any papers included in your vitae where you

6 have studied and measured the electrical activity of the

7 brain?

8 A Yes.

9 Q And how many of those papers are included on the vitae?

10 A There is over 20.

11 Q In the case we are talking about today, doctor, where we

12 are concerned with the electrical response called the P300

13 brain wave, are you familiar with that?

14 A Yes.

15 Q Have you published papers on the P300?

16 A Yes.

17 Q If you can recall, how many papers have you published?

18 A Maybe a dozen.

19 Q These were used specifically referring to the P300?

20 A Well, that would be one element of these publications,

21 yes.

22 Q Have you studied certain applications of the P300?

23 A I have studied at least one application which might be

24 used to detect concealed memories. I use it to help me

25 understand other things, so if you consider that to be an

Page 157

1 application, then I guess I have used it in multiple

2 applications.

3 Q You say you used it to study certain things. What would

4 be one example of the things you would study using the P300?

5 A Well, I have used it to -- I have studied its

6 heritability. So I have conducted twin research to show how

7 genes are likely to influence the amplitude and latency of

8 the P300 response. I use it to study young people who have

9 psychopathology disorders like attention deficit

10 hyperactivity disorder and conduct disorder. I use it to

11 study who are at risk for developing antisocial personality

12 and alcoholism. As you mentioned, I have used it to study

13 memory and amnesia and used it to study whether or not we can

14 detect if people have concealed memories that they are trying

15 to hide, by measuring P300.

16 Q You mentioned one of your studies had to do with the

17 twins. I believe you have a study that is very famous; is

18 that correct, using twins?

19 A I don't know how famous it is, but the University of

20 Minnesota is generally famous for twin research, and one of

21 the largest twin studies ongoing there is one that I'm

22 co-director of.

23 Q Have you developed your own procedure for using the P300

24 to assess concealed memories?

25 A Yes. Several years ago, a student of mine named John

Page 158

1 Allen, who is now a colleague and professor at the University

2 of Arizona, he and I pursued a program of research on this

3 topic.

4 Q Did you receive any special recognition for that work?

5 A I haven't personally received special recognition, but

6 Dr. Allen has. This year he received the Society for

7 Psychophysiological Research Distinguished Scientific

8 Contribution Award, and he made as his address to the society

9 meeting, which was held about a month ago, a review of his

10 work in this area.

11 Q Have your studies in forensic psychophysiology, the

12 guilty knowledge test or the assessment of guilty knowledge

13 and your work with the P300, received any funding from the

14 government or from government agencies?

15 A They have. When I was in Canada, I received research

16 support from Canadian Science Funding Agencies. In this

17 country, I received research support from the National

18 Institute of Health and I have also obtained support from the

19 Department of Defense Polygraph Institute to fund some of

20 these studies.

21 Q Dr. Iacono, how did you come to be involved in Terry

22 Harrington's case?

23 A I found out about the case from the 60 Minutes news crew

24 who called me in the middle of the summer to ask me questions

25 about what was happening with this particular case and the

Page 159

1 assessment that was done and what my thoughts were about it

2 and they had sent me materials related to the case to review

3 so I could discuss them with them on the phone.

4 Q So you did not learn of this case from Dr. Lawrence

5 Farwell?

6 A No, I did not.

7 Q Can you tell the Court what is your -- what is the

8 nature of your relationship with Dr. Farwell?

9 A I really don't have a relationship with Dr. Farwell. I

10 had met him, I think, on several occasions when he was a

11 graduate student at the University of Illinois, and these

12 were brief meetings at the annual meetings of the Society for

13 Psychophysiological Research where we would both show up, but

14 I don't believe I have seen Dr. Farwell probably since

15 he's -- since he obtained his Ph.D. from the University of

16 Illinois. I have talked to him briefly on the phone to get

17 reprints of his work, and he has requested reprints of my

18 work, but I really don't have any meaningful professional or

19 personal relationship with him.

20 Q This may be a sore subject, but are you receiving any

21 compensation for coming here today?

22 A No, I'm not. I'm responsible for covering my own cost.

23 Q Thank you. Do you have any patents or inventions or any

24 financial interest at all in the technology used in the case

25 that was presented here today?

Page 160

1 A No.

2 Q Dr. Iacono, regarding the science of the P300, you made

3 reference to a brain event related potential. Can you

4 explain what that is?

5 A When you are presented with an event like a word phrase

6 that flashes on a computer screen, the neural activity in

7 your brain is synchronized and it produces an electrical

8 wave. That can be measured by attaching sensors on the head

9 and amplifying them, and this electrical wave is called an

10 event related potential.

11 Q How well established is the measurement of event related

12 potentials as a scientific procedure?

13 A It's very well established. They have been measured for

14 many decades. There are thousands of publications on the

15 measurement of event related brain potentials.

16 Q I believe you said you reviewed the tests on

17 Terry Harrington published by Dr. Farwell; is that correct?

18 A I reviewed some of the material, yes.

19 Q And the test that was given to Mr. Harrington was based

20 on the P300 brain response. Can you explain what that

21 response is?

22 A Well, the brain elicits -- When you record an event

23 related brain potential, you are actually recording a complex

24 wave form that has many components, and one component that

25 occurs late, meaning after most of the other ones have

Page 161

1 already appeared, is this P300 wave that we have been talking

2 about in this case. It has a latency that is typically 300

3 milliseconds or longer. It's a wave that is elicited, and as

4 a partial response to evaluating a stimulus, because it's

5 significant, it's recognized, it's somehow meaningful to the

6 person who sees the event or hears the event.

7 Q How well established is the P300 as a measure of brain

8 information processing, in your opinion?

9 A It's very well established. There is -- must be

10 hundreds, if not thousands, of publications on this topic.

11 Q And can the P300 be used to index memory?

12 A Yes, it can.

13 Q How well established is that science?

14 A It's not as well established as these other more general

15 uses of P300, but there are a substantial number of studies --

16 I'm not sure what number -- maybe 50 or more that would

17 indicate that it is sensitive to memory processes, and in

18 particular, recognition memory.

19 Q You explained before that the guilty knowledge test

20 assesses concealed memories. And I think you explained that

21 formerly the results were gauged by the galvanic skin

22 response; correct?

23 A Yes.

24 Q And that has changed to the normal way of doing that,

25 and correct me if I'm wrong, is to now measure brain

Page 162

1 electrical activity?

2 A I don't think there is a normal way to do this. And I

3 don't think it's changed. What I would say is that the --

4 when a guilty knowledge test is administered, you can measure

5 any physiological response as long as it's one that's

6 sensitive to memories that the person might have, and

7 historically, it's been the galvanic skin response that's

8 been used for this purpose, but more recently in about the

9 last 10 years or so, scientists have been measuring the event

10 related brain potential, but even today, people doing guilty

11 knowledge test research, would still publish papers using the

12 galvanic skin response.

13 Q How accurate would the research show that the guilty

14 knowledge test has been?

15 A The guilty knowledge test, there have been a number of

16 publications on it that have examined its accuracy. David

17 Lykken, who was actually the person who first published on

18 this technique back in 1959, wrote a recent review of the

19 accuracy of the procedure summed over all the publications

20 that exist to date, and this review of is was published in

21 1998, and he concluded from that review that it had

22 approximately 88 percent accuracy with guilty people and

23 around 97 percent accuracy with innocent people.

24 Q Have you yourself done research on the guilty knowledge

25 test?

Page 163

1 A Yes, I have.

2 Q And in your research, how accurate has this research

3 been shown to be?

4 A In my published work, it's been -- the accuracy rates I

5 have obtained are similar to those that have been generally

6 reported in the literature, and that were summarized in

7 Dr. Lykken's review. For instance, the first study I did on

8 this topic, we obtained I think 88 percent accuracy for

9 guilty people and 100 percent accuracy for innocent people in

10 that particular investigation.

11 Q Have you done research specifically to determine if the

12 guilty knowledge test was accepted as a valid science?

13 A Yes, I have.

14 Q Would you tell the Court about that?

15 A I have been concerned for some time about what types of

16 procedures that are used in forensic psychophysiology are

17 broadly embraced by the scientific community, because this is

18 an issue that comes up in many trial proceedings. It is an

19 empirical question that can be answered simply by asking

20 scientists what their opinions are about these techniques.

21 So several years ago, David Lykken and I did a survey of

22 scientific opinion that ultimately was published in the

23 Journal of Applied Psychology, which is one of the top

24 journals in the field. In the survey, we sampled members of

25 the Society for psychophysiological Research. This is the

Page 164

1 organization that contains scientists who should be

2 knowledgeable in these types of procedures, and we also

3 surveyed members of the American Psychological Association

4 who were elected to the status of Fellow and were members of

5 the division of that association that's called General

6 Psychology. So these are psychologists who are broadly

7 informed about psychology. And for each of these

8 organizations, we surveyed a little over 200 people, and for

9 members of the Society of Psychophysiological Research, the

10 response rate we got in terms of the surveys being returned

11 was about 90 percent. It was a little bit higher than 90

12 percent, and for the American Psychological Association, we

13 got a response rate of about 75 percent. In these surveys,

14 we asked questions about conventional polygraph tests and we

15 also asked questions about the guilty knowledge technique.

16 In particular, members of both organizations were asked if

17 they believed that the guilty knowledge technique was based

18 on scientifically sound psychological principles and theory,

19 and in both organizations, approximately 75 percent indicated

20 that they thought it was scientifically sound. We also

21 included a hypothetical example based on real life cases that

22 were current in the news at the time that these surveys were

23 done. And the surveys were done actually at two different

24 points in time.

25 The first survey was done when the O. J. Simpson trial

Page 165

1 was in the news, and so we included an item on the survey

2 that said if O. J. Simpson showed guilty knowledge to 80

3 percent of the memories that were assessed on a guilty

4 knowledge test, would it be reasonable to assume that he in

5 fact did have knowledge of the crime that was hidden and

6 concealed. And approximately 75 percent of the respondents

7 said they felt that was a reasonable inference to make under

8 those circumstances.

9 In the second survey, the Unibomber was in the news. At

10 that time, we didn't know that Ted Kazinski was the Unibomer.

11 The suspect was still at large. So we asked scientists if

12 the FBI had identified a suspect and he was given a guilty

13 knowledge test that related to how the bombs were constructed

14 and where they were sent and these sort of things, the type

15 of things that the perpetrator of this crime would be

16 expected to know, and he was able to show a physiological

17 response to 80 percent of of the memories, would it be

18 reasonable to infer that that person in fact knew this

19 information. And again, about 75 percent of the respondents

20 said that they thought that was a reasonable inference.

21 So from this, I conclude that the guilty knowledge test

22 is broadly embraced by knowledgeable scientists as a

23 technique that is scientifically sound and that does provide

24 valuable information to determine if people possess crime

25 relevant memory.

Page 166

1 Q The survey that you just described so well, has it been

2 published in any peer review journals?

3 A Yes. It's been published in the Journal of Applied

4 Psychology.

5 Q Dr. Iacono, in your opinion what is the relationship

6 between guilty knowledge test and the procedures used by

7 Dr. Farwell to test Terry Harrington?

8 A Dr. Farwell's procedure is essentially an elaboration

9 and extension of the guilty knowledge test. It's based on

10 similar types of principles. It doesn't have the exact

11 format that a guilty knowledge test has where questions are

12 asked and then multiple choice items are presented. That's

13 in part because when we record autonomic nervous system

14 responses such as the galvanic skin response, we get a clear

15 response to a single presentation of a question. But when

16 you measure the brain event related potential, you have to

17 present the -- In this case, Dr. Farwell used words and

18 phrases. You have to present these stimuli over and over

19 again. So that's the difference, but if you look at the way

20 Dr. Farwell's test is constructed, it's essentially very much

21 like a set of multiple choice items, each associated with a

22 different theme or question, and in that context it's

23 basically an extension of the guilty knowledge test.

24 Q You mentioned that the technique used by Dr. Farwell in

25 his technique, you would have to show those words or phrases

Page 167

1 multiple times. Why is that?

2 A You have to do that, because you can't get a clear index

3 of the brain potential in most people from a single stimulus

4 presentation. When you present a person with an event like a

5 word phrase on a computer screen, that's presented while they

6 are awake and alert and they are attending to other things

7 that are going on at the same time, and they are breathing

8 and their eyes are blinking and they are hearing sounds, and

9 these sorts of things and all of these other stimuli in the

10 environment are producing brain activity, and the brain has a

11 resting activity, and all of this electrical energy that is

12 generated by the brain tend to mask the brain's response to

13 the presentation of a word phrase on a computer screen. That

14 phrase might elicit a brain potential that is 50 percent the

15 size or even smaller of the ongoing background brain

16 electrical energy. So in order to see the response to the

17 word presented on a computer screen, you need to present that

18 word over and over and over again, and average the response.

19 And because all of the other electrical activity is not

20 specifically associated with the presentation of the word on

21 the screen, all that other electrical activity, when you

22 average it, it cancels itself out and basically just

23 disappears.

24 In other words, the background activity at the time a

25 word is presented might be positive for the first time the

Page 168

1 word is presented, and negative for the second time the word

2 is presented. When you average that naturally varying

3 positive waver of the first presentation with the negative

4 waver of the second presentation, the result might be a zero.

5 But the event related potential that's generated by the word

6 on the computer screen is going to be more or less the same

7 each time that word is presented, so by averaging you are

8 able to pull out the event related potential to the word and

9 all the other brain activity is canceled. It's through this

10 technique that you are actually able to see visually the

11 event related potential such as the ones that Dr. Farwell

12 projected on the screen this morning.

13 Q So in your view, the technique described by Dr. Farwell,

14 which is in your view an application of the guilty knowledge

15 test, does this have a high degree of acceptance in the

16 scientific community?

17 A In my view, it does, yes.

18 Q Could you explain how the measurement of brain waves in

19 a guilty knowledge test is likely to affect its accuracy?

20 A I think it's likely to improve it's accuracy. There are

21 a number of problems measuring the galvanic skin response in

22 a traditional guilty knowledge test. They have to do with

23 the fact, for instance, that the galvanic skin response is

24 not always emitted by somebody when they are presented with

25 the stimulus. Sometimes people are just under-aroused and

Page 169

1 under-reactive and you get very little information to work

2 with. It's also the case that autonomic nervous system

3 responses are strongly affected by emotions and other types

4 of extraneous factors, such as if somebody moves in their

5 chair, takes a deep breath or thinks a stressful thought.

6 All of these sorts of things might generate a galvanic skin

7 response that is inappropriate in the context of trying to

8 determine if that response is linked to a particular memory

9 that a person has when they see a word. It's also the case

10 that people can deliberately manipulate their galvanic skin

11 responses by doing things like biting their tongue. So if I

12 stop for a second in my response to your question and lightly

13 bit my tongue, that would produce a galvanic skin response.

14 So when you use the galvanic skin response, there is a lot of

15 noise and a lot of problems that make it so it's not an ideal

16 index of memory, even though it works reasonably well as the

17 studies suggest.

18 When we are using the brain event related potential,

19 this is a phenomenon that occurs in the brain that is not

20 easily affected by all of these other types of factors. It's

21 reliably elicited by these types of stimuli and it's

22 difficult for people to understand what they could do to have

23 any kind of an effect on it, which is not true of the

24 galvanic skin response.

25 It's also the case that we did a study where we looked

Page 170

1 at comparatively the effectiveness of the galvanic skin

2 response in the guilty knowledge studies that have been done

3 to date, and compared them to the effectiveness of the P300

4 wave when used with the guilty knowledge paradigm for

5 classifying people as guilty or innocent, and you can

6 calculate a statistic that summarizes the effectiveness of

7 these techniques for classifying on people as guilty or

8 innocent. I won't go into the complex nature of that, but

9 it's a number that varies between zero and one. And for the

10 guilty knowledge test, one is highly accurate. For the

11 guilty knowledge test, over the literature where people have

12 looked at its classification accuracy, it's accuracy ranges

13 from about point 77 to point 85. But when you look at this

14 P300 type extension of the guilty knowledge test, its

15 accuracy is point 98 point 99, which is very close to one;

16 one being perfect accuracy.

17 So it seems that the use of the P300 and the elaboration

18 of the guilty knowledge test is likely to be more accurate

19 and an improvement over the conventional applications that

20 have been used for much of the last 40 years.

21 Q Can you tell the Court about the procedure you and

22 Dr. John Allen developed to assess concealed memories using

23 the P300?

24 A Yes. It's actually very similar to the procedure that

25 Farwell and Donchin developed. We used targets, probes and

Page 171

1 irrelevants much like they did, although in our work, we call

2 them different things, and the nature of the stimuli, that

3 is, the words that people were trying to conceal, the items

4 that they were trying to conceal, was different than what

5 they used in their study. Another thing that was different

6 in our research was that we used a different statistical

7 procedure for making the determination of whether the probe

8 event related potential, the probe P300 response, resembled

9 more of the target or more of the irrelevant items that

10 people were exposed to. So our procedure actually led us

11 to -- forced us to classify every single person, for

12 instance, and then we were able to make a probability

13 determination as to the likelihood that a person was either

14 aware of concealed information or not aware of the concealed

15 information. This was just a number between zero and one.

16 But when we used our particular approach, because we didn't

17 have inconclusives the way that Dr. Farwell does -- He used

18 the word indeterminate in his testimony today. Indeterminate

19 meant that in his research, the finding that he got was not

20 clearly indicative of whether the person had recognition

21 memory or not. In our research, we had no inclusive or

22 indeterminate category. We classified every single person,

23 and when we did that, we got approximately 95 percent

24 accuracy using our classification procedure.

25 Q Aside from the classification procedure you just

Page 172

1 described and the statistics, are there any other differences

2 between your procedure and that of Dr. Farwell's?

3 A Well, I think I covered the major differences.

4 Q Have you published a comparison to your work to

5 Dr. Farwell's?

6 A Yes, we did. We actually did three studies, each with

7 20 people in it, and in the studies, we tried to vary

8 people's motivation to conceal the information that we were

9 trying to detect with the P300 wave. So people were told to

10 conceal the information, to conceal the information and lie

11 about it, and then we also told them to conceal the

12 information and lie about it and we would pay them if they

13 could prevent us from detecting their having this information

14 using this P300 procedure. So that was the original study

15 that we published, and because our technique and

16 Dr. Farwell's statistical technique for evaluating the data

17 were different, we thought it would be interesting to use the

18 technique that Farwell and Donchin used on our study, which

19 was actually three studies; as I indicated, three

20 experiments, each one with a different level of motivation

21 for people to conceal this information. We used the Farwell

22 and Donchin statistical procedure which Dr. Farwell referred

23 to earlier today as bootstrapping, to analyze our data.

24 What we found when we did this was that the

25 bootstrapping procedure worked very well. It did produce

Page 173

1 indeterminate outcomes, but it produced no errors. So in our

2 60 cases that we studied, we had no errors at all with their

3 procedure, as long as we had allowed for these indeterminates

4 that Dr. Farwell mentioned earlier.

5 Q In this work that you just described, did you uncover

6 any evidence to suggest that Dr. Farwell's procedure would

7 work better in real life, in real life situations rather than

8 work better in real life -- rather than in the laboratory?

9 A We did, in the sense that the classification procedure

10 worked best when people were told to conceal the information,

11 lie about it, and they were given a monetary incentive to do

12 that. So when people were just told to conceal the

13 information, there were more indeterminates, but when they

14 were told to conceal it, lie about it and they could get

15 money if they hid it from us, there were no indeterminates,

16 and all the cases were correctly identified. So that would

17 suggest that in real life, where these things that we are

18 trying to assess, these memories that we are trying to

19 assess, are even more meaningful to the person than they are

20 in these laboratory simulations that we do, they might be

21 easier to detect.

22 Q What's the significance in that the recognition memory

23 procedure you used with Dr. Allen, you used different

24 procedures from Dr. Farwell and yet got very similar results?

25 A I think that's interesting, because it shows that you

Page 174

1 can get very high classification accuracies even though the

2 procedure is varied. There is certain elements of the

3 procedure that are very important, and the elements in common

4 to the studies, include having targets, probes and

5 irrelevants, and having these types of stimuli appear in

6 certain proportions, because stimuli that are presented less

7 frequently, but that are recognized by a person, are ones

8 that elicit P300 waves, but it shows that the P300 approach

9 to detecting guilty knowledge is relatively robust and that's

10 not affected by minor variations in procedure, at least to

11 the extent that there are variations across our studies.

12 Q And explain your use of the word robust?

13 A What I mean by that is that our procedure which,

14 although similar, was different from Farwell and Donchin's

15 procedure. When we used their statistical technique, it

16 produced essentially identical results. In their original

17 report, they made no classification errors, and when we used

18 their technique in our three experiments, we had no

19 classification errors either. By robust, what I'm referring

20 to is the fact that the finding might overwhelm procedural

21 errors and minor deviations from one study to the next,

22 because the P300 affect is strong enough to make up for these

23 kind of differences of cross study so that minor differences

24 of cross study are probably not important in determining

25 one's ability to get the result that we are interested in.

Page 175

1 Q If a person claims to have amnesia but really doesn't,

2 would it be possible to get at their concealed memory using

3 the guilty knowledge test?

4 A It would, and that type of work has been done.

5 Q Other than the guilty knowledge test, could you do the

6 same thing with using the P300?

7 A You could, and that work has been done.

8 Q Is it widely accepted, been widely done, the work with

9 the P300 in the amnesia?

10 A No. That's work that several investigators have done

11 and we have done. In particular, this is a program of

12 research that Dr. Allen has been following at the University

13 of Arizona and that we still collaborate on.

14 Q What would you conclude about the effectiveness of this

15 procedure using your studies or your knowledge of the studies

16 of amnesia?

17 A Well, in our studies, we studied different types of

18 amnesia and other people have as well. I would need to say a

19 little bit about the studies to explain the conclusions. For

20 instance, some of the studies that were done, and that got us

21 interested in this line of research, were done on people who

22 have a lesion that is an area of their brain that is damaged.

23 They have a lesion in their brain that causes prosopagnosia.

24 This is a disorder where people cannot recognize faces. So

25 these are people who you can take x-rays or brain images of

Page 176

1 their brain and it's unquestioned that they have damage in

2 part of the brain that is responsible for a human's ability

3 to recognize a face. In these studies, that use the guilty

4 knowledge test, what they were -- There were different types

5 of studies that were done with these kind of patients, but an

6 example would be to show a person pictures of people, and

7 before they were hooked up to the recording equipment, they

8 would be asked do you recognize any of these people, and

9 these people might be friends, relatives, parents, children,

10 as well as the President of the United States and key figures

11 that you would expect most people to be able to recognize,

12 and they would be asked do you recognize these people. And

13 of course, they don't, because they have this disorder. But

14 when their GSR was measured, their galvanic skin response,

15 when they were shown these pictures, they gave large

16 responses to people that they would have been expected to

17 recognize; for instance, pictures of their friends and

18 relatives when mixed in with pictures of other people that

19 they wouldn't be expected to know. There was also a P300

20 application of this procedure and that study showed that P300

21 was large; in one case actually with a person with

22 prosopagnosia that showed that that person also had

23 recognition of these faces. So in these people who have

24 brain damage, the P300 shows that they recognize faces that

25 they themselves don't even realize they recognize.

Page 177

1 We extended this type of research by studying hypnotic

2 amnesia. So what we did was we identified people who were

3 highly hypnotically susceptible and we exposed them to

4 information while they were hypnotized, then gave them the

5 hypnotic suggestion to not remember any of the material. And

6 we then tested them to see if they did remember any of the

7 material by asking if they recognized or could recall any of

8 it, and they could not do that. When we then measured their

9 P300 wave using this type of procedure that we have been

10 talking about similar to the Farwell and Donchin procedure

11 and similar to the guilty knowledge test, it clearly showed

12 that these highly hypnotizable individuals who reported

13 profound amnesia, nonetheless gave the P300 recognition

14 response to the material they were supposed to know.

15 There were also people in the study who were called

16 simulators. These were people who were pretending to have

17 amnesia, and they had to not only conceal the information

18 from us, but they also had to conceal the fact that they were

19 pretending. And the concealed information that these people

20 possessed was also easily detected with the P300.

21 Since then, Dr. Allen has studied people with what's

22 called multiple personality disorder. These are individual

23 who have different identities, and each identity is claimed

24 to have its own memories, and there is no sharing of memory

25 across identities. So what we did was we probed one identity

Page 178

1 for the memories that another identity was supposed to have,

2 and these people, when you -- you know, talk to Identity A,

3 let's say Identity A is Bill, and Identity B is Tom, we would

4 use the P300 test to probe Tom for Bill's memories, and it

5 turned out that we could show that Tom did indeed have Bill's

6 memories. I should say this is actually Dr. Allen's

7 research. I wasn't involved in it, although it was actually

8 to do this study of multiple personality disorder that was

9 the impetus for he and I starting our collaboration in the

10 first place. So these studies showed that even when people

11 deny that they have these kind of memories when they don't

12 seem to be aware that they themselves have these memories,

13 nonetheless you can record P300 and show that the memory is

14 indeed residing in their brain.

15 Q You reviewed some materials prior to coming here today

16 that were provided by Dr. Farwell; correct?

17 A Yes.

18 Q Do you recall what you reviewed?

19 A Be hard pressed to give an exact list, but I reviewed

20 his in-press paper in the Journal of Forensic Scientists that

21 he mentioned, some summaries of technical reports and

22 unpublished work that he has done. I reviewed his materials

23 related to the Harrington tests that he did and I got a copy

24 of the supplemental report that he provided to the Court

25 today. It was a draft copy, and I didn't get a chance to

Page 179

1 read the whole thing completely, but I did get to look at it

2 a little bit before this hearing.

3 Q Okay. And I don't know, Dr. Iacono, do you have

4 Exhibit 2 up there?

5 A No.

6 MS. KENNEDY: May I approach?

7 THE COURT: You may.

8 Q I will show you what we have admitted as Exhibit 2. Did

9 you --

10 A Yes.

11 Q Did you review Table 1 that is listed there in Exhibit 2

12 which would be Dr. Farwell's supplemental report?

13 A Yes.

14 Q And what does that table appear to be?

15 A It's a list of the technical procedures and -- Well,

16 it's a list of the various aspects of the technical procedure

17 that he uses in this line of research.

18 Q And do you think that -- In your opinion, does this list

19 reflect the use of standard and accepted procedures that

20 would be used and are generally accepted by the scientific

21 community?

22 MR. CROWL: Your Honor, I am going to make sure

23 that my standing objection pertains to this witness as well;

24 object to any testimony relative to the Harrington report

25 under 402, 403 and 702. I would like a standing objection

Page 180

1 for this witness as well.

2 THE COURT: Your objection is noted. Ruling on

3 that objection is reserved.

4 Doctor, you can answer the question.

5 A Okay. The listed procedures here are all within what

6 would be considered scientifically acceptable. There is no

7 set procedure that every single scientist would use. There

8 would be variations in procedure, but the procedures that

9 Dr. Farwell uses are within the realm of what most scientists

10 would use and would consider acceptable.

11 Q In your review of Dr. Farwell's reports that you

12 reviewed prior to coming here today, how would you say that

13 Dr. Farwell's extension to guilty knowledge test would differ

14 from a convention polygraph test?

15 A It's very different. There is really -- The only thing

16 they share in common is that a conventional polygraph test

17 and this extension both involve recording physiological

18 signals from humans, but beyond that, they really deal with

19 different topics. Dr. Farwell's technique is intended to

20 assess memory, in particular, recognition memory and whether

21 or not people recall or not recall, but recognize information

22 that's relevant to a crime. In conventional lie detection

23 techniques, we are measuring particularly autonomic nervous

24 system response; not brain responses, to determine whether or

25 not people are lying or trying to deceive in their response

Page 181

1 to different questions. And the lie detector technology is

2 very subjective and based on a number of assumptions that are

3 made about how the procedure works that are not generally

4 accepted in the scientific community. The survey that I

5 mentioned earlier, a number of the questions that we included

6 on that survey had to do with scientists' appraisal of

7 conventional polygraph procedures, and we ask a large number

8 and a wide variety of questions to scientists about these

9 procedures, and they express widespread skepticism that

10 conventional polygraph procedures are valid. They did not

11 believe they were based on sound science and sound

12 psychological principles. They felt that the accuracy rates

13 recorded by proponents of polygraphy are greatly overstated

14 and they felt there were many problems with trying to admit

15 polygraph evidence in court that relate to the way these

16 tests are administered and presented before courts. So they

17 are very different.

18 The P300 technique is one that has a solid scientific

19 basis and foundation conventional polygraphy doesn't. The

20 P300 technique is designed to assess memory, and the

21 conventional polygraphy is designed to assess whether or not

22 people are lying, and it doesn't do it very well.

23 Q Dr. Farwell, uses the term MERMER in his reports. Is

24 the MERMER a well established scientific phenomenon?

25 A No.

Page 182

1 Q How is the MERMER different from the P300?

2 A Essentially, when one looks at an event related

3 potential, it's an electrical, complex electrical wave that

4 has a duration of perhaps a second or longer, and what the

5 trick is with this application is to figure out what portion

6 of that complicated wave do you actually think contains

7 information about whether or not the individual has

8 recognition memory. What we have been talking about for the

9 most part is the fact that the portion of this wave that

10 identifies the P300 response in fact does contain this

11 information, and that's the area that most people have been

12 focused on in this line of research. What people don't look

13 at, information that precedes the P300 wave that is the early

14 variations and electrical activity that occur, for instance,

15 in the first two-tenths of a second after an event is

16 presented, because this information is not as interesting, is

17 not as useful for determining whether or not people have

18 recognition memory. But what Dr. Farwell has done is

19 basically looked beyond the P300 wave; that is, in the time

20 interval that follows it, to see if there is also information

21 there that might be useful to determine whether or not a

22 person has recognition memory. In his overheads or projected

23 power point presentation from this morning, he showed that

24 there was a positive going wave that is the P300 we have been

25 talking about, and that that was then followed by an

Page 183

1 electrical dip or negativity. The MERMER involves taking

2 into account information that's contained in the P300

3 response as well as in this negativity that follows it.

4 Q Would you say that the MERMER is a reasonable extension

5 of the valid science that preceded it?

6 A Yes, I would. In my own research, for example when I

7 look at the -- using this technique where we look at people's

8 responses to targets, probes and irrelevants, just as was

9 done in this Harrington case and in Dr. Farwell's work, we

10 can see the same negative dip that occurs after the P300

11 wave. So essentially he's trying to take advantage of the

12 additional information that's contained in the later part of

13 the response.

14 Q Did you review that part of Dr. Farwell's reports where

15 he reanalyzed the data using just the P300?

16 A Yes.

17 Q And how did those results compare?

18 A His results using just P300, and not using what he calls

19 the MERMER, were essentially identical.

20 Q Would there be anything about Dr. Farwell's use of the

21 MERMER that would invalidate the results from the P300?

22 A Not that I am aware of.

23 Q You have stated, Dr. Iacono, that the detection of

24 concealed information is a well established science using the

25 P300. You have also stated that the CQT is well accepted in

Page 184

1 the scientific community. If you look at Dr. Farwell's

2 technique in detecting concealed information, would the

3 Harrington case fall within the realm -- the use of the

4 science in Dr. Farwell's reports on the Harrington case,

5 would that fall within the realm of those well established

6 principles?

7 A You know, that was an awful lot a of words.

8 Q It was an awful lot of words; really bad question.

9 Why don't I just move on. I think our time is getting

10 pretty short here. Is there a fundamental difference between

11 information stored in the brain regarding a crime and

12 information stored in the brain regarding anything else?

13 A No. The only difference might be that information

14 stored in the brain related to a crime is likely to be, you

15 know, very significant to the person in a special way, but

16 basically how the information got in there, if it's recalled,

17 I mean recognized in the context that we are talking about

18 here, I don't think it's going to make any difference.

19 Q In your opinion, could -- If Dr. Farwell were using this

20 technique on criminals, would there be a way they could beat

21 this system or somehow get around his tests?

22 A If there is a way for criminals to beat the test, it's

23 not known what it would be. You could imagine people trying

24 different things, but since it's a novel technique and

25 criminals certainly wouldn't know about it, it would be very

Page 185

1 difficult for them to decide what to even try to try and

2 defeat the technique. One of the advantages of this

3 procedure over the version of the guilty knowledge tests

4 where autonomic nervous system measures are recorded like the

5 galvanic skin response, is that it's easy for people to

6 manipulate autonomic nervous system responses. They can

7 tighten up a muscle in their body, hold their breath, curl

8 their toes. All these sorts of things will affect an

9 autonomic nervous system response. But the brain event

10 related potential is not something that people can easily

11 understand how to manipulate, and because it has a certain

12 timing and a certain shape, a certain characteristic shape,

13 it's also the case that if people probably tried to do

14 something to manipulate their brain electricity, it would

15 distort the signal in a way that would make it look odd and

16 might make it stand out. So I think at this point it's

17 pretty unlikely that people would know what to do. They

18 certainly wouldn't have a good idea going into it about how

19 successful anything they tried would be and it's highly

20 likely if they tried a bunch of things, it would be detected

21 in some way as producing some sort of odd-looking response,

22 that that would produce an inconclusive result or some kind

23 of concern that the person wasn't cooperating.

24 Q What can we conclude with reasonable scientific

25 certainty about the results that Dr. Farwell obtained on the

Page 186

1 tests given to Mr. Harrington?

2 A I think you can conclude several things. In the first

3 place, you can conclude that Mr. Harrington's brain works

4 like most people's brains in the sense that when he is

5 supposed to recognize stimuli related to the crime, he does.

6 We know that because the target stimuli in his tests

7 generated a strong P300 wave. We can also conclude that he

8 does not generate a large P300 wave in response to the probe

9 stimuli. The probe stimuli are the stimuli that deal with

10 guilty knowledge, or the hidden or concealed information we

11 might expect him to have. What this test shows is that there

12 is no P300 response to this information, which makes it

13 unlikely that he is recognizing the probe stimuli. We can

14 also conclude that, I think, that the -- by comparing the

15 test that was given on crime related information to the test

16 that was given on the alibi relevant information, that he was

17 capable of remembering things from 23 years ago, including

18 things that don't seem very important, because the alibi

19 probes dealt with items that, you know, didn't seem

20 overwhelmingly significant, yet nonetheless he showed the

21 P300 response to that information. So those are the things

22 that I think you can conclude from these -- the Harrington

23 tests.

24 Q Do the results provide relevant information

25 considering -- in the consideration of Harrington's guilt or

Page 187

1 innocence?

2 A I think they do provide relevant information, because I

3 think they clearly indicate that he doesn't recognize the

4 memories related to the probes and he does recognize the

5 memories related to the alibi information. I don't think

6 that from that you would conclude that he was guilty or

7 innocent. That would require going further and making

8 inferences about how reasonable it is that he should actually

9 know this sort of information about the crime, but the basic

10 test that assesses whether or not he has recognition memory,

11 I think the results of that test is clear. Why he doesn't

12 show recognition memory, I don't think that this procedure

13 explains why he doesn't. One possible reason why he wouldn't

14 is because he's innocent.

15 Q And one last question, Dr. Iacono. Could the test that

16 Dr. Farwell administered to Terry Harrington be given as long

17 as 23 years ago?

18 A It's pretty -- It's inconceivable, really, that it would

19 have been given then, based on the state of the field at that

20 point in time. Twenty-three years ago, even the development

21 of the guilty knowledge test was fairly preliminary, and so

22 in the time elapsing since then, we have not only done a lot

23 more research on the guilty knowledge test, but this whole

24 elaboration of the guilty knowledge test which involves the

25 recording of event related potentials has just essentially

Page 188

1 come on the scene starting with papers that Farwell and

2 Donchin presented at the Society for Psychophysiological

3 Research meetings in the 1980's, and in particular, in the

4 late 1980's, and others have followed up, including us and

5 Dr. Rosenfeld, who has been mentioned a couple of times, and

6 some other investigators.

7 Q One last question. Based on your knowledge of

8 Dr. Farwell's reputation and his publication and research,

9 would you consider him an expert in the area of

10 psychophysiology?

11 A Yes.

12 MS. KENNEDY: Thank you, doctor. No further

13 questions.

4 EMANUEL DONCHIN

5 being produced as a witness on behalf of the defendant,

6 and having been first duly sworn by the Court, was examined

7 and testified as follows:

8 DIRECT EXAMINATION

9 By Mr. Crowl:

10 Q Doctor, you have in front of you Exhibit 9. Can you

11 identify that?

12 A Yeah. That's my curriculum vitae.

13 MR. CROWL: State offers Exhibit Number 9.

14 MR. FRERICHS: No objection.

15 THE COURT: 9 is received.

16 Q What have you reviewed for today, doctor, what

17 documents?

18 A Oh, I have seen the report on the Harrington case, I

19 have seen the -- I have actually seen all the documents that

20 are available on the Farwell site, whatever it's called, and

21 I have downloaded the patents and read them, not in greatest

22 detail, and I have seen this morning when I arrived the

23 question and answer format addendum report from Dr. Farwell.

24 Q That addendum report on the Harrington case?

25 A Right.

Page 203

1 Q Do you agree with the conclusions reached by Dr. Farwell

2 in the Harrington report?

3 A Well, I don't think I can agree or disagree with those

4 conclusions, because I don't know the case. There are two

203, 1 - 203, 4

Page 203

1 Q Do you agree with the conclusions reached by Dr. Farwell

2 in the Harrington report?

3 A Well, I don't think I can agree or disagree with those

4 conclusions, because I don't know the case.

208, 7 - 208, 10

7 Q And again, that's part of the art to all this, how you

8 interpret the results?

9 A That's interpretation, yeah. That's right. That's just

10 like fingerprints.

209, 4 - 209, 6

4 . . . It was somewhat of an art to come up with

5 irrelevants, none of which would be like the probes, and that

6 required some work,. . .

209, 21 - 210, 16

P. 209

22 Q As I understand your testimony, you simply can't agree

23 or disagree with the conclusions reached by Dr. Farwell in

24 his reports; is that correct?

25 A Yeah.

Page 210

1 Q In order to disagree, you would have to know a lot more

2 about the kind of stimuli that he used as he conducted the

3 test; is that correct?

4 A Yes. I say, from my knowledge, I can interpret the wave

5 forms, and if I had seen all the data, I might be able to

6 interpret the data. I'm talking about the P300. The memory

7 stuff is a completely different story, but the P300 I could

8 interpret, but the meaning of that within any context is it

9 has to be interpreted within the context in which the test

10 was made and whether the probes were good probes or bad

11 probes, or whether he rehearsed the alibi in his mind for 17

12 years or not, is a matter of judgment.

13 Q And basically, so what you are saying is you have not

14 prepared or looked at the case to the same extent Dr. Farwell

15 did when he prepared the test?

16 A I should hope so.

17 Q And as a result, you are not in a position to say that

18 his test is either inaccurate or accurate?

19 A I'm testifying as an expert on psychophysiology,

20 cognitive neuroscience in the P300; not the Harrington case.

210, 21 - 211, 8

21 Q Fine. Let's stick with that then, because when you talk

22 about the P300 and the testing procedures, a lot of that

23 basically grew out of a report that you authored with

24 Mr. Farwell; is that correct?

25 A That's correct, yeah.

Page 211

1 Q And in that report, that report itself was subjected to

2 peer review and all of the types of things that would be

3 necessary for it to receive the type of publication --

4 A Yeah.

5 Q And as a result, that study has been widely accepted

6 within the scientific field of psychophysiology; is that

7 correct?

8 A I would say so, yeah.

216, 23 - 217, 1

23 What science does, it says if you give me these probes,

24 the science says if you get a P300, they are discriminated,

25 they are recognized. If you don't get a P300, they are not

Page 217

1 recognized.

217, 24 - 218, 3

23 all. What science does, it says if you give me these probes,

24 the science says if you get a P300, they are discriminated,

25 they are recognized. If you don't get a P300, they are not

Page 217

1 recognized. Whether those probes allow you to make any kind

2 of legal inferences or inferences about what somebody did 23

3 years ago, that's not science. That's art.

220, 2 - 220, 14

Page 220

2 A What I'm saying is that the procedure -- the procedure

3 which Dr. Farwell calls brain fingerprinting, which is

4 essentially identical to the procedure we published in 1991,

5 has a scientific component to it which is generating --

6 running an odd-ball paradigm, which is the technical term for

7 the kind of stimulus Dr. Farwell is using, and this odd-ball

8 paradigm, the rarest and frequents, or rarers and

9 irrelevants, and we embed in this technique in the

10 irrelevants, probes, and if the probes are responded to, the

11 science of P300 would agree that means that the subject has

12 clearly identified the probes as different from the

13 irrelevants. That's where the science of psychophysiology

14 ends.

222, 19 - 223, 22

19 Q Then when you are saying that the Farwell-Donchin study,

20 which included both Experiment 1 and Experiment 2, is widely

21 accepted within the scientific community, you are including

22 within that acceptance of the scientific community, the art,

23 as you termed it, that was included in Experiment 2; aren't

24 you?

25 A Well, they are aware that if you skillfully develop the

Page 223

1 probes and the skillfulness depends on your art, you will get

2 useful interpretation of P300. I didn't say that using art

3 is illegal. I just say that -- or I just said it's not

4 scientific. It is a component of the work that depends on

5 skill and judgment and subjective evaluation, and not on

6 scientific procedure.

7 Q But that art, as you have termed it, as witnessed in the

8 Farwell-Donchin study, has been accepted within the

9 scientific community; hasn't it?

10 A Yes, because those particular probes made sense relative

11 to the case we were discussing.

12 Q Just as the probes that Dr. Farwell may have used in the

13 Harrington case make sense; is that correct?

14 A It may or may not. I have no idea.

15 Q Right. And the only way you would be able to know that

16 would be to know the case better than you actually do. Is

17 that fair to say?

18 A Absolutely. That's -- That's true. For any application

19 of this technique, there is going to be an art component,

20 subjective judgment component, and to the extent that --

21 That's not important. That's fine. I'm just giving you the

22 realities of the situation.

225, 24 - 226, 3

24 Q But let's get this. You would agree; wouldn't you, with

25 respect to Experiment 2, since you included it in your

Page 225

1 publication, that Dr. Farwell was skillful in developing

2 those probes; is that correct?

3 A Well, yeah.

BACK HARRINGTON LINKS HOME