Opinion for the Court filed by Circuit Judge MIKVA.
Dissenting opinion filed by Circuit Judge SCALIA.
MIKVA, Circuit Judge:
This is an appeal from a jury verdict that awarded Geraldine V. Carter $10,000 as compensatory damages for discriminatory activities proscribed by section 1981 of the Civil Rights Act of 1870, 42 U.S.C. § 1981. Carter had alleged that throughout her employment, her employer, Duncan-Huggins, Ltd. had intentionally discriminated against her because of her race. Following the jury verdict for Carter, appellant Duncan-Huggins timely moved for a judgment non ob-stante veredicto (judgment n.o.v.). The trial court denied the motion and this appeal followed. Appellant here challenges the denial of the judgment n.o.v. motion and the trial court’s failure to give certain jury instructions. Finding no merit in appellant’s arguments, we affirm.
I. Standard of Review
In reviewing a motion for a judgment n.o.v., we ask the same question that the district court asked initially in considering the motion. Because a motion for judgment n.o.v. intrudes upon the jury’s domain, that question is very narrow. The jury’s verdict must stand unless “the evidence, together with all inferences that can reasonably be drawn therefrom is so one-sided that reasonable men could not disagree on the verdict.” Coburn v. Pan American World Airways, Inc., 711 F.2d 339, 342 (D.C.Cir.1983) (quoting Murphy v. United States, 653 F.2d 637, 640 (D.C.Cir.1981)), cert. denied, -- U.S. --, 104 S.Ct. 488, 78 L.Ed.2d 683 (1983). See also Vander Zee v. Karabatsos, 589 F.2d 723, 726 (D.C.Cir.1978), cert. denied, 441 U.S. 962, 99 S.Ct. 2407, 60 L.Ed.2d 1066 (1979). The motion should be denied unless
“there can be but one reasonable conclusion” drawn from the evidence viewed “in the light most favorable to the [plaintiffs] ..., giving them the advantage of every fair and reasonable inference that the evidence may justify”.
Metrocare v. Washington Metropolitan Area Transit Authority, 679 F.2d 922, 924-25 (D.C.Cir.1982) (quoting Foster v. Maryland State Savings and Loan Association, 590 F.2d 928, 930 (D.C.Cir.1978), cert. denied, 439 U.S. 1071, 99 S.Ct. 842, 59 L.Ed.2d 37 (1979)). If fair-minded people may differ as to the conclusion, or if there is substantial conflicting evidence, the judgment n.o.v. motion must be denied. Necessarily, the reviewing court must consider all the evidence offered. In evaluating the evidence, however, the court should not decide the motion based on which side has the “better of the case.” Bohrer v. Hanes Corp., 715 F.2d 213, 218 (5th Cir.1983) (quoting Maxey v. Freightliner Corp., 665 F.2d 1367, 1371 (5th Cir.1982)), petition for cert. [1228]*1228filed, 52 U.S.L.W. 3512 (U.S. Jan. 10, 1984) (No. 83-1017). Nor should the court assess witness credibility or weigh the evidence. Coburn v. Pan American World Airways, Inc., 711 F.2d at 342. Those are functions reserved for the jury.
II. Background
We rehearse the testimony in detail so that we can evaluate what inferences a reasonable juror may have drawn. Appellee Geraldine V. Carter, a black woman, secured employment with Duncan-Huggins, Ltd., in January 1979. Duncan-Huggins operates a wholesale showroom of fabric and furniture for architects and designers. At all times relevant to this case, Duncan-Huggins hired less than fifteen people and thus did not fall within the ambit of Title VII of the Civil Rights Act of 1964, 42 U.S.C. §§ 2000e to 2000e-17 (1976). Instead, this case falls within the purview of section 1981, which imposes as an additional prerequisite to a finding of liability that the plaintiff prove intentional discrimination. General Building Contractors Association v. Pennsylvania, 458 U.S. 375, 102 S.Ct. 3141, 73 L.Ed.2d 835 (1982).
Carter was the first, and only, black employee. When Carter applied for the job, Perez, Duncan-Huggins’ president, informed her that her primary duties would be as sample librarian, with some opportunities to make sales. Transcript (Tr.) at 81. As sample librarian, Carter would retrieve carpet samples that customers wanted to borrow and reshelve those that had been returned. Carter remained at Duncan-Huggins until June 1980, when she voluntarily left to begin a job elsewhere.
During Carter’s initial period of employment, she shared desks with the other, white employees in the Company’s showroom. Eventually, however, Carter was given a desk behind the tall shelves in which the carpet samples were kept. Because of these shelves, the desk was isolated from the showroom and from customer contact. In contrast, all the white employees used desks located in the showroom. Duncan-Huggins suggested at trial that the placement of the desk reflected Carter’s duties as sample librarian. Id. at 37. Moreover, the employee responsible for this work arrangement testified that Carter had asked for the desk. Id. at 262. Carter denied having made this request and instead testified that she had been instructed to work in the back. Id. at 123-24. No explanation was proffered, however, as to why Carter’s desk could not have been placed closer to the shelves, but not in such an extremely isolated position.
In June, 1979 Duncan-Huggins moved to a new facility. That facility had a sample library that was separated from the showroom by a wall with two doorways. For a short period of time following the move, Carter allegedly occupied a desk in the showroom. Id. at 128, 265-66. Thereafter, one of Carter’s de facto supervisors carried her desk into the separate sample library. This library, according to Carter, had inadequate lighting and ventilation. Not surprisingly, the parties all suggested different reasons for this relocation. Carter alleged that the move was motivated by a discriminatory intent. The employee who actually moved the desk testified that he did so at the direction of one of Duncan-Huggins’ owners. Id. at 266. That owner, who admitted having responsibility for the placement of furniture, and who indicated that because sample work was messy the sample room was to be secluded from the showroom, never specifically explained whether, and if so why, he ordered the desk to be moved. Id. at 289-93. Moreover, the owner denied saying anything about where Carter was to be stationed. Id. at 292. Perez, Duncan-Huggins’ president, suggested that the desk was relocated because construction work in the showroom had not been completed. Id. at 234-35. Yet, when the construction work was completed, Carter’s desk was not returned to the customer area. The desk was not returned to the showroom, according to Perez, because its placement in the back gave Carter better access to the samples. Id. at 235. As evidence that the placement did not reflect Carter’s race, Duncan-Huggins pointed to the fact that the current librarian, a white woman, [1229]*1229spends all her time at the same remote desk. This does not alter the fact, however, that Carter, Duncan-Huggins’ only black employee, was separated from both other employees and the customers for the overwhelming majority of her work experience.
Carter’s isolation from the customers extended beyond the opportunity to make personal contact. The same owner who allegedly instructed that Carter’s desk be moved, admitted that he directed Carter to not answer the phones. Id. at 293. To justify this additional isolation, the owner indicated that answering phones was not within Carter’s responsibilities, but instead was to be the responsibility of others. Id. He offered no explanation of why the responsibility was so allocated. Indeed, if sales personnel in the showroom were allowed to answer the phone, logic might suggest that Carter, who was also hired to have some sales responsibilities, also should have been expected to answer the phone.
An additional facet of Carter’s unique treatment was her level of compensation.
This disparity in compensation was aggravated by Carter’s consistently lower incentive payments. Duncan-Huggins had a program in which the president awarded incentive payments to employees twice annually. For the same three periods, two white employees received total incentive payments of $4,926 and $4,200 compared to Carter’s $1,550. Id. at 22-32, 334; Appellant’s Brief at 10; Appellee’s Brief at 7. Indeed, Carter once received the same incentive bonus as a white employee who had left the Company’s employment substantially earlier in the pay period. In response to this showing of disparate levels of compensation, Duncan-Huggins stated that the discrepancies reflected the employees’ seniori[1230]*1230ty and sales quantity. Tr. at 96. The Company’s contemporaneous announcement of the incentive plan, however, specified neither seniority nor sales as determinative of bonus payments. Id. at 108-09; Plaintiff’s Deposition Exhibit H (also introduced at trial). Moreover, the jury might have seen some inconsistency between these putative factors and Perez’s alleged promise to Carter of significant incentive payments. Id. at 161-62. That is, if the amount of sales were the measuring gauge, Carter, who at best would have been given limited sales opportunities, could never have qualified for substantial incentive payments.
The jury also heard evidence clearly establishing that Carter suffered unequal treatment in her day-to-day existence at Duncan-Huggins. Testimony at trial focused on four subjects: access to staff meetings; access to a parking space; access to the facility itself; and a racially derogatory anecdote.
Carter testified that although white employees were allowed to attend staff meetings, her presence was prohibited. Although other employees could participate, Carter testified that her de facto supervisor instructed her to not attend staff meetings. Id. at 163-64. That supervisor did not challenge the accuracy of Carter’s testimony. Id. at 259-79. Because he did not address the issue, he proffered no explanation for allegedly prohibiting Carter’s attendance.
An additional uncontroverted fact is that while Perez and two white employees were given parking spaces, Carter continuously was denied such a privilege. When Duncan-Huggins was located at its old showroom, parking was scarce. Carter, who drove to work, asked Perez for a space. Perez, according to Carter’s testimony, promised her a space at the new building. Id. at 158. Perez denied having made such a promise. Id. at 97. In any event, when Duncan-Huggins moved to its new showroom the only spaces allocated were given to three white workers. Perez explained that the decision was based on a scarcity of space and on seniority. When one of these employees left, however, Carter apparently was still not assigned a space. The parties disagreed about the limitations imposed by the parking lot’s size.
In addition to being denied a parking space, Carter also was denied a key to the new showroom. When the move was first made, the same three white employees who had been assigned a parking space received keys to the showroom. After the move, but before Carter was so entrusted, a white employee with less seniority than Carter also received a key. According to Carter, the employee taunted her about the key. The employee denied the allegation. Id. at 218-19. Carter also testified that the lack of a key created a hardship because she arrived at work early and had to' wait outside until a white employee arrived with a key. Testimony from other Duncan-Huggins staff suggested Carter was never the first to arrive, and thus she had no need for a key.
The jury also heard undisputed evidence about a racially derogatory anecdote told in Carter’s presence. The offensive joke was told by the same white de facto supervisor who later moved Carter’s desk into the back room. Both appellee and appellant agree that the joke was told and that the anecdo-tist, when confronted by Carter, apologized. The evidence was silent as to whether the narrator was reprimanded for this episode. The parties vigorously contest, however, the presence of certain employees during the joke’s narration. Carter testified that a fellow employee (white) and Duncan-Huggins’ president Perez were present. Id. at 166-68. She further testified that at the punch line, they all looked at her and laughed. In rebuttal, the employee allegedly in the audience testified that she did not remember if she were present. Id. at 75. Perez, according to her own testimony, was absent. Id. at 231. The narrator testified, somewhat contradictorily, that although he could not remember who was present, he did remember that Perez was absent. Id. at 279.
The final principal element of Carter’s argument was the difference between the job responsibilities promised and those actu[1231]*1231ally assigned. When Carter accepted her job, she was promised some sales responsibilities. Indeed, she had such opportunities in the early months of her employment. Id. at 272. After the move, however, her sales responsibilities became non-existent. All her time was spent in the sample room. In contrast, full-time white employees who subsequently were hired, and who thus were junior to Carter, were given greater responsibilities for sales. To explain the difference between the promised and actual job, the Company suggested that Carter’s inability to make sales reflected the all-consuming responsibilities of being a sample librarian. The Company pointed to the fact that the white librarian currently employed at Duncan-Huggins has no sales responsibilities. The relevance of this observation, however, may be minimal because the evidence does not clearly indicate whether she was promised anything more than library work.
At the end of the trial, the jury returned a verdict for Carter and awarded her $10,-000 in compensatory damages. Duncan-Huggins made a timely motion for a judgment n.o.v. The district court denied the motion and further concluded that defendant’s challenge to certain jury instructions was without merit.
III. Analysis
This case arises under section 1981 of the Civil Rights Act of 1870. That section provides:
All persons within the jurisdiction of the United States shall have the same right in every State and Territory to make and enforce contracts, to sue, be parties, give evidence, and to the full and equal benefit of all laws and proceedings for the security of persons and property as is enjoyed by white citizens, and shall be subject to like punishment, pains, penalties, taxes, licenses, and exactions of every kind, and to no other.
42 U.S.C. § 1981 (1976). As interpreted in General Building Contractors Association v. Pennsylvania, 458 U.S. 375, 102 S.Ct. 3141, 73 L.Ed.2d 835 (1982), section 1981 prohibits only intentional discrimination. The discriminatory animus, however, can be demonstrated through either direct or indirect evidence. In an analogous disparate treatment case under Title VII, the Supreme Court stated: “As in any lawsuit, the plaintiff may prove his case by direct or circumstantial evidence. The trier of fact should consider all the evidence, giving it whatever weight and credence it deserves. [T]he district court should not have required [the plaintiff] to submit direct evidence of discriminatory intent.” United States Postal Service Board of Governors v. Aikens, -- U.S. --, 103 S.Ct. 1478, 1481 n. 3, 75 L.Ed.2d 403 (1983). And, in Metrocare v. Washington Metropolitan Area Transit Authority, 679 F.2d at 929 (D.C.Cir.1982), a section 1981 case, we noted that: “The absence of overly racist statements or actions by this quasi-government agency is not sufficient to overturn the jury.” See also Denny v. Hutchinson Sales Corp., 649 F.2d 816 (10th Cir.1981).
Contrary to this precedent, the dissent would seem to restrict severely the use of circumstantial evidence in those section 1981 cases that involve small employers. At a minimum, the dissent would have us hold that discriminatory treatment has no weight as circumstantial evidence of discriminatory motivation. This is the logical meaning of the dissent’s assertion that the circumstantial evidence introduced in this case was “not circumstantial evidence of racial motivation, but only ... of an intent to disfavor Carter.” Implicit therein is the proposition that proof of discriminatory treatment and proof of discriminatory animus are completely unrelated. To say the least, this is a novel theory. The dissent, not content with limiting the use of circumstantial evidence, apparently would have us go so far as to require direct evidence of discriminatory animus. Clearly, that is the underlying thesis in the dissent’s statement that “the only shred of evidence ... that could conceivably suggest any racial animus” was the racially derogatory joke. Dissenting opinion, infra, at 1245-1246. See also Racozy illustration, id. at 1246 (implying that circumstantial evidence irrel[1232]*1232evant). An imposition of a direct evidence requirement is without precedent, and, in this time of sophisticated employers would effectively eviscerate section 1981. Moreover, a requirement of direct evidence would fly in the face of Aikens where, in an analogous situation, the Supreme Court found error in the district court’s requirement of direct evidence. 103 S.Ct. at 1481 n. 3.
To allocate the burden of production in section 1981 cases, courts have borrowed the Title VII guidelines announced in McDonnell-Douglas v. Green, 411 U.S. 792, 93 S.Ct. 1817, 36 L.Ed.2d 668 (1973). See, e.g., Metrocare v. Washington Metropolitan Area Transit Authority, 679 F.2d at 925. Under these standards, the plaintiff has the initial burden of establishing a prima facie case. Once this prima facie case is established, a rebuttable “presumption that the employer discriminated against” the employee arises. Texas Department of Community Affairs v. Burdine, 450 U.S. 248, 254, 101 S.Ct. 1089, 1094, 67 L.Ed.2d 207 (1981). To rebut this presumption, the defendant either can disprove the prima facie case or can articulate some legitimate, nondiscriminatory reason for the action. Plaintiff then has the opportunity to demonstrate that the proffered reasons are pretextual. United States Postal Service Board of Governors v. Aikens, 103 S.Ct. at 1482 n. 15. One method for the plaintiff to attack the defendant’s justifications is to show that the alleged explanation is unworthy of credence. Id. at 1483 (Blackmun, J., concurring).
The McDonnell-Douglas standards, however, were “never intended to be rigid, mechanized or ritualistic. Rather, [they are] merely a sensible, orderly way to evaluate the evidence in light of common experience as it bears on the critical question of discrimination.” Id. at 1482 (quoting Furnco Construction Corp. v. Waters, 438 U.S. 567, 577, 98 S.Ct. 2943, 2949, 57 L.Ed.2d 957 (1978)). In a case requiring discriminatory animus, the three-step procedure is merely one method to adduce evidence relevant to the central question of intentional discrimination. Once the case has been fully tried, the court thus should not narrowly focus on a party’s performance at any one stage. Each separate stage merges into the ultimate question of whether the plaintiff has established sufficiently a case of intentional discrimination. As the United States Court of Appeals for the Eleventh Circuit recently stated in an age discrimination suit:
[W]hen the parties have developed their full proof it is unnecessary to analyze the evidence according to the “ebb and flow of shifting burdens” of proof under McDonnell-Douglas.... At that point, the trier of fact must determine whether plaintiff has met the ultimate burden of showing that the employer intentionally discriminated against him in violation of [the statute].
Krieg v. Paul Revere Life Insurance Co., 718 F.2d 998, 999 (11th Cir.1983) (quoting Smith v. Farah Manufacturing Co., 650 F.2d 64, 68 (5th Cir.1981)).
Because of the posture of this case, our task is straight-forward: we need only ask whether the evidence, in the light most favorable to Carter, is such that a reasonable juror could have concluded that Carter established a case of racial discrimination. In applying this standard of review, we are mindful of Justice Rehnquist’s recent comment that “the question facing triers of fact in discrimination cases is both sensitive and difficult.” United States Postal Service Board of Governors v. Aikens, 103 S.Ct. at 1482 (1982).
The record, taken as a whole and read in the light most favorable to Carter, can support a finding that a discriminatory motivation was present. We are required to evaluate the evidence under the presumption that the jury resolved all factual disputes in favor of the plaintiff. Moreover, we must give the advantage of every possible inference to Carter. Under these standards, we are compelled to agree with the district court’s conclusion that:
The Court cannot say that the evidence was so lacking in support for plaintiff’s allegations as to require that the verdict be annuled or set aside. Questions of [1233]*1233intent are often necessarily decided on the basis of circumstantial evidence, and it would have been possible for a reasonable person to have concluded, from the evidence adduced at trial, that plaintiff was treated differently because she was black.
Carter v. Duncan-Huggins, Ltd., No. 81-546, slip op. at 2 (D.D.C. Dec. 23, 1981). We therefore affirm.
Sufficient evidence was introduced from which the jury could have inferred a discriminatory motive. In the testimony concerning the anecdote, the jury heard evidence of an overtly racist statement. The defendant admitted that the joke was told. Assuming, as we must, that the jury accepted Carter’s recollection as accurate, Duncan-Huggins’ president was present during the narration and found the joke humorous. Moreover, the narrator apparently was never reprimanded. The plaintiff also introduced a significant amount of circumstantial evidence which could have bolstered the jury’s perception of discrimination. The plaintiff demonstrated that she was singled out for unique treatment in all aspects of her employment — from salary, to customer contact, to her work station, to participation in staff meetings, to fringe benefits such as parking and keys. Additionally, despite the fact that it was part of her job description, she was denied the opportunity to make sales. Yet, white employees with less seniority were assigned to the showroom and given sales opportunities. From this evidence, the jury could have concluded that Carter, the Company’s only black employee, consistently suffered conduct and conditions that were worse than those imposed upon her fellow white employees. While each piece of circumstantial evidence alone might have been an insufficient basis from which to infer discriminatory animus, the cumulative impact of the many discrete instances of disparate treatment and broken promises reasonably may have led to an inference of discriminatory intent. Applying common sense and general experience to the hard facts presented, a reasonable juror could have found that the Company treated Carter differently because she was black.
The dissent suggests that because Carter had some unique responsibilities, a comparison between Duncan-Huggins’ treatment of Carter and its treatment of other employees must be completely excluded from our analysis. As an initial matter, the dissent’s assertion that Carter was in a unique position is an overstatement since she was hired to have some sales responsibilities. Thus, a comparison between Carter’s treatment and the treatment accorded other sales personnel is reasonable. Moreover, the analytical framework suggested by the dissent would immunize, to a large extent, small employers from discrimination suits. Such employers, it is worth remembering, are too small to be within the ambit of Title VII. And under section 1981 — if the dissent’s position were adopted — employers could dis-criminatorily set a minority employee’s salary and job responsibilities with impunity, so long as that employee was assigned job responsibilities that were even slightly different than those assigned other employees. Absent a “smoking gun” or other direct evidence of racial animus, a court would be hard-pressed to ever discern discrimination. Because we believe that section 1981 should be more than a hollow shell, we decline to accept the dissent’s suggestion.
To meet plaintiff’s allegations, Duncan-Huggins challenged certain factual representations made by Carter. The dispute between the parties often concerned facts that were related closely to the ultimate question of discrimination. For example, the employer disputed Carter’s allegations about the audience present for the offensive joke. This disagreement was significant since the presence or absence of Duncan-Huggins’ president directly bears on the employer’s involvement in the alleged discrimination. If Perez were present, the inference of discriminatory animus may be stronger. Or, to take another example, the question of whether Carter was instructed to sit in the sample room or did so on her own volition is a factual issue of import. If Carter voluntarily moved into the sample library, the inference of discrimination that [1234]*1234otherwise might arise from the relocation is weakened. In some instances, the disputed fact was secondary but nonetheless could have had an impact on a finding of discrimination. The issue over Carter’s arrival time at work illustrates this category since her arrival time was tied to her need for a key. If in fact she arrived later than those employees with a key, the inference of discriminatory motive is weakened.
The Company’s articulation of a different version of the facts, however, is not a sufficient reason for a court to grant a judgment n.o.v. To hold otherwise would require us to conclude that the Company’s version was more credible than Carter’s version. Clearly, this is beyond the scope of our review. The weighing of conflicting evidence and the evaluation of witness credibility is exclusively within the jury’s province.
In addition to raising factual disputes, Duncan-Huggins offered business reasons for some of its actions. The Company explained, for example, that Carter was in the separate room because she needed easy access to the samples. The Company now appears to argue that its mere articulation of any putatively legitimate business rationale requires the court to take the case from the jury. We reject that argument. See, e.g., Metrocare v. Washington Metropolitan Area Transit Authority, 679 F.2d at 928 (jury could believe employer’s articulated reasons to be overborne). Duncan-Huggins’ alleged justifications for the disparate treatment were presented through the testimony of past or current employees, officers, or owners. For the most part, the only evidence of intent was oral testimony: the explanations were without support of extrinsic evidence. For example, there was no internal memorandum indicating that Carter’s assignment to the back room was to provide easier access to the sample library, or evidencing that seniority and sales were the basis for incentive payments. Contra Coburn v. Pan American World Airways, Inc., 711 F.2d 339 (D.C.Cir.) (employer had a detailed, complex plan to determine who was fired and such plan was not challenged by employee), cert. denied, -- U.S. --, 104 S.Ct. 488, 78 L.Ed.2d 683 (1983). The weight to be given to these offered business justifications thus required an evaluation of witness credibility. And, an evaluation of witness credibility is the exclusive function of the jury. Indeed, where the only evidence of intent is oral testimony, a jury could always choose to discredit the proffered explanation. As we stated in Metrocare v. Washington Metropolitan Area Transit Authority, 679 F.2d at 926-27: “The jury, whose province it was, could disbelieve the ... supervisors’ testimonial explanation of why they acted against the [employee] ... and thus could find their asserted justifications to be pretextual.” Because the strength of the purported business reasons — and indeed the strength of the Company’s case — was thus inextricably interrelated to questions of witness credibility, the case had to go to the jury.
To reach the opposite conclusion, that this case could not go to the jury, the dissent forgets that we here review a jury’s verdict and that we thus are required to draw all inferences favorable to the plaintiff. The dissent seems to reverse this requirement and accordingly draws all inferences least favorable to the plaintiff. One example illustrates the dissent’s modification of our standard of review. The dissent emphasizes that Carter never filed an EEO complaint or otherwise complained to her supervisors. Dissenting opinion, infra, at 1245. From this, the dissent seemingly infers that Carter did not suffer discrimination. Yet, there is nothing in the record to support this inference and much to support a far different inference. In explaining her decision to not complain to Perez directly, Carter identified a thought process that also could have motivated her decision to not file an EEO complaint, or to not complain to any other supervisor.
Why didn’t I go to [Perez]? Because of the treatment that was already displayed. She allowed harassment — she allowed and participated in laughter with the racist joke....
[1235]*1235Tr. 190-91. And later, she further amplified this rationale.
I couldn’t — I didn’t have any power to complain to someone in authority. I had to remain back in the back. I felt and was isolated and, therefore, there was nothing that I felt that I could go to Mrs. Perez to talk about.
Tr. 197. It would not have been unreasonable for the jury to infer that the same feelings of frustration and futility motivated Carter’s decision to not file an EEO complaint or to not complain to her supervisors. This inference of frustration may have been bolstered by the fact that Carter’s previous EEO complaint had been unsuccessful. Clearly then, the dissent’s inference was not the only inference possible, and certainly was not the inference most favorable to the plaintiff.
This case was properly left to the jury. If all factual disputes were resolved in plaintiff’s favor, Carter introduced sufficient evidence from which the jury could have deduced discrimination. Moreover, the jury could have found the employer’s purported justifications to be incredulous. A reasonable juror thus could have found for Carter. Accordingly, we find no error in the district court’s decision to deny the motion for a judgment n.o.v.
IV. Other Challenges
Duncan-Huggins also challenges other aspects of the district court’s disposition of this case. Specifically, the Company contends that the trial court should have instructed the jury that: (1) no statistical proof of discrimination could be drawn from a difference in treatment in such a small work force; and (2) a single joke is not evidence of race discrimination. Duncan-Huggins also argues that the district court should have directed a verdict for nominal damages or ordered a remittitur of the jury’s $10,000 verdict. We find none of these arguments to be meritorious.
Duncan-Huggins argues that the district court erred in refusing to give the following requested instruction:
Defendant’s work force is too small to make any statistical analysis of the treatment of black and white employees useful in deciding whether it purposefully discriminated against plaintiff because she was black.
Appellant’s Brief at 26. In support of this instruction, the Company argued that the smallness of the sample size rendered any “statistical proof” inherently unreliable. The district court rejected this argument and therefore chose not to give the instruction. In its Memorandum and Order denying defendant’s motions, that court explained its rationale, stating in part:
A plaintiff attempting proof of disparate treatment in a small work force is entitled to offer evidence and argument, be it persuasive or not, involving comparison of treatment accorded different workers. The “statistical” inferences to be drawn are for the trier of fact .... The cases defendant has now cited in support of its motion, on the unreliability of statistical evidence, go to other points....
Carter v. Duncan-Huggins, Ltd., No. 81-546, slip op. at 2. We agree with the underlying reasoning, and thus conclude that the district court properly denied the motion.
The fatal aspect of Duncan-Huggins’ argument that the court should have limited the jury’s use of statistics is the simple fact that this case did not involve “statistical” proof of discrimination. Although she compared dollar figures of her salary, pay raises, and incentive payments with the dollar figures of other employees’ salaries, raises, and payments, Carter did not abstract from these figures any statistical pattern or practice or calculate any subtle but discriminatory impact. Carter’s simple comparison of several employees’ salaries is not the kind of “statistical” proof that requires the mathematical safeguards that the defendant seeks to impose. The distance between this suit and the “statistical proof” cases upon which the defendant attempts to rely is revealed merely by examining what those cases have identified as proper statistical proof. For example, in Davis v. Califano, 613 F.2d 957 (D.C.Cir.1980), the employee in [1236]*1236a Title VII case used statistical evidence to demonstrate, in part, that too few women were in upper-level employment positions. After noting that statistics could be used to establish a prima facie case, the court explained the contours of statistical proof: “The proper comparison is between the composition of the relevant work force and the qualified population in the relevant labor market.” Id. at 963. See also Hazelwood School District v. United States, 433 U.S. 299, 308, 97 S.Ct. 2736, 2742, 53 L.Ed.2d 768 (1977) (“proper comparison was between the racial composition of Hazel-wood’s teaching staff and the racial composition of the qualified public school teacher-population in the relevant market”). Unlike the plaintiff in Davis, Carter made no attempt to use any type of general societal comparison or to suggest that anyone other than herself was subject to discrimination. The appellant’s labelling the proof as “statistical” therefore is inappropriate and the cases appellant cites thus are inapposite.
Moreover, we find no error in the trial court’s treatment of this numerical evidence. Evidence of this type clearly is relevant to the proof of discrimination. See, e.g., Minority Employees at NASA v. Beggs, 723 F.2d 958 at 962 (D.C.Cir.1983). The court here allowed the parties to argue about the proper weight to be given to the comparisons of salaries, raises, and incentive payments. As with any other circumstantial evidence, the jury was the proper body to weigh the evidence and to decide whether any inferences could be drawn. Accordingly, we affirm the district court’s decision to not give Duncan-Huggins’ proffered limiting instruction.
Duncan-Huggins also argues that the district court erred in refusing another proposed instruction that would have stated:
[A]n isolated racial joke or slur told by another employee, even if told in the presence of a supervisor and no matter how tasteless, is not evidence of race or color discrimination by the employer.
Carter v. Duncan-Huggins, Ltd., No. 81-546, slip op. at 2. In refusing to use this instruction, the district court reasoned that the jury had been presented with evidence that the joke’s narrator acted as Carter’s de facto supervisor and that it was for the jury to decide what weight to give the evidence of the anecdote.
We hold that the district court did not err in refusing this instruction. Indeed, at oral argument, appellant’s attorney conceded as much when he admitted that a racial slur may be “some evidence, taken with other things.” In contrast, the proposed instruction would have indicated to the jury that the single slur could not be considered as evidence. Were this concession absent, we nevertheless would uphold the district court’s decision.
An employee could be subject to many racially derogatory jokes, only one, or some number in between. The law at the extremes is relatively clear. A pervasive pattern of racial jokes of which the employer is aware can give rise to liability. See, e.g., Taylor v. Jones, 653 F.2d 1193 (8th Cir.1981). At the other end, a single isolated racial joke, unattended by any indicia of discriminatory animus, may not be sufficient in itself to establish a violation. See, e.g., Bundy v. Jackson, 641 F.2d 934 n. 9 (D.C.Cir.1981). The instant case, however, falls somewhere between the two extremes. Here, the single racial joke was presented to the jury together with numerous other pieces of circumstantial evidence. Logic suggests no reason why this piece of evidence should be singled out for separate treatment. It was for the jury to evaluate all the evidence and to decide the weight properly allocable to any one piece. Simply, no support exists for appellant’s proposed instruction that the joke could not be evidence of the employer’s alleged discriminatory motivation. Accordingly, we affirm the district court’s decision to deny this instruction.
In its motion for judgment n.o.v., Duncan-Huggins also argues that the jury’s $10,000 award was so speculative that the district court should have overturned it, directed a verdict for nominal damages, or ordered a remittitur. In fact, the district court had instructed the jury that plaintiff [1237]*1237was not to be awarded speculative damages. Tr. at 362. Appellant, who does not specifically challenge this instruction, argues that the award must have been speculative since appellee offered “no evidence of actual damages.” Appellant’s Brief at 30. In its post-verdict Memorandum, the district court disagreed. Because Carter had presented extensive evidence of variations in compensation levels, the court concluded that the jury’s award was not so unreasonable as to warrant disturbance by the court. Carter v. Duncan-Huggins, Ltd., No. 81-546, slip op. at 3-4. We affirm.
Appellant places substantial reliance on Carey v. Piphus, 435 U.S. 247, 98 S.Ct. 1042, 55 L.Ed.2d 252 (1978). At issue in Carey were damages awarded to students who had been suspended from school without procedural due process. As the district court in Carey had observed, the record was barren of any evidence upon which even a speculative measurement of damages could be based. 435 U.S. at 251 — 52, 98 S.Ct. at 1045 — 46. On appeal, the Supreme Court held that in the absence of any such evidence, actual damages could not be presumed. Id. at 247-48, 98 S.Ct. at 1044. In Halperin v. Kissinger, 606 F.2d 1192 (D.C.Cir.1979), aff’d per curiam by an equally divided Court, in part, cert. dismissed, in part, 452 U.S. 713 (1981), a case involving a substantive fourth amendment search and seizure claim, this circuit explained the reach of Carey. After noting that Carey prohibited a presumption of actual damages, this court explained: “To the extent that the plaintiffs here can establish actual though intangible injury, damages would clearly be appropriate even under [Carey v.] Piphus.” 606 F.2d at 1207 n. 100. In Doe v. District of Columbia, 697 F.2d 1115 (D.C.Cir.1983), separate statements published at 701 F.2d 948, a case involving inter alia a substantive eighth amendment cruel and unusual punishment claim, we further refined our analysis of Carey.
It is equally apparent that the [Carey] Court intended to require, not just that plaintiffs demonstrate actual injury before they secure more than a nominal recovery, but that damage awards be limited to sums necessary to compensate plaintiffs for actual harm.
Id. at 1123 (emphasis in original). In Doe, this court found error in an instruction that permitted the jury to compensate plaintiffs for the “intrinsic value” of their constitutional rights. Plaintiffs thus could recover only actual damages.
In the present case, the district court clearly and specifically instructed the jury that it could base an award only on actual evidence of injury and further cautioned against speculation in the calculation of damages.
Now, if you find a verdict for the Plaintiff, you will award her such actual or compensatory damages as you find from the preponderance of the evidence was proximately caused by the actions of the Defendant. If you find for the Plaintiff but do not find that she suffered any actual damages, you may then award her some nominal sum such as one dollar as damages.
The burden of proof is upon the Plaintiff to establish all elements of her damages. The amount of your verdict must be based on the evidence presented as to her losses. You are not to award to the Plaintiff speculative damages. That is compensation for present or future detriment which although possible is remote or guesswork.
Transcript at 361-62. Appellee introduced evidence of variations in salary and incentive payments that ranged in the thousands of dollars, and further proffered evidence of intangible harm. Because the plaintiff here received recompense for actual injuries that were demonstrated at trial, this case is a far cry from Carey, where the court of appeals had presumed a compensable injury, or from Doe, where the district court had instructed the jury that it could award damages for the somewhat nebulous “intrinsic value of a constitutional right.” Here, the plaintiff merely was compensated for actual, demonstrated injuries.
Because we conclude that Carter presented evidence of actual, but intangible injury [1238]*1238we need not decide whether this is a proper case in which to infer the infliction of emotional distress. As we stated in Doe v. District of Columbia:
It may well be possible, in an appropriate case, to infer the infliction of emotional distress from the circumstances of the violation.... While the court in Carey seemed to frown on such a mode of proof in the context of procedural process ... such inferences likely would be much more reliable when the plaintiff has been subjected to “cruel and unusual punishment.”
697 F.2d at 1124 n. 24 (citation omitted). Similarly, the inference might be much more reliable when the plaintiff has been subjected to open racial discrimination.
Appellant’s argument that Carter offered no evidence of actual damages misstates the evidence and collapses upon even a cursory examination of the record. Much testimony was adduced that specified the precise dollar amounts of the compensation differential and the length of Carter’s employment. From these two figures, the jury could have calculated a substantial part of the award. Because the difference was in the thousands of dollars, Carter’s reduced earnings reasonably could have accounted for the overwhelming majority of the award.
We are satisfied that the remainder of the award reasonably reflected injuries which Carter actually suffered despite their intangibility. Throughout the trial, the jury heard testimony evincing the individualized intangible injury that appellee suffered — namely, humiliation and other emotional harm. Carter testified that she “felt isolated,” that her “mind just flew out” when she heard the “disrespectful” racist anecdote, and that she was taunted about not having a key. She further testified that she was shocked by an accusation of incompetence and that the atmosphere was one of harassment. These statements constitute the stuff of which a claim for humiliation and emotional harm is composed. See, e.g., 1 T. Sedgwick, Measure of Damages § 47, at 75-81 (1920); see also Annot., Recovery of Damages for Emotional Distress Resulting from Racial, Ethnic, or Religious Abuse, or Discrimination, 40 A.L.R.3d 1290 (1971), distinguished in Carey v. Piphus, 435 U.S. at 264 n. 22, 98 S.Ct. at 1052 n. 22. And, humiliation is a cognizable and compensable injury under section 1981. See, e.g., Williams v. Trans World Airlines, Inc., 660 F.2d 1267 (8th Cir.1981); Garner v. Giarrusso, 571 F.2d 1330 (5th Cir.1978); see also Seaton v. Sky Realty Co., 491 F.2d 634 (7th Cir.1974) (allowing compensation for humiliation under section 1982), cited in Doe v. District of Columbia, 697 F.2d at 1124 n. 24. In Metrocare v. Washington Metropolitan Area Transit Authority, 679 F.2d at 927, 929, a section 1981 case, we noted without comment that the jury had awarded damages for pain, suffering, and emotional distress. The mere fact that Carter did not recite the specific terms “emotional harm” or “humiliation” while on the stand does not wash away the evidence of her intangible injury. To place a premium on conclusionary terms such as these would distract attention away from the central question of whether plaintiff established actual, but intangible harm. A thorough review of the record convinces us that Carter here established such a harm. The jury, then, applied its experience and common sense to evaluate this injury and reasonably calculated the amount necessary to compensate Carter. See generally Williams v. Trans World Airlines, Inc., 660 F.2d 1267, 1272 (8th Cir.1981). Since the jury’s award for lost compensation may properly have been as high as seven or eight thousand dollars, that portion which corresponded to the intangible injury may have been as low as two or three thousand dollars. We cannot conclude that this is unreasonable compensation for what the jury may have believed to have been seventeen months of incessant humiliation, harassment, and feelings of isolation.
In reviewing the actual amount of a jury’s award, our task is limited and a reluctance to interfere is our touchstone. This limited role reflects the obvious fact that we are not. privy to the jury’s deliberations. In reviewing the amount of the [1239]*1239jury’s award, we thus need not — and indeed cannot — reconstruct the precise mathematical formula that the jury adopted. Nor need we explore every possible quantitative analysis or compute the basis of each penny and dollar in the award. Our inquiry ends once we are satisfied that the award is within a reasonable range and that the jury did not engage in speculation or other improper activity. As we stated in Taylor v. Washington Terminal Co., 409 F.2d 145 (D.C.Cir.), cert. denied, 396 U.S. 835, 90 S.Ct. 93, 24 L.Ed.2d 85 (1969): “Where the jury finds a particular quantum of damages and the trial judge refuses to disturb its findings on the motion for a new trial ... an appellate court should be certain indeed that the award is contrary to all reason before it orders a remittitur or a new trial.” Id., at 148 (emphasis added). The heavy burden of establishing the award as “contrary to all reason” firmly rests on that party which challenged the jury’s award. In the case sub judice, Duncan-Huggins has failed to meet this burden. Given that the standard of review we must apply is extremely deferential, the fact that the district court admonished the jury to award damages only for actual injuries, and that credible evidence existed to establish these actual injuries, we cannot conclude that the award was excessive. Accordingly, we affirm.
V. Conclusion
For the reasons indicated above, we affirm the district court’s denial of Duncan-Huggins’ request for specified jury instructions and its motion for a judgment n.o.v.
Affirmed.
The dissent’s attempt to minimize the extent of the compensation differential culminates in the presentation of comparative charts. See dissenting opinion, infra at 1241, 1242. These charts, however, distort the evidence as they paint only part of the picture. For example, the dissent emphasizes that Carter’s salary was, at times, commensurate with the salaries of Barlow and Sales. That observation misses the mark. Because Barlow and Sales were hired after Carter, they were her junior and, applying the analysis the dissent itself suggests in its attempt to justify Carter’s initial lower salary, thus should have received lower, not equal salaries.
Moreover, the “aggregate compensation” chart is misleading since it incorrectly implies that all employees listed thereon could participate in the incentive program. Employees Hook and Barlow were employed only for a short period at Duncan-Huggins, and employee Sales was hired only on a temporary basis. Transcript at 104. Because a Duncan-Huggins employee qualified for the incentive plan only upon six month’s employment, the aggregate compensation for Hook, Sales, and Barlow may have reflected only a failure to qualify for the program or a failure to participate in the program for any meaningful period. The comparison of incentive payments, which the dissent implicitly included in its comparison of aggregate compensation chart, should be limited to those employees who participated in the program.