delivered the opinion of the Court.

In this action we address a longstanding conflict among the Federal Courts of Appeals over whether Federal Rule of Evidence 803(8)(C), which provides an exception to the hearsay rule for public investigatory reports containing “factual findings,” extends to conclusions and opinions contained in such reports. We also consider whether, on the facts of this litigation, the trial court abused its discretion in refusing to admit, on cross-examination, testimony intended to provide a more complete picture of a document about which the witness had testified on direct.

I

This litigation stems from the crash of a Navy training aircraft at Middleton Field, Alabama, on July 13, 1982, which took the lives of both pilots on board, Lieutenant Commander Barbara Ann Rainey and Ensign Donald Bruce Knowlton. The accident took place while Rainey, a Navy flight instructor, and Knowlton, her student, were flying “touch-and-go” exercises in a T-34C Turbo-Mentor aircraft, number 3E955. Their aircraft and several others flew in an oval pattern, each plane making successive landing/takeoff maneuvers on the runway. Following its fourth pass at the runway, 3E955 appeared to make a left turn prematurely, cutting out the aircraft ahead of it in the pattern and threatening a collision. After radio warnings from two other pilots, the plane banked sharply to the right in order to avoid the other aircraft. At that point it lost altitude rapidly, crashed, and burned.

Because of the damage to the plane and the lack of any survivors, the cause of the accident could not be determined with certainty. The two pilots’ surviving spouses brought a product liability suit against petitioners Beech Aircraft Corporation, the plane’s manufacturer, and Beech Aerospace Services, which serviced the plane under contract with the Navy.1 The plaintiffs alleged that the crash had been *157caused by a loss of engine power, known as “rollback,” due to some defect in the aircraft’s fuel control system. The defendants, on the other hand, advanced the theory of pilot error, suggesting that the plane had stalled during the abrupt avoidance maneuver.

At trial, the only seriously disputed question was whether pilot error or equipment malfunction had caused the crash. Both sides relied primarily on expert testimony. One piece of evidence presented by the defense was an investigative report prepared by Lieutenant Commander William Morgan on order of the training squadron’s commanding officer and pursuant to authority granted in the Manual of the Judge Advocate General. This “JAG Report,” completed during the six weeks following the accident, was organized into sections labeled “finding of fact,” “opinions,” and “recommendations,” and was supported by some 60 attachments. The “finding of fact” included statements like the following:

“13. At approximately 1020, while turning crosswind without proper interval, 3E955 crashed, immediately caught fire and burned.

“27. At the time of impact, the engine of 3E955 was operating but was operating at reduced power.” App. 10-12.

Among his “opinions” Lieutenant Commander Morgan stated, in paragraph 5, that due to the deaths of the two pilots and the destruction of the aircraft “it is almost impossible to determine exactly what happened to Navy 3E955 from the time it left the runway on its last touch and go until it impacted the ground.” He nonetheless continued with a detailed reconstruction of a possible set of events, based on pilot error, that could have caused the accident.2 The next two paragraphs stated a caveat and a conclusion:

*158“6. Although the above sequence of events is the most likely to have occurred, it does not change the possibility that a ‘rollback’ did occur.

“7. The most probable cause of the accident was the pilots [sic] failure to maintain proper interval.” Id., at 15. .

The trial judge initially determined, at a pretrial conference, that the JAG Report was sufficiently trustworthy to be admissible, but that it “would be admissible only on its fac*159tual findings and would not be admissible insofar as any opinions or conclusions are concerned.” Id., at 35. The day before trial, however, the court reversed itself and ruled, over the plaintiffs’ objection, that certain of the conclusions would be admitted. Id., at 40-41. Accordingly, the court admitted most of the report’s “opinions,” including the first sentence of paragraph 5 about the impossibility of determining exactly what happened, and paragraph 7, which opined about failure to maintain proper interval as “[t]he most probable cause of the accident.” Id., at 97. On the other hand, the remainder of paragraph 5 was barred as “nothing but a possible scenario,” id., at 40, and paragraph 6, in which investigator Morgan refused to rule out rollback, was deleted as well.3

This action also concerns an evidentiary ruling as to a second document. Five or six months after the accident, plaintiff John Rainey, husband of the deceased pilot and himself a Navy flight instructor, sent a detailed letter to Lieutenant Commander Morgan. Based on Rainey’s own investigation, the letter took issue with some of the JAG Report’s findings and outlined Rainey’s theory that “[t]he most probable primary cause factor of this aircraft mishap is a loss of useful power (or rollback) caused by some form of pneumatic sensing/fuel flow malfunction, probably in the fuel control unit.” Id., at 104, 111.

At trial Rainey did not testify during his side’s case in chief, but he was called by the defense as an adverse witness. On direct examination he was asked about two statements contained in his letter. The first was to the effect that his wife had unsuccessfully attempted to cancel the ill-fated training flight because of a variety of adverse factors including her student’s fatigue. The second question concerned a portion of Rainey’s hypothesized scenario of the accident:

*160“Didn’t you say, sir, that after Mrs. Rainey’s airplane rolled wings level, that Lieutenant Colonel Habermacher’s plane came into view unexpectedly at its closest point of approach, although sufficient separation still existed between the aircraft. However, the unexpected proximitely [sic] of Colonel Habermacher’s plane caused one of the aircrew in Mrs. Rainey’s plane to react instinctively and abruptly by initiating a hard right turn away from Colonel Habermacher’s airplane?” Id., at 75.

Rainey admitted having made b'oth statements. On cross-examination, Rainey’s counsel asked the following question: “In the same letter to which Mr. Toothman made reference to in his questions, sir, did you also say that the most probably [sic] primary cause of this mishap was rollback?” Id., at 77. Before Rainey answered, the court sustained a defense objection on the ground that the question asked for Rainey’s opinion. Further questioning along this line was cut off.

Following a 2-week trial, the jury returned a verdict for petitioners. A panel of the Eleventh Circuit reversed and remanded for a new trial. 784 F. 2d 1523 (1986). Considering itself bound by the Fifth Circuit precedent of Smith v. Ithaca Corp., 612 F. 2d 215 (1980),4 the panel agreed with Rainey’s argument that Federal Rule of Evidence 803(8)(C), which excepts investigatory reports from the hearsay rule, did not encompass evaluative conclusions or opinions. Therefore, it held, the “conclusions” contained in the JAG Report should have been excluded. One member of the panel, concurring specially, urged however that the Circuit reconsider its interpretation of Rule 803(8)(C), suggesting that “Smith is an anomaly among the circuits.” 784 F. 2d, at 1530 (opinion of Johnson, J.). The panel also held, citing Federal Rule of Evidence 106, that it was reversible error for the trial court *161to have prohibited cross-examination about additional portions of Rainey’s letter which would have put in context the admissions elicited from him on direct.5

On rehearing en banc, the Court of Appeals divided evenly on the question of Rule 803(8)(C). 827 F. 2d 1498 (CA11 1987). It therefore held that Smith was controlling and consequently reinstated the panel judgment. On the Rule 106 question, the court unanimously reaffirmed the panel’s decision that Rule 106 (or alternatively Rule 801(d)(1)(B)) required reversal. We granted certiorari to consider both issues. 485 U. S. 903 (1988).

II

Federal Rule of Evidence 803 provides that certain types of hearsay statements are not made excludable by the hearsay rule, whether or not the declarant is available to testify. Rule 803(8) defines the “public records and reports” which are not excludable, as follows:

“Records, reports, statements, or data compilations, in any form, of public offices or agencies, setting forth (A) the activities of the office or agency, or (B) matters observed pursuant to duty imposed by law as to which matters there was a duty to report,... or (C) in civil actions and proceedings and against the Government in criminal cases, factual findings resulting from an investigation made pursuant to authority granted by law, unless the sources of information or other circumstances indicate lack of trustworthiness.”

Controversy over what “public records and reports” are made not excludable by Rule 803(8)(C) has divided the federal courts from the beginning. In the present litigation, the Court of Appeals followed the “narrow” interpretation of Smith v. Ithaca Corp., supra, at 220-223, which held that the *162term “factual findings” did not encompass “opinions” or “conclusions.” Courts of Appeals other than those of the Fifth and Eleventh Circuits, however, have generally adopted a broader interpretation. For example, the Court of Appeals for the Sixth Circuit, in Baker v. Elcona Homes Corp., 588 F. 2d 551, 557-558 (1978), cert. denied, 441 U. S. 933 (1979), held that “factual findings admissible under Rule 803(8)(C) may be those which are made by the preparer of the report from disputed evidence . . . .”6 The other Courts of Appeals that have squarely confronted the issue have also adopted the broader interpretation.7 We agree and hold that factually based conclusions or opinions are not on that account excluded from the scope of Rule 803(8)(C).

*163Because the Federal Rules of Evidence are a legislative enactment, we turn to the “traditional tools of statutory-construction, ” INS v. Cardoza-Fonseca, 480 U. S. 421, 446 (1987), in order to construe their provisions. We begin with the language of the Rule itself. Proponents of the narrow view have generally relied heavily on a perceived dichotomy between “fact” and “opinion” in arguing for the limited scope of the phrase “factual findings.” Smith v. Ithaca Corp. contrasted the term “factual findings” in Rule 803(8) (C) with the language of Rule 803(6) (records of regularly conducted activity), which expressly refers to “opinions” and “diagnoses.” “Factual findings,” the court opined, must be something other than opinions. 612 F. 2d, at 221-222.8

For several reasons, we do not agree. In the first place, it is not apparent that the term “factual findings” should be *164read to mean simply “facts” (as opposed to “opinions” or “conclusions”). A common definition of “finding of fact” is, for example, “[a] conclusion by way of reasonable inference from the evidence.” Black’s Law Dictionary 569 (5th ed. 1979). To say the least, the language of the Rule does not compel us to reject the interpretation that “factual findings” includes conclusions or opinions that flow from a factual investigation. Second, we note that, contrary to what is often assumed, the language of the Rule does not state that “factual findings” are admissible, but that “reports . . . setting forth . . . factual findings” (emphasis added) are admissible. On this reading, the language of the Rule does not create a distinction between “fact” and “opinion” contained in such reports.

Turning next to the legislative history of Rule 803(8)(C), we find no clear answer to the question of how the Rule’s language should be interpreted. Indeed, in this litigation the legislative history may well be at the origin of the dispute. Rather than the more usual situation where a court must attempt to glean meaning from ambiguous comments of legislators who did not focus directly on the problem at hand, here the Committees in both Houses of Congress clearly recognized and expressed their opinions on the precise question at issue. Unfortunately, however, they took diametrically opposite positions. Moreover, the two Houses made no effort to reconcile their views, either through changes in the Rule’s language or through a statement in the Report of the Conference Committee.

The House Judiciary Committee, which dealt first with the proposed rules after they had been transmitted to Congress by this Court, included in its Report but one brief paragraph on Rule 803(8):

“The Committee approved Rule 803(8) without substantive change from the form in which it was submitted by the Court. The Committee intends that the phrase ‘factual findings’ be strictly construed and that evaluations or opinions contained in public reports shall not be *165admissible under this Rule.” H. R. Rep. No. 93-650, p. 14 (1973).

The Senate Committee responded at somewhat greater length, but equally emphatically:

“The House Judiciary Committee report contained a statement of intent that ‘the phrase “factual findings” in subdivision (c) be strictly construed and that evaluations or opinions contained in public reports shall not be admissible under this rule.’ The committee takes strong exception to this limiting understanding of the application of the rule. We do not think it reflects an understanding of the intended operation of the rule as explained in the Advisory Committee notes to this subsection. . . . We think the restrictive interpretation of the House overlooks the fact that while the Advisory Committee assumes admissibility in the first instance of evaluative reports, they are not admissible if, as the rule states, ‘the sources of information or other circumstances indicate lack of trustworthiness.’

“The committee concludes that the language of the rule together with the explanation provided by the Advisory Committee furnish sufficient guidance on the admissibility of evaluative reports.” S. Rep. No. 93-1277, p. 18 (1974).

Clearly this legislative history reveals a difference of view between the Senate and the House that affords no definitive guide to the congressional understanding. It seems clear however that the Senate understanding is more in accord with the wording of the Rule and with the comments of the Advisory Committee.9

*166The Advisory Committee’s comments are notable, first, in that they contain no mention of any dichotomy between statements of “fact” and “opinions” or “conclusions.” What was on the Committee’s mind was simply whether what it called “evaluative reports” should be admissible. Illustrating the previous division among the courts on this subject, the Committee cited numerous cases in which the admissibility of such reports had been both sustained and denied. It also took note of various federal statutes that made certain kinds of evaluative reports admissible in evidence. What is striking about all of these examples is that these were reports that stated conclusions. E. g., Moran v. Pittsburgh-Des Moines Steel Co., 183 F. 2d 467, 472-473 (CA3 1950) (report of Bureau of Mines concerning the cause of a gas tank explosion admissible); Franklin v. Skelly Oil Co., 141 F. 2d 568, 571-572 (CA10 1944) (report of state fire marshal on the cause of a gas explosion inadmissible); 42 U. S. C. § 269(b) (bill of health by appropriate official admissible as prima facie evidence of vessel’s sanitary history and condition). The Committee’s concern was clearly whether reports of this kind should be admissible. Nowhere in its comments is there the slightest indication that it even considered the solution of admitting only “factual” statements from such reports.10 *167Rather, the Committee referred throughout to “reports,” without any such differentiation regarding the statements they contained. What the Committee referred to in the Rule’s language as “reports . . . setting forth . . . factual findings” is surely nothing more or less than what in its commentary it called “evaluative reports.” Its solution as to their admissibility is clearly stated in the final paragraph of its report on this Rule. That solution consists of two principles: First, “the rule . . . assumes admissibility in the first instance . . . .” Second, it provides “ample provision for escape if sufficient negative factors are present.”

That “provision for escape” is contained in the final clause of the Rule: evaluative reports are admissible “unless the sources of information or other circumstances indicate lack of trustworthiness.” This trustworthiness inquiry — and not an arbitrary distinction between “fact” and “opinion” — was the Committee’s primary safeguard against the admission of unreliable evidence, and it is important to note that it applies to all elements of the report. Thus, a trial judge has the discretion, and indeed the obligation, to exclude an entire report or portions thereof — whether narrow “factual” statements or broader “conclusions” — that she determines to be untrustworthy.11 Moreover, safeguards built into other portions of *168the Federal Rules, such as those dealing with relevance and prejudice, provide the court with additional means of scrutinizing and, where appropriate, excluding evaluative reports or portions of them. And of course it goes without saying that the admission of a report containing “conclusions” is subject to the ultimate safeguard — the opponent’s right to present evidence tending to contradict or diminish the weight of those conclusions.

Our conclusion that neither the language of the Rule nor the intent of its framers calls for a distinction between “fact” and “opinion” is strengthened by the analytical difficulty of drawing such a line. It has frequently been remarked that the distinction between statements of fact and opinion is, at best, one of degree:

“All statements in language are statements of opinion, i. e., statements of mental processes or perceptions. So-called ‘statements of fact’ are only more specific statements of opinion. What the judge means to say, when he asks the witness to state the facts, is: ‘The nature of this case requires that you be more specific, if you can, in your description of what you saw.’” W. King & D. Pillinger, Opinion Evidence in Illinois 4 (1942) (footnote omitted), quoted in 3 J. Weinstein & M. Berger, Weinstein’s Evidence ¶701[01], p. 701-6 (1988).

See also E. Cleary, McCormick on Evidence 27 (3d ed. 1984) (“There is no conceivable statement however specific, detailed and ‘factual,’ that is not in some measure the product of inference and reflection as well as observation and memory”); R. Lempert & S. Saltzburg, A Modern Approach to Evidence 449 (2d ed. 1982) (“A factual finding, unless it is a simple report of something observed, is an opinion as to what more basic facts imply”). Thus, the traditional requirement that lay witnesses give statements of fact rather than opinion may *169be considered, “[l]ike the hearsay and original documents rules ... a ‘best evidence’ rule.” McCormick, Opinion Evidence in Iowa, 19 Drake L. Rev. 245, 246 (1970).

In the present action, the trial court had no difficulty in admitting as a factual finding the statement in the JAG Report that “[a]t the time of impact, the engine of 3E955 was operating but was operating at reduced power.” Surely this “factual finding” could also be characterized as an opinion, which the investigator presumably arrived at on the basis of clues contained in the airplane wreckage. Rather than requiring that we draw some inevitably arbitrary line between the various shades of fact/opinion that invariably will be present in investigatory reports, we believe the Rule instructs us-as its plain language states — to admit “reports . . . setting forth . . . factual findings.” The Rule’s limitations and safeguards lie elsewhere: First, the requirement that reports contain factual findings bars the admission of statements not based on factual investigation. Second, the trustworthiness provision requires the court to make a determination as to whether the report, or any portion thereof, is sufficiently trustworthy to be admitted.

A broad approach to admissibility under Rule 803(8)(C), as we have outlined it, is also consistent with the Federal Rules’ general approach of relaxing the traditional barriers to “opinion” testimony. Rules 702-705 permit experts to testify in the form of an opinion, and without any exclusion of opinions on “ultimate issues.” And Rule 701 permits even a lay witness to testify in the form of opinions or inferences drawn from her observations when testimony in that form will be helpful to the trier of fact. We see no reason to strain to reach an interpretation of Rule 803(8)(C) that is contrary to the liberal thrust of the Federal Rules.12

*170We hold, therefore, that portions of investigatory reports otherwise admissible under Rule 803(8)(C) are not inadmissible merely because they state a conclusion or opinion. As long as the conclusion is based on a factual investigation and satisfies the Rule’s trustworthiness requirement, it should be admissible along with other portions of the report.13 As the trial judge in this action determined that certain of the JAG Report’s conclusions were trustworthy, he rightly allowed them to be admitted into evidence. We therefore reverse the judgment of the Court of Appeals in respect of the Rule 803(8)(C) issue.

Ill

Respondents also contended on appeal that reversal was required because the District Court improperly restricted the cross-examination of plaintiff Rainey by his own counsel in regard to the letter Rainey had addressed to Lieutenant Commander Morgan. We agree with the unanimous holding of the Court of Appeals en banc that the District Court erred in refusing to permit Rainey to present a more complete picture of what he had written to Morgan.

We have no doubt that the jury was given a distorted and prejudicial impression of Rainey’s letter. The theory of Rainey’s case was that the accident was the result of a power failure, and, read in its entirety, his letter to Morgan was fully consistent with that theory. While Rainey did discuss problems his wife had encountered the morning of the accident which led her to attempt to cancel the flight, and also agreed that her airplane had violated pattern integrity in turning left prematurely, the thrust of his letter was to chal*171lenge Morgan’s theory that the crash had been caused by a stall that took place when the pilots turned sharply right and pitched up in attempting to avoid the other plane. Thus Rainey argued that Morgan’s hypothesis was inconsistent with the observations of eyewitnesses, the physical findings in the wreckage, and the likely actions of the two pilots. He explained at length his theory of power failure and attempted to demonstrate how the various pieces of evidence supported it. What the jury was told, however, through the defendants’ direct examination of Rainey as an adverse witness, was that Rainey had written six months after the accident (1) that his wife had attempted to cancel the flight, partly because her student was tired and emotionally drained, and that “unnecessary pressure” was placed on them to proceed with it; and (2) that she or her student had abruptly initiated a hard right turn when the other aircraft unexpectedly came into view. It is plausible that a jury would have concluded from this information that Rainey did not believe in his theory of power failure and had developed it only later for purposes of litigation. Because the court sustained defense counsel’s objection, Rainey’s counsel was unable to counteract this prejudicial impression by presenting additional information about the letter on cross-examination.

The common-law “rule of completeness,” which underlies Federal Rule of Evidence 106, was designed to prevent exactly the type of prejudice of which Rainey complains. In its aspect relevant to this litigation, the rule of completeness was stated succinctly by Wigmore: “[T]he opponent, against whom a part of an utterance has been put in, may in his turn complement it by putting in the remainder, in order to secure for the tribunal a complete understanding of the total tenor and effect of the utterance. ” 7 J. Wigmore, Evidence in Trials at Common Law §2113, p. 653 (J. Chadbourn rev. 1978).14 The *172Federal Rules of Evidence have partially codified the doctrine of completeness in Rule 106:

“When a writing or recorded statement or part thereof is introduced by a party, an adverse party may require the introduction at that time of any other part or any other writing or recorded statement which ought in fairness to be considered contemporaneously with it.”

In proposing Rule 106, the Advisory Committee stressed that it “does not in any way circumscribe the right of the adversary to develop the matter on cross-examination or as part of his own case.” Advisory Committee’s Notes on Fed. Rule Evid. 106, 28 U. S. C. App., p. 682. We take this to be a reaffirmation of the obvious: that when one party has made use of a portion of a document, such that misunderstanding or distortion can be averted only through presentation of another portion, the material required for completeness is ipso facto relevant and therefore admissible under Rules 401 and 402. See 1J. Weinstein & M. Berger, Weinstein’s Evidence ¶ 106[02], p. 106-20 (1986). The District Court’s refusal to admit the proffered completion evidence was a clear abuse of discretion.

While much of the controversy in this suit has centered on whether Rule 106 applies, we find it unnecessary to address that issue. Clearly the concerns underlying Rule 106 are relevant here, but, as the general rules of relevancy permit a ready resolution to this litigation, we need go no further in exploring the scope and meaning of Rule 106.15

Unfortunately for the clarity of the proceedings, the defendants’ objection to the question put by Rainey’s counsel was couched not in terms of relevance but rather as calling *173for an opinion.16 While the question put to Rainey indeed inquired about an opinion Rainey had earlier expressed, it should have been obvious from the context that the purpose of the question was not to elicit Rainey’s opinion on the cause of the accident. Rather, Rainey was asked, in effect, whether he had made a certain statement in his letter. That was a question he was eminently qualified to answer.17 Counsel’s objection that Rainey was not entitled to give opinion evidence could not avail in view of the obvious purpose for which the statement was offered.18

*174Petitioners have also objected that Rainey waived the right to pursue this issue on appeal because he did not properly raise it in the trial court. We disagree. Rule 103(a)(2) requires, in the first place, that to preserve an argument that evidence was wrongly excluded the proponent must make known the substance of the evidence sought to be admitted by an offer of proof unless it “was apparent from the context within which questions were asked.”19 Here the nature of the proposed testimony was abundantly apparent from the very question put by Rainey’s counsel. The proponent must also comply with Federal Rule of Civil Procedure 46, which requires that a party seeking to preserve an objection to the court’s ruling must “mak[e] known to the court the action which the party desires the court to take or the party’s objection to the action of the court and the grounds therefor.” Although, as is frequently the case in the heat of a trial, counsel did not explain the evidentiary basis of his argument as thoroughly as might ideally be desired, we are satisfied that he substantially satisfied the requirement of putting the court on notice as to his concern. In the colloquy following the defense objection to his question,20 and before he was cut off first by defense counsel and then by the judge, Rainey’s counsel began to articulate the argument that his question should be allowed because the defense had been able to question Rainey concerning his letter. Moreover, the judge’s response21 suggests that he perceived the completeness argument. We cannot say that the point was not sufficiently *175made.22 Rainey therefore was not barred from pursuing this issue on appeal.

IV

We hold, first, that statements in the form of opinions or conclusions are not by that fact excluded from the scope of Federal Rule of Evidence 803(8)(C). We therefore reverse the judgment of the Court of Appeals in that respect. Second, we hold that on the facts of this litigation the District Court abused its discretion in restricting the scope of cross-examination of respondent Rainey by his counsel, and to that extent we affirm the Court of Appeals’ judgment. The case is remanded for further proceedings consistent with this opinion.

It is so ordered.

with whom

I join Parts I and II of the Court’s opinion, but dissent from Part III. I do not believe the District Court abused its discretion in refusing to admit this particular testimony. The Court concedes that “counsel did not explain the evidentiary basis of his argument'as thoroughly as might ideally be desired ...” ante, at 174, but I would go further and say that counsel’s brief presentation to the District Court was ambiguous at best.

Rainey’s attorney was faced with an objection to testimony he wished to elicit from his client based on opposing counsel’s perception that it would be nonexpert opinion.1 He responded by saying “[o]n the basis that this letter constitutes an admission by Commander Rainey, he has been asked to answer every single question [opposing counsel] had respecting — .” App. 77. At that point the court cut in with an explanation of why that answer was insufficient. The judge explained:

“I don’t recall going into anything except the matter about the right turn and so forth, and that’s all he went into. He did express that opinion and that came in as an admission against him, I suppose, but that doesn’t mean you can[’t] qualify him for the questions you are now asking. The objection is sustained.” Id., at 78.

Rainey’s lawyer seems to have been arguing that, because no one objected to Rainey’s answers to defendant’s questions about the letter as nonexpert opinion, Rainey should be able to answer similar questions put by his own attorney without that objection. The argument looks more like one based on *177fairness or waiver (often known as “opening the door”2) than one based specifically on completeness. That is how the judge understood it. He explained his ruling sustaining the objection by noting that although the defense questioning had elicited some opinion, it was admissible on other grounds and then suggested that Rainey’s lawyer qualify Rainey as an expert. Here the trial judge ruled on the basis of a reasonable understanding of respondents’ stated reasons for allowing the evidence to be admitted, and the trial judge made this understanding clear to respondents’ counsel. The evidence was not admissible under this view, and counsel made no attempt to clarify his position.

Today, the Court offers sound reasons for the admission of the testimony in question, but they are reasons which it has adduced from briefs and careful research, not the reasons expressed by counsel at trial.

“If counsel specifies a purpose for which the proposed evidence is inadmissible and the judge excludes, counsel cannot complain of the ruling on appeal though it could have been rightly admitted for another purpose.” E. Cleary, McCormick on Evidence §51, p. 125 (3d ed. 1984).

Trial judges do not have the luxury of briefs or research when making a typical evidentiary ruling, and for this reason we have traditionally required the proponent of evidence to defend it against objection by showing why it should be admissible. Federal Rule of Evidence 103(a)(2) requires an “offer of proof” in order to preserve for review a perceived error excluding evidence.3 Most courts and treatises have *178interpreted the need for an “offer of proof” as requiring a specific and timely defense of the evidence. See 1 J. Weinstein & M. Berger, Weinstein’s Evidence ¶ 103[03], pp. 103-36 to 103-38 (1988); 21 C. Wright & K. Graham, Federal Practice and Procedure §5040, pp. 209-211 (1977); United States v. Peters, 732 F. 2d 1004 (CA1 1984); United States v. Grapp, 653 F. 2d 189, 194 (CA5 1981); Huff v. White Motor Corp., 609 F. 2d 286 (CA7 1979). The need for a showing of evidence is the same, whether it is an essential part of the “offer of proof,” or, as the Court agrees, required by Federal Rule of Civil Procedure 46.4

The disagreement in these cases is not about applicable Rules of Evidence, but how a trial judge should fairly have understood an offer of proof under these circumstances. This Court, far removed from the factual context and on the basis of a cold record, is in no position to say that the trial court’s ruling in this situation was an abuse of discretion. Cf. Anderson v. Bessemer City, 470 U. S. 564, 575 (1985).