Dr. Lawrence Cranberg filed these diversity actions for product disparagement and libel against Consumers Union of United States, Inc., (Consumers Union), the publisher of Consumer Reports magazine, and for libel against the Associated Press (AP), and Cox Enterprises, Inc., d/b/a Austin American-Statesman (American-Statesman ). These cases were consolidated for trial on the issue of liability alone. In a nonjury trial, the district court found all defendants not liable. We affirm the finding of the district court.

* This law suit was sparked by an article that appeared in the January 1981 issue of Consumer Reports magazine. Dr. Cranberg claims this article libeled him and disparaged his product, the Texas Fireframe. Cranberg is the sole proprietor of the Texas Fireframe Company, which manufactures and sells the Texas Fireframe, a grate he invented for use in building the "slot fire." Cranberg, a physicist, discovered and patented the slot fire, also known as the "physicist's fire." This fire is built by stacking logs in a double-decked arrangement with a larger log in the back. This "C" formation creates a cavity that faces the room and acts as a natural radiator, generating intense heat which is directed out into the room rather than up the chimney. The Texas Fireframe is a double-decked grate developed and patented by Cranberg to hold the logs in the proper stacked formation.

Along with promoting his Fireframe, Cranberg has fervently waged a continuous battle against what he calls the "anti-fireplace hoax"--the belief that fireplaces are inefficient home heating sources because they draw centrally heated air out of the room and allow more heated air to escape through the flue than is sent back into the room. The record contains voluminous examples of Cranberg's activities and writings directed toward dispelling this hoax. Throughout these efforts, Cranberg has challenged the scientific basis underlying the notion that fireplaces waste energy and has asserted that his product, the Texas Fireframe, greatly improves fireplace efficiency.

In its January 1981 issue of Consumer Reports, Consumers Union reviewed a number of fireplace devices to determine their efficiency as domestic heat sources. As stated in the article, Consumers Union set out "to see if a fireplace could be modified in order to pull its weight as an auxiliary heat-provider." One of the devices tested and reported on was the Texas Fireframe.

To conduct its tests in a controlled environment, Consumers Union built a calorimeter room, a room specially designed to measure gain or loss of heat energy. Each device was tested in this room under two scenarios. One test estimated the net heat output in a "log cabin" situation, where the air supplied to the fire is not preheated. The "log cabin" output supposedly shows how much heat is provided when the fireplace is the only source of warmth. The second test dealt with the more common situation, in which the fireplace is operating in a centrally heated home. Unlike the log cabin test, this scenario supposedly accounted for the energy loss occurring from the fire's being fueled by air heated by another source.

As its benchmark for data comparison, Consumers Union built what it called a simple or "standard" fire. This arrangement had the logs placed directly on the hearth, abutting the back wall, and the front of the fireplace covered with a wire-mesh screen. The measurements collected from this base case provided a point for relative comparison of the heat production data collected from the ten devices Consumers Union tested and reviewed. Upon completing its testing, Consumers Union found this standard fire, without a grate or add-on device of any kind, out-performed three of the devices it tested, including the Texas Fireframe.

The discussion of the Texas Fireframe appeared under a subheading "And sometimes you lose."1 The discussion noted that the Texas Fireframe, while better than ordinary grates, was not as good as Consumers Union's standard fire built directly on the hearth. The article stated the Texas Fireframe would increase the efficiency of a fireplace as compared to a regular one-level grate, but would decrease the efficiency of a fireplace in a centrally heated home as compared to a fireplace with no grate and no device.

Cranberg believed the Consumer Reports article to be completely in error, the testing procedures wholly inaccurate and Consumers Union a perpetrator of the anti-fireplace hoax. He alleged the article disparaged his product and libeled his business and himself.

Cranberg attempted to rectify the perceived injustice perpetrated by Consumers Union against him and his grate. Cranberg prepared a paper disputing Consumers Union's findings and testing methods and emphasizing that the grateless fire, touted by Consumers Union as more energy efficient, would not even ignite and burn effectively. Cranberg wrote to Consumers Union on January 19, 1981 to inform them he planned to present his paper, tentatively titled "Plagarism, Suppression of Evidence, and Misrepresentations in Consumer Reports, January 1981," at the April meeting of the American Physical Society in Baltimore. Cranberg further stated in his letter to Consumers Union that if he did not receive an appropriate response he would mail the paper to the Society with copies to representatives of the media. When Cranberg had not received what he deemed an appropriate response by January 27, he sent an abstract of his paper to news organizations and media personalities to publicize his side of the dispute with Consumers Union. In a letter accompanying the abstract, Cranberg indicated he was challenging the publication and intended to ask the Society to initiate a thorough study of Consumers Union. Dr. Cranberg's express purpose in sending these abstracts, and later the full text of his paper, to these news organizations was to encourage the broadcast of his paper "at least as widely as the story that had been written about [his] product by Consumers Union, [and], preferably much more."

Prior to the convening of the American Physical Society, Cranberg continued his media campaign with mixed success.2 On the day of the Society meeting, before his presentation, Cranberg was interviewed by various news representatives. As one of these interview sessions was being conducted in the hallway outside the main meeting room, David Berliner, Director of the Office of Public Information for Consumers Union, approached Cranberg. Consumers Union had sent Berliner to the meeting to respond to Cranberg's paper. A verbal confrontation began between Cranberg and Berliner. During the commotion, a fellow Society member indicated to Cranberg that he was being tape recorded by Consumers Union. Although it is uncertain whether the tape recorder was turned on, Cranberg, believing it was, took the recorder and began recording a message to Consumers Union denouncing surreptitious recording. Berliner tried to take the recorder away from Cranberg. Both men tugged at the machine; eventually, one let go. Immediately following this confrontation, Cranberg presented his paper to the Society. Berliner followed with Consumers Union's statement. Both men were given opportunities for rebuttal.

Associated Press reporter, Chris Sullivan, was covering the Society meeting at the request of the New York AP office. Sullivan testified that he spoke to both Cranberg and Berliner before the meeting; he witnessed the struggle over the tape recorder, attended the meeting and heard the presentations by both Cranberg and Berliner. He interviewed both Cranberg and Berliner again after the meeting and was given written copies of the presentations made by each. Sullivan's story went over the AP wire and was published by the American-Statesman.

Following these events, Cranberg filed suit against Consumers Union, the AP and the American-Statesman. These suits were consolidated for trial and heard without a jury on the question of liability only. Dr. Cranberg appeared pro se at the trial and throughout this appeal. In oral findings of fact and conclusions of law issued at the conclusion of the trial, the court credited Sullivan's account of what occurred between Cranberg and Berliner. The court found that the AP story, as written by Sullivan and reprinted by the American-Statesman, was a fair and accurate rendition of what occurred, that Dr. Cranberg was a public figure for the purposes surrounding these events and that Cranberg failed to prove actual malice by either the AP or the American-Statesman.

The district court also found the statements made by Berliner before the American Physical Society were substantially true. In addition, the district court found Cranberg had voluntarily asserted himself into a public controversy and had attempted to influence its resolution, thus, making him a public figure within the limited context of the controversy with Consumers Union. Because Cranberg was a public figure, the court held he must prove actual malice which he failed to do.

As to the claims against Consumers Union, the court found Cranberg failed to disprove the accuracy of the tests reported in Consumers Report. The district court also found there was no malice.

In addition to appealing the libel findings, Cranberg asserts here that the trial judge was not an Article III judge, that the court abused its discretion in denying his motion to amend his complaint, and that he was denied a fair trial. We address the libel issues first.

These events have spawned four separate libel claims: 1) a claim of libel and product disparagement against Consumers Union for a review of the Texas Fireframe in the January 1981 issue of Consumers Report; 2) a claim of libel and slander against Consumers Union for statements made by its representative, David Berliner, before the American Physical Society; 3) a claim of libel against the AP for its wire story about the events at the Society meeting; and 4) a claim against the American-Statesman for libel in reprinting the AP story.

* We deal first with the claims against the AP and the American-Statesman for the alleged libel against Cranberg in the AP wire story. Cranberg's complaints against the articles were numerous. However, he objected most strenuously to the account of the struggle with Berliner over the tape recorder, which Cranberg denies ever occurred. He also objected to the headline, "Firegrate inventor fights with consumer group official," added by the American-Statesman. The court found both stories "fair and accurate rendition[s]" of what transpired.

The court's findings were based in part on credibility decisions made after hearing conflicting testimony as to whether a struggle over the tape recorder actually occurred. Sullivan testified that he witnessed a physical struggle between Cranberg and Berliner over the tape recorder. Berliner testified that he and Cranberg exchanged heated comments and struggled over the tape recorder. Cranberg disputed these accounts. He insisted no struggle occurred. He introduced deposition testimony of a reporter from Discover magazine and of the deputy executive secretary of the Society to corroborate his version of the events. Although both witnesses mentioned a lively verbal exchange, only the Society member positively stated he never saw Cranberg's hands on the recorder. The reporter from Discover said he could not recall seeing a struggle, but it could have happened while he was turned away. The court chose to believe Sullivan, an impartial eyewitness, who "called it like he saw it." The court found Sullivan had no motive to be untruthful or to distort his observations. Sullivan testified that he received no directives as to what he would encounter at the meeting or what to write; he had no preconceived ideas, and he lacked any ties or bias for or against Cranberg or Consumers Union. In making this credibility choice, the court stated that "when things happen in the heat of the moment, the participants, especially the one who seems the most aggrieved" is seldom the one best able to recollect what happened.

Testimony from the American-Statesman indicated the AP story was reprinted with only minor changes. The city editor of the American-Statesman testified that upon receiving the AP wire story he made some minor editorial changes. He made no effort to contact Cranberg to verify the story because he had an eyewitness account from the AP, a reliable news service, which he believed was accurate. He passed the story to the copy desk where it was given the headline "Firegrate inventor fights with consumer group official." Cranberg objected to the word "fights" as an irksome misrepresentation of the events. The court looked up the word in the dictionary and found it meant, among other things, "to oppose physically, to strive for the mastery of ... struggle, quarrel...." The court's acceptance of the AP article as accurate validates the almost identical reprinting by the American-Statesman and the added headline.

A reviewing court should respect credibility choices made by the fact finder. Levine v. CMP Publications, Inc., 738 F.2d 660, 672 n. 19, (1984), reh'g en banc denied, 753 F.2d 1341 (5th Cir. 1985). "[D]ue regard shall be given to the trial judge's opportunity to observe the demeanor of the witnesses." Bose Corp. v. Consumers Union, --- U.S. ----, 104 S. Ct. 1949, 1959, 80 L. Ed. 2d 502 (1984). The record discloses no basis for disturbing the credibility choice made below. Furthermore, a comparison of the article with the documents on which Sullivan relied reveals no inaccuracies. Thus, the trial court could have found the articles' rendition of the events were fair and accurate, and there is ample support for that conclusion. See Time, Inc. v. Firestone, 424 U.S. 448, 457-59, 96 S. Ct. 958, 967-68, 47 L. Ed. 2d 154 (1976).

Under Texas law, truth is an absolute defense in defamation actions. Tex.Rev.Civ.Stat.Ann. art. 5431 (Vernon 1958); Mitcham v. Board of Regents, University of Texas, 670 S.W.2d 371, 373 (Tex.App.--Texarkana 1984, no writ); Traweek v. Radio Brady, Inc., 441 S.W.2d 240, 243 (Tex.Civ.App.--Austin 1969, writ ref'd n.r.e.); Gulf Construction Co. v. Mott, 442 S.W.2d 778, 784 (Tex.Civ.App.--Houston [14th Dist.] 1969, no writ) (substantial, not literal, truth is sufficient). Because the articles were found to be true, and because truth is an absolute defense, it is unnecessary to reach the issues of whether Cranberg was a public figure or whether actual malice existed. We affirm the district court's findings as to libel by the AP and the American-Statesman.

Next, we deal with the alleged defamation resulting from Berliner's remarks before the American Physical Society. The complaint centers on Berliner's statement that "Dr. Cranberg demanded payment of $1000 a day in damages and a complete retraction. Even before we could respond to his claims, Dr. Cranberg warned us that if we failed to meet his demands, he would give a speech attacking us at this prestigious meeting." Cranberg asserts this was a misrepresentation of his actions which unfairly labeled him a blackmailer.

The court found Berliner's comments accurately described Cranberg's communications to Consumers Union as shown in letters written by Cranberg and admitted into evidence. This finding is amply supported by the evidence. A review of the record establishes that Berliner's comments were substantially accurate representations of Cranberg's written communications with Consumers Union. Again, because the statements were found to be true, and because truth is a defense, Tex.Rev.Civ.Stat.Ann. art. 5431 (Vernon 1958), we need not determine whether Cranberg was a public figure or whether actual malice was proven.

The claim against Consumers Union, for libel and product disparagement in its review of the Texas Fireframe, is a more complicated issue. It is uncertain whether Texas recognizes a cause of action for product disparagement or trade libel. Jetco Electronic Industries v. Gardiner, 325 F. Supp. 80, 84 (S.D.Tex.1971), rev'd on other grounds, 473 F.2d 1228 (5th Cir.1973). All actions for libel in Texas are governed by statute, which defines defamation only in terms of injury to a person.3 No other definition may be considered in determining whether a particular publication is libelous. Id.; Deen v. Snyder, 57 S.W.2d 338, 340 (Tex.Civ.App.--Fort Worth 1932, no writ). Although Cranberg was never mentioned by name in the Consumer Reports article, he may still bring a defamation action under Texas libel law. Texas courts have allowed such suits on the grounds that a false publication concerning a business or product could tend to injure the reputation of the company's owner by subjecting him to financial injury. Jetco, 325 F. Supp. at 84; see also Bell Pub. Co. v. Garrett Engineering Co., 141 Tex. 51, 170 S.W.2d 197, 202 (1943). In Newspapers, Inc. v. Matthews, 161 Tex. 284, 339 S.W.2d 890 (1960), the Texas Supreme Court held that under Texas libel law the owner of a business can be defamed, even though the publication does not mention him by name, when the asserted libel refers to some ascertained or ascertainable person. Id., 339 S.W.2d at 893. Therefore, it is not necessary that Cranberg be named in the article because those who knew and were acquainted with him understood from reading the publication that it referred to him. See id., 339 S.W.2d at 894. Cranberg was entitled to bring this libel suit in his own right. This case is governed by Texas libel law applied pursuant to the dictates of New York Times v. Sullivan, 376 U.S. 254, 84 S. Ct. 710, 11 L. Ed. 2d 686 (1964), and its progeny. We find no error in the legal standard applied below.

The district court ultimately found Consumers Union's tests and supporting data were reasonable; nothing pointed to their being unreasonable. The court acknowledged that the analysis done by Consumers Union could have been better, and that they could have performed more than one test for each device. However, no tests or evidence were admitted to refute Consumers Union's findings. Cranberg presented the testimony of two experts, a professor at the University of Texas and a consultant in explosives. Both said it is extremely difficult to obtain consistent results in measuring the heat output of fireplaces. They derided the fact Consumers Union tested each device only one time for each scenario. Although their testimony labeled Consumers Union's test results as unreasonable and totally meaningless, neither had performed any actual tests with the Texas Fireframe or "standard" fire. Cranberg introduced tests by an organization called Terralab whose heat output numbers on the Texas Fireframe were actually greater than Consumers Union's, but consistent when accounting for alternative testing factors. The court found the results of Consumers Union's tests were consistent with logic and observed that Consumers Union "are the only people that have tested [the Texas Fireframe] and given me the results and explained the test with engineers in this court room."Although suggesting that Consumers Union could have clarified certain aspects of the article, the court held the burden was on Cranberg to prove that the article was inaccurate and that he failed to carry that burden. Examples of statements pointed out by Cranberg which may have warranted further clarification include the fact Consumers Union's results refer to heat outputs in a log cabin type environment and in a centrally heated home. The article does disclose that the tests were not actually conducted in these environments, but rather under conditions which Consumers Union claimed simulated these scenarios. Although these conditions could have been more fully described if presented as a precise scientific experiment, the disclosure made in this layman's advisory publication obviously was considered sufficient by the trier of fact. That conclusion of fact is supported by the record. Cranberg criticizes the measurements obtained by Consumers Union as partly the result of circumstances it imposed. Consumers Union admitted at trial that individual consumers may observe different results under different circumstances, such as, using more or bigger logs (Consumers Union only used half of the fuel stipulated in the Texas Fireframe's instructions), outside temperature variations and height of the chimney stack. Furthermore, the "standard" fire, against which Consumers Union compared all its data, was never adequately described in the article so that it could be reproduced. The court held that neither these defects in the article, nor others mentioned in Cranberg's brief, impinged on the truth of the Consumers Report article. This ultimate finding is supported by the evidence.

The trial court had sufficient evidence before it to conclude that Consumers Union's tests were an accurate approximation of the performance of fireplace add-ons for consumers. Cranberg presented no scientific tests to compare the Consumers Union findings and refute their contention that the Texas Fireframe produces less heat than a fire built on the hearth. Furthermore, the trial court correctly noted that the article does not say the Texas Fireframe is a bad grate. To the contrary, the article praises its performance in comparison to other grates. "[F]or people who are going to use grates in their fireplaces, I think Consumers Union said yours is the best.... Consumers Union supports your claim that that is the best firegrate in the world. They just say that building the fire not on the grate but on the hearth is better." Dr. Cranberg testified as to the energy efficiency of his device by relying on his own reduced gas bills. This did not show that the Fireframe performs better than a grateless fire built on the hearth. In this regard, the district court aptly noted that it was Cranberg's practice to turn down the thermostat when using his Fireframe. The court found: "There has not been sufficient test in this record in my judgment--to test it against a fire built right on a hearth." No proof met this single, crucial point.

The record supports the district court's conclusion that Cranberg failed to prove falsity. Whether Cranberg was a public figure for purposes of the Consumers Report article is irrelevant to our review of this case, as Texas law holds substantial truth is a defense to a libel claim. Gulf Construction Co. v. Mott, 442 S.W.2d 778, 784 (Tex.Civ.App.--Houston [14th Dist.] 1969, no writ); Walker v. Globe-News Publishing Co., 395 S.W.2d 686, 689 (Tex.Civ.App.--Amarillo 1965, writ ref'd n.r.e.).

This case was tried on September 13 and 14, 1983. Two weeks later the trial judge announced his intention to leave the bench and return to private practice on January 1, 1984. Final judgment was entered in this case on October 19, 1983. Cranberg filed Motions for New Trial, New Judge and for Hearing of Motions by New Judge, arguing that the judge was no longer an Article III judge. The court denied these motions. Cranberg's challenge to this ruling is frivolous.

Because Cranberg is a pro se litigant, we have also looked for the substance of his claim, rather than to its label as an Article III concern. Cranberg argues that "lame-duck" judges are subject to the outside pressures that life-time tenure under Article III is designed to alleviate. Cranberg's theoretical assertion is unsupported by any proof that when the trial judge tried this case, or when he entered final judgment, any real conflict of interest existed.

Cranberg's claim seems to assert prejudice in the hearing and deciding of his case due to the fact that the judge was contemplating a return to private practice. Cranberg suggests the judge may have been prone to side with defendants' lawyers in order to curry favor with soon-to-be-colleagues of the bar. Even when viewed as a claim of prejudice, the record is wholly devoid of proof that the judge was false to his oath of office for any period of his service. No specific incident of prejudice or favortism is advanced, or can be found, to support this assertion. Our review of the entire record establishes that Cranberg lost a fairly tried case on an honest resolution of contested fact.

Cranberg's brief also evinces an unwarranted concern that his case was unfairly expedited to reduce the workload of an exiting judge. The announcement from the bench of findings and conclusions did include a statement that the court did not want to file written findings of fact or conclusions of law. This is not improper. The announced findings are supported by the evidence; they are complete in that they address all the questions at issue, and they are correct as to the legal standards applied. Cranberg received a proper decision promptly rendered. He is entitled to no more.

Cranberg claims the court conducted the proceedings so as to favor the defendants and deny him a fair trial. A number of trial procedures were unusual. For example, the judge requested that the two Consumers Union engineers who conducted the tests at issue be the first witnesses. He then directed the defendants to proceed first in examining these witnesses. The court asked all the witnesses a significant number of questions, at times dominating parts of the examination. According to Cranberg, the judge's participation in the trial comprises 23% of the transcript. At one point, the court announced, "Everyone has a running objection to hearsay and they are all overruled." On the other hand, the court also permitted the parties to make a liberal supplementation of the record upon conclusion of trial; Cranberg admits his supplementation was "massive." Rather than demonstrating prejudice, however, the record reflects that the judge's participation and evidentiary rulings tended to assist Cranberg, a pro se litigant who represented himself well, though he lacked familiarity with procedural rules.

The conduct of a fair trial is vested in the sound discretion of the trial judge. Excel Handbag Co. v. Edison Bros. Stores, 630 F.2d 379, 388 (5th Cir.1980). His duty extends far beyond ruling on objections and preserving decorum in the courtroom. Though the trial judge must be neutral, he need not be a passive spectator. Ruiz v. Estelle, 679 F.2d 1115, 1129, modified, 688 F.2d 266 (5th Cir.1982), cert. denied, 460 U.S. 1042, 103 S. Ct. 1438, 75 L. Ed. 2d 795 (1983). He has an affirmative obligation under Federal Rule of Evidence 611(a) to "exercise reasonable control over the mode and order of interrogating witnesses and presenting evidence so as to (1) make the interrogation and presentation effective for the ascertainment of the truth, (2) avoid needless consumption of time, and (3) protect witnesses from harassment or undue embarrassment." Furthermore, as a common law judge, he has an overall responsibility to see that the trial is just and not subject to delay. Ruiz, 679 F.2d at 1129. He may question witnesses, elicit facts, clarify evidence and pace the trial. Id. On review, this conduct will be measured against a standard of fairness and impartiality. Nordmann v. National Hotel Co., 425 F.2d 1103, 1109 (5th Cir.1970).

Greater latitude should be allowed in the conduct of a bench trial than would be permitted in a trial conducted with a jury. In the former there is no possibility that the judge's actions might improperly influence jurors. Ruiz, 679 F.2d at 1130. In certain other cases, the trial judge may need to assume a greater responsibility for the direction of the trial. See Ruiz, 679 F.2d at 1130 (complex and protracted trial). When a pro se litigant, even one as skillful as Cranberg, goes to trial against a party represented by a member of the bar, the responsibility of the trial judge may warrant participation which differs markedly from what would be appropriate to a trial between adversaries represented by counsel. The record of this trial reveals a judge conscious of his duty and conscientious in its discharge. Rather than prejudicing Cranberg's position, we see the judge's efforts as attempts to enlighten it. We are satisfied that the measures taken below to develop and control the proof were reasonable and directed towards reaching a just decision.

The trial court did not abuse its discretion in refusing to allow Cranberg to file a supplemental complaint against Consumers Union raising issues under the Lanham Act. 15 U.S.C. Sec. 1125(a). The trial court has discretion to grant or deny motions to amend the pleadings. Fed.R.Civ.P. 15(a). Cranberg's motion to supplement was filed five days after the trial on the merits was concluded. In exercising its discretion, the trial court may consider such factors as undue delay, prejudice to the opposing parties and futility of the proposed amendment. Foman v. Davis, 371 U.S. 178, 182, 83 S. Ct. 227, 230, 9 L. Ed. 2d 222 (1962); Gregory v. Mitchell, 634 F.2d 199, 203 (5th Cir.1981).

Undue prejudice resulting from delay in seeking leave to amend is sufficient grounds to deny the motion. Cranberg's post trial motion came more than 18 months after his original complaint, after discovery was closed and after all evidence was presented. See Addington v. Farmer's Elevator Mutual Insurance Co., 650 F.2d 663, 667 (5th Cir.), cert. denied, 454 U.S. 1098, 102 S. Ct. 672, 70 L. Ed. 2d 640 (1981). Cranberg's Lanham Act claims do not involve issues completely tried in the course of the trial. His motion did not seek to conform the pleadings to the proof. In substance, his proposed amendment would raise a new cause of action. Moreover, the proposed amendment appears futile as the Lanham Act requires false statements of fact about one's own product. See Bernard Food Industries, Inc. v. Dietene Co., 415 F.2d 1279, 1282-84 (7th Cir.1969); Markel v. Scovill Mfg. Co., 471 F. Supp. 1244, 1253 (W.D.N.Y.), aff'd mem., 610 F.2d 807 (2d Cir.1979). The court did not abuse its discretion in refusing to allow the commencement of another action at the conclusion of this one.

The findings of fact which support resolution of the libel issues are supported by the record. The remaining issues presented are without merit. The judgment appealed from is

AFFIRMED.