What Can Linguistics Say About Criminal Intent?
Roger W. Shuy
Georgetown University
European Forensic Linguistics Conference, Barcelona, April 2003
In late October, 2002, the Washington DC area was being terrorized by men who were referred to as the Beltway Snipers. The snipers left a couple of notes for the police, which were kept from the public for a time. The press was impatient to know what the notes said, so they called on profilers to explain what they might learn if they ever got their hands on the messages. As one of the people who told the press what linguistic evidence might be able to deduce, I was astonished to read the headline of the article in which I was quoted. The headline proclaimed, "Reading the Mind of a Killer." But were the FBI profilers and I really reading anybody's minds? Of course not. No science would make such an audacious claim. All we could do is to explain the language and behavioral clues that the messages contained.
Clues are what language evidence provides. Nobody would claim that broken pieces of pottery found by archeologists can provide all the information necessary to describe an ancient culture. But it does provide important clues. So linguistic analysis of language offers clues to such things as the age, gender, ethnicity, race, dialect, education level, religious orientation, political views, and occupation of the writer or speaker. Such clues do not usually identify the origin of the message, but they can narrow down the suspect list for investigators to follow. This is a valuable contribution, but linguists certainly do not read the mind of the suspect.
Let's move a bit to the evidence in criminal cases with tape recorded evidence. One of the factors often stressed in an investigation of criminal activity is the intention of the target. This is particularly important in cases that I have called "language crimes" (Shuy 1993). Such crimes are committed through saying or writing things that are illegal, such as soliciting a hit man to kill someone, offering or accepting a bribe, agreeing to purchase or sell drugs or stolen property, and making stock investments that are considered illegal. There is seldom any physical harm done to people in language crimes. Nobody is hit over the head or robbed. Everything is done with language.
Actions and intentions are obviously not the same thing. We can act with or without intentions to act. The most obvious type of criminal case grows out of the action of committing a criminal activity. Searle's example of raising ones arm is illustrative (1983). This action consists to two components: the intention to raise ones arm and the subsequent movement of the arm. If you have no intention to move your arm and there is movement anyway, this is an unintended action. If there is no arm movement, there may be only a failed intention or no intention at all. The problem in language crime cases is to discover the intention, whether or not an act is committed.
The US government recognized the need to capture "white collar crime" in the sixties and seventies, through the convening of various panels and the subsequent enactment of the Omnibus Crime bill of 1971. Out of such legislation came many other specific laws relating to such things as money laundering (18 USC 1956A3), which once again tried to get at the issue of intention:
(3) Whoever with the INTENT
(A) TO PROMOTE the carrying on of specified unlawful activity;
(B) TO CONCEAL OR DISGUISE the nature, location, source,
ownership or control of property believed to be the
proceeds of specified unlawful activity; or
(C) TO AVOID a transaction reporting requirements under
State or Federal law, CONDUCTS OR ATTEMPTS TO
CONDUCT a financial transaction involving property
REPRESENTED by a law enforcement officer to be the
proceeds of specified unlawful activity, or property
used to conduct or facilitate specified unlawful
activity....
This law seems clear enough, except for the way that the intention of the action can be proven. On the surface it would seem reasonable that anyone who commits money laundering would have first intended to do so. But it is not that simple. Remembering Searle's example of raising ones arm, it is possible that the arm could have been raised unintentionally. That is, a person who does not intend to commit money laundering yet somehow does so, could not be found guilty according to this law. It would also seem reasonable that a person who has the intention to commit money laundering might have this intent without ever actually doing it. As such, he still could be found guilty. To prove this, however, law enforcement would need to show that he had such intent even though the actual act of laundering never actually took place.
It is obviously much easier to prove that a money laundering violation took place than it is to prove that a person intended for it to take place. The action of laundering money is a physical event, one that can be shown through accounting records, video tape of the transaction, or even from the testimony of eye-witnesses and undercover agents. Getting into the target's mind to find his intentions, however, is much more difficult.
Although in much of everyday life we have the right to infer the intentions of others, in legal issues the standards must be much higher, simply because the possibility of an incorrect inference is so great and the stakes are so high. So how can we determine such things? One way is to ask the target whether or not it was his intention to launder money. This is unlikely to succeed at finding truth, since there is every reason for the target to deny it and no reason for him not to try to protect his self-interest. Another way is to ask the cooperating witness or agent whether he understood that the target intended to launder money. This is equally unlikely to succeed at finding truth since it is obvious that it is in this person's best interest to believe that he has inferred the target's intentions properly. A third way is to carefully examine the message to see what clues to intentions it might provide.
In the US at least, language crimes are best captured on tape, either audio or video. Since the early seventies, law enforcement agencies have developed the capacity to wiretap or make consensual recordings of the events which are then used as evidence. The success of this process was recently reported by a national commission set up to evaluate it. The commission took testimony from law enforcement, prosecutors, defense counsel, and judges at all levels. All agreed that recorded conversations made since the laws were enacted created a strong, if not indelible, impression on jurors. These groups from the legal field said that tape recorded evidence was among the most reliable evidence that exists and that it was the most powerful evidence that the prosecution could offer. It has led to fewer spurious prosecutions and defenses, credibility contests, and jury debate.
Enter the linguist. Linguistics deals more directly with the message than its sender and receiver. What does the language of the message indicate? What clues to the intentions of the speakers can be found? There is no magic here; no inferencing. Linguistics can't get inside the mind of the speaker or writer, but, like the archeologist, the linguist examines the language equivalent of archeological pots and shards and then shares this analysis with the triers of the fact for them to use when they make their own determination of guilt or innocence. Those not trained in archeology can't see the same things that the specialist sees. Like the skilled physician who looks at an x-ray and sees things that a patient cannot see, the linguist examines the language and sees things that are not obvious to a juror, a lawyer, or a judge.
All professions have bags of tools to work with. For linguists, this bag contains the tools of phonetics, phonology, morphology, syntax, semantics, pragmatics, discourse analysis, sociolinguistics, dialectology, and language change. Different tools are used for different tasks. For example, we use our phonetic skills to transfer the sounds of spoken language on tape to a written transcript that jurors can follow more easily. We also use phonetics as an essential tool in speaker identification cases. The tools of sociolinguistic sand dialectology come into play in cases of authorship identification. Semantics tools are used to disambiguate complex sentences and to determine ambiguity in referencing. Discourse analysis tools are often the most prominent ones used in cases where tape recordings provide the evidence, largely because the structure of the conversation is often the crucial determiner of the speaker's intentions. Carrying out a discourse analysis, of course, means that the other tools also come into play, since the larger the unit of analysis, the more a range of tools is involved.
The discourse analytical routines that relate closely to issues of intention are topic analysis, response analysis, and conversational strategy analysis.
I. How Topic Analysis Reveals Clues to Intentions
One of the best clues to our intentions is what we choose to talk about. Conversation is a series of topics brought up by both speakers who then elaborate on those topics, saying what is on their minds (Shuy 1982). Except when responding to the questions or topics of the other speaker, people choose what they want to say. The topics we introduce are the best reflection of our overall conversational agenda.
In a criminal case in which tape recordings of conversations constitutes major evidence, the first thing the linguist can do is to carry out a topic analysis. Not only does it point out clearly who brought up the subjects that frame the accusations in the case but it also provides a skeletal view of the flow of the entire conversation.
For example, if an important topic is recycled, the meaning of the recycling has several possibilities. One meaning is that, to that point at least, the topic has not been resolved in a manner acceptable to the speaker, or he would not have recycled it. People don't usually bring up topics that have been resolved. Another meaning is that the person who recycles the topic thinks it so important that it bears repeating, another clue to the agenda.
I have found that when the agent recycles a topic bearing on the alleged offense, it is usually good news for the defendant. This means that the government hasn't gotten the response it wanted and it usually means that the defendant is okay to this point at least.
When the defendant recycles a topic that does not bear on the offense, it is also good news for him, since it gives evidence of what he really wanted out of the conversation. It is commonplace for the prosecution to focus only on the places in the conversation that it believes capture a crime and to ignore all other places as irrelevant for their purpose.
Another advantage in preparing and using a topic analysis grows out of a serious problem that most juries have with tape recorded evidence. They are overwhelmed with language and tend not to keep the threads of the conversation straight in their memories. A topic analysis chart, presented as testimony, enables the jury to see a skeletonized version of the entire conversation (or set of conversations). It enables them to get a picture of the developments that led up to the passages that the prosecution focuses on. It goes without saying that such analysis also helps the defense counsel in much the same way.
It's time to give a brief example of what a topic analysis chart looks like. The following example is not the whole conversation in the case in point. I use it in reference to the case of US v. John DeLorean (Shuy 1993) only as a visual partial example.
SAMPLE TOPICS: DELOREAN CASE
topic page Agent DeLorean
1 1 Nice to see you
2 1 Had lunch yet?
3. 2 Nice building here
4. 2 We could meet them in DC
5. 2 Want a drink?
6. 3 I wanted to meet you
7. 3 Where does your company
stand?
8. 4 We've had delays
9. 4 Prior to interim financing?
10. 5 My group has the ability--
30 million
11 5. Do they have interest?
12. 6 They're Colombian folks--
dope.
etc. etc.
One of the first things we can see about this chart is that it visually separates the speaker, something like a cartoon strip with balloons over their heads indicating who is talking about what. One problem with a conventional transcript is that such visual separation is not present. Listeners find it difficult to recall who said what. This type of chart alleviates this problem. It also shows that the major topic introducer is the agent, not DeLorean. The fact that DeLorean's only topic introductions so far are questions suggests that he is somewhat of an outsider to the content of the talk.
This conversation turned out to be one of the most critical of the 63 conversations over a nine month period in the DeLorean case.
It occurred near the end of the investigation. DeLorean Motors was in financial trouble and much in need of investors or loans. The agent, pretending to be a banker, after several months of offering to help DeLorean get a loan or find investors, finally told him that he couldn't arrange a loan but that he was also in the drug business and that he could get quick money that way. But he did not say that he would stop trying to find investors. In this later conversation, the agent went all out and was more explicit about the drug business than ever before. At the conclusion of this meeting, DeLorean was arrested. The entire conversation can be summarized as follows:
TOPIC SUMMARY: DELOREAN CASE
topic DeLorean FBI agent
1-6 (greetings and small talk)
7-8 Where do you stand with
the company?
9 Prior to interim financing?
10 My group has 30 million
11 Do they have an interest?
12-15 Colombian folks. Dope.
16 It'll be dangerous
17 An 800,000 investment
returns 40 million
18-27 Two ways to go: interim
financing or this: buy
100 kilos, 300 investment, get back 14 million in ten days
28 I'm getting money through
an Irish group
29 It's got to be legitimate
30-33 We'll go ahead or stop.
34-36 I want to do it but they
need to confirm it
37-44 We'll move in 2 to 3 days. Is there any way to accelerate funding?
45-46 I'd like to do this anyhow
47-48 We can get you 30 million in 4 days
49-50 Is their interest as a loan
or an equity investment?
51-53 Their interest is coke,we
move money every 90
days, so their interest is
stock.
54-70 Can you turn loose
a few cars? We're ready
to move. In the future
you'll deal with my boss.
71 (telephone call interruption)
72-76 (talks about his company)
77 (closing)
From this topic summary chart it is clear that the topics introduced by the FBI agent all focus on his own drug business and a possible drug deal with DeLorean. The agent is finally explicit about his business, his intentions, and his offer to involve DeLorean.
DeLorean's topics are tellingly different. He raises two major topics which ultimately led to his acquittal.
1. He still speaks of "interim financing," and asks "do they still have an interest?" These expressions have two possible meanings here: (a) the continuing interest of the agent's people in providing interim financing that would derive from investments or loans or (b) interim financing that would derive from DeLorean investing money in a quick turn-around drug deal. The government obviously thought it meant the latter. But even after the agent describes how his drug business operates, DeLorean still asks him if any money he might get from them is a loan or an equity investment.
2. He creates a fiction about an Irish group (the IRA) who he falsely claims are already investing in DeLorean Motors and who insist that any DeLorean investment be legitimate and honest.
When DeLorean says, "I'd like to do this anyhow," the government thought it had him for sure. However, they first needed to determine the reference to "this." What is it that DeLorean would like to do? Note that his preceding topics are all about the Irish group that will provide him money, their insistence on his being legitimate, but who still require a confirmation from him, all three topics providing a reasonable referent to DeLorean's "this."
Undeterred, the agent continues to describe how his group operates and the conversation is interrupted by a telephone call from another agent who was monitoring the conversation in an adjacent room. We never learned what was said in that phone call, but since the conversation soon ends with no further effort to represent illegality, we can assume that the government believed that they had hooked DeLorean into a drug deal.
The topics in this case reveal very important clues to the intentions of both DeLorean and the FBI agent. The ensuing trial was based largely on the inferences made by both parties. DeLorean inferred that "investment" meant that the agent would invest in DeLorean Motors. The agent inferred that "investment" meant that DeLorean would invest in their drug scheme. The government could not show that DeLorean committed the speech act of agreeing here. And since no money was exchanged, the only crime that could have been proved was based on language, the intention to commit a crime. Thanks to topic analysis, the government's effort to prove this also failed and DeLorean was acquitted.
II. How Response Analysis Reveals Clues to Intentions
Another important way to determine clues to the intentions of participants in a criminal case is through the responses they offer to the overtures of the agents (Shuy 1993).It is useful to list the possible responses that anyone might give to the topics of the other speaker.
RESPONSE OPTIONS
Resolve the topic
Resolve part of the topic
Request clarification or amplification
Give an off-topic response
Change or redirect the topic
Give vague, ambiguous or no response at all
Offer clarification denying or changing the premise of topic
Deny the premise of the topic
Resolving the topic, or even part of it, in the way that the agent suggests can usually mean that the target is in deep trouble but the other response strategies listed here open the door to other clues to the respondent's intentions. Requesting clarification or amplification signals that the target isn't clear about the topic and needs further information, usually not damaging to the defense. Off-topic responses usually indicate that the target either doesn't understand the overture or that he is not very interested in it.
Changing the subject is a pretty good clue that the target is more interested in something else. Vague or ambiguous responses do not further the government's case. Denying or changing the premises of the topic work only for the defendant.
When the criminal case provides a multitude of tape recorded conversations, it is sometimes useful to categorize all of the target's responses, as illustrated by the case of US v. Harrison Williams, in the following fashion (Shuy 1993):
SENATOR WILLIAMS' RESPONSE TYPES
date none feedback change defer full positive negative
June 28 13 15
June 29 2 2 5
Aug. 5 30 2 4
Sept 11 9 2 3 4 2
Oct. 7 8 2 1 4 3 2 1
Jan 15 3 6 1 2
TOTALS 30 54 6 19 6 8 3
24% 42% 5% 15% 5% 7% 2%
The six tape recorded meetings in which Senator Williams was one of the many participants show that two-thirds of his responses were to give no response at all or to offer only feedback responses. One inference that we can make of this behavior is that he was by no means a leader in these conversations. In fact, he was more of an outsider to the information that other participants already shared. When specific actions were suggested, Williams suggested waiting, or deferring any action until more information was made available (15% of his responses). His changes of the subject (5%) in no way suggest his agreement with them. He gave negative responses 2% of the time. This leaves the 12% responses in which he offered a full response or a positive one. For the defense, these responses were the only ones that could possibly damage his case. Analysis was then done on the relationship of the topic to his responses in these 28 responses. In none of his responses in this type did Williams offer anything inculpatory.
In cases in which there are only one or two tapes, it is easier to represent to the jury the critical topics and responses in the same chart, as in the bribery case involving a Texas legislator (Shuy 1993):
Agent Joe Hauser and L.G. Moore Politician Billy Clayton
JH: There will be a savings of
about a million dollars
Any time you can save the
state money, I'm for it
LGM: We want to make a
contribution to your campaign
Let's take care of this thing
first and then let's think about that
LGM: Could I give you a contribution?
I'll give whatever you want to
run. $100,000 going in and we
can prepare to put in half a
million.
Any time you an show me
where you can save the state
money, I'll go to bat for you.
That's what my job is.
JH: There's $600,000 every
year.I'm keeping 600 and 600
for whatever you want to do
with it to get the business.
Our only position is we don't
want to do anything illegal
and you don't either. This is
legitimate as it can be because any time you can show me how to save the state some money, I'm going to bat for it. But you know it will be reported.
JH: Why do you have to report it?
Well, I don't want to get into
a damn tax-- (interrupted)
JH: You can report it later on,
a year from now.
No, no, no.
In this case, the undercover agent, Joe Hauser, had already corrupted the head of the Texas Operating Engineer's Union, L.G. Moore. Hauser represented himself as an agent of the Prudential Insurance Company. His sting operation was to get important figures in various businesses to reopen the bidding on their organizations' insurance policies so that Prudential could come in with a lower bid and secure the business. In the process, Hauser would offer to split with his targets any savings that his policies might bring. In short, this operated as a bribe. Moore had already bitten on this deal and he told Hauser that an old friend of his, Billy Clayton, was the speaker of the Texas State House of Representative, an influential position that could open the door to further business for Prudential.
It is true that Moore and Clayton were old friends and it was not difficult to set up a meeting. It was almost time for politicians to begin their campaigns for reelection, so Moore suggests that they approach Clayton with this in mind.
The approach begins with the perfectly legal offer to save the state a million dollars with the Prudential proposal. Quickly, Moore also offers a campaign contribution. Clayton says they should separate these things, one a time. Moore moves right back to the campaign contribution topic and Clayton's response is to saving the state money, not the campaign. Hauser then leaps in, upping the ante to 600,000 every year. He also makes explicit the quid pro quo, "to get the business." Smelling a rat, Clayton then offers his exculpatory statement, adding that he'd report any campaign contribution.
Despite this, Hauser placed a check for $5000 on Clayton's desk as the two men departed. Clayton put the check in his unlocked desk drawer and never cashed it. Nevertheless, he was indicted and went to trial for accepting a bribe.
What did the language tell us about the intentions of the participants? Clearly, agent Hauser wanted it to be a bribe transaction. Moore focused on it as a campaign contribution. Hauser's words, "whatever you want to do with it to get the business," point to his understanding of it as a quid pro quo. Clayton's focus on saving the state some money, not wanting to do anything illegal, and advising that he will report it as campaign contribution point to his own intentions. The ensuing trial was about intentions.
This simple chart of what was actually said by the speakers was influential in Clayton's acquittal.
Response analysis also played an important part in the case of two Nevada brothel commissioners who met with a woman from San Francisco, wanting to set up trailers in the state (Shuy 1993). During their initial meeting with the woman, the commissioners suspected that she was Mafia based and they didn't know quite what to do about it. Stupidly, they decided that in their next phone contact with her they would suggest that she give them a bribe. If she agreed, they reasoned, this would tell them that she was from the Mafia. Upon hearing this, the woman went to the FBI and they wired her up for her following meeting with the commissioners. The meeting took place in a noisy restaurant in Nevada. The following chart shows her topics with their responses.
Agent's topics Brothel Commissioner's Responses
I need something more
reasonable
(changes subject)
How about paying in two
to six months?
Why don't you put it in a trust?
(defers action)
$50,000 is too high.
(no response)
I want to come in right.
What's right?
(requests clarification)
How about $35,000?
No. I wouldn't take a bribe, would
you?
(denies)
There better not be
any extras.
We'll come by and see you once
in a while.
(vague response)
Just take it! Here!
What do you mean?
(requests clarification)
The commissioners' responses change the subject, defer taking any action, offer no response at all, request clarification, and introduce humor. They offer no agreement to accept a bribe. As they get up to leave, however, the woman puts $5000 in cash on the seat of one of the men, then rushes to the ladies bathroom. Seeing the cash left there on the chair, one of the men picks it up and walks out of the restaurant, where he was immediately arrested.
This is another case in which the intentions of the participants is crucial. Having agreed with each other not to take a bribe, then picking up the money on the vacant chair, were they guilty of accepting a bribe?
Was their initial telephone request for a bribe sufficient to identify their intentions? The prosecution thought so. The defense, aided by this chart of the exact words on the tape, went with the meaning of their responses and got a hung jury.
As both of these case examples show, the responses strategies made by the participants revealed a great deal about their intentions that might not be as easily discovered had there not been linguistic analysis to guide the process.
III. How Conversational Strategies Reveal Clues to Intentions
Most people use language so naturally and habitually that they are not often conscious of the fact that they employ strategies to accomplish their conversational goals (Shuy forthcoming). We tend to think of strategies as deliberate, conscious efforts. Sometimes they are and sometimes they aren't. Whether conscious or unconscious, however, conversational strategies are discoverable if we know what to look for. Like boxing, conversational participants throw punches as their offense and block punches or dance away from them as their defense. How they do this constitutes their strategies.
In the previous discussion of topic and response analysis, we have already seen some of these strategies. Speakers try to get their own agendas on the table and do various things with the agendas of the other speakers, depending on how they think or feel about them. Along with these strategies, one can include a number of others, some of which are clearly intentional maneuvers.
When the evidence in a criminal case is a tape recording made to show the guilt of a participant, the level of consciousness about strategies often raises a notch or two. Agents whose intent is to capture crime are sometimes tempted to stack the deck a bit since they know that they are creating a documentary for later listeners, such as juries, to hear. There may be many motivations for law enforcement to do this. They may know or feel that the target is guilty and try to shortcut the laborious path of getting at truth. Such strategies have been noted not only in undercover tape recordings but also in taped interrogations and confessions. Eight such strategies are the following:
1. Securing the appearance of agreement strategy:
A few years ago, the prosecution commonly made outrageous claims that "uh-huh" always indicated agreement. Having been stung a few times by linguistic testimony to the contrary, the government now uses this argument less frequently. Nonetheless, it has to be made by the defense, whether or not it is claimed by the prosecution, simply because the jury usually is not aware of what a speaker intends by his feedback markers.
2. Coaching the target strategy:
This strategy is sometimes used by social workers interviewing children who are alleged to have been sexually abused. It is also used in certain types of undercover stings in which the agent tells the target what to say when he meets another participant in a crime, such as a drug dealer or a hit man. In one celebrated case, the FBI agent tape recorded his conversation telling a politician what to say when he met the man wanting a favor. Unfortunately for the prosecution, the taped meeting showed that target used none of the words or expressions suggested by the agent and the strategy backfired badly. The intentions of the target negated the intentions of the coaching in this case.
3. Camouflaging strategy:
In this strategy the agent tries to disguise the illegality of the event into which he is trying to ensnare the target. He does this with ambiguity and vagueness, euphemisms or outright lies. The point is to get the operation jump-started so that it will get to a later point at which the target is motivated to go on to the next illegal step. At times, however, the camouflage is almost purely linguistic, as the following chart, from a stolen property case (Shuy 1993) shows:
Rubin (wearing a mike): I bought a bunch of this stuff out
East. It's stolen, you know, but I bought it back East
and, uh, I bought a lot of it. So instead of sending it
to you by mail, I was worried that it would ever get
here or anything. I brought this back myself and the
rest is coming UPS this week and I'll have it here
along with a lot of silver.
Jeweler: UPS is not the greatest way to ship.
Rubin gives six propositions without pausing before the Jeweler has a chance to respond. The recency principle avers that people tend to respond to the last, or most recent, proposition in a series such as this. The real damage, however, was in his second proposition, when Rubin tries to report that the goods were stolen. Careful listening to the tape shows that Rubin lowered his voice and deleted the "l" sound of the word, "stolen," camouflaging in the highest way, especially to a target who was hard of hearing in the first place.
normal intonation: stol
It's en
Rubin's intonation: It's
sto en.
Rubin's intention was clearly to confuse the jeweler. In my testimony in this case, I showed how the critical word was distorted to the point of inaudibility even to an audience with good hearing.
4. The Criminalizing Strategy:
The point of criminalizing a tape is just the opposite of camouflaging.
Agents who put onto the tape suggestions or references to bad things not germane to the conversation do this intentionally to make the tape look worse than it is. The issue in the case of Senator Williams was whether he intended to follow the legal practice of declaring any money he might make from a business transaction or, instead, hide any financial gain in some way. The following contrasts the way the agent tried to criminalize the tape along with the actual words used by the Senator:
FBI Agent DeVito Senator Williams
Keep it a secret. I'll Pay the taxes.
Everything is going to be hidden. I'm going to find a way
to protect myself with
some kind of declaration
Everybody can declare. You can't. I'm going to have to go
public with something or
other.
You were going to declare it We can blind trust me.
but in some other way.
When the Senator said he wanted We have it under the
to declare, he was trying to trust.
protect himself by coming up
with some kind of gimmick.
Here it is clearly the FBI agent who is uttering the bad words, not the target. This creates the contamination principle (Shuy 1993), in which later listeners (juries, prosecutors, or judges) hear the bad things but do not discriminate about who said them. It is easy for the memory to play such tricks on them and later they erroneously recall that it was the target who said these things.
Sometimes such criminalizing is so frequent that quantitative measures can be made, as the following chart shows in a money laundering case:
expression used by target by agent
confidentiality 14 0
no names will appear 0 4
protection 10 8
can't trace my name 0 2
people start looking at me 0 5
have to insulate myself 0 4
Note here how the benign business term, "confidentiality," is used frequently by the target, while that agent never uses this word at all. Instead, he says, "no names will appear," "people will start looking at me," "they can't trace my name," and "I'll have to insulate myself," all suggesting a furtive and possibly illegal venture. Context clearly shows that "protection" is used by the target as guarding against the money being stolen or lost while the agent uses the word to mean protecting himself from visibility. Clues to the intentions of both speakers can be found in this case.
5. The Blocking Strategy
Language has a number of ways to block what the other person may be trying to say, including interrupting, talking at the same time as the target (overlapping), and answering on behalf of the target.
The following example is from a Federal bribery case in which the mayor of Camden, New Jersey, Angelo Erichetti, had told the undercover FBI agent that his friend, Ken McDonald, a member of the New Jersey Casino Commission, would take a bribe for passing on the application of a new casino which was then in the planning stages (Shuy 1993). This turned out to be untrue, since Erichetti was actually planning to pocket the money for himself. But in front of the agent, he had to make it look as though McDonald's intention was to agree to the bribe. So Erichetti intentionally interrupted McDonald and answered the agent's questions on McDonald's behalf.
FBI agent Mayor Erichetti Ken McDonald
I hope, Ken, that there
won't be any problem
with you--
(interrupting)No, there's no problem.
(finishing his sentence)
--licensing or anything
in, uh, Atlantic City as
a result of this.
(answering on
behalf of Ken)
Okay, in regards
to licensing, if I
may just bring
that point out,
just recently I
talked to him on
the phone, so
there's no question
about that. You're
in first place.
Thank you very much,
Ken.
Good to see you and I'm sure--
(interrupting) I'm
sure we'll do
alright, huh?
I don't--
(interrupting) Won't
be any problems.
No problems?
I have nothing
to do with it.
Another example of the blocking strategy is illustrated in the investigation of convenience store owners in a Southern US city (Shuy forthcoming). The US Drug Enforcement Agency officer claimed to have money that he was ready to invest in something so the store owner suggested that he purchase a nearby motel that was for sale. The owner waxes eloquently about this property while the agent overlaps him with his statement about money laundering. The store owner speaks at the same time on the topic of the property. The government claimed that a representation of the intent to launder money was made here. The defense used the actual tape to show that this representation was unrecognizable. We cannot prove that the agent intentionally tried to overlap the target's speech but, since he made no other representations of illegality, there is every reason to believe that what he did was intentional. The crucial passage is the following (all capital letters indicate the part of the conversation in which both men were talking at the same time).
DEA Agent Store owner
If you're interested I might look
at doing something with you,
invest in something if you want to
show me some things and give me
some legitimate.
There is a property near
here. He wants $650,000
20% down payment and
the rest they carry at
cheap interest rates, 5
and a quarter percent a
year, no points.
I'm not interested in actually
getting into the business, doing
it.
No, but--
I'm just putting some money
somewhere
AND CLEANING IT UP AND SEE HE WANTED 650 SO
MAKING IT LOOK LIKE I HAVE YOU'RE LOOKING AT
SOME LEGITIMATE STUFF 20,000. YOU PUT DOWN
AS DOWN PAYMENT
and the rest of it the
bank will give it, have a
bank repayment on a ten
year schedule.
Yeah, that wouldn't be too bad.
6. The hit and run strategy
The hit and run strategy is also sometimes used in undercover operations. This strategy simply means that after a bad thing is inserted into the conversation, the speaker quickly moves on to something else before the target has a chance to respond. This appears to be an intentional strategy. It is especially effective if the last utterance in the series requires the target to answer. Questions are particularly good in this regard. Linguists refer to this as the recency principle.
7. The strategy of preventing the target's access to information shared by others
In cases where there are many meetings between many people over long stretches of time, one obvious way to prevent the target from access to information shared by others is to exclude him from some of the critical meetings. Such a strategy gives us every reason to suspect that it is intentional. This is exactly what happened to Senator Williams in the FBI's Abscam investigations two decades ago. When he did manage to attend a meeting, Williams' role was more to ask questions and to try to find out what the others already knew. Asking information questions is, by definition, an admission of previous ignorance. Executives and others in roles of power are often more inclined to fake it, to let their ignorance pass, hoping that they will figure it all out later. This leaves them in a very vulnerable position in a criminal investigation. But if Williams didn't know what was going on, why didn't he just ask? In hindsight, he should have, but life is never that simple. We often preserve the image of knowing when we really don't.
Besides keeping the target from necessary information, another version of this strategy is to put the bad stuff on the tape outside of the target's attention or even outside of his physical presence.
In a sensational solicitation to murder case over two decades ago, Cullen Davis, a wealthy oilman from Ft. Worth was accused of hiring a hit-man to kill his wife (Shuy forthcoming). He had asked an employee named McCrory to spy on his wife during their divorce proceedings. McCrory then went to the police telling them that Davis wanted him to find a hit-man. They wired him up and several conversations were taped as they talked in Davis' car in a parking lot. Before the trial even began, the prosecution released to the press a transcript of a portion of one of the tapes between Davis and the cooperating witness, David McCrory:
McCrory: I got Judge Eidson dead for you.
Davis: Good.
McCrory: I'll get the rest of them for you. You want a bunch of
people dead, right?
Davis: Alright, but--
McCrory: Help me too.
Davis: I've got to have an alibi ready for Art when the subject comes
up. So give me some advance warning.
McCrory: I will. I've got to go.
To say the least, this didn't look good for Davis. After listening to the tapes many times, I concluded that they seemed to be talking over each other about different topics. As it turns out, the police had also made a video of this event from a van across the parking lot. Not everything was audible in the audio portion of the tape from that location, but enough could be heard to provide markers for the physical movements of the speakers. The bad stuff that McCrory is reported to have said in the newspaper release needed to be set in the context of the video tape. When this was done, my new transcript emerged, one that includes the topic being discussed, with the overlapped speech in all capital letters, as follows:
Topic: Art (McCrory's boss)
Davis McCrory
I told him that, uh, to treat
you like any other employee
and, uh, so don't give me too
much pressure in that regard.
I can't, uh, say you're going to
be gone a day or two every
week or so.
Well look, this fucking
murder business--
You better--
--is a tough son of a
bitch
Alright,
Now you got me
GIVE ME INTO THIS GODDAMN
DEAL
GIVE ME RIGHT?
a little
NOTICE, ADVANCE NOTICE NOW I GOT JUDGE
I got Judge Eidson
dead for you
good
I'll get the rest of
them dead for you.
You want a bunch of
people dead, right?
alright, but I,
UH, YOU KNOW DID, DID
Help me too
(INAUDIBLE) OKAY?
I got to have an alibi ready
for Art when the--
Okay?
when the subject comes
UP ALRIGHT
So give me some advance--
I will
--warning
I've got to go.
Note how topic analysis was used in this case. The segment begins with them talking about the topic of Art, who was McCrory's immediate supervisor. Art had complained that McCrory was missing too much work. McCrory asked Davis to intervene, since Davis' request for McCrory to spy on his wife was the reason for his absence from work. At this point, the video shows Davis getting out of the car and walking around to the trunk to get his sun glasses. McCrory, in his effort to convince the police that his story about hiring a hit-man was true, took advantage of the distance between them and muttered the words about murder into his body mike. The defense contended that these words were never heard by Davis. The fact that Davis continued his topic about Art bolstered the defense position, and Davis was subsequently acquitted. The jury believed that McCrory had intentionally said this when Davis couldn't hear him.
McCrory's manipulation of the physical context is a clear example of how one person can prevent another from learning important information. If there had not been a video tape, however, we might never know the distances the two men were from each other and from the microphone.
8. The strategy of using a different language and culture to promote ambiguity and the appearance of the target's cooperation.
Perhaps to try to give authenticity to the crime, undercover agents sometimes take on the role of a foreigner or a person of relatively low socio-economic status. African American agents put on their best vernacular English and pretended to be drug dealers buying autos in a DC area sting operation in the eighties. In one instance, African American Postal Inspectors pretended to be uneducated people feigning injuries in an operation netting false medical claims and lawsuits in Kansas City.
This may seem well and good, but something must be said about the intended disadvantage such roles place on targets. Most humans are sympathetic to the inarticulate efforts of foreigners as they try to speak a new language. We forgive their mispronunciations and grammatical errors, recognizing that they are still learners of our language and culture. Sometimes foreigners misunderstand Western ethics and rituals. Thus, when the fake Arab sheik offered a bribe to Senator Williams, the Senator was not enraged or indignant. He simply said, "No, no, no, no," and explained how the legal process of obtaining citizenship works in America. The Senator was criticized for not raising the roof in anger. But he felt that foreign dignitaries are not to be treated this way. They may simply not understand our culture. His language revealed his intention to be polite while denying the offer.
For Senator Williams, however, the damage was done. Other Senators asked him why he didn't throw a fit, even though they might have maintained the same level of protocol themselves if they had been in his shoes. The prosecutor used this against the Senator at the trial, apparently hungry for a conviction. Even though no bribe was accepted, he was convicted of bribery by a jury that could never seem to figure this out.
A store-front lawyer in Kansas City was met one day by a Black couple who claimed they had been injured in an accident. They fumbled and stumbled with the details and the attorney, used to inner city clients, was not suspicious. They said things about their medical treatment and injuries that would not be acceptable from most clients. The attorney didn't condemn them, accuse them, or call them names. He did the best he could with the sparse information they could provide. Tape recordings of these sessions were used to indict the attorney and the prosecution did its best to convict him. But in the end they failed, since the judge recognized the unfairness of the intentional investigatory techniques and rendered one of those rare defendant's verdicts even before the defense got the opportunity to put on its case.
I have worked on cases where the agent playing the role of a lower class citizen acted the part well and captured a crime on tape in a professional manner. But I have also worked on cases where the agents take advantage of the lowered guard that the average citizen puts up when someone appearing to have inferior knowledge and status has trouble saying things appropriately. One such undercover agent got an Oklahoma judge indicted when she came to him in tears asking if there wasn't some way she could "buy a pardon" for her recently convicted husband. The judge didn't choose to correct this distraught woman's legal gaff. He went right on explaining the law to her but, to the police, it appeared that he was intentionally agreeing with her effort to "buy a pardon."
In other such sting operations, Black or Hispanic agents have played the role of relatively audacious and inarticulate drug dealers, trying to catch auto dealerships in money laundering by offering to pay cash for a number of vehicles at the same time. Shortly after the new money laundering law was enacted that required all cash transactions over $10,000 to be reported, some DC area auto dealerships were slow to catch on and were convicted in this sting operation. They were, of course, quite wrong to accept this payment in cash without reporting it, but some have argued that a certain amount of slack might have been given because of the newness and unfamiliarity of the regulation. Even some of the auto dealers who reported the transactions were convicted because of the agent's intentionally inarticulate effort to represent that he was a drug dealer buying the cars as a way of laundering the money. I am pleased to report that linguistic analysis of this inarticulate representation of illegality enabled a few of the dealers to gain an acquittal.
Conclusion:
The role of intention is huge in most criminal cases, even though there is no way that anyone can get into the minds of the people who are said to have such intentions. We can ask them what their intentions were, of course, but their response may or may not be accurate, especially when the stakes are highest. The best tool we have is found in the clues to intention that are found in the words they use. The analytical procedures described here, topic analysis, response analysis, and conversational strategy analysis, are not perfect but they provide more solid evidence that the guesswork normally offered about intentionality. The topics a person brings up offer strong clues about their agendas and what they are concerned about. Their response strategies tell us much about what they think of the topics introduced by others. Their attempts to camouflage, criminalize, block the other speaker, hit and run, isolate the other person from important facts, or to role play in a dialect that introduces new cultural and comprehension difficulties are reasonably clear indicators of their intentions. All such factors can work to the advantage of determining justice in a criminal case that has tape recorded language as evidence.
References Cited
Searle, John R. 1983. Intentionality. Cambridge: Cambridge University Press.
Shuy, Roger W. 1982. Topic as the unit of analysis in a criminal case. In Deborah Tannen (ed.) Analyzing Discourse: Text and Talk. Washington DC: Georgetown University Press.
Shuy, Roger W. 1993. Language Crimes. Oxford: Blackwell.
Shuy, Roger W. 2005. Creating Language Crimes. Oxford U Press.