Jump to ContentJump to Main Navigation
The Language of Bribery Cases$

Roger W. Shuy

Print publication date: 2013

Print ISBN-13: 9780199945139

Published to Oxford Scholarship Online: January 2014

DOI: 10.1093/acprof:oso/9780199945139.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (oxford.universitypressscholarship.com). (c) Copyright Oxford University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 24 November 2020

The coded bribery event of Federal Judge Alcee Hastings

The coded bribery event of Federal Judge Alcee Hastings

(p.184) [11] The coded bribery event of Federal Judge Alcee Hastings
The Language of Bribery Cases

Roger W. Shuy

Oxford University Press

Abstract and Keywords

After being acquitted in a bribery trial, the US Congress believed (but couldn’t prove) that the evidence tapes in his case contained coded references to the bribery. I was asked to determine whether that their recorded language showed evidence of being coded. The speech event took the form of a progress report in which the speakers’ topics were consistent in their schemas about getting some reference letters supporting a mutual friend. However, the way they communicated their topics and responses, including excessive use of pause fillers, checks on confirmation of mutual understanding, missing follow-ups, and unusual anaphoric referencing, were key to the conclusion that the conversations indeed were conducted in a hastily constructed, partially disguised code. Hasting then was impeached and removed from the bench.

Keywords:   progress report speech event, schemas, strategies, smoking guns, agendas, partially disguised code, pause fillers, confirmation checks, discourse follow-ups, referencing

Alcee Hastings was appointed to the federal bench by President Jimmy Carter in 1979. Two years later, the bothers Thomas and Frank Romano were tried and convicted for stealing a million dollars from a union pension fund. Before the Romanos were sentenced, a known criminal named William Dredge contacted the FBI, probably seeking some personal benefit, and informed them that Judge Hastings had solicited a bribe in this case and that the judge's good friend and D.C. lawyer William Borders was to be the intermediary who would pass the money along to Hastings. After the FBI interviewed Dredge, the bureau set up an undercover operation in which an agent would pretend to be Frank Romano and approach Borders and tell him that he wanted to pay off Judge Hastings in order to receive reduced jail time and get his forfeited union money returned to him.

Borders agreed and ultimately received twenty-five thousand dollars as a down payment for Hastings, who subsequently dismissed the judgment against the Romano brothers. A while later, Borders was arrested while he was receiving the remaining $125,000 promised by the agent posing as Frank Romano. Hastings, who made a suspiciously sudden trip to Florida, was arrested there in December 1981 on the charge of conspiring with Borders to extort the bribe. At trial, where no linguistic analysis of the tapes was used, Borders was convicted, but to the surprise of many, Hastings was acquitted.

(p.185) Despite Hastings's acquittal, the Eleventh Circuit Court of Appeals was not convinced of his innocence, and because they were allowed to use their authority to discipline fellow judges, they hired prominent lawyer John Dore to investigate the matter. Dore's subsequent detailed report strongly suggested that Judge Hastings was guilty of extorting bribery as well as lying under oath at trial. This information was enough for the Subcommittee on Criminal Justice of the Committee on the Judiciary of the U.S. House of Representatives to bring impeachment charges against Judge Hastings (Shuy 1997).

Critical evidence in the case against both Borders and Hastings included twelve intercepted telephone conversations between them recorded by the FBI between October 5 and October 8, 1981. The lead lawyer for the House Subcommittee on Criminal Justice asked me to review in particular the very brief October 5 tape and determine whether or not it contained coded language. Even if I were to conclude that it did contain some kind of code, I was not to opine on the code's meaning, because that was the proper sole task of the trier of fact, the U.S. House of Representatives.

Not all of the linguistic tools were applicable to this very short conversation. For example, the ostensible speech event was clear from the start, the speakers had the same purported schemas, and they didn’t need to employ any known conversational strategies on each other. I represent the usual approach to the analytical bribery speech event sequence slightly out of order here because the major issue was the participants’ agenda, which comprises most of the analysis.

Speech event

This speech event took the appearance of a progress report, which has the following structure:

  1. (p.186) 1. Report of progress made to this point

  2. 2. Questions or discussion about that progress

  3. 3. Mention of future related activities

  4. 4. Questions or discussion about future related activities

This purported speech event, unlike many others described in this book, remained constant throughout the twenty-three mostly short sentences of the participants. The question before the subcommittee was whether or not what looked like a progress report speech event was actually a disguised bribery speech event that was carefully coded by the participants.


Likewise, the speakers had the same schemas in common, which were consistent from beginning to end. They were in agreement about the ostensible goal of their conversation: getting information for the purported letter that Hastings was to write. Again, the subcommittee wondered whether the participants’ schemas were coded and instead related to bribery.

Conversational strategies

This was a conversation in which the participants had no need to persuade each other about anything or trick each other into admitting something illegal, because they both were on the same page in terms of their intentions and goals.

Smoking gun expressions

Hastings and Borders were careful to avoid using any apparent smoking gun expressions in this conversation, but if this conversation (p.187) actually was in code, their agendas, speech acts, and discourse use of words and sentences offered strong analytical possibilities to the contrary.


I discuss agendas out of the usual sequence here because the way the speakers talked about their topics and responses was the key to whether or not the conversation was in coded language.

The speakers’ ostensible agenda was Hastings's plan to write support letters for a South Carolina lawyer named Hemphill Pride, who had been disbarred and was now trying to get his license reinstated. If this was the speakers’ real agenda, it was accomplished in an odd manner, which must have been what attracted the attention of the subcommittee members, who were unable to determine why it sounded so odd. The answer to their bewilderment, I concluded, was that the speakers carefully coded what they were saying.

I reported two reasons for my conclusion. The first was based on the way their code was structured. The second was based on my discourse analysis of their talk. There are three types of codes: totally obvious, partially obvious, and partially disguised.

“Totally obvious” codes intend to be so unclear to outsiders that they can’t decode them because virtually every word and symbol is intended to be unclear to anyone but the participants. Professional cryptologists commonly are called upon for tasks such as this.

“Partially obvious” codes are illustrated in Oliver North's book Taking the Stand (1989, 143), in which he quotes himself in the Iran Contra operation as having said, “If these conditions are acceptable to the banana, then oranges are ready to proceed.” Another example of a partially obvious code was used in another case I worked on years (p.188) ago in which two middle-aged women who ran a chain of women's clothing stores from their Los Angeles base used such a code. In what became known as the Grandmother Mafia case, these women did a brisk drug business under the guise of a retail women's clothing chain in and around the Asian Pacific area. The government eventually intercepted their conversations and found that they were using “blouses” to refer to one type of narcotics, “dresses” for another type, and “skirts” for still another. In both of these examples of partially obvious codes there was no effort to disguise the fact that a code was being used to refer to something that only the receiver of the message could understand.

“Partially disguised” codes are subtler in that they contain a series of discourse-connected words that are consistent with a story that is very different from what the participants might want any eavesdroppers to understand. Such codes are comprised of commonly used words or sentences that are carefully selected to veil what both parties understand in order to give the impression to outsiders of an entirely different meaning. The participants tell their story in a way that potential eavesdroppers can think that the normal English lexicon and syntax they use is about a very different and coherent event than the one that the coders were leading them to believe.

In this case I concluded that the October 5 conversation was a partially disguised code that only the insiders, Hastings and Borders, could understand. At the same time, any outsiders who might have been listening could have thought that it referred to the purported surface story. The subcommittee was bothered because on its surface the conversation didn’t seem to look at all like code. Instead it appeared to be consistent and relevant to the real-life situation of gathering information for Hastings to use in his support letter for Hemphill Pride. It was specific and it didn’t slip up by using different code words for the same referent. Because partially disguised codes often are constructed spontaneously, they require direct and frequent (p.189) confirmation from the other participants that they understand the disguised meaning. The more spontaneous the code, the more confirmation is necessary. The subcommittee was unable to detect this signal as well.

The reasons for my conclusion that the October 5 conversation was conveyed in a partially disguised, spontaneously developed code were found primarily in my analysis of the participants’ responses to the topics introduced by the other speaker. First, an undisputed transcript of the entire conversation.

The October 5 conversation between Hastings and Borders

There were nineteen turns of talk in this entire conversation, numbered here for later reference as follows:


  • Yes, my brother.
  • 2.HASTING:

  • Hey, my man.
  • 3.BORDERS:

  • Uh-huh.

  • I’ve drafted all those, uh, uh, letter, uh, for Hemp.
  • 5.BORDERS:

  • Uh-huh.

  • And everything's okay. The only thing I was concerned with was, did you hear if, uh, you hear from him after we talked?
  • 7.BORDERS:

  • Yeah.

  • Oh, okay.
  • 9.BORDERS:

  • Uh-huh.
  • 10.HASTINGS:

  • All right, then.
  • 11.BORDERS:

  • See I had, I talked to him and he, he wrote some things down for me.
  • 12.HASTINGS:

  • I understand.
  • 13.BORDERS:

  • And then I was supposed to go back and get some more things.
  • (p.190)


  • All right, I understand. Well then, there's no great big problem at all. I’ll, I’ll see to it that, uh, I communicate with him. I’ll send the stuff off to Columbia in the morning.
  • 15.BORDERS:

  • Okay.
  • 16.HASTINGS:

  • Okay.
  • 17.BORDERS:

  • Right.
  • 18.HASTINGS:

  • Bye-bye.
  • 19.BORDERS:

  • Bye.
  • Agenda topics and responses


    The conversation contained seven topics, four introduced by Hastings and three by Borders.

    1. 1. Greetings, introduced by Borders, who answered the phone (1, 2, and 3)

    2. 2. I’ve drafted letters, introduced by Hastings (4 and 5)

    3. 3. Did you hear from him, introduced by Hastings (6–10)

    4. 4. He wrote some things down, introduced by Borders (11 and 12)

    5. 5. I was to go back and get some more things, introduced by Borders (13)

    6. 6. I’ll communicate with him, introduced by Hastings (14, 15, 16, 17)

    7. 7. Closing, introduced by Hastings (18 and 19)

    As with the speech event and schemas, there was nothing particularly remarkable or suspicious about the topics themselves or the sequence in which they occurred. The participants’ responses were the key to identifying this conversation as coded.

    (p.191) Responses

    Although Sacks (1972) and Schegloff (1968) had carried out research on the rules of sequencing in conversations, it was Labov and Fanshel (1977) who described sequential units more precisely: “Sequencing rules do not appear to relate to words, sentences, and other linguistic forms but rather from the connections between abstract actions such as requests, compliments, challenges, and defenses” (Labov and Fanshel, 25).

    It was not the purported topics that sounded odd to the subcommittee, since these sounded normal to them. The apparent oddness they could not detect was located in the speakers’ discourse responses. Although code is most noticeable at the level of vocabulary, its use also affects the discourse structure of a conversation, the sequencing rules in particular. Spontaneously disguised language causes the participants to speak hesitantly and strive to be careful.

    Pause fillers

    Hastings used pause fillers at the very places where he apparently was trying to be careful about his wording (twice in 4 before “letter,” once in 6 before “hear from him,” and once in 14 before “communicate”). Speakers use pause fillers for one of four possible reasons: to find the right word, to remember something, to be very cautious about how to say something, or to prevent the other speaker from usurping their turn of talk. In partially disguised codes, one can expect to hear the “uh” pause filler at exactly those places where the speaker is trying to be careful to use an appropriate code word that the listener could understand. Here Hastings appeared to use pause fillers before the key codable words “letter,” “hear from him,” and “communicate” in an apparent effort to be careful not to slip into uncoded language.

    (p.192) Checks on confirmation of mutual understandings

    One of the characteristics of partially disguised, spontaneously constructed codes is that using such codes encourages and maybe even requires the participants to make confirmation checks about whether they understand each other's coded meaning. Examples of this occurred in turns 6 to 10:


  • And everything's okay. The only thing I was concerned with was, did you hear if, uh, you hear from him after we talked?
  • 7.BORDERS:

  • Yeah.

  • Oh, okay.
  • 9.BORDERS:

  • Uh-huh.
  • 10.HASTINGS:

  • All right, then.
  • The subcommittee may have felt something odd about Hastings’ “Okay” (in 8) to Borders's “Yeah” (in 7) and their continued confirmations in 9 and 10, but they didn’t know why this sounded odd. Repetitive confirmations of understanding responses to responses also occurred in turns 11 to 17, where the same routine was used:


  • See I had, I talked to him and he, he wrote some things down for me.
  • 12.HASTINGS:

  • I understand.
  • 13.BORDERS:

  • And then I was supposed to go back and get some more things.
  • 14.HASTINGS:

  • All right, I understand. Well then, there's no great big problem at all. I’ll, I’ll see to it that, uh, I communicate with him. I’ll send the stuff off to Columbia in the morning.
  • 15.BORDERS:

  • Okay.
  • (p.193)


  • Okay.
  • 17.BORDERS:

  • Right.
  • Even more specific checking on the other participant's understanding of the code occurred at 12 and again at 14, when Hastings responded to Borders's statements with “I understand.” The first was to Borders's “See I had, I talked to him and he, he wrote some things down for me,” containing two false starts that suggest strongly that Borders was struggling for a way to code this while avoiding saying anything by which an eavesdropper could know what he really meant. False starts work like pause fillers to point to the most likely codable elements. Hastings's “I understand” here normally confirms the truth or existence of facts presented by the other speaker. It is hard to understand why Hastings needed to confirm that Borders's “talked to him and he, he wrote some things down for me.” On the other hand, his response is appropriate as an indication that Hastings had understood that Borders was referring to something other than the conventional meaning of his words.

    Although the words here are conventional, the discourse structure and sequencing rules differ enough from normal dialogue to stand out as peculiar, because there is nothing in the content of the facts they report that would require such frequent confirmation. Both speakers give strong indications of checking to see that they were catching on to each other's real meaning, one of the signs of partially coded language.

    Speech acts

    The conversation consisted of two types of speech acts—reporting facts and requesting information—which in themselves sound pretty much like normal exchanges in everyday conversations.

    • (p.194) Reporting facts

    Hastings reported four facts:

  • 4.I’ve drafted all those, uh, uh, letter, uh, for Hemp.
  • 6.And everything's okay. . . .
  • 14.I’ll see to it that, uh, I communicate with him. I’ll send the stuff off to Columbia in the morning.
  • Borders reported three facts:

  • 11.I talked to him and he, he wrote some things down for me.
  • 13.And then I was supposed to go back and get some more things.
    • Requesting information

    Hastings requested information one time:

    1. 6. did you hear if, uh, you hear from him after we talked?

    It is easy to see how the subcommittee could find nothing unusual about the conventional speech acts used in this conversation.

    Missing discourse follow-up

    The follow-up sequences in this conversation might have sounded a bit peculiar to the subcommittee, but again they could not pinpoint why. In (6), when Hastings asked, “you hear from him after we talked?” Borders responded, “Yeah,” but oddly enough Hastings did not follow up by asking what Borders heard. In conventional everyday (p.195) discourse, one might expect Borders to have said something like “yeah, I saw him yesterday.” But here Hastings made no effort to probe about what Borders heard, instead simply following up with, “Oh, okay” and again with “All right, then.” The individual words are normal, but in everyday conversation it is not likely that speakers confirm and then reconfirm that the other person heard from somebody without asking what it was that they heard. Hastings never asked this, and Borders never volunteered it, both of which suggest that something else was going on.

    Even Borders's turn 3 “uh-huh” during the greetings topic sounded odd. One might expect Borders to have said something like “what's up?” “how ya doin’?” or “why are you calling?” in his turn of talk. In fact, their October 8 recorded call contrasted with the October 5 call in which Borders did ask “how you doing?” in this part of the exchange. This didn’t happen in the October 8 conversation, when if Borders had something to tell Hastings, he might have said, “I’m glad you called because I have something to tell you,” following normal sequencing rules. Since Borders said none of these things, we can easily conclude that he knew or expected Hastings's actual agenda for the call and he also knew that he should try to stick with the code.

    Anaphoric referencing

    In turns 11 and 13 Borders speaks of what Hemp gave him as “things” and that there were “more things” that he was to go back for, to which Hastings both times responded, “I understand.” If these “things” were really points, ideas, or information for Hastings to use in his letter of support, it would not have been difficult for Borders to use these common words. Likewise, in turn 14 Hastings said that he would send the “stuff” off to Columbia, an equally odd-sounding way to refer to the support letter he purportedly was writing.

    (p.196) Conclusion of the case

    The subcommittee also gave me eleven other tape-recorded conversations between the two men that I could use to compare with this critical October 5 conversation. In none of them did I find any of the characteristics outlined above. Using my findings, I testified before the Subcommittee on Criminal Justice of the Committee on the Judiciary, House of Representatives on May 18, 1988, telling them that based on this linguistic analysis, it was my opinion that the October 5, 1981, recorded conversation between Borders and Hastings was a hastily constructed and partially disguised coded communication.

    The House of Representatives brought seventeen articles of impeachment against Hastings, sixteen of which were based on the Borders/Romano case. Article 1 said that he “engaged in a corrupt conspiracy to obtain $150,000 from defendants in United States v. Romano, a case tried before Judge Hastings, in return for the imposition of sentences which would not require incarceration of the defendants.” Articles 2 through 15 charged Hastings with lying under oath about specific elements of the case. He was convicted of seven charges that he lied under oath. The House members voted to impeach him by a vote of 413 to 3. The matter then went first to a select U.S. Senate committee, before which I gave my same testimony, and then to the full body of the Senate in 1989, which voted 69 to 26 for impeachment. Hastings became the sixth federal judge in the history of the United States to be so removed from office. Interestingly, the Senate neglected to forbid Hastings from ever seeking a federal office again. Subsequently, in 1992, he ran for and was elected to the U.S. House of Representatives, representing a district in his home state of Florida.