On Appeal From the United States District Court For the Eastern District of Pennsylvania District Judge: Honorable Harvey Bartle, III[Copyrighted Material Omitted][Copyrighted Material Omitted][Copyrighted Material Omitted][Copyrighted Material Omitted][Copyrighted Material Omitted]

Paul J. Hetznecker, Esquire (argued) 1420 Walnut Street, Suite 911 Philadelphia, PA 19102 Counsel for Appellant Terrence Gibbs.

Christopher D. Warren, Esquire (argued) DeStefano & Warren The Lafayette Building 437 Chestnut Street Philadelphia, PA 19106 Counsel for Appellant Antjuan Sydnor Mark C. Levy, Esquire (argued) Saul, Ewing, Remick & Saul 1500

Market Street 3800 Centre Square West, 38th Floor Philadelphia, PA 19102 Counsel for Appellant Earl Lamont Brown.

Michael R. Stiles, Esquire United States Attorney Walter S. Batty, Jr., Esquire Assistant United States Attorney Chief of Appeals Robert E. Courtney, III, Esquire Deputy United States Attorney Chief, Organized Crime Strike Force Frank A. Labor, III, Esquire (argued) Assistant United States Attorney 615 Chestnut Street Philadelphia, PA 19106 Counsel for Appellee United States of America

Before: Becker, Chief Judge, Cowen, Circuit Judges, and Stagg, District Judge.*

This opinion addresses the appeals of Antjuan Sydnor, Earl Lamont Brown, and Terrence Gibbs, who were convicted of conspiring to distribute cocaine and to possess cocaine with the intent to distribute in violation of 21 U.S.C. § 846. Sydnor's appeal requires us to explore the scope of conspiracy liability for a defendant whose sole involvement in a conspiracy consists of buying drugs from another member of the conspiracy. Brown's appeal compels us to consider the limits on a government agent's testimony about the meaning of coded drug conversations. In Gibbs's appeal, we must determine whether the introduction of evidence of a conspiracy's use of violence as part of its modus operandi violates Fed. R. Evid. 404(b). In an attack on their sentences, all three defendants ask us to examine whether the District Court erred in attributing various amounts of crack and powder cocaine to each of them, where those attributions were largely based on a government agent's interpretation of coded drug conversations.

Under our existing caselaw, in order to prove a defendant's membership in a conspiracy when that defendant has only been in a buyer-seller relationship with a member of the conspiracy, the government must prove both that the defendant purchased drugs from the conspiracy and that the defendant knew that the individual from whom he purchased the drugs was part of a larger drug operation. Since the government produced sufficient evidence that Sydnor was more than a one-time buyer of drugs from the conspiracy and that, in buying drugs from Gibbs, he was aware of part of the scope of the conspiracy of which Gibbs was a leader, we will affirm Sydnor's conviction under 21 U.S.C. § 846.

We further conclude that, in Brown's case, some of the testimony of the government expert should have been excluded because, in interpreting language that the jury needed no assistance in interpreting, that testimony violated the dictates of Fed. R. Evid. 702. However, we reject Brown's contention that the agent's testimony violated Rule 704(b), for it merely translated the coded drug language, and did not opine on Brown's intent. Because we are satisfied that there was sufficient evidence of Brown's role as an enforcer for the conspiracy without the improper testimony, and that the error in admitting the testimony was harmless, we will affirm his conviction. We also conclude that the evidence of violent acts by the conspiracy, the introduction of which is now contested by Gibbs, did not violate Rule 404(b), because such violence did not constitute an act separate from the conspiracy itself. Hence his conviction too must be affirmed. On the sentencing issues, we will affirm the sentences of all three defendants in their entirety. In making this determination, we conclude that an enforcer for a drug conspiracy may be held responsible for the amount of drugs transacted by the conspiracy during the time he acts in that capacity.

I. Background Facts Relevant to All Defendants

Sydnor, Gibbs, and Brown are three of sixteen co-defendants who were charged with conspiracy to possess and distribute cocaine in violation of 21 U.S.C.§ 846. The indictment charged that the conspiracy ran from May 1992 until April 1995; Count 1 of the indictment identified Darryl Coleman and Gibbs as the leaders of the conspiracy and listed Sydnor and Brown (among others) as working for the conspiracy by distributing cocaine and collecting drug proceeds.1 Count 1 also averred that the members used coded and cryptic language when discussing the cocaine business on the phone, and that some members of the conspiracy, including Gibbs and Brown, used and attempted to use acts of violence to further the conspiracy.

At trial, the government established that Coleman and Gibbs ran a drug organization that obtained cocaine from various suppliers and resold the cocaine in Philadelphia in both powder and crack form. Coleman and Gibbs processed the cocaine into crack at different houses in Philadelphia. After Coleman was arrested on state charges in April 1994 and subsequently imprisoned, Gibbs assumed primary responsibility for the organization, which meant that he supervised and managed the business, and recruited individuals to distribute cocaine and to collect drug proceeds from that distribution.

In addition to oral testimony by former members of the alleged enterprise, the government introduced a large number of tape-recorded conversations between various co-defendants, the recording of which was authorized by a district Judge. Five tapes contain conversations between Gibbs and Sydnor. Four tapes, recorded over a three-week period, contain Discussions between Gibbs and Brown. At least thirteen tapes contain conversations between Gibbs and Coleman. The government also introduced conversations in which other defendants discussed Brown's and Sydnor's roles in the conspiracy. However, much of the language on the tapes was in code and is virtually incomprehensible to the untrained ear. The government therefore called FBI Agent Jesse Coleman to interpret the coded language. Agent Coleman has been a narcotics investigator for eighteen years, and the District Court qualified him as an expert in the analysis and interpretation of drug conversations. None of the defendants challenged Agent Coleman's qualifications as an expert in analyzing and interpreting the intercepted 771conversations. The jury convicted each of the three defendants of the conspiracy charge. We have jurisdiction to review their appeals pursuant to 28 U.S.C. § 1291.

At trial, the government attempted to prove that Sydnor processed powder cocaine into crack for distribution to others, and that he worked as a distributor for the conspiracy. The government's evidence in this regard consisted of five tape-recorded conversations between Gibbs and Sydnor, as well as a conversation between Gibbs and another confederate named Robert Saunders. The government also introduced Agent Coleman's testimony in which he interpreted those conversations. Finally, the government proffered a list of names and numbers found in Gibbs's apartment. A confederate testified that Gibbs often listed on pieces of paper names of people who owed Gibbs money. Among the initials and names on the proffered list were the initials "ANT," which, the confederate testified, referred to Antjuan Sydnor.

Sydnor's defense at trial was that the government proved only that Sydnor had a buyer-seller relationship with Gibbs. The defense pointed out that the government introduced no evidence proving that Sydnor assisted the conspiracy in acquiring cocaine, processing cocaine into crack, collecting or laundering cash proceeds, or maintaining stash houses in which cocaine was stored. In addition, there was no evidence that Gibbs ever paid Sydnor for drug-related activities or did anything involving Sydnor other than sell him distribution quantities of drugs. Therefore, the defense contended, there was no proof that Sydnor knowingly and voluntarily joined the conspiracy.

In support of his theory, Sydnor introduced evidence at trial that Gibbs had tried to have him killed because he believed Sydnor had tried to rob him. Sydnor argued that this refuted the contention that he and Gibbs were working together toward a common goal. The jury, which was instructed that a buyer-seller relationship between two people cannot by itself establish a conspiracy, rejected Sydnor's defense and convicted him of the conspiracy charge.

The Presentence Investigation Report ("PSI") found that Sydnor could be held responsible for a total of 1.8 kilograms of crack and two kilograms of powder cocaine based on the intercepted conversations. Sydnor objected to the PSI's Conclusion, claiming that Agent Coleman's interpretations of the phone conversations were inconsistent and therefore unreliable. In addition, Sydnor argued that none of the drugs sold to him were ever seized, observed, or subjected to chemical analysis, though he admitted that he could be held responsible for two kilograms of powder cocaine (in contrast to crack). The District Court, rejecting his argument, found that Sydnor was involved with and reasonably foresaw involvement with 1.5 kilograms of crack and two kilograms of powder cocaine.

On appeal, Sydnor raises two issues. First, he submits that the government failed to offer sufficient evidence to support the jury's finding that he was guilty of conspiring to distribute cocaine and to possess cocaine with the intent to distribute. When a defendant challenges the sufficiency of the evidence supporting a verdict, we must review the evidence in the light most favorable to the government. See United States v. McGlory, 968 F.2d 309, 321 (3d Cir. 1992). Sydnor bears a heavy burden, for we must uphold his conviction if any rational trier of fact could have found the essential elements of the offense beyond a reasonable doubt. See id. at 321-22. Sydnor also contends that the District Court erred in finding, in connection with his sentence, that he was involved with at least 1.5 kilograms of crack and two kilograms of powder cocaine. We review for clear error the District Court's findings of fact regarding the relevant quantities of cocaine and crack attributable to Sydnor, see United States v. Miele, 989 F.2d 659, 663 (3d Cir. 1993), as well as the types of drugs at issue, see United States v. Roman, 121 F.3d 136, 140 (3d Cir. 1997), cert. denied, 118 S. Ct. 722 (1998).

To prove a conspiracy, the government must establish a unity of purpose between the alleged conspirators, an intent to achieve a common goal, and an agreement to work together toward that goal.2 See United States v. Robinson, 167 F.3d 824, 829 (3d Cir. 1999). The government may prove these elements entirely by circumstantial evidence. See McGlory, 968 F.2d at 321 (citing United States v. Kapp, 781 F.2d 1008, 1010 (3d Cir. 1986)). The existence of a conspiracy "can be inferred from evidence of related facts and circumstances from which it appears as a reasonable and logical inference, that the activities of the participants . . . could not have been carried on except as the result of a preconceived scheme or common understanding." Kapp, 781 F.2d at 1010 (internal quotation omitted). The government need not prove that each defendant knew all of the conspiracy's details, goals, or other participants. See United States v. Theodoropoulos, 866 F.2d 587, 593 (3d Cir. 1989), overruled on other grounds by United States v. Price, 13 F.3d 711, 727 (3d Cir. 1994). However, the government must proffer sufficient evidence from which a jury could have concluded that each drug transaction in which Sydnor was involved was "a step in achieving the conspiracy's common goal of distributing cocaine for profit." Theodoropoulos, 866 F.2d at 593.

It is well-settled that a simple buyer-seller relationship, without any prior or contemporaneous understanding beyond the sales agreement itself, is insufficient to establish that the buyer was a member of the seller's conspiracy. See McGlory, 968 F.2d at 324-25; Kapp, 781 F.2d at 1010; see also United States v. Kozinski, 16 F.3d 795, 808 (7th Cir. 1994). As the Kozinski court explained: "[A] conspiracy requires an agreement to commit some other crime beyond the crime constituted by the agreement itself." Id. (internal quotation omitted). We have endorsed that logic, concluding that if the only agreement is for the seller to sell and the buyer to buy an amount of cocaine, no conspiracy exists. See United States v. Price, 13 F.3d 711, 727 (3d Cir. 1994) (noting that this is true even if the buyer buys a "distribution quantity" of drugs). This precept follows as a matter of common sense as well as basic agency law.

However, we also reasoned in Price and Theodoropoulos that "even an occasional supplier (and by implication an occasional buyer for redistribution) can be shown to be a member of the conspiracy by evidence, direct or inferential, of knowledge that she or he was part of a larger operation." See Price, 13 F.3d at 728; Theodoropoulos, 866 F.2d at 594. Despite the misgivings of the opinion writer, which are explicated in the margin, this precept is the law of the circuit, by which this panel is bound.3 Often that knowledge is evidenced by the defendant's agreement to process cocaine into crack, or collect or launder drug proceeds. This case is made more difficult by the fact that there is no evidence that Sydnor ever did anything to further the conspiracy other than buy and sell drugs. In cases where the defendant's only involvement in the conspiracy appears to be drug purchases, courts have looked to the surrounding circumstances to determine whether the defendant is a mere buyer who had such limited dealings with the conspiracy that he cannot be held to be a conspirator, or whether he has knowledge of the conspiracy to the extent that his drug purchases are circumstantial evidence of his intent to join that conspiracy.

Among the factors courts have considered in making that evaluation are: the length of affiliation between the defendant and the conspiracy; whether there is an established method of payment; the extent to which transactions are standardized; and whether there is a demonstrated level of mutual trust. See United States v. Hach, 162 F.3d 937, 943 (7th Cir. 1998), cert. denied, 119 S. Ct. 1586 (1999). While these factors are not necessarily dispositive of the issue, their presence suggests that a defendant has full knowledge of, if not a stake in, a conspiracy: when a defendant drug buyer has repeated, familiar dealings with members of a conspiracy, that buyer probably comprehends fully the nature of the group with whom he is dealing, is more likely to depend heavily on the conspiracy as the sole source of his drugs, and is more likely to perform drug-related acts for conspiracy members in an effort to maintain his connection to them.

Courts also have examined whether the buyer's transactions involved large amounts of drugs. See United States v. Flores, 149 F.3d 1272, 1277 (10th Cir. 1998), cert. denied, 119 S. Ct. 849 (1999); Kozinski, 16 F.3d at 808. A large transaction or an accumulation of deals suggests more trust, garnered over a period of time, as well as a greater likelihood that the parties have "put their heads together" to figure out planning, organization, and ways to conceal their activities. Whether the buyer purchased his drugs on credit may also be relevant. See Price , 13 F.3d at 728; United States v. Dortch, 5 F.3d 1056, 1065-66 (7th Cir. 1993); United States v. Carbone, 798 F.2d 21, 27 (1st Cir. 1986). A credit relationship may well reflect the kind of trust that is referenced supra, and often evidences the parties' mutual stake in each other's transactions. By extending credit to a buyer, the seller risks the possibility that the buyer will be unable to resell the drugs: even if the buyer does successfully resell the drugs, in this generally thinly capitalized "business," the seller will likely have to wait until the buyer collects the money from his resale before he can pay the seller back for the initial purchase. In addition, the buyer has a vested interest in the seller's ability to maintain a good working relationship with his supplier, since the buyer will not profit unless the drugs continue to flow from the seller's supplier to the seller.

Though no one of these factors alone will necessarily be sufficient--without more--to establish a mere buyer's agreement to join the conspiracy and his intent to achieve a common goal with that conspiracy, the presence of one or more of these factors furthers the inference that the buyer knew that he was part of a larger operation and hence can be held responsible as a co-conspirator.

The government contends that it has sufficiently proven that Sydnor intended to join and in fact joined the Gibbs/Coleman conspiracy, with its attendant goal of distributing cocaine and crack for profit. We therefore examine closely the intercepted conversations (which make up the bulk of the government's proof against Sydnor) with an eye towards whether the conversations reflect Sydnor's interest or stake in the success of the operation or whether they evidence a simple buyer-seller relationship between Sydnor and Gibbs.

The first relevant telephone conversation occurred on November 28, 1994. During the conversation, Gibbs told Robert Saunders that Gibbs "did something for Antjuan." Agent Coleman explained that "doing something" meant making cocaine into crack. Gibbs continued, "It was one funny looking jawn and I gave it to him. [It] . . . came out to . . . eight seven five and nine the other one came out like eight . . . twenty three . . . ."4 Agent Coleman interpreted this as Gibbs talking about a kilogram of cocaine that had a funny color. Coleman also testified that Gibbs was talking about two half-kilograms of cocaine, and that the numbers referred to the weight of the half-kilograms (875 and 823 grams, respectively) after Gibbs had cooked them into crack. Because we must construe all facts in favor of the government, and because Agent Coleman has proffered a reasonable interpretation of the facts discussed in Gibbs's conversation, this conversation indicates that Sydnor purchased what was clearly a distribution quantity (1.7 kilograms) of crack from Gibbs.

On December 1, 1994, the government recorded a conversation between Sydnor and Gibbs. They first discussed an incident that occurred during a craps game outside the Elmwood Skating Rink. Sydnor thought someone had cheated Gibbs, and he told Gibbs he had been about to beat that person up. Sydnor, as if talking to the person about Gibbs, continued, "That's my bread and butter. You leave that alone." Gibbs replied, "Boy, I got a lot of bodyguards, you, Boo, Derrick." Talking about himself, Gibbs said, "Here everybody that's the connect you can't fuck with." Sydnor agreed with him. Gibbs then told Sydnor that he had left the Rink and "was making like deliveries outside. I was telling like everybody to meet me up there." Agent Coleman testified that "the connect" meant "the drug supplier," that by "bread and butter" Sydnor meant that Gibbs was the one who supplied him with cocaine, and that Sydnor expressed that he did not want anything to happen to Gibbs.

Sydnor's statement that Gibbs was his "bread and butter" implies that Sydnor purchased a significant amount of drugs from Gibbs, and his assent to Gibbs's description of himself as a "connect" suggests that he knew Gibbs sold drugs to people other than Sydnor. That Gibbs told Sydnor he had been "making deliveries" furthers that inference. In addition, Sydnor's representation to Gibbs that he was willing to protect Gibbs as his drug connection manifests his agreement to work together with Gibbs in the distribution of cocaine.

In the same conversation, they began to speak in more intricate code, but they confused each other. Sydnor asked Gibbs what he usually "got back" when he did "a whole log up." It later became clear that Sydnor was asking Gibbs how much money Gibbs typically made from a kilogram of powder cocaine after he cut it with baking soda, but at the time, Gibbs thought Sydnor was asking how much crack Gibbs got from melting down a kilogram of cocaine.

In an effort to explain to Gibbs what he was talking about, Sydnor referenced another deal Gibbs had done. Once Gibbs figured out what Sydnor was asking, he explained his confusion: "I'm thinking that you telling me that you did your other one that I didn't do. . . . And you lost on it and I'm ready to say what the fuck is wrong with you." Sydnor indicated that he relied on Gibbs to melt down (or "cook") cocaine into crack for him and that he would not "experiment" on his own. Gibbs then laughed. In other words, testified Agent Coleman, Gibbs thought Sydnor had cooked a kilogram of powder cocaine (the kilogram that Gibbs had not cooked for Sydnor), and Sydnor explained that he would have called on Gibbs to cook it for him if he had wanted it cooked. Sydnor further explained that he had been selling "nicks," which Agent Coleman testified meant five dollar bags of cocaine or crack.

This Discussion evidences a familiarity and a working relationship between Gibbs and Sydnor that goes beyond an arm's length buyer-seller relationship. Sydnor solicited advice on the commercial aspects of cocaine distribution, which Gibbs (after some initial confusion) was able to provide. Sydnor admitted that he had done some stupid things, but he repeatedly asked Gibbs to "give[him] some credit," thus trying to assure Gibbs that he was a competent cocaine distributor who could be relied on in future business transactions. While Gibbs's relative lack of concern about Sydnor's mistake (as evidenced by the fact that Gibbs laughed at the possibility that Sydnor lost money on a deal) could be interpreted to mean that Gibbs had no stake in Sydnor's success in reselling the drugs, we think that on balance this conversation furthers the inference that Sydnor knew of the larger conspiracy and intended to participate therein.

On December 8, 1994, Sydnor and Gibbs talked again. Sydnor told Gibbs, "I got fifteen cents for you." Agent Coleman testified that Sydnor was saying he had either $1,500 or $15,000 for Gibbs. A reasonable jury could infer from the fact that Sydnor owed Gibbs money that Gibbs was willing to sell drugs to Sydnor on credit.

The next conversation occurred on December 19, 1994.5 According to Agent Coleman, Sydnor wanted to buy 4.5 ounces of powder cocaine from Gibbs to resell to a customer. Gibbs informed Sydnor that he was unable to process small amounts of powder cocaine into crack and so would not cook the cocaine into crack for him, adding that when Sydnor cooked it into crack he was not going to make much crack or much money, since it would produce such a small rock. Sydnor then told Gibbs that once he put some money together he needed to see Gibbs.

This conversation demonstrates that Sydnor resold his drugs to customers, that he continued to seek out Gibbs for advice or assistance (including cooking cocaine into crack for him), and that he wished to continue his involvement with Gibbs as soon as he was able to put together sufficient funds.

On January 20, 1995, Gibbs called Sydnor and asked him why he had been paging him earlier. Sydnor told Gibbs that he had been waiting for him. Gibbs's response (we give Agent Coleman's interpretation infra) was as follows:

"I know, I wait on my man, I told you I only had like a little bit what I had . . . you know, that shit come in I mean he give it to me like that I can feed you but otherwise, I mean like if I get a little bit I be giving you two and shit, I can't, I can't do nothing with it."

After Sydnor confirmed that Gibbs was not able to give him anything, Gibbs explained:

"I don't have anything . . . . I get a jawn, I fry the whole thing, you know what I mean, I be given one like the Peanut, I get nine the fuck, that ah, Derrick and shit, so I'll get my extras off it. That's all I be having, you know. That shit ain't doing me no good, I get to give you like two jawns and shit."

Sydnor thereupon told Gibbs that anything was better than nothing and that Gibbs should call him "whenever."

Agent Coleman testified that "feed" meant "give cocaine to," that Peanut and Derrick were cocaine customers, and that Gibbs was telling Sydnor that Gibbs could make more money dealing with customers like Derrick and Peanut because they bought smaller amounts of crack, allowing Gibbs a greater markup. Sydnor purchased kilograms (also known as "jawns"), which only gave Gibbs $500 or $1,000 profit margins. This conversation illustrates that Sydnor bought drugs in larger quantities than at least two of Gibbs's other customers. It also highlights that Sydnor was aware that Gibbs supplied a number of buyers other than Sydnor. In addition, Sydnor expressed a continuing and future interest in purchasing cocaine from Gibbs.

The final recorded call between Gibbs and Sydnor took place on January 25, 1995. Sydnor asked Gibbs, "[W]hat was the numbers on that?" Gibbs: "Uh one." Sydnor: "Huh?" Gibbs: "One." Sydnor: "Two one." Gibbs: "One yeah." Coleman testified that "two one" referred to the price of cocaine, which was $21,000 per kilogram. A jury could well have inferred that Sydnor was either asking Gibbs what price Sydnor should expect to get on the street for a kilogram of cocaine or was confirming what he owed Gibbs from a kilogram he had just purchased from Gibbs.

Viewing all this evidence in the light most favorable to the government, as we must, we believe that a reasonable jury could have concluded beyond a reasonable doubt that Sydnor knew that he was dealing with a larger drug operation when he purchased his drugs from Gibbs. See United States v. Padilla, 982 F.2d 110, 114 (3d Cir. 1992); Theodoropoulos, 866 F.2d at 594. Not only did Gibbs tell Sydnor that he preferred to sell to other buyers, but he also let Sydnor know that he had to wait on his "man"--his supplier--before he received a new shipment from which he could sell Sydnor drugs. Therefore, Sydnor knew that Gibbs was working with people on either end of the drug chain. In addition, Sydnor evidenced a familiarity with the dealings of Gibbs and with the coded drug language. Sydnor's repeated purchases from Gibbs advanced the conspiracy's goals, since it was only through distributors like Sydnor that Gibbs was able to unload the cocaine he had received from his supplier (and, in doing so, make his own profit). Finally, there was evidence that Gibbs sold Sydnor drugs on credit. Under our jurisprudence, described above, this evidence is sufficient to support the Conclusion that Sydnor intended to join the conspiracy and shared the conspiracy's goal of distributing cocaine for profit. Therefore, we will affirm Sydnor's conviction.6

Sydnor argues that, if we uphold his conviction, the District Court clearly erred at sentencing when it found that he was responsible for in excess of 1.5 kilograms of crack. He concedes that he is responsible for the two kilograms of powder cocaine that he and Gibbs discussed on December 1, but powder cocaine is evaluated very differently from crack under the Sentencing Guidelines. Sydnor objects to both the District Court's quantity determination and its determination that some of the drugs were crack. The heart of Sydnor's argument is that the District Court erred in relying on Agent Coleman's testimony as the sole basis on which to calculate the drug types and quantities. He points out that the drugs were never seized or chemically analyzed and, relying on United States v. Roman, 121 F.3d 136, 141 (3d Cir. 1997), argues that the evidence that the processed cocaine was crack was weaker than the "barely" sufficient evidence found to establish crack in that case. We discuss the drug quantity and drug type arguments separately.

A district court may rely on intercepted drug conversations to estimate drug quantities. See United States v. Collado, 975 F.2d 985, 999 (3d Cir. 1992) (relying on intercepted conversation about "one" to attribute one-eighth of a kilogram of heroin to defendant); see also United States v. Ramirez, 1998 WL 514284, at *3 (2d Cir. July 2, 1998) (stating that use of wiretap conversation to determine drug quantities is acceptable); United States v. Green, 40 F.3d 1167, 1175 (11th Cir. 1994) (same). We have recognized that in calculating the amount of drugs involved in a particular operation, a degree of estimation is sometimes necessary. See United States v. Paulino, 996 F.2d 1541, 1545 (3d Cir. 1993) (noting that the covert nature of the drug trade precludes seizure and precise measurement of the drugs that flow through a drug distribution conspiracy); see also U.S.S.G. § 2D1.1 application note 12 (stating that the quantity of drugs can be estimated when no drug seizure occurs or the amount seized does not reflect the scale of the offense).

We review factual findings, including drug quantity determinations, for clear error, see United States v. Williams, 917 F.2d 112, 113 (3d Cir. 1990), and a preponderance of the evidence must support the District Court's determination, see Miele, 989 F.2d at 663. The evidence, which need not be admissible at trial, must possess "sufficient indicia of reliability to support its probable accuracy." Id. (internal quotation omitted).

The government argues that the November 28 conversation (in which Gibbs told Saunders that he "did something" for Antjuan) reflected that Gibbs had processed two half-kilograms of cocaine into crack for Sydnor; the total amount of crack produced was 1,698 grams (1.7 kilograms) of crack.7 The crack weighed more than the cocaine because Gibbs added baking soda to the cocaine powder before cooking it. Based on Agent Coleman's testimony that Gibbs was talking about cooking powder cocaine into crack and that the numbers Gibbs mentioned to Saunders represented the weight of the crack, it was not clearly erroneous for the District Court to attribute 1.7 kilograms of drugs to Sydnor based on this conversation. The Court attributed an additional kilogram of cocaine to Sydnor based on the December 1 conversation. Although Sydnor concedes that he is liable for two kilograms of powder cocaine based on this conversation, the District Court erred on the side of caution in attributing only one kilogram to Sydnor. Gibbs mentioned "your other one that I didn't do"; the District Court acknowledged that the one Gibbs "did" could be the same kilogram that Gibbs had cooked for Sydnor in the November 28 conversation. It was not clearly erroneous for the District Court to attribute one kilogram of powder to Sydnor based on this conversation.

However, the District Court attributed another kilogram of powder to Sydnor based solely on the January 25 call to Gibbs in which Sydnor asked what "the numbers" on a kilogram of powder cocaine were. We must determine whether a district court may make an estimate of drug quantity based on a simple price quote, without more. Courts have estimated drug quantities based on the amount of money the defendant is carrying, see United States v. Hicks, 948 F.2d 877, 882 (4th Cir. 1991); the amount of "cutting agents" found on the defendant, see United States v. Lucas, No. 97-30325, 1998 WL 708776, at *1 n.1 (9th Cir. Oct. 6, 1998), cert. denied, 119 S. Ct. 1283 (1999); testimony by a co-defendant about the weight of drugs he and the defendant transported; and testimony about average amounts sold per day multiplied by length of time sold, see United States v. Maggard, 156 F.3d 843, 848 (8th Cir. 1998), cert. denied, 119 S. Ct. 1372 (1999). However, there appear to be no instances in which drugs were credited to a defendant based solely on a pricing conversation without further negotiations confirming the sale.

The government cites two cases for the proposition that it need not prove that amounts under negotiation were actually distributed. See United States v. Layeni, 90 F.3d 514 (D.C. Cir. 1996); United States v. Williams, 994 F.2d 1287 (8th Cir. 1993). However, in Layeni, the district court noted that the amounts attributed to Layeni that were not actually distributed by him only included: (i) amounts that he offered to an undercover agent but that the agent did not purchase; and (ii) amounts that the agent agreed to purchase and that Layeni promised to produce but did not. See Layeni, 90 F.3d at 521. Likewise, in Williams, the defendant actually negotiated a sale of cocaine: he offered to obtain and sell the informant the quarter-pound of cocaine that the agent requested, at a price he knew the agent could pay. See Williams, 994 F.2d at 1293. No such negotiations were present in the January 25 conversation.

We think it too speculative to conclude that the January 25 pricing call meant that Sydnor had one kilogram in his possession and was ready to resell it at the price designated by Gibbs; it is as viable--if not more viable--to assume that Sydnor was simply obtaining price information in general or checking to see how much he would have to pay Gibbs to buy his next kilogram. Since, as Agent Coleman admitted, there was no evidence that a kilogram ever changed hands after this conversation, we conclude that the District Court clearly erred in attributing a kilogram of cocaine to Sydnor based on this conversation. However, because Sydnor's offense level was not affected by the attribution to him of this additional kilogram of cocaine, this error was harmless.

In sum, it was proper to attribute to Sydnor at least 1.5 kilograms of crack and one kilogram of powder cocaine, and any error that occurred in attributing another kilogram of powder cocaine to Sydnor was harmless, as it had no effect on the length of Sydnor's sentence.

In addition to objecting to quantity, Sydnor contends that the District Court erred in concluding, based on Agent Coleman's testimony, that the 1.7 kilograms of cocaine discussed in the November 28 conversation were kilograms of crack, rather than powder cocaine or a different cocaine base. The Sentencing Guidelines describe crack as "the street name for a form of cocaine base, usually prepared by processing cocaine hydrochloride and sodium bicarbonate and usually appearing in a lumpy, rocklike form." U.S.S.G. § 2D1.1 note D. At trial it became quite clear that all of the testifying defendants believed that what Gibbs was making when he cooked the cocaine was crack.

First, the government introduced the testimony of Collier and Ellis, two witnesses to Gibbs's crack-making process. Ellis stated that Gibbs put powder cocaine into a pot, put baking soda in it, put water in it, and started cooking it. After it became liquid, Gibbs put ice into it, pulled it out of the pot, broke it into pieces, and dried it. Collier testified that Gibbs had performed the same process in his presence.

Second, Agent Coleman, whose qualifications have not been challenged, described what those witnesses had seen as a common way to manufacture crack. He stated, "One process they use when they make crack cocaine, they put cocaine and baking soda in the water and they cook it until the cocaine dissolves. Then they put ice in it to cool the mixture and it becomes oily. They remove that, that's the crack cocaine . . . and they dry it." Sixteen years as an undercover narcotics agent renders Coleman's testimony about this process reliable. Third, two witnesses testified that Gibbs distributed both powder cocaine and crack. These pieces of evidence make clear that crack was an integral part of Gibbs's larger conspiracy.

The question, then, is whether the District Court clearly erred in concluding that the 1.7 kilograms of cocaine that Gibbs told Saunders he had given Sydnor was in crack form. Although it is a close question, we conclude that the District Court did not clearly err, based on Gibbs's consistent use of the expression "doing up" or "doing something," which Coleman interpreted to mean cooking cocaine into crack, and on Coleman's testimony that "doing something for Antjuan" meant turning cocaine into crack and selling the resulting crack to Sydnor.

Sydnor contends that Coleman misinterpreted the November 28 conversation: he claims that "doing something" does not necessarily mean converting cocaine to crack. However, the December 1 conversation supports the Conclusion that Gibbs understood "doing up" cocaine to mean cooking cocaine into crack. On December 1, Sydnor asked Gibbs how much he "got back" when he "did a whole log up"; Gibbs interpreted this to mean, "How much crack do you get when you cook a kilogram of cocaine?" while Sydnor intended it to mean, "How much cocaine do you get from a kilogram of powder after you mix in the baking soda?" Therefore, when Gibbs heard the expression "done up" or some version thereof, he interpreted it as cooking cocaine into crack. And when he used the same expression in his conversation with Saunders, it is likely that he meant that he cooked a kilogram of cocaine into crack for Sydnor.

A few further points support the inference that what Sydnor had received from Gibbs on November 28 was crack. It is clear that Sydnor knew how to cut his own cocaine with baking soda; it appears that Sydnor only asked Gibbs to prepare cocaine when Sydnor wanted the cocaine cooked into crack. This is evidenced in the"4.5 ounce" Discussion on December 19, and in the exchange on December 1 about letting Gibbs do Sydnor's "cooking" for him and about Sydnor's reluctance to experiment. Sydnor repeatedly asked Gibbs to give him "some credit" when Gibbs suggested that Sydnor had cooked for himself. Sydnor appears to be telling Gibbs that he was not going to start cooking drugs on his own. Therefore, when Gibbs discussed "doing something" to the drugs he was selling Sydnor, it appears to have meant cooking the powder into crack.

In addition, Agent Coleman testified about the numbers Gibbs spoke of in the November 28 conversation. Coleman stated:

"Mr. Gibbs is describing to Mr. Saunders, or telling Mr. Saunders that he cooked a kilogram of powder cocaine into crack cocaine for Mr. Sydnor. And using the process that they call whipping to increase the weight of the powder cocaine temporarily as it turned to crack cocaine, he describes the weights that he got for the two half kilograms of cocaine that he cooked. You know, a half kilogram is 500 grams. He says that he got 900 on one half kilogram . . . and he says he got either 875 or 823 grams on the other half kilogram of powder cocaine after he completed the cooking and whipping process."8

Based on the above evidence, we cannot conclude that the District Court clearly erred in finding that the 1.7 kilograms discussed on November 28 was crack. Therefore, the District Court did not err in attributing in excess of 1.5 kilograms of crack to Sydnor, as well as one kilogram of powder cocaine, and we will affirm Sydnor's sentence.

The critical issue in Brown's appeal is whether he was an "enforcer" for the conspiracy, for that is the sole basis on which he can be said to have participated in the Gibbs/Coleman operation. At trial, Brown argued that the government had presented no evidence that he was an enforcer for the conspiracy. Nevertheless, the jury convicted Brown of the conspiracy charge. At the sentencing hearing, the defense objected to the PSI's Conclusion that Brown was responsible for twenty-six kilograms of powder cocaine and 49.5 ounces of crack. The District Court rejected these arguments, attributed to Brown all of the drugs handled by the conspiracy during the time in which Brown was involved in the conspiracy (for a total of twenty-seven kilograms of powder cocaine and nine ounces of crack), and calculated an offense level of 38.9 The Court thus sentenced Brown to 324 months in prison and ordered him to forfeit $5,200 to the government. Brown timely appealed his conviction and sentence.

On appeal, Brown presents a number of arguments why his conviction should be overturned and why, in the alternative, he is entitled to resentencing. With regard to his conviction, Brown argues that the government at trial failed to offer any evidence that he joined the larger conspiracy or that he took action to further the conspiracy's ends. He also contends that the District Court erred in allowing Agent Coleman to testify about ultimate issues of law, including Brown's knowledge and intent, in violation of Fed. R. Evid. 704(b), and he implicitly argues that Coleman's testimony violated Fed. R. Evid. 702 to the extent that his explication of the meaning of certain conversations was not helpful to the jury. As to sentencing, Brown submits that the District Court erred in: (i) failing to appoint new counsel for him because of his irreconcilable differences with his attorney; (ii) calculating the amount of drugs attributable to him; (iii) increasing his criminal history category based on a juvenile conviction; (iv) adding two points to his offense level based on an earlier conviction for possession of a weapon; and (v) failing to address his request that he receive a downward adjustment to his offense level based on his status as a minor participant in the conspiracy. We address the validity of Brown's conviction, as well as sentencing issues (ii) and (iv), below, after setting forth the relevant facts. We dispose of the remaining issues in the margin.10

The government's contention that Brown was associated with the conspiracy was two-pronged: first, it claimed that he was an enforcer, and second, it alleged that he purchased drugs from Gibbs. Because the drug purchases, if any, were so minor, we focus on Brown's role as enforcer. On February 14, 1995, Gibbs was shot during an apparent robbery attempt; he came to suspect that the shooter had been either Antjuan Sydnor11 or Belvin Brickel. The government alleged, based on the recorded conversations and the testimony of Vincent Collier, that Gibbs asked Frank Fluellen and Earl Brown to hurt or kill the person Gibbs suspected had shot him. The heart of Brown's defense was essentially one of mistaken identity: there was another person in the conspiracy named Earl (Earl Packer Hunte), and Hunte was the true enforcer for the conspiracy.12 Since the conversations provide the bulk of the evidence against Brown, we discuss each relevant exchange.

The first recorded conversation arguably relating to Brown took place on March 1, 1995, between Gibbs and Fluellen. Fluellen asked Gibbs whether he paged "Earl" (without specifying which Earl) earlier that day, and Gibbs said no. Gibbs then stated that he was going to call Earl to "see if he can make any progress." Gibbs also told Fluellen that Gibbs had not planned on calling Earl "till [Gibbs] got the jawn." Fluellen ended the conversation by telling Gibbs that he would "be on post." Agent Coleman interpreted this conversation as Fluellen asking Gibbs about a car and telling Gibbs that he was "ready and waiting."

The first conversation in which Brown is recorded occurred the next day. Gibbs called Brown and told him that he was waiting for Collier to call him back, since Collier had been out the night before. Collier later testified that Gibbs had asked him to steal a car for a person or persons to use in their attempt to kill Sydnor or Brickel (an attempt referred to as a "mission"). Gibbs told Brown that he hoped Collier "did that." Brown told Gibbs, "Know what I mean, got to do it tonight . . . . Before Friday." Gibbs responded, "Yeah, I hope so. That shit getting on my nerves. . . . That [guy] be around and I'll be thinking about that shit, more and more." Gibbs told Brown, "I'm almost sure he got it. It should be somewhere sitting. So I'll just let you know where it's at so y'all can pop it right there." Agent Coleman testified that Brown and Gibbs were discussing whether Collier had stolen a car for Brown to use on his attempt to hurt Sydnor, and that "got to do it tonight" referred to that attempt.13

A few minutes after Brown's conversation with Gibbs, Gibbs called Collier. Gibbs asked Collier, "You get that?" to which Collier replied that he had not, because his hand hurt. Collier told Gibbs that as soon as it got dark that night he would go out and "grab anything [he] can." Gibbs responded, "I need it kind of early, man, cause they be trying to get on a mission." Agent Coleman testified that Collier was telling Gibbs he was going to steal a car. Collier himself testified that "going on a mission" meant going to hurt or kill someone, though the people Gibbs was sending on the mission were not identified in this conversation.

The next afternoon, March 3, at 2:02 p.m., the government intercepted another conversation between Gibbs and Collier. Collier told Gibbs, "I got this jawn sitting and shit." Gibbs announced, "I'm trying to get this boy tonight, man." Collier recounted to Gibbs the story of how he stole the car and told him where he had left it. Gibbs then said, "I gotta tell E . . . so they can move [it]."

Minutes later, at 2:15 p.m., Gibbs called Brown, telling him, "Vince got that jawn." Brown informed Gibbs that he and someone else had gone out the previous night but that the person (presumably the person they were looking for) didn't "come back out." Gibbs said, "[T]onight will be the night though," to which Brown responded, "I know." Gibbs confirmed that he had a stolen car waiting for Brown when Brown wanted it. Agent Coleman testified that "tonight will be the night" meant that it "would be the night that [they] would be able to go through with the mission that they . . . talked about earlier." Agent Coleman further testified that when Brown said that he had gone out the previous night but that the person had not come back out, he believed that Brown was referring to going out and attempting to kill Sydnor or Brickel but that Sydnor or Brickel did not come back outside.

On March 17, the government recorded Gibbs speaking to Fluellen. The Discussion went as follows:

"Gibbs: I was um, telling T, you know, when he getting ready to go inside that, that, club his pants be down, you know what I mean."

"Fluellen: That's, that's what we trying to find out now. . . ."

"Gibbs: They pull his pants down to go in there cause they don't play that in there, you know."

"Fluellen: That's what I, I tryin' to find out which jawn he go to."

"Gibbs: Yeah."

". . ."

"Gibbs: Right here, before when you get out of there he gotta take it and leave it in there with him in there, you know what I mean."

"Fluellen: A huh."

"Gibbs: It'll be in the wheel."

"Fluellen: Ah huh."

"Gibbs: I might you know that's like the perfect place and shit.""Fluellen: Yeah I know. Okay, that's what "E" was talkin' bout then."14

Agent Coleman testified that Gibbs was telling Fluellen that the club at 47th Street and Woodland Avenue had a metal detector so that guests could not enter the club with a gun. (In this instance, "jawn" referred to a nightclub.) Agent Coleman said that when people have "their pants down," it means they are unarmed and that "in the wheel" meant "in the car." Coleman further testified that he understood Gibbs and Fluellen to be discussing a plan to allow their target to enter the club and shoot him when he came out because he would not have a gun.

On March 24, in a call between Brown and Gibbs, Brown told Gibbs, in no particular context, "I was on that last night plus we're gonna be on that tonight." Gibbs responded that he wanted to talk to Brown about "that" and that they could discuss it once they got together. Agent Coleman testified that "on that last night" and "on that tonight" referred to the mission Brown was on for Gibbs.

Three days later, on March 27, Gibbs spoke to Collier. Collier told Gibbs that he had seen "Earl" the other day. Gibbs recounted, "Earl was like he was on a mission the other night . . . . Ain't never turned nothing out." Collier later testified that when he spoke about "Earl" he meant Earl Brown. Gibbs and Collier then proceeded to discuss the advantages of "getting" people in the daytime, since they would not be expecting it. Gibbs ended by saying that he knew right where to catch "that motherfucker .. . with his pants down."

Because Brown challenges his conviction, we first must decide whether the government introduced enough evidence against Brown such that a rational trier of fact could have found the essential elements of the offense beyond a reasonable doubt. The heart of the government's case against Brown lay in the inference that Brown contributed to the conspiracy as an enforcer; the pieces of evidence supporting this inference were the tape recorded conversations and Agent Coleman's interpretations thereof, in addition to Collier's testimony.

On appeal, Brown challenges the sufficiency of the government's evidence by arguing that the District Court abused its discretion in permitting Agent Coleman to testify to Brown's knowledge and intent in violation of Fed. R. Evid. 704(b). Under Rule 704(b), no expert witness "testifying with respect to the mental state or condition of a defendant in a criminal case may state an opinion or inference as to whether the defendant did or did not have the mental state or condition constituting an element of the crime charged . . . ." Fed. R. Evid. 704(b). Specifically, Brown alleges that Agent Coleman was expressing his own opinion that Brown intended to hurt Sydnor or Brickel. He also implicitly argues that the testimony of Agent Coleman, who was qualified as an expert, violated the dictates of Rule 702 when he defined words that did not require specialized knowledge to understand, since that testimony would not assist the jury.

In support of his Rule 704(b) argument, Brown points to two statements by Agent Coleman. First, Coleman testified that (in his opinion) when Gibbs told Brown, "Tonight will be the night," Gibbs meant that Brown and Fluellen could complete the mission against Sydnor or Brickel that night. Agent Coleman also testified that when Brown stated, "I was on that last night, plus we're going to be on that tonight," he believed that Brown was referring to "the mission they had talked about earlier to locate or find some individual and hurt them."

The government responds that Agent Coleman did not testify to Brown's state of mind or intent. Rather, the government contends that, in the passages cited by Brown, Coleman limited his testimony to interpreting the cryptic language, never opined on Brown's intent, and never stated that Brown was an "enforcer." Indeed, the government points to two exchanges in the record where Agent Coleman explicitly disclaimed knowledge of what Brown's intentions were with regard to the "mission." The government also notes that the District Court instructed the jury that they were not bound to believe or follow Agent Coleman's expert opinion. This was a proper instruction.15

As a first step in deciding whether the evidence against Brown was sufficient, we must decide whether Agent Coleman improperly opined on Brown's intent or knowledge in violation of Rule 704(b), or whether his testimony violated Rule 702, and thus whether part of Coleman's testimony should have been excluded. The admission of expert testimony should be reversed only for an abuse of discretion. See United States v. Bennett, 161 F.3d 171, 182 (3d Cir. 1998). The trial Judge has broad discretion to admit or exclude expert testimony, based upon whether it is helpful to the trier of fact. See id. (citing Fed. R. Evid. 702); 4 Jack B. Weinstein & Margaret A. Berger, Weinstein's Federal Evidence § 702.02[2] (Joseph M. McLaughlin ed., 2d ed. 1997).

As a preliminary matter, we note that it is well established that experienced government agents may testify to the meaning of coded drug language under Fed. R. Evid. 702. See, e.g., Theodoropoulos, 866 F.2d at 590-91; see also United States v. Plunk, 153 F.3d 1011, 1017 (9th Cir. 1998) (noting that the jargon of the narcotics trade and drug dealers' code language are proper subjects of expert opinion), cert. denied, 119 S. Ct. 1376 (1999); United States v. Delpit, 94 F.3d 1134, 1145 (8th Cir. 1996) (same); United States v. Boissoneault, 926 F.2d 230, 232 (2d Cir. 1991) (same). Because the primary purpose of coded drug language is to conceal the meaning of the conversation from outsiders through deliberate obscurity, drug traffickers' jargon is a specialized body of knowledge and thus an appropriate subject for expert testimony. See United States v. Griffith, 118 F.3d 318, 321 (5th Cir. 1997); Theodoropoulos, 866 F.2d at 591.

Such testimony is relatively uncontroversial when it permits a government agent to explain the actual meanings of coded words--that is, when the agent acts as a translator of sorts. See, e.g., Griffith, 118 F.3d at 322 (agent testified that "days of work" meant pounds of marijuana); Delpit, 94 F.3d at 1144 (agent testified that "straps" meant guns); United States v. Simmons, 923 F.2d 934, 946 (2d Cir. 1991) ("boyfriend" or "boy" meant heroin). This precept applies here to much of what Agent Coleman testified about. It was within the scope of Agent Coleman's expertise to explain, for example, in specific contexts, that "jawn" meant "cocaine," that to "hit" someone meant to page them on a beeper, that "on post" meant "ready and waiting," and that a "quarter" meant $2,500.

It is a different matter when an agent testifies that, in light of the meanings he has attributed to certain conversations, a defendant has played a certain role in, or has certain knowledge about, a conspiracy or other offense. See, e.g., Boissoneault, 926 F.2d at 233 (expressing discomfort with expert testimony that draws Conclusions as to the significance of conduct or evidence). But see United States v. Foster, 939 F.2d 445, 452 (7th Cir. 1991) (holding that although certain behavior may have an innocent explanation, it is a fair use of expert testimony to offer another explanation for that behavior); United States v. DeSoto, 885 F.2d 354, 360-61 (7th Cir. 1989) (everyday appearance of an activity is not an automatic bar to admission of expert testimony that may attribute a more sinister motive to the actions, though admission does require special vigilance to avoid unfair prejudice). Brown contends that Agent Coleman testified about Brown's intent to harm the target of the mission in violation of Rule 704(b).

However, Agent Coleman never testified to what Brown's intent was with regard to the mission. Indeed, he specifically refused to do so. Where an expert in a criminal case has not explicitly testified about a defendant's intent, courts have been reluctant to exclude the expert's testimony under Rule 704(b). See United States v. Lipscomb, 14 F.3d 1236, 1242-43 (7th Cir. 1994) (using two-part test that examines whether the actual word "intent" was used and looking to the source of the expert's opinion to determine admissibility under 704(b)); United States v. Smart, 98 F.3d 1379, 1388 (D.C. Cir. 1996) (adopting Seventh Circuit test). Similarly, in Plunk, the court noted that the defendant, who alleged that the expert's testimony violated Rule 704(b), "pointed to nothing in [the expert's] testimony that comprises an explicit opinion that Plunk intended or knew anything in conjunction with the crimes charged. Likewise, nothing in the testimony necessarily compels such an inference or Conclusion." 153 F.3d at 1018; see also Simmons, 923 F.2d at 947 ("[The agent's] testimony, which related only to the meaning of unfamiliar narcotics jargon, left to the jury the task of determining whether the decoded terms demonstrated the necessary criminal intent.").

The two sentences of Agent Coleman's testimony that allegedly offer an opinion on Brown's intent to further the conspiracy by protecting Gibbs--as the ringleader of the conspiracy--against a threat of harm do not in fact offer such an opinion. At no point did he articulate either that Brown intended to kill Sydnor or Brickel, or that Brown intended to further the conspiracy by acting as its enforcer. Coleman specifically stated, "I don't know what his intentions were," and he agreed that he had no idea "whether [Brown was] going on this mission or not." The District Court therefore did not abuse its discretion in permitting Agent Coleman to testify in this regard.

We read Brown's objection to Agent Coleman's testimony to include an objection that the District Court permitted Coleman to interpret several segments of conversation that did not require expertise to interpret, that his interpretation would thus not assist the jury, and that this evidence was prejudicial. Coleman testified that "tonight is the night" was a reference to the fact that Brown and Fluellen were going to go on their mission that night. He also testified that "got to do it tonight" meant "got to go on the mission tonight." Unlike a word like "jawn," which would not be familiar to most jurors and which is the proper subject of expert testimony, a phrase like "tonight is the night" contains no intrinsic code that a jury would be unable to understand. Testimony about such a phrase's meaning is therefore not helpful to the jury.

We have upheld the exclusion of expert testimony when that testimony ventures into areas in which the jury needs no aid or illumination. See, e.g., United States v. Dicker, 853 F.2d 1103, 1108-09 (3d Cir. 1988) ("Although courts have construed the helpfulness requirement of Fed. R. Evid. 701 and 702 to allow the interpretation by a witness of coded or `code-like' conversations, they have held that the interpretation of clear conversations is not helpful to the jury, and thus is not admissible under either rule."); see also Fed. R. Evid. 702 advisory committee notes (stating that whether the situation is a proper one for expert testimony "is to be determined on the basis of assisting the trier"); United States v. Stevens, 935 F.2d 1380, 1384 (3d Cir. 1991) ("[W]e agree with the district court's exclusion of Stevens's expert testimony on two of the three disputed points in that such testimony would not have been `helpful' --the touchstone of Fed. R. Evid. 702--to the jury."); Scott v. Sears, Roebuck & Co., 789 F.2d 1052, 1055 (4th Cir. 1986) (noting that Rule 702 makes inadmissible expert testimony as to a matter that obviously is within the common knowledge of jurors because such testimony, almost by definition, can be of no assistance; but noting that the admission of such testimony, though technical error, will almost invariably be harmless); 2 Stephen A. Saltzburg et al., Federal Rules of Evidence Manual 1218-19 (7th ed. 1998).

We conclude that the District Court abused its discretion in failing to exclude Agent Coleman's testimony explaining what "tonight's the night," "we're going to be on that tonight," and "got to do it tonight" meant. It was the function of the jury, which heard all of the relevant tape recordings, to determine what these phrases meant in the context of the surrounding sentences. Agent Coleman's testimony was not helpful to the jury; in fact, the only purpose of that testimony was to bolster the government's allegations that Brown was an enforcer.

We cannot say, however, that the three sentences of Coleman's testimony that infringed on the jury's role constituted prejudicial error under Fed. R. Evid. 103(a), in light of the other properly admitted evidence against Brown.16

We reach this Conclusion based on the strength of the remaining evidence against Brown. First, Collier testified that a mission meant an attempt to harm or kill someone, and that Gibbs had asked Collier to steal a car to be used in effectuating the mission. See supra Part III.A. Agent Coleman opined that "mission" meant "[a]n attempt to locate, find and hurt whoever Gibbs wants them to locate, find and hurt." This type of interpretation is admissible since it is an interpretation of a code word whose definition a jury would most likely need to have explained. Taken in conjunction with those definitions of "mission," Gibbs's statement that "Earl was . . . on a mission the other night" substantially furthers our Conclusion. Finally, taking the content and timing of conversations between Gibbs and Brown as a whole, a jury reasonably could have concluded that the exchanges (including Brown's statements like "got to do it tonight" and his assent to Gibbs's statement that "[t]onight will be the night") traced a plan to harm someone who was "getting on [Gibbs's] nerves," a plan that involved waiting outside a particular nightclub for the person to emerge unarmed and then injure him.

In light of our Conclusion that the vast bulk of Agent Coleman's testimony is admissible, and in view of the strength of the remaining evidence, we believe that the government offered sufficient evidence for a jury to conclude not only that Brown had acted as an "enforcer" for the conspiracy, but that Brown knew that by his acts he was furthering the aims of the conspiracy by protecting the individual whom he knew to be the ringleader. We therefore will affirm Brown's conviction.

Brown raises a number of reasons why his sentence should be overturned. We disposed of most of these supra, in the margin. We deal more fully with two of them:

"whether the District Court erred in attributing to him all of the drugs that were distributed by the conspiracy in the one-month period during which Brown acted as an enforcer; and whether the District Court properly increased Brown's offense level on the basis of Brown's possession of a firearm."

At sentencing, the District Court attributed twenty-seven kilograms of powder cocaine and nine ounces of crack to Brown. Relying on Collado, 975 F.2d at 995, Brown contends that his involvement with the conspiracy, if any, was limited to interactions with Gibbs, and that the District Court therefore should not have attributed to him amounts of drugs that were part of other transactions of the conspiracy. We review for clear error the District Court's findings of fact regarding the relevant quantities of cocaine and crack attributable to Brown, as well as the types of drugs at issue. See Miele, 989 F.2d at 663; Roman, 121 F.3d at 140.

Brown is correct that, under the Guidelines, a court may not sentence a defendant for the entire amount of drugs in a conspiracy merely because the defendant has been found guilty of the crime of conspiracy. The sentencing court can consider "relevant conduct," including the "conduct of others that was both in furtherance of, and reasonably foreseeable in connection with, the criminal activity jointly undertaken by the defendant." Price, 13 F.3d at 732. However, courts often have held enforcers for drug conspiracies responsible for the amount of drugs that pass through the conspiracy during the time the person acts as an enforcer. For the most part, courts have based their holdings on the ground that the role of enforcer implies a strong familiarity with the goings-on of the entire conspiracy. See, e.g., United States v. Cortinas, 142 F.3d 242, 250 (5th Cir. 1998) (holding enforcer responsible for all of the drugs handled by the conspiracy during the twenty months he was part of the conspiracy); United States v. Phillips, 37 F.3d 1210, 1214 (7th Cir. 1994) (affirming district court's finding that the defendant, who carried guns and enforced for the conspiracy, had an intimate relationship with the operation and could be held responsible for all the drugs handled by the conspiracy).

We think these decisions are correct. In our view, the role of enforcer is often central to the viability of the drug conspiracy, which perforce exists in a dangerous environment, and thus an enforcer who has engaged in more than peripheral enforcement activities ought to be chargeable for any drugs passing through the conspiracy during his "enforcership"--at least for those deals made within the geographic region in which he operated. We recognize that, just as there are different paradigms of conspiracies, there may be different types of enforcers in a conspiracy. Some enforcers may be close to the center of the conspiracy, while others may work on the periphery. One enforcer may enforce during the entire life of the conspiracy; another may perform one task for the conspiracy and then go his own way. One must therefore be careful to consider the scope of the putative enforcer's role. In this conspiracy, however, during the limited time in question, there was only one leader: Gibbs. As we discuss below, the government proved that Brown served directly as an enforcer for Gibbs, who was widely known to be a major drug dealer, and that Brown's job was to protect Gibbs by hurting or killing someone who Gibbs perceived as a threat, thus facilitating Gibbs's ability to deal drugs. We are therefore comfortable in denominating Brown an enforcer for the conspiracy throughout the time in which Brown worked for Gibbs.

We also acknowledge the danger of attributing to an enforcer the entire quantity of drugs that pass through a conspiracy when that enforcer has acted for only a short period time relative to the conspiracy's total length. Here, Brown was involved in the conspiracy for about a month as the enforcer for the head of the conspiracy, and the amount attributed to him at sentencing reflected only the amount of drugs that passed through the conspiracy during that month. We have thus avoided the referenced danger.

In sum, in light of the number of conversations between Brown and Gibbs and the District Court's specific finding that Brown was an enforcer for "a major drug dealer," we do not think it was clearly erroneous for the District Court to conclude that Brown was sufficiently involved with the conspiracy, as a protector of the leader of a large trafficking organization, so as to be charged with the twenty-seven kilograms of powder cocaine and nine ounces of crack that passed through the conspiracy during the time he served as an enforcer.17

In July 1996, while the investigation of the conspiracy was ongoing, an informant told the government that Brown had participated in an armed robbery during which a security guard was shot. The government searched Brown's apartment, discovered a shotgun, and prosecuted Brown under 18 U.S.C. § 922(g), which criminalizes possession of a weapon by a convicted felon. The jury convicted him of this offense. At sentencing in the instant case, the government introduced as evidence the sawed-off shotgun, which was at the heart of his earlier conviction, on the ground that Brown had possessed this shotgun during the relevant time periods for which he was an enforcer for the instant conspiracy.

The District Court took Brown's possession of the shotgun into account in adding two points to his offense level, in accordance with U.S.S.G. § 2D1.1(b)(1), which directs a district court to increase a defendant's base offense level by two points when a defendant possesses a firearm during a drug offense. Brown alleges that this two-level increase constituted double-counting and violated double jeopardy, since he had already been punished for possessing the shotgun. We review this argument, which he did not raise below, for plain error. See United States v. Coates, 178 F.3d 681, 683 (3d Cir. 1999).

The Supreme Court has made clear that the use of relevant conduct to increase the punishment of a charged offense does not punish the offender for that relevant conduct. See Witte v. United States, 515 U.S. 389, 395 (1995). That is, a court does not violate a defendant's protections against double jeopardy when it convicts a defendant for crime X, enhances his sentence for crime X because of conduct Y, and convicts him for conduct Y as well. In Witte, the defendant pleaded guilty to possessing marijuana with intent to distribute. In calculating Witte's offense level under the guidelines, the district court considered not only the amount of marijuana involved in the charged offense, but also an additional 1000 pounds of marijuana and 1091 kilograms of cocaine involved in uncharged criminal conduct. See id. at 393-94. Witte was later indicted for conspiring and attempting to import the 1091 kilograms of cocaine considered in the earlier sentencing. The Supreme Court rejected Witte's contention that his subsequent prosecution on the cocaine offense would subject him to double jeopardybecause the district court had considered that conduct in sentencing him on the marijuana charge.

The Court explained, "[W]e specifically have rejected the claim that double jeopardy principles bar a later prosecution or punishment for criminal activity where that activity has been considered at sentencing for a separate crime." Id. at 398 (citing Williams v. Oklahoma, 358 U.S. 576 (1959)).18 The Court also explained that "it makes no difference in this context whether the enhancement occurred in the first or second sentencing proceeding." Id. at 399. In Brown's case, the two-point sentence enhancement occurred in the second proceeding--his conspiracy conviction--based on his possession of a firearm during a drug offense, the possession for which he earlier had been convicted. Under Supreme Court precedent, this does not create a double jeopardy problem. See also United States v. Street, 66 F.3d 969, 980 (8th Cir. 1995) (holding that use of defendant's earlier state conviction to enhance federal sentence for assaulting park rangers did not violate double jeopardy clause). Because Brown's possession of a weapon was the basis for an earlier conviction but a mandatory ground for enhancement in a separate offense with different requirements, the District Court did not violate Brown's double jeopardy rights and did not double count in reaching Brown's final offense level. We therefore will affirm Brown's sentence in its entirety.

As discussed supra, Terrence Gibbs, as one of the ringleaders of the alleged cocaine conspiracy, was indicted on a number of counts. In Count 1, he was charged with conspiring to distribute cocaine and to possess cocaine with the intent to distribute. The indictment alleged that Gibbs supervised a number of associates who distributed powder and crack cocaine; that he obtained kilograms of cocaine from supplier Juan Arana; and that he used violence to protect himself in his position as leader of the conspiracy. The jury convicted Gibbs of Count 1 and the District Court sentenced him to life imprisonment on this count.19

The government established its case against Gibbs in much the same way that it did against Sydnor and Brown: through intercepted phone conversations and the interpretation thereof by Agent Coleman. It also adduced testimony from a cooperating witness named Charles Wilkes. On appeal, Gibbs objects both to his conviction and to his sentence. With regard to his conviction, he first argues that Agent Coleman's testimony went beyond mere interpretation of the coded conversations into the realm of speculation and that the District Court therefore plainly erred in allowing Agent Coleman to so testify. We find no merit in this contention, which was not raised in the District Court and which we therefore consider under a plain error standard of review, see Coates, 178 F.3d at 683, and we reject it summarily in the margin.20

Gibbs's principal contention is that the District Court erred in allowing the government to submit evidence of violent acts by members of the conspiracy and that, even if such evidence were properly admitted, the Court erred by giving the limiting instruction it did with regard to the evidence of violence. This Court has not reached this question before, and though we do not think it is a difficult question, we address it herein to offer future guidance to district courts and practitioners. As to sentencing, Gibbs argues that the District Court erred in finding that he was responsible for in excess of 150 kilograms of powder cocaine and 1.5 kilograms of crack.

The government introduced evidence at trial about the 718use of violence in the conspiracy. For example, it introduced testimony that Gibbs, who suspected that Maurice Grannum had attempted to kidnap him, sent Earl Packer Hunte and Derrick Parks to shoot Grannum, and that once Gibbs suspected that Sydnor had shot him, he sent Earl Brown and Frank Fluellen on a mission to kill Sydnor. Two police officers also testified that the attempt to shoot Grannum did occur on the date and time Gibbs and Parks had agreed upon. (The bullets passed into the police officers' apartment.) Although the defense did not object to the admission of this evidence at trial, Gibbs now complains that this evidence should have been excluded because (i) the admission of the evidence served no other purpose than to portray him as ruthless and violent; (ii) the government did not offer the evidence under any one theory of admissibility, as required by Rule 404(b); and (iii) the government failed to show that the probative value of the evidence outweighed its prejudicial effect.

We review a district court's decision to admit evidence for abuse of discretion. See Government of V.I. v. Edwards, 903 F.2d 267, 270 (3d Cir. 1990). Because the defendant did not request a limiting instruction with regard to the evidence of violence at trial, we review the District Court's jury instructions for plain error. See Price, 13 F.3d at 724; see also Henderson v. Kibbe, 431 U.S. 145, 154 (1977) (noting that "[i]t is the rare case in which an improper instruction will justify reversal of a criminal conviction when no objection has been made in the trial court.").

Rule 404(b), which proscribes the admission of evidence of other crimes when offered to prove bad character, does not apply to evidence of uncharged offenses committed by a defendant when those acts are intrinsic to the proof of the charged offense. As a prominent commentator has explained:

"In cases where the incident offered is a part of the conspiracy alleged in the indictment, the evidence is admissible under Rule 404(b) because it is not an 'other' crime. The evidence is offered as direct evidence of the fact in issue, not as circumstantial evidence requiring an inference as to the character of the accused. Such proof . . . may be extremely prejudicial to the defendant but the court would have no discretion to exclude it because it is proof of the ultimate issue in the case."

22 Charles A. Wright & Kenneth W. Graham, Federal Practice and Procedure § 5239, at 450-51 (1978).

A number of courts likewise have held that Rule 404(b) does not limit the admission of evidence of the defendant's participation in acts of violence as direct proof of a conspiracy. In United States v. Miller, 116 F.3d 641 (2d Cir. 1997), cert. denied, 118 S. Ct. 2063 (1998), for example, the court held that proof of murders by the defendants did not fall under 404(b) even though the murders were not charged in the indictment; rather, the court held that the murders were relevant to show the existence and nature of the conspiracy. See id. at 682; see also United States v. Chin, 83 F.3d 83, 87-88 (4th Cir. 1996) (holding that testimony challenged under 404(b) regarding contract killing scheme emphasized the violent and dangerous context of a heroin deal, and was inextricably intertwined with defendant's crime of selling heroin and conducting an ongoing criminal enterprise). We endorse this reading of Rule 404(b). Since the government introduced evidence of Gibbs's use of violence to further the illegal objectives of the cocaine conspiracy by removing threats to himself (since threats to Gibbs meant threats to the trafficking enterprise), the District Court did not abuse its discretion in permitting this evidence to come in.21

The government presented two witnesses at Gibbs's sentencing hearing: Agent Coleman and Charles Wilkes, a self-proclaimed contract killer facing the death penalty who struck a deal with the government. Gibbs contends that the District Court erred in reaching its totals of 150 kilograms of powder cocaine and 1.5 kilograms of crack based on this testimony.22 We review the District Court's determination of the drug quantities attributable to Gibbs for clear error. See Miele, 989 F.2d at 663.

Agent Coleman testified that the conspiracy received 111 kilograms of powder cocaine in the four-month period in which the government wiretapped Gibbs's phone. From that number, the government asserted that the conspiracy must have handled at least 150 kilograms over the life of the conspiracy, which ran from the summer of 1992 through April 1995 (approximately thirty-one months). The government notes that Agent Coleman's estimate of 111 kilograms was a conservative one: the electronic surveillance did not cover every day between November 1994 and April 1995; Gibbs used other phones that the government did not wiretap; and if the conversation involved a multi-kilogram transaction but Agent Coleman could not determine the specific number of kilograms under negotiation, he attributed one kilogram to the conversation.

Agent Coleman described his attribution method as follows:

"[W]hen I reviewed . . . the tape recordings of the conversations, after conducting the wire, I was able to determine a pattern of speech where kilograms . . . of cocaine were discussed. And most of these conversations were between Terrence Gibbs and Juan Arana or Juan Arana and Domingo Arana or others. And I would take my analysis of either the money that was discussed referring to kilograms of cocaine or the kilograms of cocaine and determine how many were discussed in each conversation. Some conversations . . . discussed, what I believe to be crack cocaine."

He compiled a list of fifty-two phone conversations and the amounts he believed were discussed in each conversation.

As we discussed supra, it is appropriate for a district court to base its quantity calculations on intercepted conversations. The government can base its calculations on the amount of drugs under negotiation, see United States v. Raven, 39 F.3d 428, 432 (3d Cir. 1994), and it need not prove that the amounts under negotiation were actually distributed, see United States v. Layeni, 90 F.3d 514, 522 (D.C. Cir. 1996). A district court may carefully estimate the total drug quantities involved in a conspiracy based on evidence of average drug transactions during the conspiracy. See U.S.S.G. § 2D1.1 application note 12. In addition, a leader of a drug conspiracy is responsible for drug quantities transacted by his subordinates in furtherance of the conspiracy. See, e.g., United States v. Russell, 134 F.3d 171, 184 (3d Cir. 1998) (attributing to defendant, who was the conspiracy's organizer, the total quantity of drugs for which defendant's co-conspirators had taken responsibility); United States v. Magana , 118 F.3d 1173, 1206 (7th Cir. 1997) (holding that defendant is liable for entire quantity of drugs attributable to conspiracy in circumstances where defendant is one of conspiracy's central figures).23

While the conversations and amounts calculated there from are too lengthy to recount here, the District Court did not clearly err in finding that the government had proved the quantities of powder attributable to Gibbs by a preponderance. In fact, it appears that Agent Coleman was conservative in Gibbs's favor in his calculations.

The government must also prove by a preponderance that the substance involved is crack. See United States v. Holman, 168 F.3d 655, 658 (3d Cir. 1999) (noting that "[i]t is the serious duty of the district court to hold the government to" the preponderance standard, "particularly because of the impact the identity determination has on sentencing"). Agent Coleman testified that twelve kilograms of crack were sold during the life of the conspiracy. First, agents seized baking soda and cooking equipment from the apartment in which Gibbs allegedly cooked powder into crack. Second, Agent Coleman interpreted Discussions involving cocaine that was "hard" or "done" as Discussions about cocaine that had been cooked into crack. Third, Collier, Ellis, Parks, and Wilkes all testified that they personally had seen crack processed and distributed while they were members of the conspiracy. In fact, Ellis testified that he helped Gibbs process kilograms of crack five or six times. Coleman testified that the process these witnesses had seen would indeed produce crack. Finally, to avoid double counting when calculating Gibbs's sentence, the District Court subtracted these twelve kilograms of crack from the 111 kilograms of powder, since the crack had been processed from the powder.

We conclude that the government proved by a preponderance that Gibbs was responsible for producing at least 1.5 kilograms of crack. In fact, it arguably proved that Gibbs had produced much more than that, in light of the taped conversations and Ellis's testimony that he had been present when Gibbs cooked five or six kilograms of cocaine into crack. The government does not need to perform chemical analysis on the seized substances in order to prove that a substance was crack, see United States v. Dent, 149 F.3d 180, 190 (3d Cir. 1998), cert. denied, 119 S. Ct. 833 (1999); Roman, 121 F.3d at 141, and other courts have held that the government may rely on the testimony of co-conspirators who distributed crack or who observed its manufacture to establish that the substance at issue was crack, see United States v. Hargrett, 156 F.3d 447, 451 (2d Cir.) (holding that testimony of co-conspirator that he cooked cocaine into crack for defendant was reasonable basis on which to charge defendant with that amount of crack), cert. denied, 119 S. Ct. 607 (1998); United States v. Cantley, 130 F.3d 1371, 1378-79 (10th Cir. 1997) (multiple police officers and lay witnesses who purchased substance from, or sold substance to, defendant testified that substance was crack), cert. denied, 118 S. Ct. 1098 (1998); United States v. Taylor, 116 F.3d 269, 273-74 (7th Cir. 1997) (drug supplier, purchasers, and assistants testified that substance was crack).

In sum, based on the amount of drugs discussed in intercepted conversations and on the reasonable estimate that if the conspiracy handled 111 kilograms of cocaine in four months, it handled 150 kilograms over thirty-one months, we conclude that the District Court did not clearly err in attributing 150 kilograms of powder cocaine and 1.5 kilograms of crack to Gibbs as a leader of the conspiracy.

The judgments of the District Court will be affirmed.