Tribunal Criminal Tribunal for the Former Yugoslavia

Page 8614

1 Monday, 19 March 2001

2 [Open session]

3 --- Upon commencing at 9.38 a.m.

4 [The accused entered court]

5 JUDGE RODRIGUES: [Interpretation] Good morning, ladies and

6 gentlemen; good morning to the technical booth and the interpreters; good

7 morning to the staff of the registry.

8 Madam Registrar, could you call the case, please. We haven't been

9 sitting for some time.

10 THE REGISTRAR: Yes. Good morning, Your Honours. This is the

11 case number IT-98-33-T, the Prosecutor versus Radislav Krstic.

12 JUDGE RODRIGUES: [Interpretation] May we have the appearances,

13 please. The Prosecution, Mr. Harmon, first.

14 MR. HARMON: Good morning, Mr. President; good morning, Your

15 Honours; and good morning, counsel. My name is Mark Harmon and assisting

16 me in the Prosecution is Mr. Peter McCloskey. Thank you.

17 JUDGE RODRIGUES: [Interpretation] Thank you, Mr. Harmon.

18 Mr. Petrusic, please, the Defence.

19 MR. PETRUSIC: [Interpretation] Good morning, Your Honours. The

20 Defence is represented by me, Nenad Petrusic, and my colleague,

21 Mr. Visnjic, Tomislav Visnjic, as has been the case hitherto.

22 JUDGE RODRIGUES: [Interpretation] Thank you very much. We have

23 two decisions, two rulings to make, and we are going to render those

24 rulings orally. And there will be no written ruling.

25 On 26 February 2001, the Defence submitted two motions. The first

Page 8615

1 seeks "to exclude alleged statements of the accused" and the second

2 regards disclosure of exculpatory evidence. The Prosecution opposed the

3 motions when it responded to them on 14 March 2001. The Defence presented

4 a reply in relation to the first motion.

5 I - Motion to exclude alleged statements of the accused

6 The issue concerns recordings of tapped conversations held at the

7 time of the facts on which the Prosecutor contends the voice of General

8 Krstic can be heard (saying in particular "kill them all") and which the

9 Prosecution would like to have admitted as evidence. The Defence objects

10 to such admission on the ground that: (1) the tapped conversations

11 constitute statements of the accused within the meaning of Rule 66 of the

12 Rules; and (2) the evidence concerns the merits of the case and as such

13 cannot be admitted during the Prosecutor’s rebuttal.

14 (1) The Chamber observes that, by having the Prosecution disclose to

15 the Defence the statements of the accused and of the witnesses, the

16 purpose of Rule 66 of the Rules is to allow for a better preparation of

17 the defence and, generally, for respect for the rights of the Defence.

18 However, the Chamber points out that Rule 66 consists of two major

19 parts. Although Rule 66(A) requires disclosure of the statements of the

20 accused, Rule 66(B) makes it clear that the Prosecutor is in no manner

21 bound to disclose all the books, documents and other evidence in her

22 custody. She is bound to do so only if the Defence initiates the

23 "reciprocal disclosure" procedure laid down in Rule 67 of the Rules.

24 The Prosecution would therefore have been obliged to disclose to the

25 Defence the tapped conversations at issue only if they constituted

Page 8616

1 statements of the accused in the literal sense. Such is not the case. The

2 Chamber is of the view that "statements of the accused" within the meaning

3 of Rule 66 of the Rules are to be understood as the statements made by the

4 accused, spontaneously or not, only when he is aware that what he is

5 saying is being recorded somehow during any type of police or judicial

6 procedure.

7 Such is evidently not the case here. The "tapped

8 conversations" at issue must be dealt with as "documents"

9 within the meaning of Rule 66.

10 The motion is therefore dismissed in this regard.

11 (2) As regards the tardiness of the Prosecutor’s

12 use of the tapped radio conversations, the

13 Chamber regrets that the Prosecution did not

14 present the documents at a more timely moment

15 for the conduct of the whole trial.

16 The documents are incontestably relevant and,

17 assuming that they are admitted, they are

18 among those which clearly concern the merits of the

19 case at least as much as the credibility of the accused.

20 That said and whatever their ultimate probative value,

21 the Chamber observes without however sharing all of the

22 Prosecutor’s reasons that the Defence had a reasonable

23 time to present any objections to the documents.

24 Indeed, the Defence was informed

25 at the beginning of General Krstic’s


Page 8617

1 cross-examination by the Prosecutor that she intended to

2 request that the documents be admitted during her rebuttal.

3 As such, the Defence had sufficient time to prepare

4 itself, in such a way that the principle of balance between

5 the parties cannot be considered to have been breached.

6 Since the documents are therefore relevant,

7 the Chamber will decide on whether to admit them

8 after examining all of the explanations provided

9 by the Prosecution and the Defence during the rebuttal

10 and rejoinder.

11 The Defence’s motion on the tapped radio conversations

12 of General Krstic is thus dismissed.

13 So those are the rulings of the Chamber with respect to the

14 Defence motion to exclude the assumed statements by the accused. We have

15 also stated that it will be necessary to rule on the motion on the

16 disclosure of exculpatory evidence, but before doing so, the Chamber would

17 like to have some clarification on the part of the Defence to see whether

18 we need or not to make a ruling on that question.

19 The information has to do with a note at the bottom of the page of

20 the response of the Defence regarding the motion to the exclusion of

21 statements by the accused, and the footnote, I have the text in English so

22 I'm going to read it in English: [In English] "The Defence also filed a

23 motion for disclosure of exculpatory evidence. It is satisfied from the

24 Prosecutor's disclosures subsequent to filing the motion and from the

25 Prosecutor's representations in their response that the Defence has

Page 8618

1 received the essence of what it sought, evidence that no Muslims were

2 killed as a result of General Krstic's alleged statements on 2 August."

3 [Interpretation] Therefore, I'm going to ask the Defence whether that

4 means that the motion has been withdrawn, is no longer valid, or any other

5 relevant information.

6 Mr. Visnjic, please.

7 MR. VISNJIC: [Interpretation] Good morning, Your Honours; good

8 morning my learned friends opposite.

9 Mr. President, from the moment the Defence submitted this request

10 or this motion to the present, the Prosecution has in the meantime

11 disclosed to the Defence a certain quantity of documents which would meet

12 almost all the requests of the Defence as listed in our motion. The only

13 thing that remains substantially disputed between the Defence and the

14 Prosecution are the statements of two witnesses, and they are witnesses DB

15 and Obrenovic, with whom the Prosecution has conducted interviews. And in

16 connection with our motion for disclosure, we were informed by the

17 Prosecution that both witnesses during their interview were not questioned

18 at all in connection with the critical intercepts of the 2nd of August.

19 If the Prosecution confirms that to us, in that case we consider that the

20 majority of our requests have been fulfilled and we would therefore

21 withdraw this motion.

22 JUDGE RODRIGUES: [Interpretation] Very well, Mr. Visnjic.

23 Mr. Harmon, have you any comments to make in response to this

24 statement regarding the withdrawal of the motion?

25 MR. HARMON: Mr. President, we understood footnote 1 to be a

Page 8619

1 motion to withdraw the application that had been filed with the Trial

2 Chamber, but in response to the question raised by my colleague

3 Mr. Visnjic, we addressed that issue in paragraph 19 of our motion. I

4 have been informed that in the course of the interviews of two individuals

5 who have been referred to by my colleague, that neither was asked about

6 the conversation in question, and based on that information I then

7 included that in paragraph 19 and confirm it orally today. Thank you.

8 JUDGE RODRIGUES: [Interpretation] Very well. Thank you very

9 much. So we don't need to make any ruling about this since the Defence

10 has confirmed that it has withdrawn its request.

11 It is now time for us to resume our case. As you have seen, after

12 the Prosecution's rebuttal and the rejoinder of the Defence, we will then

13 decide on the admission of this document which was the object of

14 discussion between the parties.

15 I see that Mr. Visnjic has a question to raise.

16 MR. VISNJIC: [Interpretation] Mr. President, I just wish to notify

17 the Chamber that the Defence intends to appeal this ruling by the Chamber,

18 and therefore the Defence suggests that the proceedings be suspended until

19 a final ruling on appeal. However, in the meantime, we can continue with

20 some other evidence that has been part of the Prosecution's rebuttal

21 case. So I'm referring to the decision rejecting the motion for the

22 elimination of the alleged statements of the accused.

23 JUDGE RODRIGUES: [Interpretation] Thank you very much,

24 Mr. Visnjic.

25 Mr. Harmon, do you have any response to make? Not regarding the

Page 8620

1 appeal, but regarding the suggestion for suspension.

2 MR. HARMON: Yes. Mr. President, we oppose the motion to suspend

3 these proceedings. We have, I think, 15 or 13 witnesses who are coming,

4 who have been scheduled. We are prepared to present their testimony. We

5 have a finite amount of time - one week - in which to present our case and

6 we're prepared to move forward. It would be a huge disruption in the

7 orderly proceedings at this point to suspend the proceedings that have

8 been scheduled to allow the Defence an opportunity to file an appeal.

9 It's their right to file an appeal, they can do so when they want, but to

10 suspend these proceedings would not be a very prudent decision at this

11 point in time and we oppose the application to suspend these proceedings.

12 JUDGE WALD: I would just like to -- Mr. Visnjic, I would just

13 like to make sure that you understand -- your understanding is the same as

14 mine of the President's ruling, and that is that thus far his only ruling

15 has been that these intercepts do not fall within Rule 66 definition of a

16 statement. It is yet -- the question of authentication of those, which

17 would be prior to any admission, has not been reached, or indeed the

18 evidence hasn't been put in by the Prosecution; and secondly, we have not

19 yet ruled on whether, despite -- assuming they are properly authenticated,

20 they are indeed appropriately put in in rebuttal as opposed to the case in

21 chief. So those rulings, which will be forthcoming, have not yet been

22 made. I just want to make sure that we all understand that.

23 [Defence counsel confer]

24 MR. VISNJIC: [Interpretation] Your Honours, I must admit that that

25 was not my understanding of the ruling. I have just consulted with my

Page 8621

1 colleague, Mr. Petrusic. So my understanding was that our motion was

2 rejected as a whole. So I now understand it's only the interpretation

3 according to Rule 66 (A). So our motion to suspend would not be

4 appropriate. We are just announcing our intention to file an appeal, and

5 we may perhaps later on make another suggestion, depending on how we

6 proceed.

7 JUDGE RODRIGUES: [Interpretation] Very well. Thank you very much,

8 Judge Wald. Thank you, Mr. Visnjic, too. So we are going to continue our

9 work and our proceedings. And of course you are fully entitled to appeal.

10 I would now like to give the floor to the Prosecutor to present,

11 in line with Rule 85, its rebuttal.

12 MR. McCLOSKEY: Good morning, Mr. President, Your Honours. Our

13 first witness today is Dr. Peter French. He has been asked to review the

14 tape for a voice comparison matching analysis and a voice tamper analysis.

15 We call him as our first witness slightly out of order because he has

16 court engagements in the UK on Wednesday, and so as opposed to the

17 intercept operators that took down this material and listened to these

18 tapes, we would ask him to be the first witness.

19 And because it's slightly out of order, I think I should give you

20 a little bit of background because there is a new exhibit in this case.

21 And as you may recall from the motions that went back and forth in, I

22 think, especially the footnote material, the Prosecution at about shortly

23 before General Krstic's testimony, another team had a consent search of

24 the Bosnian Muslim army archives and where they recovered from the

25 archives in Sarajevo numerous reel-to-reel tapes, much like the tapes that

Page 8622

1 we've already played in evidence.

2 After General Krstic was through testifying, our analysis and

3 review found another version of the so-called August 2nd "kill them all"

4 tape, and so we sent that tape as well as the original tape that you heard

5 to Dr. French for his overall analysis and comparison. So he will be, in

6 speaking to you today, be referring to both tapes, as well as will the

7 other witnesses that you'll be hearing from.

8 JUDGE WALD: Does the Defence have the second tape? Do they know

9 about it?

10 MR. McCLOSKEY: Yes, Your Honour. As soon as we found out about

11 it, we gave them a copy, a transcript, and fully informed them of that

12 tape. And we have actually been working with the Defence and their expert

13 and our expert to make sure they had a full chance to review the material.

14 In fact, their expert did go to the UK and spend two days with our expert

15 reviewing the material, and we have not received a report from their

16 expert yet, but if there is one, we hope to get it.

17 So as far as I know, they are fully apprised of all the material,

18 the notebooks, the printouts, the tape.

19 JUDGE RODRIGUES: [Interpretation] Mr. McCloskey, allow me to ask

20 the Defence whether they have any objections outside this order.

21 MR. VISNJIC: [Interpretation] No, Mr. President.

22 JUDGE RODRIGUES: [Interpretation] Very well, then. We can then

23 call the witness, Mr. McCloskey.

24 MR. McCLOSKEY: Dr. French, please. And perhaps we could have his

25 CV ready for him. It's Exhibit 840/A.

Page 8623

1 [The witness entered court]

2 JUDGE RODRIGUES: [Interpretation] Good morning, Dr. Peter French.

3 Can you hear me?

4 THE WITNESS: I can. Thank you, Mr. President.

5 JUDGE RODRIGUES: [Interpretation] You will now read the solemn

6 declaration given to you by the usher, please.

7 THE WITNESS: I solemnly declare that I will speak the truth, the

8 whole truth, and nothing but the truth.


10 JUDGE RODRIGUES: [Interpretation] Please be seated.

11 THE WITNESS: Thank you.

12 JUDGE RODRIGUES: [Interpretation] Thank you very much for coming,

13 Dr. French. You will first be answering questions put to you by the

14 Prosecutor. It is Mr. Peter McCloskey, who is standing to your right.

15 Mr. McCloskey, your witness.

16 MR. McCLOSKEY: Thank you, Mr. President.

17 Examined by Mr. McCloskey:

18 Q. Dr. French, could you first tell the Court what your professional

19 position is, and then if you could tell the Court your educational

20 background and your training background that led you to that profession.

21 A. Certainly. I'm an independent forensic consultant specialising in

22 the analysis of tape recordings, speech, and language samples. I have a

23 higher degree in linguistics with phonetics - phonetics being the science

24 of speech sounds - and a doctoral degree - that's a PhD - in the analysis

25 of tape-recorded conversation.

Page 8624

1 I am a member and a fellow of a number of professional

2 associations. I'm currently chairman of the International Association For

3 Forensic Phonetics, and that is the professional association which takes

4 responsibility for work in the area in which I appear today. And I'm -- I

5 was awarded a fellowship of the Institute of Acoustics, acoustics being

6 the physics of sound.

7 I wonder, Mr. McCloskey, rather than actually going through my

8 curriculum vitae a line at a time, I think for purposes of Mr. President

9 and Their Honours, there's probably nothing worse than just listening to

10 someone rattling off their own achievements. Perhaps if they have this in

11 writing, they could take it away and we would save court time.

12 Q. And that's Exhibit 840/A that you should have in front of you, and

13 the Court will also have access to it.

14 A. That's correct.

15 Q. I think that's a good idea. Can you briefly, though, tell us,

16 have you testified in court before on your profession?

17 A. Yes, I have. I say in the written CV that it's about 200 times.

18 I think it's probably around 300 now. I've been consulted in 3.000 or

19 more cases which have involved many more thousands of examinations of

20 individual tapes. In criminal cases, I act for both the Prosecution and

21 Defence I like to say in roughly equal proportions, but I probably do just

22 slightly more Prosecution work.

23 Q. Now, you mentioned phonetics and linguistics. Can you give us a

24 brief, a little more detailed brief description of what it is you really

25 do in practical terms for the Prosecution and Defence.

Page 8625

1 A. Yes. The work that I do in relation to tape recordings falls into

2 a number of different types. Firstly, if tape recordings are produced

3 which are of bad quality, which are contaminated by noise, I'm asked to

4 filter them, to enhance them, to remove the noise in order to make the

5 conversation more audible or intelligible. I also examine tape recordings

6 for any evidence of their having been tampered with or edited before

7 they're put into court as evidence.

8 I compare voices in recordings. This is called forensic speaker

9 identification work where there will be the voice of a known suspect,

10 usually on one tape, and the voice which is questioned in a criminal

11 recording, so I would compare the voices in order to attempt to identify

12 or eliminate the suspect.

13 In addition to that, I'm sometimes asked to determine what's said

14 on a tape. These are cases where perhaps the person has a very

15 non-standard dialect or they speak English with a very heavy foreign

16 accent or the tape is in some way problematic. What I would do is to

17 undertake a study of those speech patterns and, based on that, examine

18 what's said in the tape and give a view.

19 Q. So I take it your voice expertise is mainly in English, your

20 mother tongue?

21 A. Yes, it is.

22 Q. Now, when you are analysing the voice or language of someone in a

23 foreign tongue, do you generally bring on a consultant that's an expert in

24 that foreign tongue?

25 A. Yes. I would want someone who was -- who have thorough knowledge

Page 8626

1 of that language, either as a native speaker or had research knowledge of

2 that language. I would also want that person to be a linguist or a

3 phonetician. The reason that I need someone in that capacity is that, in

4 attempting to identify speakers, one is looking for idiosyncratic or

5 personally distinguishing features of speech, in other words, departures

6 from the norm. In order to identify a departure from the norm, one must

7 know what the norm is for that language or the variety of that language in

8 question.

9 Q. And in this particular case, I believe you consulted with a

10 witness who we will have as a protected witness, so we would refer to him

11 as Witness EE. Is that right?

12 A. I did, yes.

13 Q. And what did we ask you to do in this particular case, and then

14 what did you review and what did you do?

15 A. I was asked to undertake two main tasks. The first one was to

16 examine a tape recording which occurred on two different open-spool

17 tapes. And I don't know whether we should identify these for the Court

18 now, Mr. McCloskey.

19 Q. Yes, that's fine.

20 A. Yes. These are the tapes in front of me; one known as, I think,

21 tape 48, and one known as tape 32.

22 MR. McCLOSKEY: And for the record, tape 48 is Exhibit 789, and

23 that is the new tape you'll be hearing; and tape 49 was Exhibit -- excuse

24 me. Tape 48 was the old tape, the first one you heard, and tape 49 is

25 Exhibit 835, and that is the new tape.

Page 8627

1 A. Right. My attention was drawn to a recording of conversation

2 which occurred on both of these tapes, and I was given a transcript of

3 this in which three speakers had been identified, one as General Krstic,

4 one as Lieutenant Colonel Obrenovic, and one as Lieutenant Colonel

5 Jevdjevic.

6 JUDGE RODRIGUES: [Interpretation] Excuse me, Dr. French, for

7 interrupting, but I think there's something we must clear up now. I think

8 that you mentioned a tape 48 and 32, I think you mentioned, and

9 Mr. McCloskey spoke of tape 49 and 48. So could we clear that up, please,

10 and the correspondence between those numbers and exhibit numbers.

11 So Mr. McCloskey, please, could you clear that up.

12 MR. McCLOSKEY: You're right, Mr. President, and I'm sorry. The

13 witness is well familiar with the different numbers that have been

14 assigned to these tapes.

15 Q. So perhaps you could -- tape 32 is also tape 49, which is Exhibit

16 835.

17 A. That's correct, yes.

18 JUDGE RODRIGUES: [Interpretation] Please continue. Thank you.

19 MR. McCLOSKEY: And that is the new tape.

20 A. As I said, there was a transcript of the conversation in question

21 and, on the transcript, speech had been variously attributed to General

22 Krstic, Major Obrenovic, and Major Jevdjevic. I was asked to consult

23 other tapes that were provided to me - these were tape recordings of UN

24 interview proceedings in which the three officers had been interviewed -

25 and to compare the voice and speech patterns found on the questioned

Page 8628

1 recordings with the known voice and speech patterns of the three officers

2 from their interview recordings, the purpose being to try to identify or

3 eliminate them from participation in the questioned conversation, which

4 was an intercepted military radio transmission.

5 I consulted Witness EE over this, Witness EE being an expert in

6 Slavian (sic) languages and in B/C/S in particular. We worked together on

7 this and carried out a number of tests.

8 I should say that in forensic speaker identification and

9 elimination cases, the tests one carries out are normally looked on as

10 falling into two groups. On the one hand, there are tests which are known

11 collectively as auditory phonetic tests, which is perhaps a rather

12 highfalutin way of saying listening tests. They are listening tests which

13 are based on university ear training in phonetics.

14 What one listens for is, firstly, how the individual consonant and

15 vowel sounds are pronounced. So just to give an example, in the ordinary

16 Roman alphabet, which we all use to write with, or most of us use to write

17 with, we have only 26 letters. But in carrying out an auditory phonetic

18 analysis, one has available an extended alphabet which has been developed

19 by the International Phonetic Association for capturing the fine-grained

20 nuances of pronunciation. This would allow one to find discriminations

21 which the Roman alphabet would not.

22 For instance, irrespective of how one pronounces a "T" sound, or

23 "ta", using the Roman alphabet one would have available only one letter

24 "T" for recording how it had been pronounced, whereas using the IPA

25 system, the International Phonetic Association system of symbols, one

Page 8629

1 could discriminate between different subvarieties of "ta" sound; some

2 perhaps made with the tip of the tongue further forward against the upper

3 teeth, some with the tip of the tongue curled backwards, some with the tip

4 of the tongue on the tooth ridge, and so on.

5 And I pick out "T" just as an example. The same would apply to

6 any of the other consonant or vowel sounds.

7 So one part of the auditory phonetic testing involves going

8 through the questioned recordings and the known samples - in this case,

9 the interview recordings - and making a note of how the consonants and

10 vowels are pronounced to see if they match across. Now, that part of the

11 work I think is perhaps better reported on by Witness EE, because there

12 are obvious accent and dialect considerations there and he takes primary

13 responsibility for that element of the report.

14 Also under the auditory phonetic testing one would be listening to

15 the rhythms of the speech, the intonation - that's the rise and fall of

16 pitch across sentences - and to the voice quality. Voice quality refers

17 to the resonances of the speech, whether it's a creaky voice, a husky

18 voice, a breathy voice, a nasal voice, or perhaps, like my voice, an

19 adenoidal voice. People ask me if I've got a cold when I haven't.

20 At the end of the auditory phonetic testing, one would then bring

21 into play another set of tests known as acoustic tests. Nowadays they

22 involve re-recording the samples, the interview samples and the questioned

23 samples, within a computer system and running various programmes to take

24 physical measurements of elements of the speech. So for instance, in this

25 case, we measured the average fundamental frequency. The fundamental

Page 8630

1 frequency refers to the rate at which the vocal cords are vibrating when a

2 person speaks. Behind the Adam's apple, here in the larynx, there are two

3 folds of skin -- two folds of muscle, sorry, known as the vocal folds or

4 the vocal cords, and when we speak, when we produce vowel sounds and

5 certain consonants, these vibrate backwards and forwards very, very

6 rapidly. The more rapid the rate of vibration, the higher the pitch of

7 the voice as we hear it. And this was calculated for both the known

8 samples and the questioned speech attributed to the three officers.

9 It was also possible to undertake a very limited amount of what's

10 known as spectrographic analysis. This involves looking at the

11 distribution of sound energy across voice frequencies across time. But

12 the only person, for reasons I'll come on to at the moment, that it was

13 possible to do any spectrographic work with at all, and that was very

14 limited, was the speech attributed to Lieutenant Colonel Obrenovic.

15 I should say in this case the comparisons that we undertook were

16 made rather difficult by two main factors. One was that the radio

17 transmitted speech - this is the questioned recordings on the military

18 intercepts - was a very restricted frequency range. And I don't know

19 whether I may, Mr. President, Your Honours, make references to my notes

20 and report on this matter.

21 JUDGE RODRIGUES: [Interpretation] Mr. McCloskey, I think it is up

22 to you, perhaps, to answer that question.

23 MR. McCLOSKEY: Yes. If that would assist your testimony, please

24 do.

25 A. Thank you.

Page 8631

1 MR. McCLOSKEY: And I may note for the record, he's written two

2 reports, one involving the examinations of speech, which is OTP Exhibit

3 839/A, and then the matter which he'll get to a little later, the report

4 on the audio recordings, which is Exhibit 838/A.

5 A. Yes. As I mentioned earlier, the recording occurred in two forms,

6 one on each of two tapes. The one on the first tape, as it's been

7 referred to, that's tape 48, there was little speech energy at frequencies

8 above 2.3 kilohertz. That's 2.3 vibrations per second. I used that tape

9 to look at the questioned voice of Major Obrenovic.

10 The other tape, tape 32, there was a slightly better signal on

11 that tape than there was on tape 48 of the speech attributed to General

12 Krstic and Major Jevdjevic. And on that tape there was little speech

13 energy at frequencies higher than 1.7 kilohertz. Now, these are very

14 thin, narrow, restricted frequency bands, and that will become clear if

15 one considers that a telephone conversation, which is also frequency

16 limited, extends normally up to about 3.4 kilohertz. So these were very

17 narrowly restricted in frequency band, which detracts from the quality and

18 it detracts from the amount of analysis that can go into them.

19 So the spectrographic tests could only be done on the voice of

20 questioned Obrenovic on tape 48, and then only a small amount of work

21 looking at energy at the low speech frequencies.

22 The second factor which restricted the amount of testing on the

23 strength of the conclusions was the fact that in the military

24 interceptions, the people appeared to be shouting. This was less so

25 perhaps with General Krstic than with the other two speakers, but the

Page 8632

1 quality was very much like "dobro! dobro! dobro!" It was strained and it

2 was shouted as though people were transmitting over long distances. And

3 again, that limited the amount of testing because one was comparing those

4 voices with voices in interviews where people were speaking in a rather

5 slow, deliberate and measured way.

6 At the end of the testing, the results were inconclusive. None of

7 the tests that we carried out would allow us to identify any of the three

8 officers, even as a matter of probability. The most that can be said is

9 that we couldn't eliminate them. There were some similarities in voice

10 quality, but these were difficult to assess because of the shouting versus

11 non-shouting. The shouting also affected the pitch of the voice, the

12 fundamental frequencies. One would expect the military communications to

13 be higher, which they were. At the end of it, the most is -- we could say

14 is that it remains possible, an open possibility, that the three speakers

15 in the military intercepted communication were Obrenovic, Krstic, and

16 Jevdjevic.

17 For completeness, I should perhaps say that I normally work with a

18 scale of opinion which has five positive points and five negative points.

19 On the positive -- in the middle of the scale, there is a neutral point

20 which is point zero where I rested in this particular case. I'm saying

21 it's possible, point zero, that it's the questioned speakers.

22 Above that there are five positive points where I say point plus

23 1, it's rather more likely than not to be the same person; point plus 2,

24 fairly likely; 3, likely; 4, very likely; and in a very, very small

25 minority of cases where one has copious high-quality recordings and

Page 8633

1 perhaps speech impediments or something like that, there's point plus 5

2 which I perhaps evoke in 0.05 per cent of cases - one case in 200 - where

3 I'll say I'm sure. So effectively, it's a four-point scale; the fifth

4 point is seldom evoked.

5 On the negative side, the scale is a mirror image of itself, so

6 point minus 1, rather more likely than not that they're different

7 speakers, down to minus 5 where I'm saying I'm sure.

8 Q. All right, [redacted]. Thank you for that analysis on the voice

9 comparison. On the technical side of the comparison, the spectrogram and

10 the other technical aspects, you did that yourself, all by yourself; is

11 that correct?

12 A. That's correct. For limited parts of it, looking at, for example,

13 fundamental frequency, Witness EE was present, and I would describe to

14 that witness what was there on the computer screen and we would discuss it

15 together.

16 Q. All right. Let's go into your report and your analysis of the --

17 between the two tapes and whether or not there was any indications

18 evidence of tampering or authentication analysis, as we've called it.

19 And we're both obviously speaking the same language. We need to

20 slow down a bit so the interpreters can keep up.

21 A. I do apologise. Perhaps, Mr. McCloskey, just before coming to, I

22 can point out that there is an error in the report that we've just

23 referred to, if I may.

24 Q. Please do.

25 A. This is in the speaker identification report, 840/A, the one

Page 8634

1 that's just been produced -- no, sorry, that's the curriculum vitae. One

2 moment.

3 Q. I believe you would be looking to --

4 A. Page 2 of that report.

5 Q. -- 839/A, "Examination of Speech Samples."

6 A. It could be. I don't have one with a number on.

7 Q. You should have one with a number right in front.

8 A. I think not.

9 Q. Well, one is on its way.

10 A. Thank you very much. Yes, this is 839/A. I turn to page 2 of

11 that report, page 2 being the first text page after the cover, looking

12 about a quarter of the way down the page to the second major heading,

13 which says: "2. Material Examined and instructions," five lines down

14 under (i), I say: "An Agfa brand open spool audio tape..."; and on the

15 next line, under (ii), I say that: "A BASF brand open spool audio tape..."

16 In fact they should be reversed; it's tape 32 which is the BASF brand and

17 tape 48 which is the Agfa brand.

18 Q. Thank you.

19 A. The second exercise that I was asked to undertake involved

20 examining the two versions of the recording for any evidence of them

21 having been edited or tampered with.

22 I just wonder if it's possible to have an indication from the

23 interpreters as to whether this is the appropriate speed for delivery?

24 Q. We hear from them that it's not, so I think you're doing just

25 fine.

Page 8635

1 A. All right, thank you.

2 To assist me with these examinations, I was provided firstly with

3 the two tapes themselves; secondly with a transcript in B/C/S of the

4 conversation and also with a translation of it into English. In addition

5 to that, a Mr. Jack Hunter, who works for the visual information branch of

6 the Bureau of Alcohol, Tobacco, and Firearms in the United States, sent me

7 various things, including a preliminary report that he had made concerning

8 his inspections of radio interception equipment, and also Uher tape

9 recordings which I understand had been used to record from the intercept

10 equipment; a further report which concerned his, if you like, slightly

11 more detailed investigations of those items, some photographs of those

12 items, and also three spools, open spools of audio recording tape which

13 he'd used to make some experimental recordings on from Uher tape recorders

14 which I understand had been used in Bosnia.

15 Q. Just to clarify --

16 A. Sorry.

17 Q. -- how much of Jack Hunter material did you rely on in your tamper

18 analysis?

19 A. Very little, very little. I was given certain information to help

20 with the examinations. Firstly, I was told that the first recording to

21 come to light had been the one on tape 48. Once the recording on tape 48

22 had come to light, the transcript had been made of it, and it was only at

23 a later date, I think in some archives in Sarajevo, that it was explained

24 to me that tape 32 had been found during the execution of a search

25 warrant, and it also transpired that tape 32 contained the same recording

Page 8636

1 as tape 48.

2 It was explained to me, and this was consistent with the

3 information contained in Jack Hunter's reports, that at the time of the

4 war, different intercept stations would be intercepting sometimes the same

5 conversations, and it appeared from examining the tape, the two tapes,

6 that this is what had happened here, and I'll explain why in a moment.

7 When I examined the tapes, I found that the questioned recording

8 on tape 48, that was the first one to come to light, was to be found at

9 around 20 minutes and 15 seconds into the running time of the tape. The

10 version which was on the second tape, tape 32, appeared at around 12

11 minutes and 50 seconds into the running time of the tape.

12 On both tapes, the questioned recording appeared partway through a

13 series of other recordings of a similar kind; so in other words, both

14 tapes appeared to contain, if you like, a succession of military radio

15 intercept recordings, and in both cases, the questioned recording appeared

16 partway through that succession.

17 When I looked at the recordings in a bit of detail, it appeared

18 that they had been recorded firstly at a running speed of 4.7 centimetres

19 a second, that's 1 7/8 inches a second, which is speed that Uher

20 tape-recorders do record at, and both of them were in monophonic rather

21 than stereophonic, half-track format. In other words, the magnetic track

22 on the tape extended from near to the edge to about the halfway point.

23 The physical dimensions of the tracks were similar to those on the

24 test recordings supplied by Mr. Hunter. The tapes both showed evidence of

25 quite a lot of wear. The appearance was that they'd been used and reused

Page 8637

1 and used and reused, and tape 32 in particular, when it was examined under

2 a microscope, showed quite a lot of scratching to the surface of the tape.

3 When I compared the tapes with -- the recordings with one another,

4 I found firstly that there was a slight difference in running speed, that

5 if I synchronised the two recordings and recorded them within a

6 two-channel system of a computer, tape 32, say, in channel 1 displayed at

7 the top of the screen, and tape 48 in channel 2 displayed at the bottom of

8 the screen, and I synchronised the starting points of the recordings and

9 the waveforms, by the end of the recording, it was clear that tape 32 was

10 slightly ahead. After the 2 minutes 29 seconds running time of the

11 recording, it appeared that tape 32 was ahead by just under 2 seconds,

12 1.94 seconds, which suggested that tape 32 had been recorded on a machine

13 with a slightly slower running speed than tape 48. So if it's recorded in

14 playback slightly slowly [sic], it will appear slightly speeded up in

15 playback.

16 This is quite common, that you'll find two instances of the same

17 make and model of the machine which are supposed to run at, say, a nominal

18 running speed of 4.7 centimetres a second, but in reality, particularly

19 when they've had quite a lot of wear and tear and use, they -- one will

20 start to run more quickly or more slowly than the other.

21 Another difference between the two recordings was the strength of

22 the voice signals. If we remember, these are two-channel radio

23 communications. In Mr. Hunter's report it's suggested that what happened

24 was speech would be broadcast on one transmission frequency, and the

25 acknowledgement or the reply would come back on another frequency. And the

Page 8638

1 scanners, the receivers that were being used to intercept them, would be

2 tuned to one frequency only, and one would achieve a good, clear quality

3 or louder signal from the channel or the frequency that one was tuned

4 into. The replies coming from the other channel would either be

5 non-existent or they would be less loud, the signal strength would be

6 less, and this is consistent with experiments I've done on radio

7 frequencies for a previous case.

8 In the two recordings in question, the one on tape 48, the speech

9 attributed to Obrenovic was stronger, was of greater amplitude than that

10 attributed to Jevdjevic and Krstic, their being at the other part, the

11 other side of the communication; and in tape 32, the position was

12 reversed. In other words, Krstic was louder and so was Jevdjevic, and

13 Obrenovic was the more quiet of the two voices.

14 There were differences in the opening sections of the two

15 recordings. In the tape 48 version -- it would help me if I might use

16 the, I think you call it an ELMO to project from the transcript and to

17 explain what I mean on this point. May I do so?

18 Q. Yes. I think you'll get some help from the usher, and that's

19 Exhibit 789/A, which is --

20 THE INTERPRETER: Microphone, please, Mr. McCloskey.

21 MR. McCLOSKEY: Excuse me.

22 Q. Yes, that's Exhibit 789/A which is tape 48, which is, again, the

23 first -- the tape that first came to light.

24 A. Yes. This is the transcript of the tape 48 version of the

25 recording, and here --

Page 8639

1 MR. McCLOSKEY: We're going to need a little adjustment for the

2 ELMO so we can get in the relevant parts. Thank you.

3 A. Sorry, can you just give me a moment to look at another version of

4 the transcript before I, before I move on with that?

5 Would it be possible to do this from the B/C/S version of tape 48

6 as opposed to this transcript?

7 Q. Certainly. And that should be underneath it as 789/B, or close by

8 somewhere.

9 A. That's the one. Yes, in this version here, the tape 48 version,

10 it wasn't entirely clear to me whether the recording began at that point

11 there just before "Obrenovic sa Palme," or whether it began two lines

12 later, around ten seconds later, because we have here at the beginning of

13 the recording what could be the switching transient, the click sound for

14 the recorder going on; but here at this line there are three further

15 possible transients, three further possible clicks, which could indicate

16 that that is not the beginning of the recording but this is.

17 Q. I'm sorry, Dr. French, for the record, when you point, if you

18 could just explain where you're pointing so the record will accept it as

19 opposed to "this" or "that."

20 A. I do apologise. I'll go through that again.

21 At this point here we're about 20 minutes and 15 -- sorry, at line

22 1 of the transcript we're at about 20 minutes and 15 seconds into the

23 running time of the tape. Just prior to line 1 there's a click sound

24 which could be a switching transient. A switching transient is a click

25 resulting from a surge in electrical energy introduced onto the tape by

Page 8640

1 the turning on, in this case, of a tape-recorder or by the release of a

2 pause button. At line 3 of the transcript, some ten seconds later, there

3 are three further click sounds which again could be switching transients.

4 They could be on clicks of some sort or another.

5 Without access to the specific recorder which had been used to

6 make this recording, I really couldn't say whether the beginning is here

7 at -- that's just prior to line 1 at 20 minutes and 15 seconds, or some

8 ten seconds later here at line 3 of the transcript.

9 With regard to tape 32, the position is a little more complex.

10 What appears to be the case with tape 32 is that the tape-recorder has

11 been left running in the record mode. In other words, the record button

12 has been pressed, but the pause button has also been pressed, too, if you

13 like, possibly in readiness for the making of a recording, so that if one

14 were to hear a recording that one wished to go onto the tape, a

15 conversation that one wished to go onto the tape, all one would need to do

16 would be to lift the pause button. So it appears that the recorder has

17 been left in the record mode but with the pause button depressed, and

18 about midway on the page, just prior to the word "momenat," the pause

19 button has been lifted, thereby recording the word "momenat" as the first

20 word in the tape 32 version of the conversation.

21 Q. Okay. Could we just, for the record, establish that tape 32 is

22 Exhibit 835. And perhaps on that third arrow that you've written down

23 there, which is where you're telling us tape 32 started, could you just

24 put a little 32 there so that will give us a reference that this

25 is -- instead of bringing out the other transcript right now, this will

Page 8641

1 just help us.

2 A. Okay. I have done that. What appears to happen, then, is that

3 the pause button is lifted, which allows the word "momenat" to be

4 recorded. But after the recording of the word "momenat," the pause

5 button is again depressed, it's pushed down, stopping the tape. It's only

6 stopped very briefly, I think for 1.35 seconds, which is the time in which

7 the word "halo" occurs, which is attributed to Krstic here, just

8 below "momenat" on the transcript. But by pressing the record button,

9 what happens -- sorry, by pressing the pause button, what happens is when

10 that person is saying "halo," the recorder is stopped; therefore,

11 the "halo" doesn't go onto the tape. The pause button is lifted

12 immediately after "halo" and the recording continues as on the tape 48

13 version. So just to be clear on that, on the tape 32 version the "halo"

14 doesn't occur, and I'm going to strike it out just for the record. The

15 tape has the word "momenat" on, and then the next thing that one hears is

16 the person claimed to be Obrenovic saying this line here, which I can't

17 pronounce because I'm not a B/C/S speaker, "sta" -- whatever.

18 Q. Could you, just for clarity, if we see Exhibit 835/B, which is the

19 tape 32 section that is missing this particular part. They should have

20 it. They will bring it up.

21 A. I have a version here.

22 Q. Well, let's use the one with the exhibit on it, is our practice.

23 A. I think it is the one with the exhibit on it.

24 Q. If you've got it, great. Let's use that one. If you could just

25 go to the B/C/S section and point out what you were talking about.

Page 8642

1 A. If I could put the two side by side, it might clarify it. Yes,

2 please. That one and this one. Are they both on the screen there? Let

3 me align them. No. It would be easier if I just -- there we have it.

4 You can see them side by side.

5 This is the tape 48 version. We see "momenat," "halo," and then

6 this line here, the line below "halo." On the tape 32 version there is

7 no "halo" because the pause button has been pressed after "momenat," I

8 think, and then lifted, and then the recording continues. But by the

9 action of pressing the pause button, one bypasses the recording of this

10 word "halo." It's omitted. The recordings then continue in parallel and

11 the contents are the same until we come to the end sections.

12 On the tape 48 version, after the conclusion of the conversation,

13 there are a series of beeps and what sounds like radio equipment

14 disconnection sounds. On the tape 32 version, those don't appear. The

15 recording finishes before those occur.

16 Q. Why do you conclude that the "halo" is missing on the one version

17 because of the pause button? How are you able to make that conclusion?

18 A. On two bases: There is a click there which has a particular shape

19 when one looks at the wave form. On some recordings sent to me by

20 Mr. Jack Hunter, he had made experimental recordings in which he had been

21 lifting and putting down the pause control on a newer tape-recorder.

22 There were clicks on those recordings associated with the action of using

23 the pause button which was similar in shape to or in configuration to the

24 one that occurs at that point on the tape 32 version. The second reason

25 is I made an experimental recording on another Uher 4000 tape recorder,

Page 8643

1 the type I was given to understand had been used in Bosnia, and I too

2 found a similar shape of click; not identical, but similar.

3 Q. Did you also have a chance to briefly question one of the

4 intercept operators about their practice?

5 A. [redacted]

6 [redacted]

7 Q. Excuse me. If we could be careful of the --

8 A. I do apologise. I hope that's --

9 Q. That's okay. Keep going.

10 A. The intercept operator that I spoke to gave me the information

11 that it was common practice when monitoring conversations to have the Uher

12 tape recorders running in the record mode but with the pause control down

13 so that the tape wasn't actually recording. It was at points when they

14 heard something worthwhile that they would left the pause button and

15 record onto the tape.

16 Now, very often, he tells me, once radio contact between two

17 people had been established, the person initiating the contact or

18 initiating the transmission, would say, "Could I speak to Major X?" and

19 the voice at the other end would say, "momenat," or "moment, please."

20 What they would then do, they would press the pause button again, because

21 sometimes it might take 10 or 15 minutes, he told me, go and get the

22 person concerned, and then when that person came on the line, or they

23 could hear them coming on the line, they would lift the pause button. So

24 in the light of what he told me, the pausing of the tape recorder

25 after "momenat" would be exactly where one would expect it to be paused,

Page 8644

1 perhaps in anticipation of a long wait before the person he's asked for

2 comes to the transmission equipment. But in this case, there wasn't a

3 long wait; the person was almost immediately there, hence one explanation

4 for why the "halo" is missing.

5 Q. Is there any other sort of reasonable, more suspicious explanation

6 for these clicks and the various things you've talked about?

7 A. I couldn't really, on a common-sense basis, conceive of why

8 somebody would want to edit the word "halo" out of the tape. It would

9 seem to me to be one of the -- to be a benign feature of the conversation.

10 Q. And if you could just continue your analysis.

11 A. Certainly, yes. In the position of Mozart; too many notes.

12 With the exception of the click and the missing "halo" in tape

13 32 --

14 Q. I think we're through with the ELMO for a while.

15 A. Yes.

16 Q. Thank you.

17 JUDGE RODRIGUES: [Interpretation] Mr. McCloskey, would this be a

18 convenient moment to take a break?

19 MR. McCLOSKEY: I'm sure it is, Mr. President. Thank you.

20 JUDGE RODRIGUES: [Interpretation] Let us take a 20-minute break.

21 We adjourn for 20 minutes.

22 --- Recess taken at 10.56 a.m.

23 --- On resuming at 11.24 a.m.

24 JUDGE RODRIGUES: [Interpretation] Mr. McCloskey, please continue.

25 MR. McCLOSKEY: Thank you, Mr. President.

Page 8645

1 Q. Dr. French, if you could just continue where you left off in your

2 tamper analysis.

3 A. I left off almost, I think --

4 THE INTERPRETER: Microphone, please, Doctor.

5 MR. McCLOSKEY: I'm sorry, Dr. French. The microphones got turned

6 off.

7 A. I was just arriving at my conclusions, I think, before the short

8 adjournment, which were, firstly, that with the exception of the -- what I

9 believe as a pause transient after the word "momenat" in tape 32, the

10 recording signals in both tapes appeared to be continuous. In short, I

11 can find no evidence of either of them having been tampered with or

12 edited.

13 Now, having said that, I should just add a caveat, which is that

14 nowadays it's usual for an audio expert to say, and this is quite correct,

15 that the failure to find evidence of editing doesn't necessarily imply

16 that the tape definitely has not been edited. And the reason I say this

17 is because there are nowadays available digital systems, computers,

18 whereby one could record a conversation from a tape, re-record it from a

19 tape to the computer, and then, within the specialised sound analysis or

20 sound processing or editing programme of the computer, one could edit up

21 the recording, deleting sections perhaps, whatever, and do this more or

22 less without trace. The edits are almost seamless if they're done in a

23 very competent way. And the material could then be re-recorded back from

24 the computer to a tape and passed off as an unedited version of the

25 recording. In other words, it would be possible, I believe, and I think

Page 8646

1 most other experts working within the field would nowadays say so, that it

2 would be possible to produce an edited recording which has defied forensic

3 detection.

4 However, there's a very unusual aspect to the present case, in my

5 experience, and that is that two versions of the same recording have

6 become available at different points in time. One is not a copy of the

7 other and they can't both be copies of the same original recording. In

8 view of that, I have to say that the possibility of digital editing having

9 occurred, I must say I see it as much, much lower, significantly lower

10 than, one would have to say, in a normal case. For it to have occurred,

11 both of these recordings - I can't see any alternative to this - would

12 have had to have been edited independently of one another and in exactly

13 the same ways. And I think the work and skill that would go into that

14 would be extremely great.

15 Q. If one was to create a conversation like this, if you were asked

16 to create a conversation like this, how could you do it?

17 A. With the assistance of actors, perhaps, or people who were simply

18 going to adopt the roles of those they were projecting into. I must say,

19 in the past I've had a number of tapes which have involved actors who've

20 got -- not actors, but people who have got together in order to improvise

21 conversations, perhaps reading from scripts, perhaps just improvising as

22 they went along, in order to manufacture evidence for a tape-recorder

23 which is later being presented in court. In those cases, it has always

24 seemed to me perfectly obvious that the performance, the language has been

25 wooden, it has been stilted.

Page 8647

1 In the present conversations, there are at least three areas of

2 overlapping speech, and because of that, this appears to me - and I'll

3 come on to the issue of being a non-speaker of B/C/S in a moment - but

4 this appears to me to have the sort of messiness that one would associate

5 with natural conversation, that is, two people talking at once or

6 overlapping in their speech.

7 In addition to that, having worked with Witness EE on this, who is

8 a speaker of this language and who has research knowledge of the dialects,

9 I did consult him on this issue and asked whether the conversation

10 appeared wooden or whether it appeared stilted or play acted, and the

11 answer came back no. I also did ask a native speaker and a linguist of

12 this particular language as well, a different person.

13 Q. Besides just excellent actors overlapping speech, what else would

14 have to be created to give the same kind of tonal qualities that we hear

15 on these tapes?

16 A. Well, it would have to be done over radio transmission channels

17 and recorded over them, or at least over some kind of electrical and

18 electronic circuitry that mimic their characteristics. It would be a very

19 elaborate hoax.

20 Q. Is there a possibility that if a person had a library of several

21 conversations of particular people that they could take individual words

22 and phrases and splice them all together on this computer and make

23 completely fake sentences? Is that possible?

24 A. I have seen in a previous case a tape which was -- exactly that

25 had been done with it. When one read the transcript of the conversation,

Page 8648

1 the conversation appeared entirely normal; it appeared to flow. When one

2 listened to the tape, however, it was a joke. I mean, there were two

3 aspects to it: One is having the right words in the right order, but the

4 other aspect is making it sound like phonetically and phonologically that

5 the words are on the right pitch, the intonation is correct, that the

6 cadences -- and so on and so forth, and that is very difficult indeed.

7 It can be manipulated within a computer system. One can

8 manipulate the pitch, the intonation, et cetera, but it is still a

9 difficult task, in my opinion.

10 Q. So what is your final opinion about these two particular tapes

11 regarding whether they've been tampered or faked?

12 A. I can find no evidence of tampering or faking, and I think those

13 comments must be bore -- interpreted in the light of what I've just said

14 about the difficulties of play acting and montage.

15 MR. McCLOSKEY: I have no further questions.

16 JUDGE RODRIGUES: [Interpretation] Thank you very much,

17 Mr. McCloskey.

18 Mr. Visnjic, I see it's Mr. Visnjic who is going to cross-examine.

19 Cross-examined by Mr. Visnjic:

20 Q. [Interpretation] Good morning, Dr. French.

21 A. Good morning.

22 Q. Dr. French, in your report, exhibit number 838, on page 2 you say

23 that the tapes that you examined were received in your laboratory on the

24 9th of November, 2000.

25 A. Yes.

Page 8649

1 Q. Was that the first time you saw those tapes?

2 A. I believe so, but if I can just consult my records. I've

3 certainly had copies of the tapes -- copies of the conversations before

4 that. Is your question these physical tapes, these actual ones rather

5 than copies?

6 Q. My first question would relate to the actual physical existence of

7 the tapes that you now have before you.

8 A. "Momenat." Yes, I believe so.

9 Q. Did you receive copies of those tapes earlier on?

10 A. Yes.

11 Q. Could you tell the Trial Chamber when you received those copies?

12 A. Yes. The 3rd of February in 1999. And that was a copy only of

13 the recording from tape 48, plus some other material.

14 Q. Did you prepare any preliminary reports on the basis of the

15 copies, those copies, those that you received in February 1999?

16 A. No.

17 Q. Was it possible for you to work on some preliminary findings on

18 the basis of those copies?

19 A. It was possible to do some preliminary work. At that point in

20 time, I was working not with Witness EE but with another linguist who was,

21 in fact, a native speaker of B/C/S. But as time went on, it became clear

22 that I needed a different collaborator.

23 Q. On the same page, page 2, you say, in the last paragraph, that

24 joint investigation between you and [redacted] was carried out at your

25 laboratory in York in the week from the 19th to the 26th of November,

Page 8650

1 2000.

2 A. Yes. I think that is the week when myself and Witness EE examined

3 the recordings.

4 Q. Are you aware that Witness EE -- do you know that he saw the tapes

5 for the first time then?

6 A. I don't know. I don't recall. He may have had a copy of the

7 recordings prior to our meeting in York. It's possible.

8 Q. Dr. French, did you know whether these tapes had been given to any

9 other expert; and if so, did anyone else work on an expert finding, a

10 forensic examination, and are you familiar with the results of those

11 examinations?

12 A. As far as I know, no such examinations took place, unless I wasn't

13 told about them.

14 Q. On November 1st, 2000, at a hearing, my learned friend

15 Mr. McCloskey stated on pages 6803, 6804 of the transcript, that this copy

16 of the audiotapes had been given to an expert and that the Prosecution has

17 for months been trying to get hold of a final response from the expert.

18 Bearing this in mind, do you assume that that perhaps relates to you or to

19 somebody else? Because this was on the 1st of November, 2000.

20 A. I really feel it isn't a matter for me to say what was in

21 Mr. McCloskey's mind, but it could well have related to me. I know of no

22 other expert that's been involved.

23 Q. Thank you. Dr. French, on page 6 of the English version of your

24 report on the audiotapes, Prosecution Exhibit 838, paragraph 4, "Editing

25 and Tampering," you explained that modern devices make it possible to do

Page 8651

1 some editing or tampering. My question is the following: Could digital

2 technology, which you assume might have been used, did such technology

3 exist in 1998 and prior to that period?

4 A. Certainly it existed in 1998.

5 Q. On condition that all the technical requirements existed, those

6 that you referred to regarding editing and tampering, on the basis of your

7 own experience, would a period of six weeks be sufficient to do such

8 montage or editing?

9 A. Well, the problem with it is that you would have to do it on two

10 versions of the same recording in exactly the same way, and if one were to

11 be chopping out sections and inserting material from other recordings, one

12 would have to do it in exactly the same way in two versions of an original

13 recording, getting the timings exactly, precisely the same in both -- or

14 not the same, but relatively the same. If you remember, one tapes runs

15 slightly faster than the other. Am I saying too much?

16 Q. Dr. French, perhaps we didn't quite understand one another. I

17 quite understand the procedure that you just explained to the Trial

18 Chamber. My question is: Could that whole procedure have been completed

19 in a period of six weeks?

20 A. I think the question maybe should be whether it could be completed

21 at all, rather than within six weeks, to produce this. I really

22 don't -- can't comment on the time scale that would be needed. All I can

23 say is that, in my view, to produce these two recordings would be an

24 extremely difficult feat, in six weeks or any other time scale.

25 Q. Could six weeks be used as some kind of framework for doing this,

Page 8652

1 or not?

2 A. It would depend upon the sophistication of the technology, the

3 skills of the manpower, how many people were involved, whether or not one

4 had sound engineers, perhaps phoneticians, computer operators. I really

5 can't answer the question in the abstract without knowing about the

6 circumstances of that sort, really. It's very difficult.

7 Q. Thank you. Dr. French, in your report -- or rather, you received

8 a Mr. Hunter's report?

9 A. Yes.

10 Q. I'm referring to the report dated the 23rd of March, 1998. I

11 don't know if you have it at hand.

12 A. I think I do. Yes.

13 Q. Dr. French, before I put my question to you, may I just say or ask

14 whether Mr. Hunter is also familiar with the actual technology of tape

15 collecting, examination, and so on? Is he an expert in the area? Am I

16 right in saying that?

17 A. I don't know. I don't know what his background or expertise or

18 experience is.

19 Q. In his report of the 23rd of March, 1998, page marked 00889128,

20 this would be the page 2 of his report, actually, in paragraph 4, the last

21 sentence says that all recordings were of good quality.

22 A. Could you just give me a moment to reread this report in full.

23 Q. Yes, yes.

24 A. I'm not clear which recordings he's referring to in the end of

25 paragraph 4. You see, the recordings that he refers to there clearly are

Page 8653

1 not these tapes because these tapes were recorded at a running speed of

2 4.7 centimetres per second, and none of the recordings that he refers to

3 there were recorded at 4.7, so he is clearly talking about other

4 recordings. If he is, I haven't seen them.

5 Q. But it is a report which refers to the reception of the first 19

6 audio open spool tapes, one of which is in front of you, the tape that you

7 received first marked as 48.

8 A. I really can't be accountable for Mr. Hunter's work, but the

9 recordings on that tape are not at the speeds he lists here. They're at

10 4.7 centimetres per second, and they're not of good quality.

11 Whether Mr. Hunter has overlooked that or not listened to all the

12 recordings or perhaps -- I don't know want to impugn him in any way. I

13 really don't know what he's seen. If you're telling me that this was

14 among the first -- sorry, among the first 19, or what you tell me or what

15 I read here doesn't square with what I found on the tape.

16 Q. And if I may repeat what you say in your report on page 3.3, that

17 the tapes were of poor quality. This is the report?

18 A. Yes, they were of poor quality.

19 Q. Thank you. Dr. French, in your report on the audiotapes, on page

20 3 of the English version, under (vii), you say that Mr. Hunter gave you

21 three audiotapes, open spool Ampex brand audiotapes containing

22 experimental recordings which Mr. Hunter had made on Uher tape recorders?

23 A. Yes.

24 Q. On the other hand, also in Mr. Hunter's report, dated the 24th of

25 April -- we have a large number of documents, but we'll go through them

Page 8654

1 slowly.

2 A. Yes, I have his report in front of me.

3 Q. I'm referring to paragraph 7 from the top, page 2.

4 A. Could you read the first words of that paragraph to me?

5 Q. "At the site I requested to see --"

6 A. I have it, yes, yes. I have it.

7 Q. Mr. Hunter says that he took experimental tapes for all the

8 machines that he had, and then he identified those tape recorders with

9 numbers from K1 through K7.

10 A. Yes.

11 Q. He did the same on the next page, page 3 --

12 A. Yes.

13 Q. -- where at the other position from which allegedly conversations

14 were intercepted, he also took specimen tapes and marked another three

15 Uher tape recorders.

16 A. Yes.

17 Q. My question, Dr. French, is, on the basis of the samples provided

18 to you by Mr. Hunter --

19 A. Yes.

20 Q. -- were you able to determine the tape recorder on which any one

21 of these tapes was taped?

22 A. No. In fact, that would have been a very large enterprise to try

23 and -- to try and do that. I think I still wouldn't be here with my final

24 report. No.

25 Q. But each of these machines has a specific result on the tape;

Page 8655

1 isn't that so?

2 A. Not necessarily so. There are a couple of factors here. One is

3 that it is by no means -- research has by no means established that each

4 individual tape recorder has its own unique signature or fingerprint.

5 That's the first aspect.

6 The second aspect is that I did have a conversation with

7 Mr. Hunter, and he did tell me over the telephone that it was possible

8 that none of these recorders had recorded the tape or either of the tapes,

9 that in fact machines might have been scrapped. The other thing that he

10 told me is that the recorders that he looked at, that some of them - and I

11 think he based this on his interviews with personnel in Bosnia - had had

12 repairs to them and perhaps even to the heads of the machines. So even if

13 one of those machines had been used and even if it did have some

14 individual characteristics, there would have been no guarantee that at the

15 point at which he saw it or found it, it had the same characteristic that

16 it had when these tapes were made.

17 It really would have been such an open-ended exercise to try and

18 tie either one of these tapes down to any of the recorders that it just

19 was not embarked upon.

20 Q. Dr. French, from the scientific point of view, is it easier to

21 identify the machine, the tape recorder, or the speaker? Which is more

22 precise?

23 A. Neither is precise, and I couldn't given a weighting of the

24 precision to one relative to the precision of the other.

25 Q. Dr. French, let us go back to the report on the audiotapes,

Page 8656

1 Exhibit 839.

2 MR. McCLOSKEY: Excuse me. 839 is the examination of speech

3 samples; 838 is the report on audio recordings.

4 THE INTERPRETER: The interpreter corrects herself. It was speech

5 samples that Defence referred to.

6 MR. VISNJIC: [Interpretation] Maybe it was a misinterpretation,

7 but I was referring to the report on speech samples, OTP Exhibit 839.

8 Q. Dr. French, on page 4 of the English version of your report --

9 A. Yes.

10 Q. -- the last paragraph -- it is page 5 of the Serbian version,

11 paragraph 2. Allow me to read this sentence:

12 "Owing to the non-equivalence of the known and questioned samples

13 (normal voice vs raised voice), the most one can say from the voice

14 quality comparisons is that they failed to eliminate any of the three

15 known speakers."

16 Dr. French, could the same conclusion be the same as saying that

17 not one of the known speakers was confirmed?

18 A. Yes. It's a question of the Prosecution perhaps saying that my

19 glass is half full and you would say it's half empty.

20 Q. Dr. French, we're still dealing with the same report, but let me

21 ask you: Do you know, are you aware of the International Association for

22 Identification? Are you familiar with that institution and the standards

23 adopted by that institution for speech comparisons?

24 A. I'm aware of some of those standards in outline. I'm aware of the

25 existence of the institution.

Page 8657

1 Q. On page 3 of the English version of your report, under 3, under

2 the heading "Observations and Results," you refer to the technical quality

3 of both versions of the tape.

4 A. Yes.

5 Q. You also, and quite rightly so, stated that there was very little

6 speech energy found at frequencies above 1.7 kilohertz.

7 A. Correct.

8 Q. Are you aware that some institutions -- or rather, that the

9 standards of the International Association for Identification is such that

10 tapes that have a lower frequency, below 2 kilohertz, are considered

11 absolutely unsuitable for making any kind of comparisons because they are

12 technically of poor quality?

13 A. I'm not aware of that, but I would say two things in relation to

14 it. One is, firstly, that the organisation that you refer to is by no

15 means an internationally accepted one, particularly within the European

16 forensic speech community. Indeed, its activities have caused a few

17 eyebrows to be raised.

18 Secondly, I would point out that in order to -- what one is not

19 concerned with is necessarily identification. As an objective, impartial

20 scientist, my task in this case was not to identify your client or his

21 colleagues; it was to determine whether they were on the tape. In other

22 words, when I do these examinations, I have in my mind the possibility of

23 elimination just as much as the possibility of identification. And the

24 importance of that remark is this: that it is sometimes possible to

25 eliminate speakers if they have very different voices from the ones on the

Page 8658

1 recording with very poor quality material indeed. One needs rather good

2 material sometimes for identification, but for elimination one can often

3 work with recordings of this type.

4 For instance, if someone has a very different voice quality or

5 dialect features or whatever, or one person has a speech impediment or a

6 pathology of their voice or something of that sort, even on very poor

7 quality recordings, that can show through. And if that is not on the

8 other samples that one is comparing it with, it is very important, even

9 with poor quality material, to go through with the possibility of

10 elimination.

11 So I would not accept, I would not accept that standard, and nor

12 would, I think, my colleagues in the European context, or most of them.

13 Q. Are you aware, Dr. French, that, for instance, the FBI

14 laboratories do not do these comparisons with tapes whose quality is below

15 2.5 kilohertz?

16 A. Well, I'm not aware of that, but if that is the case, I can't see

17 how they can be proceeding with the view of elimination in mind, as well

18 as identification, as equal possibilities.

19 Q. In your expert report, on page 5 of the English text, under the

20 heading of "Pitch/Fundamental Frequency Measurements," you list some of

21 the elements which lead to a change in the height of the tone of the main

22 frequency, or pitch of the main frequency, and one of those elements would

23 be a raised voice. Am I correct in making that assumption?

24 A. That is correct.

25 Q. Anger, fear, or sorrow, in one and the same person or speaker, can

Page 8659

1 they produce characteristic differences in fundamental frequency?

2 A. They are just a few of the things that can.

3 Q. Is that also true of stress, emphasis, semantics, intonation of

4 words and sentences, that kind of thing?

5 A. Are you asking whether those emotions can affect intonation?

6 Q. Yes.

7 A. Emotions do affect intonation. I don't think it affects

8 semantics. Semantics is the study of word meaning.

9 Q. The way in which words and sentences are enunciated, the speed of

10 speech, et cetera?

11 A. Would be subject to emotional factors, yes.

12 Q. With one and the same person, would there be a difference in

13 fundamental frequency of speech, from morning till night? That is to say,

14 if we look at it over the course of one day, would fundamental frequency

15 be the same throughout for a given speaker?

16 A. It might or it might not.

17 Q. If the same text were repeated again and again by the same person,

18 could there be a difference in fundamental frequency?

19 A. There could.

20 Q. May we then conclude that part of your report measuring

21 pitch/fundamental frequency measurements, shows in actual fact that there

22 is no similarity between the basic sample and the comparative sample, and

23 that there are numerous reasons for that or that numerous reasons can

24 exist?

25 A. Yes, one could conclude that. The reason for doing the

Page 8660

1 fundamental frequency test in this particular case is, as you quite

2 rightly say, when a person has raised their voice, they raise their

3 fundamental frequency. The two go hand in hand. This was established in

4 the earlier part of this century first by a French phonetician known as

5 Etienne Lombard, and Lombard noted the phenomenon whereby if a person

6 raised their voice, perhaps against a background of noise, they raised

7 their pitch as well. So with raised voices, one finds raised pitch or

8 raised fundamental frequency.

9 Now, the importance in this case was simply as an elimination

10 test, as stated. We have General Krstic talking in a reasonably calm and

11 deliberate way in his interview recording. We have a man shouting in a

12 radio recording. One would expect the radio recording, if it were the

13 same person, to be higher, but also a variety of other people would do it

14 higher as well. But crucially, if the person -- if the known sample, the

15 interview recording of your client was higher than the person with

16 obviously raised voice in the radio recording, the intercept recording,

17 then that would be at least a basis or a partial basis for eliminating

18 him; hence, the necessity to carry out the test with the view not only of

19 identification, but elimination in mind, too, you know?

20 Q. On condition that the known sample was performed under conditions

21 in which that person was not under stress or anger, emotional changes or

22 similar things. Am I right in saying that?

23 A. No. You can tell from the known sample that the person -- I don't

24 know what internal stress they were suffering. I mean people, some people

25 have enormous internal stress and emotion, and they don't reflect it much

Page 8661












12 Blank page inserted to ensure pagination corresponds between the French and

13 English transcripts.













Page 8662

1 in their voice. Others might be under a small amount of stress and

2 emotion, and they do. All I can say is that I look at the interview

3 recording and the person is obviously not raising -- it's a not "dobro!

4 dobro!" You know, it's not like that, whereas the person in the military

5 recording clearly is talking in that way.

6 If the necessity to do the test is the person who has clearly

7 raised their voice, going back to Lombard's research, if they turned out

8 to have a lower fundamental frequency than the interview sample, then that

9 would be very important. It would be very important for your client.

10 Whatever the effects, the emotional effects of the interview, your client

11 had not raised his voice in that way; hence the necessity to do the

12 fundamental frequency check.

13 Q. But the conclusion of the test, the fundamental frequency test is

14 nonetheless inconclusive; am I correct in saying that?

15 A. Yes, exactly, exactly.

16 Q. Dr. French, on the last page of your report or the addition to the

17 report, you give a scale of judgements --

18 A. Yes.

19 Q. -- and it follows a certain logic of determining degree.

20 A. Yes.

21 Q. Under -- or rather, can you tell me what the difference is, or

22 rather, is there a scientifically defined difference between, say, 2 and 3

23 on your scale? Is it a percentage of similarity or on the basis of

24 similar words, or what are the elements that go towards determining this

25 scale?

Page 8663

1 A. The first thing that I should say is it is not an interval scale.

2 It is not a numerical linear scale with equal points between each of the

3 numbers; rather, it is a scale which is shared by the majority of forensic

4 phoneticians working in the UK context. It is similar to the one in use

5 in the state laboratory here in Rijswijk and in the Bundes Criminalsamt

6 Laboratory in Germany, and so on and so forth. But at the end of it, it

7 is really a measure of personal confidence in one's judgement, and that's

8 what it is. It is subjective in that sense.

9 Q. Following on from the logics of this scale of judgements, if we

10 look at the zero position, it states, if I'm right, that it is possible

11 that X and Y are the same speaker. Now, bearing in mind one of my

12 previous questions, would you agree with me that a possible definition is,

13 "I think it is possible that X and Y are not the same speaker," by the

14 same token?

15 A. With the corollary of that statement at point zero, it is, in my

16 view, possible that X and Y are the same speaker would be as a corollary

17 that they are different speakers. That would be possible too. I would

18 agree.

19 Q. Dr. French, may we then conclude that your eight-page report in

20 fact leads to a conclusion which in English could be termed inconclusive?

21 A. An inconclusive conclusion.

22 MR. VISNJIC: [Interpretation] Thank you, Mr. President. The

23 Defence has no further questions of this witness.

24 JUDGE RODRIGUES: [Interpretation] Thank you very much,

25 Mr. Visnjic.

Page 8664

1 Mr. McCloskey, your witness again.

2 MR. McCLOSKEY: Thank you, Mr. President.

3 Re-examined by Mr. McCloskey:

4 Q. Dr. French, did a Defence expert have a chance to come to your

5 laboratory and review this material, and if so, when?

6 A. I don't have the specific date in front of me, but a Defence

7 expert did come to my laboratory and examine the material in the presence

8 of the gentleman who has just been cross-examining me. I could get the

9 particular date. My recollection is it was about two to three weeks ago.

10 Q. So Mr. Visnjic came along with the expert to your laboratory?

11 A. That's right. And the expert, a Mr. Koenig, or Koenig as

12 Anglicised here, from the United States, an ex-FBI examiner who is now a

13 private expert, came along with Mr. Visnjic and carried out examinations

14 of the recordings and made copies of them and took them away as well.

15 Q. How long were they with you?

16 A. From 10.00 in the morning until 5.30, with possibly a gap for

17 lunch, an hour or something like that.

18 Q. You're aware that Mr. Koenig has testified in this Tribunal as an

19 expert for the Defence in a case?

20 A. Yes, I am, yes.

21 MR. McCLOSKEY: Nothing further, Mr. President.

22 JUDGE RODRIGUES: [Interpretation] Thank you very much,

23 Mr. McCloskey.

24 Judge Fouad Riad has the floor.

25 JUDGE RIAD: [Interpretation] Thank you, Mr. President.

Page 8665

1 Questioned by the Court:

2 JUDGE RIAD: Dr. French, good morning.

3 A. Good morning.

4 JUDGE RIAD: I've been listening with great interest to your very

5 knowledgeable expos‚, and I would just like to have some clarification

6 which you can perhaps give to a layman.

7 A. Certainly, Your Honour.

8 JUDGE RIAD: As I see, you mentioned that there is an open

9 possibility that the three speakers were the same.

10 A. That's right.

11 JUDGE RIAD: And when discussion with the Defence lawyer, you said

12 that of course you can always say that a glass is half empty or half full,

13 of course if it is exactly half, half, 50-50.

14 Now, then, you can say that also there is an open possibility that

15 the three speakers are not the same?

16 A. One could say -- I couldn't say it was an equal possibility. I

17 could say that it's possible that it is the three people it concerned, but

18 there is an indeterminate number of other people, other males within that

19 population. I couldn't put my -- I couldn't specify how many that it

20 could also be.

21 JUDGE RIAD: Now, there are possibilities of identification, and

22 there are possibilities of elimination?

23 A. That's right.

24 JUDGE RIAD: What is more accessible to you? Is it easier to

25 eliminate, or is it easier to identify?

Page 8666

1 A. One can eliminate on the basis of less material and poor-quality

2 material if it obviously is not the person. For example, if there was a

3 very poor quality recording of my voice and Your Honour's voice, both

4 speaking in English but very poor quality and very brief, I don't think a

5 forensic examiner would have any difficulty in distinguishing between us.

6 If, however, our voices sounded very like, we spoke English with

7 the same accent and the same sorts of resonances, a forensic examiner on

8 the basis of a poor-quality recording or a brief recording would be rather

9 circumspect about producing an identification. So it is easier to do an

10 elimination on the basis of poor material or scant material if the people

11 have different voice characteristics.

12 JUDGE RIAD: And when you spoke about the fundamental frequency

13 test which -- does this -- could this test lead sometimes still to

14 identification?

15 A. Not in itself, Your Honour, no. It would -- it's only one test

16 which is used as an indicator with -- among a much larger battery or

17 package of tests.

18 JUDGE RIAD: But in our case it was inconclusive?

19 A. It was inconclusive.

20 JUDGE RIAD: But in many cases it is conclusive?

21 A. It can be conclusive in elimination. For example, I've just

22 concluded a case for the moment for the Prosecution in England where there

23 is -- there are recordings of a youth speaking over a telephone line,

24 reporting an arson. He set fire or somebody -- they believe the caller

25 has set fire to a building and then phoned the emergency services for a

Page 8667

1 fire engine to come.

2 He's clearly speaking on raised voice. One can hear noise outside

3 the telephone kiosk and his voice is raised. The suspect in interview is

4 speaking in non-raised voice, yet the fundamental frequency in the

5 interview was higher, and on the basis partially of that, I would

6 eliminate.

7 JUDGE RIAD: So your tools for elimination are more at hand?

8 A. In this case one could say that, yes.

9 JUDGE RIAD: Now speaking of the possibility of a montage, you

10 mentioned that one tape would run slightly faster than the other. Would

11 this be any -- indicative of any tampering?

12 A. Oh, no, Your Honour. One could take several different instances

13 of the same make and model of tape recorder which had been subject to

14 different degrees of use and wear. None of them would run at exactly the

15 same speed spot on.

16 JUDGE RIAD: And you mentioned that certain features were benign

17 features, like the difference in clicks and the differences in the opening

18 sections, but then you mentioned something about the strength of the voice

19 signals. And in that regard, you -- comparing, comparing the two tapes 48

20 and 32, you said that the speech of General Krstic was louder than

21 Obrenovic, and in tape 48, the speech of Obrenovic was of greater

22 amplitude.

23 A. The same thing.

24 JUDGE RIAD: All right. Now, does this entail that there could be

25 two different people?

Page 8668

1 A. Oh, no, no. It's clearly the same recording, Your Honour. No,

2 no, I think not, no.

3 JUDGE RIAD: So there was nothing in the timbre of the voice,

4 difference in timbre or --

5 A. Nothing that I noticed at all. It was simply we have two

6 recordings. On one, one end of the radio channel is louder, but on the

7 other recording, the second end of the radio channel is louder.

8 JUDGE RIAD: But what you call the voice quality is the same?

9 A. As far as one can assess, it's the same recording.

10 JUDGE RIAD: And the rhythm of speech?

11 A. Yes.

12 JUDGE RIAD: I think I have no other questions for you, but

13 definitely there is a lot to learn from what you said. Thank you very

14 much.

15 A. You're very flattering, Your Honour. Thank you.

16 JUDGE RODRIGUES: [Interpretation] Thank you very much, Judge

17 Riad. Madam Judge Wald has the floor.

18 JUDGE WALD: Dr. French, is there any way to tell whether two

19 tapes, like tape 48 and 49, were made in roughly the same time period, or

20 is that just totally outside the --

21 A. No. I couldn't tell you that, Your Honour, no.

22 JUDGE WALD: Yes. Okay.

23 A. There's no way that I would know that.

24 JUDGE WALD: All right. You said that -- you gave the zero on the

25 rating scale, evaluation for recognition of the voices of all three of the

Page 8669

1 participants: General Krstic, Jevdjevic and Obrenovic. Do you know, or

2 could you tell us, or was there any -- within that scale that you gave

3 them all zero, were any of those voices -- was the degree of confidence in

4 any of those voices higher than in the other two or the other one?

5 A. Your Honour, it wasn't. It was about equivalent for all three.

6 JUDGE WALD: Okay. Now, let me ask you a hypothetical, with which

7 I'm sure you're very familiar, having testified in court. You told us all

8 the difficulties there would be in trying to recreate a tape like this

9 after the fact and how your experience had led you to believe that -- or

10 in some other instances it was kind of easy once you listened to the

11 tape -- not kind of easy, but it was not impossible to discern that it was

12 indeed a play-acting. But let me just hypothesise and see if you would

13 have this -- if it would be as easy in the following instance.

14 Suppose this was -- I say, again, totally suppose. Suppose it was

15 not a recreation, but suppose somebody wanted to set somebody else up

16 originally, and they initiated a regular intercept, and then, by

17 prearrangement, they had somebody else take the place of one of the

18 people - we'll use as an example General Krstic - just had another person

19 step in there and take. And I suppose the motive might be something like

20 disinformation. At some point in the intercept, they say, "The Turks are

21 probably listening to us. Let them listen." Would that be as easy to

22 discern as the later created or would it all depend on how well the actor

23 was playing that one role?

24 A. Your Honour, if one had a reasonably good actor who sounded rather

25 like General Krstic, I think it would be possible, yes, for that person to

Page 8670

1 play-act this conversation with the other person, possibly one person

2 believing that -- one person acting out a real conversation or thinking it

3 was real and the other one just fitting in. You would need someone with a

4 similar sounding voice and also somebody who knew, I suppose, the sorts of

5 things that people said in these military radio conversations, and I

6 suppose one would have to get them into a position where they could

7 broadcast and ...

8 JUDGE WALD: Okay. Thank you very much.

9 A. Thank you.

10 JUDGE RODRIGUES: [Interpretation] Thank you very much, Judge Wald.

11 Dr. French, I would like to go back to the question raised by my

12 distinguished colleague Madam Judge Wald a moment ago. If you were asked

13 to recreate a conversation like the one we have, how could you do

14 it - and this is what Mr. McCloskey was asking - but in giving us -- that

15 is to say, could this be done in a completely confidential manner? That's

16 the first point. How many people would have to intervene to accomplish

17 that task, and could all that be done in the same laboratory or would you

18 have to involve other participants from different places, if I can put it

19 that way? Well, those are my questions for you, Doctor.

20 A. I think the answer to that is -- would be dependent on how one was

21 going to construct it. One possibility that Mr. McCloskey put to me was

22 that there was a sort of library of tape recordings of General Krstic, of

23 utterances, and that the -- how likely would it be that someone could pick

24 a sentence out here and a word out there, and yet another sentence from a

25 third tape and five words from a fourth tape and six words from a

Page 8671

1 sixth -- and so on and so forth, and put them together to arrive at this

2 conversation. Now, to do that would really be a mammoth task. I mean,

3 once we start getting into that area, I think one is moving nearer to

4 American, with due respect, Your Honour, detective fiction than the

5 forensic reality of it. But to, say, have a recording of this sort, where

6 one was simply deleting the odd word and then rejoining it within the

7 computer and re-recording the tape would be much -- considerably easier

8 and it would involve less in the way of resources, manpower, personnel,

9 et cetera, et cetera. But the thing is, one would have to recreate the

10 two tapes which have the same sequence of words, they have the same time

11 base except one runs slightly faster than the other, so on and so forth.

12 I really don't know how many people one would have to involve or -- I

13 imagine the second scenario that we're talking about now could be done

14 within one laboratory. Yes, well, I could do it. A gamekeeper turned

15 poacher.

16 JUDGE RODRIGUES: [Interpretation] How much time would you need to

17 do that?

18 A. Just to delete the odd word and then re-record it to a tape? It

19 would depend how much I deleted, but to do that I could perhaps do it

20 within a day. Except in this case one would have to do other things, such

21 as, well, intersperse it within a whole series of other military

22 conversations on two different tapes. No, it would take several days,

23 perhaps. I don't know. I really -- it's open-ended.

24 JUDGE RODRIGUES: [Interpretation] Okay. Very well. Thank you,

25 Dr. French. We thank you very much for coming here to the Tribunal. You

Page 8672

1 have clarified many points to the parties and to the Chamber as well. We

2 should like to thank you profoundly for coming and wish you a safe journey

3 back to your place of residence, and every success in your work, which is

4 fascinating.

5 THE WITNESS: Thank you very much indeed, Mr. President, Your

6 Honours.

7 MR. McCLOSKEY: Mr. President, this may be a good time to take a

8 break. Our next witness will require some protective measures.

9 [The witness withdrew]

10 JUDGE RODRIGUES: [Interpretation] Before taking a break, if we

11 may, Mr. Petrusic, could I ask you, with respect to General Krstic's

12 health, would you like to have shorter breaks, that is to say, shorter

13 working sessions and shorter breaks - for example, 45 minutes of work with

14 a 15-minute break - or do you prefer us to go on working for one hour, 20,

15 and then have a 20-minute break, or one and a half hours with a 30-minute

16 break? How would you like us to work, in view of General Krstic's

17 health? If you don't wish to answer now, you could talk to the General,

18 perhaps, ask him during the break, and then come back to us with your

19 answer after the break.

20 MR. PETRUSIC: [Interpretation] Yes, Your Honour. After the break,

21 we shall inform the Chamber.

22 JUDGE RODRIGUES: [Interpretation] Very well, then. For the time

23 being, let us take a 50-minute lunch break at this point. Let us adjourn

24 for 50 minutes for lunch.

25 --- Recess taken at 12.33 p.m.

Page 8673

1 --- On resuming at 1.30 p.m.

2 [The witness entered court]

3 JUDGE RODRIGUES: [Interpretation] Witness, can you hear me?

4 THE WITNESS: Yes, I can.

5 JUDGE RODRIGUES: [Interpretation] You're going to make a solemn

6 declaration now. I'm not quite sure how you're going to do it. Do you

7 have a braille edition of the solemn declaration?

8 THE WITNESS: Your Honour, I've memorised the declaration.

9 JUDGE RODRIGUES: [Interpretation] That's fine. So please make the

10 solemn declaration now.

11 THE WITNESS: I solemnly declare that I will tell the truth, the

12 whole truth, and nothing but the truth.


14 JUDGE RODRIGUES: [Interpretation] Very well, thank you. Please

15 sit down.

16 THE WITNESS: Thank you.

17 JUDGE RODRIGUES: [Interpretation] Make yourself as comfortable as

18 you can. Thank you for coming, and first you will be answering questions

19 which Mr. McCloskey is going to put to you.

20 But before we start with that, I beg your pardon for a moment, it

21 is necessary for you to confirm your name on a piece of paper, but I don't

22 know how I'm to do that.

23 Perhaps, Mr. McCloskey, you could help us.

24 As you are a protected witness, and you will be known as Witness

25 EE so we can't use your name, but you need to confirm whether we have your

Page 8674

1 name correctly. Perhaps if anybody is going to utter the name, we should

2 switch off the microphones.

3 Yes, Mr. McCloskey. I see that you wanted to say something.

4 Please go ahead.

5 MR. McCLOSKEY: Actually, I'm ready to go when you are,

6 Mr. President. I just wanted to confirm, are we in closed session?

7 Okay. We're not in closed session yet, so it's my understanding that the

8 usher whispered the witness's name in his ear.

9 JUDGE RODRIGUES: [Interpretation] Yes, but allow us to hear from

10 Witness EE whether the name that the usher whispered to him was indeed his

11 name.

12 THE WITNESS: The usher did not whisper my name in my ear; he told

13 me what my alias is. I'd be happy to write down my name.

14 MR. McCLOSKEY: That's a good idea.

15 JUDGE RODRIGUES: [Interpretation] Yes, yes. Yes, that will be

16 fine. I think that is a good way of proceeding. We'll begin with the

17 testimony by Witness EE.

18 Mr. Petrusic, have you any response to my previous question prior

19 to lunch so that we can organise our time and make the breaks?

20 MR. PETRUSIC: [Interpretation] Yes, Mr. President. Having

21 consulted General Krstic, the Defence would suggest that we preserve the

22 same rhythm that we applied up to now, that is, a break before lunch, a

23 lunch break, and then perhaps we could divide the rest of the time after

24 lunch into two as well and have another break in between.

25 JUDGE RODRIGUES: [Interpretation] Very well.

Page 8675

1 So Mr. McCloskey, we will have a ten-minute break more or less

2 halfway towards 3.00, if that suits you. Either you tell me, or I will

3 give you a sign. So please begin, your witness.

4 MR. McCLOSKEY: Thank you, Mr. President. And if we could go in

5 closed session for the first part of the testimony. We've talked to the

6 Defence about this, and it will be evident why.

7 Yes. Let us go into private session.

8 [Private session]

9 [redacted]

10 [redacted]

11 [redacted]

12 [redacted]

13 [redacted]

14 [redacted]

15 [redacted]

16 [redacted]

17 [redacted]

18 [redacted]

19 [redacted]

20 [redacted]

21 [redacted]

22 [redacted]

23 [redacted]

24 [redacted]

25 [redacted]

Page 8676













13 Page 8676 redacted – private session













Page 8677













13 Page 8677 redacted – private session













Page 8678













13 Page 8678 redacted – private session













Page 8679

1 [redacted]

2 [redacted]

3 [redacted]

4 [redacted]

5 [redacted]

6 [redacted]

7 [redacted]

8 [redacted]

9 [redacted]

10 [redacted]

11 [redacted]

12 [redacted]

13 [redacted]

14 [redacted]

15 [redacted]

16 [redacted]

17 [redacted]

18 [Open session]

19 JUDGE RODRIGUES: [Interpretation] We are in open session now, so

20 you may continue.


22 Q. And before we get into the -- some of the relevant dialects, let

23 me just ask you, do you -- based on your studies, have you travelled

24 extensively in the former Yugoslavia?

25 A. Yes.

Page 8680

1 Q. Can you just give us, oh, just a brief description of how often

2 you may have visited there?

3 A. I visited the former Yugoslavia at least -- for research at least

4 six or seven times. The most extensive stay was when I -- in 1989, 1990,

5 when I spent 13 months researching for my dissertation and travelled --

6 worked in Bosnia-Herzegovina at the institute for language and literature

7 in Sarajevo, Institut za Jecik e Knjizevnost, and also at the

8 Bosnia-Herzegovina academia of arts and sciences, and I also spent a lot

9 of time researching at the Serbian academy of arts and sciences in

10 Belgrade.

11 Since then, I had two more research trips after the war in 1997

12 and in 1998. 1997 took me to Belgrade, and 1998 took me to Bosnia again,

13 to Sarajevo and its environs, to Belgrade and to Montenegro.

14 Q. Based on your studies and your knowledge, is it possible to

15 identify indicia of ethnicity from listening to the speech of a Yugoslav?

16 A. Yes. My feeling is that you really have to be very knowledgeable

17 about the language in order to be able to do so. I think a layperson or a

18 person -- or average person might not notice all these differences, but

19 when you study and learn a lot of the dialect studies, and there were many

20 written about ethnic dialects, dialective Muslims or dialective Serbs in

21 this particular area, let's say, of Bosnia or of Croatia, then you begin

22 to note some of those small but still, I would say, audible differences

23 between -- I mean, among the various ethnic groups.

24 Q. All right. And having that in mind, can you explain to the Court

25 what the Prosecution asked you to do, and what you were able to do; and if

Page 8681

1 you can provide us with just your rough conclusions, and then you can go

2 through and support those conclusions.

3 A. The Prosecution asked me to examine tapes which -- with -- and I

4 had both the tapes and the transcripts of conversations that were

5 attributed to some officers in the army. They were attributed to General

6 Krstic [redacted]

7 [redacted]. And I was asked to analyse under

8 laboratory conditions these speech samples to see if I could identify any

9 of those characteristics which might distinguish these speakers as either

10 of Muslim or Serb or Croat, perhaps, origin, and I tried do this as

11 objectively and as honestly as I could.

12 Q. And who did you work with in that project?

13 A. I worked with Dr. French.

14 Q. And you've reviewed the overall report that you and he

15 co-authored?

16 A. Yes, I have.

17 Q. And you've seen his final evaluation for the speech part of that

18 report, reflecting his zero result?

19 A. Yes, and I would concur with it.

20 Q. All right. Now, if you could tell us, were you able to identify

21 any particular ethnic speech tendencies from any of the three speakers on

22 the two tapes that you were provided?

23 A. Yes. I have to make one caveat, that the speech sample is rather

24 short; but even in such a short speech sample, I was able to find a few

25 tendencies which make me make some -- draw some conclusions.

Page 8682

1 Q. And what were your basic conclusions, and then I think you will be

2 able to explain the tendencies you saw in the speech.

3 A. Well, my basic conclusions are that the speakers reveal some

4 features typical of some ethnic Serb, either dialects or ethnic -- or Serb

5 speech from Serbia itself. I did not find any what I would call classic

6 Muslim speech features which are spoken about in many of the dialect

7 studies.

8 Q. Could you please now just tell the Court what your examination of

9 the two tapes revealed, what tendencies did you see, and just explain it

10 as you would your class?

11 A. Okay. Well, perhaps I'll use this overhead, the ELMO for that, if

12 that's okay at this point.

13 Q. And let us make sure it's up there.

14 MR. McCLOSKEY: Okay, I think we can -- this is OTP Exhibit 847,

15 and the two or the three yellow dots are there to assist the witness and

16 don't have any significance.

17 A. May I -- is this now visible to everyone?

18 Q. Yes.

19 A. Well, the basic division in this former Serbo-Croatian speech

20 territory is the three main dialects. I'm only showing you one of the

21 main dialects; that's why there's an asterisk. I'm showing you the

22 Stokavian dialect which is the largest of the dialects of this area. The

23 other two dialects, Cakavian and Kajkavian, are purely Croatian dialects

24 and do not interest us in this particular instant.

25 The main or traditional subdivision of Stokavian is into these

Page 8683

1 three subdialects, Ikavian, and now I'll point to ijekavian, which is

2 here, ijekavian, and ekavian. These differences are based on the

3 pronunciation of a vowel that was inherited from an earlier version of

4 Slavic.

5 I will not be focussing on Ikavian in this presentation because

6 Ikavian is limited to Western Bosnia, Western Herzegovina, and Central

7 Dalmatia, and there's no signs of Ikavian at all in the speech sample that

8 I have examined.

9 What is also interesting to see from this exhibit is that the

10 ekavian form is spoken exclusively, I would say, or almost exclusively by

11 Serbs, especially in Serbia in the Federal Republic of Yugoslav. The

12 ijekavian form of the dialect is spoken by many -- all four ethnic groups

13 of the former Yugoslavia that spoke Serbo-Croatian, so we have Muslims,

14 Slavs, Croats, Serbs, and Montenegrins. However, there are some

15 differences, more complicated, and there's some more stratification among

16 these ijekavian speakers which I can go into later if need be, but there

17 are some differences, say, between the speech of Serb, ethnic Serbs from

18 Bosnia in their ijekavian speech and, say, ethnic Croats or ethnic Muslim

19 Slavs from Bosnia, and I can go into those in a bit.

20 What is most striking perhaps for me in the speech samples that I

21 reviewed is that there is a mixture of these two dialect types in the

22 speech of, I believe -- certainly of what is attributed to Obrenovic,

23 Major Obrenovic, and what is attributed to General Krstic. They often use

24 -- in some words they use the ekavian form, which would be consistent

25 with what would be used in Serbia or even among Eastern Bosnian Serbs; and

Page 8684

1 in other instances you find ijekavian forms.

2 Now, let me, just to help the Court explain where we -- how

3 significant differences of these three dialects are. I'll give one word,

4 or two or three words, perhaps, from each dialect group and show you the

5 differences. I'll try to enunciate them as clearly as possible. So the

6 word for milk, for example, in Ikavian would be "mliko," in ijekavian

7 would be "mlijeko," and in ekavian would be "mleko." The word for girl is

8 slightly different: Ikavian, "divojka"; ijekavian, "djevojka"; ekavian,

9 "dejvojka".

10 One last one because it does actually occur in our tape, in some

11 positions there's a slight other difference. Ikavian, which would be the

12 word, the past tense of to live, "zivija"; ijekavian is identical, would

13 be "ziveo" with an e-o; and ekavian would be "zivio."

14 So that's I think -- I hope I haven't bored the Court too much

15 with my analysis of these three dialects.

16 Q. So can you tell us now as you went through the tapes, were you

17 able to pick up any speech tendencies, patterns, pronunciations, and

18 identify them as possibly Serb or Muslim?

19 A. Well, one of the first things I did is I noted in some of the

20 conversations, and not in and among all the speakers simply because I

21 didn't have enough time -- enough speech material from all the speakers,

22 but I noticed some features which placed -- at least, seemed to place

23 these speakers in Bosnia.

24 One of the features in Obrenovic, what is attributed to

25 Obrenovic's speech, which is about 53 seconds on the tape, we have three

Page 8685

1 times the use of a tense called the aorist tense, which is not standard --

2 not frequent in standard use. The examples would be "izgubi," "rece" -

3 "izgubi" meaning lost, "rece" being sad - and "rekose," which would be

4 plural, they said. This would be, if you know French, this would be sort

5 of equivalent to having active use of the passe simple. So this is

6 typical especially of the speech in Bosnia, so this for me was an

7 indicator that there was some linkage to Bosnia.

8 As far as characteristics, there is a mixture here of both

9 ijekavian and ekavian in the speech samples, which are consistent with one

10 of the dialect studies I consulted specifically for the Serbs of Eastern

11 Bosnia. So, for example, in the word "here," which in Serbo -- in this

12 language is "ovde" or "ovdje," the speakers consistly use the ekavian

13 form, "ovde." Same with the word for "where"; "gde," or "gdje."

14 Consistently we see the use of the ekavian, "gde."

15 On the other hand, we do have some ijekavian forms, which would be

16 indicative of a potential origin in -- to the west of the Drina River, and

17 there we have such -- one form is "cijevi" is in the tape. That's an

18 ijekavian form. Ekavian would be "cevi." And "vjero vatno" in ekavian

19 would be "vero vatno." So these are some -- there are more features,

20 which we could go into later.

21 Q. Witness EE, would you expect ekavian speech patterns in a Muslim

22 in this situation?

23 A. Normally not.

24 Q. Is it possible, and if so, how?

25 A. It might be possible if a Muslim from Bosnia spent time, say, in

Page 8686

1 Belgrade or in Serbia, and it is possible that they might have picked up

2 some ekavian features. But my feeling is that once they would return to

3 Bosnia, they would go back to their ijekavian way of speaking.

4 Q. What specifically can you tell us about the speech on the tapes

5 attributed to General Krstic?

6 A. Yes. General Krstic has -- in the speech attributed to General

7 Krstic we have only about 28 seconds, but there are a few things that I

8 picked up which I think are significant. In the line in which he

9 says -- the speaker says, "Don't leave a single one alive," the first time

10 I believe it's, "ni jednog nemojte ostaviti zivoga."

11 We see a form of prohibition, and as I have told you earlier, this

12 was something I have researched quite extensively, which would be typical

13 of speakers west of the Drina River, meaning in Serbia - I'm sorry.

14 Strike that - in Bosnia-Herzegovina and in Croatia. It is the use of the

15 infinitive, "ostaviti," after a prohibition particle, "nemoj" or

16 "nemojte." Now, when that line is repeated, however, more emphatically,

17 it seems to me, where it's, "ni jednog zivog nemoj da ostavljate," the

18 form has changed, as you can tell, if you have the transcript, and no

19 longer are we looking at a typical west of the Drina form, but what I have

20 found in my research to be typical of Serbs in Serbia, and Montenegrins,

21 which is the loss of the infinitive, "da ostavljate," and also the

22 reduction of the prohibition particle from "nemojte," which is a plural,

23 which was in the line above, to a singular, "nemoj," and this is

24 something that I have documented and have found this same phenomenon in a

25 variety of dialects from, I would say, Serbia and Montenegro.

Page 8687

1 Q. Would this be something you would expect to see in a Muslim

2 speaker?

3 A. I had a questionnaire on specifically this form which I had passed

4 around to Muslim speakers of Bosnia, and they indicated that they would

5 not use that particular form.

6 Q. Is there anything else that you'd like to highlight as significant

7 speech tendencies of General Krstic or any other speakers in the tapes?

8 A. I think that one another feature that needs to be spoken about is

9 the status of a phoneme "h", which has been indicated in many of the

10 dialect studies as being more frequent, more frequent -- now, that doesn't

11 mean it's absolutely there all the time, but more frequent among Muslim

12 Slav speakers, and that the Serb speakers have tended more often to lose

13 that phoneme. And I have found in the speech sample that the phoneme "h"

14 is lost in several places. All three, I believe, all three speakers on

15 the tape have lost it, and especially in the word, say, "ajde," instead of

16 "hajde," which is "come on," which was also, by the way, a subject of my

17 research. I looked extensively at the dialects for that one. And also in

18 the -- there's one, I think, what we call genitive plural, "gori nema,"

19 "gori nema." I think it should have been "gorih nema," with an "h."

20 So there is evidence also in the word "hello," "alo" instead of

21 "halo," that the "h" is lost or the "h" sound is lost. This suggests

22 again that the speakers perhaps were indeed of Serb ethnicity rather than

23 Muslim ethnicity.

24 Q. Can you tell us -- you mentioned that you relied on studies,

25 dialect studies for that area. Can you just briefly tell us the time

Page 8688

1 period those studies were conducted and the ethnic origin of the authors?

2 A. Okay. That's quite an interesting subject. From around 1970

3 through the beginning of the war in 1992 in socialist Yugoslavia, there

4 were at least 18 or 19 studies that I'm aware of that are of ethnic

5 dialect rather than of geographic dialect, so that the writers of these

6 dialect studies looked specifically at the speech of particular ethnic

7 groups, be it Serb, Muslim, or Croat. So within Bosnia I believe - I can

8 check this on one of my articles - I believe I counted at least 14 such

9 what I call ethnocentric dialect studies. So I've used many of those.

10 I've looked at all 18 of those studies, actually. And I've looked also at

11 specific dialect studies for specific regions. One article I have is

12 about Western Bosnia, another article is about an ethnically-mixed area in

13 Croatia, and recently I've also looked at two dialect studies specifically

14 for Eastern Bosnia, one of which is the dialect of Serbs of a town in

15 Eastern Bosnia, the town of Kladanj. And another one is the dialect of

16 Muslims of the town of Tuholj, which is right near Kladanj. And so I've

17 been able to compare those two studies and see what are some of the

18 differences posited specifically for eastern Bosnia.

19 Q. Who authored this study of Kladanj Serbs that you spoke of?

20 A. Professor Slobodan Remetic.

21 Q. And is he a respected Serbian expert in this field?

22 A. Yes.

23 Q. And the other studies that you talked about, the pre-war studies,

24 did you find them at all jaundiced by nationalism?

25 A. I think the subtext of some of these was nationalism. However, I

Page 8689

1 think that the experts who prepared them were serious linguists who did, I

2 think -- of course, they vary in quality, but I think some of them are

3 really of high value and I think that there was a tradition, especially,

4 say, in Serbia, of high standards in dialectological research.

5 Q. You've had a chance to review a report of a Defence expert in this

6 case, a Mr. Simic. Having reviewed that report, do you find it -- have

7 you chosen to respond to anything in particular in that report that you

8 think needs to be brought out in your testimony?

9 A. To a great extent I believe we have some similarities in our

10 conclusions. One thing I might want to clarify, which might not have been

11 clear in my own report, is about the East Herzegovina dialect. What I

12 probably should have written in that report is that the Serbs of

13 Bosnia-Herzegovina, and even of Croatia, speak an East Herzegovina-type

14 dialect, which means it's not identical with what is currently spoken in

15 East Herzegovina, but it had probably gone through some local modification

16 but is based on East Herzegovina. For evidence of that, I have looked at

17 a book published 1984 in Sarajevo called "Phonological Descriptions,"

18 "Fonoloski Opisi," and specifically, two villages or two locations in

19 Eastern Bosnia, the town of Zepa and Milici, the village, I guess, or the

20 town of Milici, are said to have East Herzegovina characteristics.

21 So that's one thing I want to clarify based on that report. So

22 I'm not necessarily saying that the speakers on the tape speak the way

23 that they would speak in eastern Herzegovina, but they display some

24 eastern Herzegovina characteristics.

25 The other thing I would like to respond to is the notion that

Page 8690

1 there are only minimal or non-existent differences among the speakers,

2 especially in the context of the Yugoslav People's Army or JNA. I think

3 that certainly we're mixed nationalities in the JNA, but that most of

4 the -- many of the officer corps were of Serb ethnicity. Now, not the

5 majority, but certainly -- maybe not the majority. I'm not an expert on

6 the military, but my understanding is that a good number of the officer

7 corps were of Serb origin. It seems unlikely to me that if -- well,

8 certainly -- let me rephrase that. It seems that if these speakers were

9 members of the JNA and they have a mixed ijekavian/ekavian speech pattern,

10 the way that the Defence report speaks of it, then this actually helps my

11 thesis that the speakers on the tape were probably of Bosnian origin or

12 perhaps ijekavian dialect origin who have spent time in Serbia or in high

13 positions in Serbia where ekavian dialect was more frequently used in the

14 army.

15 Q. And did you find any indications of classic Muslim speech

16 tendencies or patterns in the tapes?

17 A. No. One thing that is apparently characteristic of ethnic Muslim

18 speech is not only the preservation of the phoneme "h," but occasionally

19 also the addition in words where it wasn't supposed to be of the phoneme

20 "h." So that, for example -- again, given the short speech sample, there

21 were no such words, but on the other hand there was also no tendency to

22 preserve "h" among the speakers.

23 In addition, the merger of the affricates or consonants which the

24 language admits, which are "che" [phoen] and "tch" [phoen], a kind of hard

25 and soft "che" [phoen], these were not -- there is quite extensive

Page 8691

1 literature which tells us that Muslim Slavs in Bosnia tend, but of course

2 not all of them do it, tend to merge these two phonemes, and there was no

3 evidence of the merging of these phonemes. These phonemes were well

4 preserved in the text, in the speech samples that I reviewed.

5 I think there was one other feature I wanted to talk about, and

6 I -- it has now slipped my mind. I might get back to it in a minute.

7 Q. That's okay. Besides the two military-type tapes, or those

8 fuzzier tapes that you were listening to, you were also provided with

9 clear tapes of the various parties. Did you use those tapes in your

10 analysis in any way?

11 A. I certainly listened to them, but the quality is so different, and

12 also I think the context was so different. Speakers in a more formal

13 context are likely to use a more standard form of the language. And in

14 the context of the tape, where this is not a formal context, they tend to

15 bring in more dialect forms. Occasionally in some of the -- well, quite

16 frequently, actually, in the speech samples from the testimonies here, I

17 did hear that "h" lost, but perhaps in some instances it was preserved,

18 and that is a difference, I think, in style. When you're trying to speak

19 a more literary form of the language, you will, from your schooling and

20 education, you will insert those elements. I think we all can do that

21 kind of code switching in any language.

22 Q. I don't know if you wanted to speak to the issues of the

23 differences of the "cha" pronunciations. If that was not the one you were

24 thinking about just ignore that question, and if you think you've got it

25 all, then if you could briefly summarise what you were able to gather from

Page 8692

1 these two tapes.

2 A. Right. My overall conclusion is that, given the limited time,

3 limited period of speech, limited -- I mean, the fact that these speech

4 samples were so short, I could notice some tendencies which made me

5 believe that the features here were more often associated with speakers

6 that are of ethnic Serb background rather than any other ethnic background

7 of Bosnia or of the former Yugoslavia.

8 Q. Thank you, Witness EE.

9 MR. McCLOSKEY: No further questions.

10 JUDGE RODRIGUES: [Interpretation] I think this would be a good

11 opportunity to have a ten-minute break, so we're going to have a

12 ten-minute break now.

13 --- Recess taken at 2.17 p.m.

14 --- On resuming at 2.29 p.m.

15 JUDGE RODRIGUES: [Interpretation] Mr. McCloskey, you are not going

16 to restart?

17 MR. McCLOSKEY: Well, Mr. President, I've spoken to Mr. Visnjic,

18 and the witness has recalled the last point he wanted to mention, and it

19 should be very brief. If he could just recall to the Court his last

20 point, and Mr. Visnjic and I agreed that that would be appropriate.

21 JUDGE RODRIGUES: [Interpretation] Very well. Please proceed.


23 Q. Witness, if you could tell us about this last point that you just

24 recalled.

25 A. Yes, it is relatively minor, but it's still something I confirmed

Page 8693

1 in the two dialect studies from Eastern Bosnia, that the Serbs of Eastern

2 Bosnia lose a J sound or a "yot" sound between two vowels, and this is

3 something I heard quite clearly on the tape. I heard the form, it is

4 good, or "dobro," without the "je," without the "yot"; while in the

5 neighbouring Muslim dialect, apparently that "yot" is preserved. So there

6 is another minor, but still if you listen for it, you can hear that little

7 difference. That would be the thing that I had forgotten to mention

8 earlier.

9 Q. Do you recall who was speaking on the tape that you heard this

10 tendency?

11 A. I think it was definitely the voice attributed to Obrenovic, and I

12 would have to review the tape one more time to make sure, but I have a

13 feeling it might have been also the voice attributed to General Krstic.

14 MR. McCLOSKEY: Thank you, Mr. President.

15 JUDGE RODRIGUES: [Interpretation] Thank you. Mr. Visnjic, your

16 witness.

17 Cross-examined by Mr. Visnjic:

18 Q. [Interpretation] Good afternoon, Witness EE. You are the only

19 expert witness who understands our language, and therefore let me ask you

20 to make pauses between question and answer to give the interpreters a

21 chance to interpret us.

22 A. Very well.

23 Q. Witness EE, have you appeared as an expert witness before, before

24 the Tribunal with respect to an analysis of voices?

25 A. Not before the Tribunal, no.

Page 8694

1 Q. Witness EE, if there are two individuals who were brought up in

2 the same environment, grew up in the same environment, they lived in a

3 similar social surrounding, would they speak differently on the basis of

4 their religion or ethnic group?

5 A. I think you would need to clarify what you mean by the same

6 environment. Are you talking about two people who are potentially

7 neighbours in the same area, or people maybe in the same town but who live

8 in different neighbourhoods?

9 Q. I am thinking of the same area. Hypothetically, let's look at

10 Eastern Bosnia. Let's start off with that, Eastern Bosnia.

11 A. Well, one of the interesting things that I found, they could be

12 from the same region but speak differently, and this is confirmed in

13 dialect studies of Eastern Bosnia that I have read, one by Senahid

14 Halilovic about the dialect of Tuholj, there could be -- and the other one

15 that I mentioned earlier about the Serb dialect of Kladanj. So there can

16 be differences. They're from the same region, and they speak somewhat

17 differently.

18 Q. I assume that they can come from different -- originate from

19 different areas but speak the same?

20 A. Can you clarify what you mean by "different areas"?

21 Q. Referring back to the study you mentioned a moment ago about the

22 Serbian dialect in Kladanj, is it possible for Serbs from Eastern Bosnia,

23 one from Kladanj, for example, and the other one from Zvornik, or two

24 individuals not necessarily Serbs, that they speak with the same

25 linguistic characteristics, speak in the same way?

Page 8695

1 A. Right. I understand your question. Well, I have looked at a more

2 general dialect overview, it's called Pregled Srpskohrvatski Dijalekata,

3 and actually in that specific dialect study, written by a professor from

4 Belgrade University, Asim Peco, there is mention of more general

5 characteristics of all of Eastern Bosnia, not just these two points, say

6 Kladanj and Tuholj, in which he does distinguish between the speech of

7 Muslims in that area and the speech of the Orthodox population in the

8 area; so, yes.

9 One particular thing that is noteworthy for Eastern Bosnia is that

10 the Muslim population can display in some parts, at least, what is known

11 in our profession as an archaic Eastern Bosnia accentuation, which means

12 they have some accents that are not, not the same as in, say, standard --

13 the standard language or in the East Herzegovina language which became one

14 of the standards of the former Serbo-Croatian language, and Asim Peco

15 addresses this in his study.

16 Q. What about social status, education, other social elements? Do

17 they lessen the gap, this distinction, or increase it?

18 A. That's a very good question. I think that in most, most cases,

19 education could have that effect indeed, but what I would like to argue

20 and what is true even of my own, let's say, American background is if I

21 speak a native dialect, let's say of Chicago, and I try to learn -- speak

22 -- through my education, I pick up a standard American dialect, not a

23 Chicago, then when I'm back in Chicago I will tend to speak like a

24 Chicagoan.

25 Similarly, I believe that even with education and even with high

Page 8696

1 education, if somebody has picked up the standard dialect, when they are

2 back home or in a more comfortable surrounding or a familiar surrounding,

3 they will code switch and use their more native dialect, and it also could

4 happen when they get more emotional or when they get into more rapid

5 speech or when they're talking to their parents, et cetera.

6 Q. Witness EE, is it possible that a Bosnian Croat, a Bosnian Serb, a

7 Bosnian Muslim speaks the same? Can they all speak the same?

8 A. It's a theoretical question. It depends where and depends in what

9 context. I think that it's entirely possible, say, in -- before the war,

10 that in Sarajevo, urban centre, you might not have been able to tell the

11 differences among the three groups; however, if you go into the provinces,

12 you tend to be able to hear more of the differences. And one of my

13 studies was particularly of Western Bosnia where I looked at Croat speech,

14 Muslim speech, and Serb speech, and I identified sort of what I called

15 seven primary speech features which are used which have been said to

16 differentiate these groups.

17 With the new reality of these three separate languages, Croatian,

18 Serbian, and Bosnian, today, for example, a Bosnian Croat is likely to use

19 more forms from standard -- from the new Croatian. So that could be in

20 terms of words or lexicon, say, they might say instead of "avion" they

21 might say "zrakoplov," which would mark them; or they will say "poduzece"

22 instead of "preduzece," et cetera, et cetera. "Avion" being airplane,

23 "poduzece" meaning -- I can't think of the word, sorry. I'm blanking on

24 the word in English, but --


Page 8697

1 A. Thank you, company. "Kazaliste" and "pozoriste", for example;

2 theatre. There's some of those.

3 As far as the Muslims, one of the features which might distinguish

4 them even in an urban setting might be their more frequent use of "h."

5 There's been a big push in Bosnia to have the Muslim speakers today, for

6 example, use "h" much more frequently even where it's not supposed to be,

7 such as in a form like widow, "udovica," where the Bosnian Muslims

8 actually say "hudovica," and that "h" has no reason for being. It's never

9 -- it never was there and should never have been there, but it is there.

10 Similarly with the word for coffee, "kafa," Muslims might say

11 "kahva," right now as in the new political realities of the Balkans.

12 Q. This last portion, I have understood as something happening now,

13 how new languages or a new language is being developed. Am I right in

14 saying that?

15 A. Yes, but if we're looking at speech samples from after the breakup

16 of Yugoslavia, we have to, to some extent, take that into consideration.

17 Prior to the breakup of Yugoslavia, there were already some --

18 quite a few differences between the speech of Croats and Serbs, and also

19 the Muslims would vacillate. They would use, say, in Bosnia they would

20 use some Croatian pronunciations, but they'd use some verbal forms or some

21 what we call lexical items or simply vocabulary words which were of Serbs

22 origin, so they had a kind of mixed speech. And -- but you could still

23 see, according to the dialect studies, and I've looked at many written in

24 the 1970s and 1980s, and even in some urban areas like Tuzla and Visoko in

25 Bosnia, where they are differentiating Muslim speech from other speech

Page 8698

1 patterns. And that's before the war.

2 Q. Let us go back to the situation in Eastern Bosnia where the vast

3 majority of the population - and we've heard statements from different

4 witnesses about this - was economically and educationally focussed on

5 Serbia. And many people who were heard here as witnesses actually worked

6 in Belgrade. They were either construction workers or work in different

7 companies.

8 So do you allow for the possibility that the use of their -- or

9 the fact that they had spent a significant portion of their time in Serbia

10 and the fact that the border is nearby, close by, and the fact that by

11 their education and training and their military service as well, that in

12 their speech, in the way they speak, the differences that you are talking

13 about have been considerably lessened?

14 A. Frankly, I believe that that would, to a great extent, depend on

15 individuals, and again, I think it's a question of code switching. If

16 these people lived extensively in Serbia or worked in Serbia but would

17 come home frequently to their parents and their family and their relatives

18 in Bosnia, that they would preserve or even return to the native -- their

19 native dialect.

20 You have to realise that I think to great extent -- I mean, I've

21 known a lot of people in Belgrade who have lived there for many, many

22 years, don't have too many relatives outside of Belgrade by now -- and I'm

23 thinking of Montenegrins, for example, who have totally adopted ekavian,

24 and certainly if it's a second generation individual whose parents might

25 be ijekavian speakers and they live within Belgrade, that the children

Page 8699

1 will adopt what's in school and speak ekavian.

2 So it really -- I think it depends on the individuals, but I

3 wouldn't be -- I wouldn't venture to guess on a general tendency. But my

4 feeling is, given what I know about socio-linguistics in many regions of

5 the world, including the former Yugoslavia, that people are good at code

6 switching, that you never do forget your -- necessarily your native

7 dialect, and that even if you have influences from other dialects, you

8 might, when you are back in your native environment, you would revert back

9 to your native dialect.

10 Q. And a native dialect for the Serbs in Eastern Bosnia, it is

11 ijekavian, am I right?

12 A. Yes, the native dialect for Serbs is ijekavian; however, however,

13 there was a lot of pressure in Republika Srpska for Serbs to adopt ekavian

14 of Belgrade. And this admittance of ekavian of Belgrade is a kind of --

15 one of those links that might link them, these Serbs, to their kin across

16 the Drina River in Serbia, and for that reason, there might be some

17 conscious reasons to want to maintain ekavian because ekavian is

18 identifiable with Serbian.

19 Q. Does that mean that people are conscious of the difference in

20 pronunciation, that is to say, whether they're speaking ijekavian or

21 ekavian, and can they correct the difference or speak differently?

22 A. I think there was an increased awareness about this over the

23 years. I'll give you a few examples from my own experiences when I would

24 travel in different parts of Yugoslavia, the former Yugoslavia, and might

25 slip into -- my own dialect that I studied was ekavian, and when I would

Page 8700

1 travel to Croatia, I would be corrected. People were aware that I was

2 speaking ekavian and identified it as Serbian.

3 When I was in Serbia, since Serbs speak both ijekavian and

4 ekavian, I didn't get into that trouble if I happened to slip and use an

5 ijekavian form, but that rarely happens. That's from personal experience.

6 Remind me, please, of your question.

7 Q. Does that mean that people are conscious of the difference in

8 pronunciation?

9 A. It would be difficult to make a generalisation, but I think that

10 many people could very well be, but this is based on my feeling rather

11 than necessarily on something that I have spent a lot of time studying,

12 and also from personal experience.

13 I've talked to a lot of people who have moved around and have

14 gone, say, to Belgrade, especially when I was there in 1998 and spoke to

15 some Serbs from Bosnia who were in Belgrade; they maintained their

16 ijekavian pronunciation quite clearly and quite deliberately. They said

17 they really don't want to switch. They feel that that -- but their

18 children are already totally ekavian speakers. So I think -- and also

19 many people in Belgrade at the time said that since there are so many

20 refugees from the other parts of the former Yugoslavia, that ijekavian is

21 used quite a lot and that they notice these differences.

22 So I think the differences are noticed. Whether it's noticed if

23 an individual makes a political -- quote unquote, political statement by

24 speaking ekavian or ijekavian, there is some precedence for that. We know

25 the examples of especially Bosnian Serb political leaders trying to force

Page 8701

1 themselves to speak ekavian, not always very successfully, but that was

2 something that happened for quite a number of -- for a few years,

3 especially during the war.

4 Q. Does that mean that if the individual wishes to give the

5 impression that he or she is a Bosnian Serb, in their speech they will, of

6 course, use partially or wholly the ekavian [Realtime transcript read in

7 error "ijekavian"] dialect? You can say yes or no, give us a yes or no

8 answer.

9 A. I'm sorry, was it ijekavian or ekavian?

10 Q. Ekavian, "E."

11 A. Could you please rephrase the question, because that made a

12 difference for me. Yes.

13 Q. Does that mean that an individual who wishes to give the

14 impression that they are talking like a Serb, will that individual in

15 their own speech consciously use elements of the ekavian, e-k, ekavian,

16 either partially or wholly?

17 A. Okay. I read -- there was a major controversy in Serb linguistics

18 about the status of ijekavian after the decision by the Bosnian Serb

19 leader at that time to declare ekavian as a language, as the dialect,

20 official dialect of the Serb-held areas of Bosnia; and this individual, I

21 think his name was -- well, I don't want to make a mistake, so I won't

22 tell you right now but I've got it in my notes. This individual said that

23 most Bosnian Serbs would speak ijekavian, but that intellectuals or those

24 who want to make a point and show that they are true Serbs, and they were

25 not a large number, would only speak ekavian, and not a mixture of

Page 8702

1 ijekavian and ekavian.

2 Now, I should point out one more thing, that the dialects of

3 Eastern Bosnia, as I said in my earlier testimony, are basically

4 ijekavian, but they have many ekavisms, many ekavian features that are

5 native to that region. So there are two possibilities here; that these

6 speakers of Eastern Bosnia had a few ekavisms in their native languages,

7 or that they picked them up through travel or through their time maybe in

8 the Yugoslav People's Army or working in Belgrade, say, or any other part

9 of ekavian-speaking Serbia.

10 Q. Does that refer to speakers who can be Muslims and Croats as well

11 but originated from Eastern Bosnia? Would I be right in saying that?

12 A. The literature shows that the Muslims of Eastern Bosnia, the

13 literature that I have read, have fewer ekavisms in their speech and often

14 they don't have -- they're certainly much less frequent in their speech.

15 And specifically, in the words that I have found in the tape recordings, I

16 found a variety of other forms that had -- did not have the ekavian

17 reflex. So such as in the word "ovde" or "gde," I found -- in some

18 villages or some areas I found "dje," which is indicative of ekavism, and

19 I found "ovdje" or "ovdi" even, but not "ovde."

20 Q. The frequency of the use of ekavian can be the result of military

21 or police training; am I right?

22 A. It's a possibility.

23 Q. Witness EE, how much time is needed, in your opinion, and you said

24 something about this at the beginning of your testimony when you were

25 explaining the phase of the disintegration of Serbo-Croatian into separate

Page 8703

1 languages, how much time, in your opinion, is necessary for this separate

2 language to be formed? I don't know if I have been clear enough.

3 A. I would like a clarification, please, yes.

4 Q. At the beginning of your testimony, you mentioned as a general

5 question for consideration the disintegration of the Serbo-Croatian

6 language into separate languages, and my question is, how long does that

7 process last until a final version of the language is formed?

8 A. Right.

9 Q. How long does this take?

10 A. It's impossible to predict, but what I can tell you is that the

11 process did not just begin in 1991.

12 The process had begun already, I would say -- well, there are two

13 things I could say. We already have an example of a separate Croatian

14 language which was part of the fascist state of Croatia, '41 to '45, which

15 of course Croatian was declared the official language, and a pre-1892

16 orthographic system was imposed on that territory. But more to our point,

17 already in the late 1960s and early 1970s, and throughout, the Croats had

18 been forming a separate language and had actually declared in the

19 declaration on the position and name of the Croatian literary language

20 that Croatian is indeed a separate language. So that process has been

21 going on for -- well, had been going on for at least 20, 25 years before

22 the breakup of Yugoslavia, and is likely to continue. And I'm not a

23 prophet, so I really can't say for how much longer.

24 With regard to the Bosnian language as a separate language, I

25 would say that that process began with the 1974 Federal Yugoslav

Page 8704

1 Constitution, which allowed for greater language rights among the

2 different republics that spoke Serbo-Croatian, and one of the things that

3 happened as a result was that the republic, the Socialist Republic of

4 Bosnia-Herzegovina, adopted a standard linguistic idiom for the republic

5 which was slightly different, say, than the western, meaning Zagreb, or

6 the eastern, meaning Belgrade, variant of the language.

7 So we sort of have an embryo of today's separate Bosnian language,

8 and as nationalist forces gained power in the '80s, those moves towards a

9 separate Bosnian language were accelerated. Now, of the three languages

10 thus far emerged - Bosnian, Croatian, and Serbian - Croatian is ahead of

11 them all in forming its own separate language, and I think we still have

12 to see where the Bosnian and Serbian different languages will go. But as

13 I said, normally these things take at least a generation because you've

14 got to train a whole cadre of school children in the "new language."

15 Q. Witness EE, during the examination-in-chief, you mentioned that if

16 somebody spent a longer space of time outside his environment, it is

17 possible that he can take on the characteristics of speech of that region;

18 but when they go back to their original region, the characteristics and

19 this influence from outside is lost. Now, my question to you is the

20 following: How much time is needed for an individual to go back to his

21 original way of speaking?

22 A. I would like to just make one correction to what you just said. I

23 didn't say that the older characteristics are immediately lost. I meant

24 to say that the person is familiar with both ways of speaking, but it

25 seems to be almost an instinctual thing for that speaker, upon return to

Page 8705

1 the native environment, to revert to what I've called the native dialect

2 of that speaker, and the influence of that other speech pattern that they

3 might have picked up lessens. It's not an absolute yes, the minute you

4 come home you lose it, but it certainly, over a short period of time, I

5 would say - and I can't venture to guess how long. I mean, it varies from

6 individual to individual - that they are likely to code switch and use

7 their native dialect. I think we all, to some extent, are capable of

8 doing that, and there's quite an extensive linguistic literature about

9 code switching and bidialectalism and how this works. Some people have

10 referred to it as home language versus work language, or home language

11 versus outside-of-the-home language.

12 Q. I assume that the period is, say, less than three years?

13 A. Do you mean the period of returning to your native dialect?

14 Q. Precisely that.

15 A. Much less than three years, I would have to say. And of course, I

16 have to add the caveat that there could be -- it does vary from individual

17 to individual. I would think that this is something that in many

18 individuals is probably just a given characteristic of code switching.

19 It's well known that speakers of a dialect are able to code switch, either

20 between their native dialect and the standard language, or their native

21 dialect and another dialect they might have picked up when they were

22 living elsewhere. I would say it's definitely -- well, most definitely in

23 much less than three years I would think there would be a return to a more

24 native form of speech. Again, also I might add, it depends. If that

25 person spends only a week at home and then returns, it's another story.

Page 8706

1 So these are all hypothetical questions which I think to some extent might

2 depend on the individual. I really can't venture to guess how long it

3 would take for any given individual. It would be wrong of me to try and

4 venture a guess.

5 Q. In the transcript of the conversation which is the subject of your

6 expert report, the word "bolan" is mentioned. This word is characteristic

7 for those belonging to all three ethnicities in Bosnia. Am I right when I

8 say that?

9 A. I'm not a hundred per cent sure about the Croats, but I certainly

10 can vouch for the Serbs and the Bosnian Muslims, yes.

11 MR. VISNJIC: [Interpretation] Thank you, Witness EE.

12 Mr. President, the Defence has no further questions of this

13 witness.

14 JUDGE RODRIGUES: [Interpretation] Thank you very much,

15 Mr. Visnjic.

16 Mr. McCloskey, any additional questions? If you have, please

17 proceed.

18 MR. McCLOSKEY: Thank you, Mr. President.

19 Re-examined by Mr. McCloskey:

20 Q. In your review of the evidence tapes, of the fuzzier tapes, did

21 you notice any bad acting, any awkward, wooden moments, anything that

22 might lead you to believe this was a put-on?

23 A. No, I did not.

24 MR. McCLOSKEY: I have no further questions.

25 JUDGE RODRIGUES: [Interpretation] Thank you very much,

Page 8707

1 Mr. McCloskey.

2 Judge Fouad Riad, please.

3 Questioned by the Court:

4 JUDGE RIAD: Good afternoon, Witness EE.

5 A. Good afternoon, Your Honour.

6 JUDGE RIAD: You highlighted the idiosyncracies between the way of

7 speech, especially between Muslims and people of Serb ethnicity,

8 indicating many words, the "h," and the "j," "yot," and so on. Suppose

9 somebody wanted -- a Muslim wanted to imitate the speech of a person of

10 Serb ethnicity. Would he find any difficulty to overcome these

11 obstacles? As much as a French would like to imitate a German, for

12 instance?

13 A. I think it would be quite difficult. I think they would have to

14 have adequate -- an expert to train them on doing that. I don't think

15 it's something that they would naturally be able to do.

16 JUDGE RIAD: So it can be detected. If one of -- one ethnicity,

17 especially Muslim, tried to imitate a Serb ethnicity, an expert can detect

18 the nuances?

19 A. I believe so. What I found in these texts is that -- in these

20 transcripts - I'm sorry - in the tapes was a kind of consistency which

21 made me believe that these people were authentic rather than acting.

22 JUDGE RIAD: So in your overall conclusion, which you said that

23 the features were more often associated with speakers of ethnic Serb

24 background, you think it is a very firm conclusion?

25 A. Well, I would have felt more comfortable had I had a longer speech

Page 8708

1 sample, but given what I had, there were, I think, enough clues to help me

2 come to that conclusion. As an academician, of course we need long speech

3 samples, but given the limitations of what I had, this is the conclusion

4 that I have reached.

5 JUDGE RIAD: Thank you very much.

6 THE WITNESS: You're welcome.

7 JUDGE RODRIGUES: [Interpretation] Thank you very much, Judge

8 Riad. I see that Madam Judge Wald has no questions, and neither have I.

9 Therefore, we wish to thank you very much, Witness EE, for coming, and we

10 wish you a safe journey and success in your future work. Please don't

11 move for the moment, because we have to bring down the blinds to protect

12 your identity.

13 We will meet again tomorrow at 9.20. For now, the hearing is

14 adjourned.

15 --- Whereupon the hearing adjourned at 3.07 p.m.,

16 to be reconvened on Tuesday, the 20th day of March,

17 2001, at 9.20 a.m.