Treating the suggestion for rehearing en banc as a petition for panel rehearing, the petition for panel rehearing is DENIED. The judges in regular active service of this Court having been polled at the request of one of said judges and a majority of said judges not having voted in favor of it (Federal Rules of Appellate Procedure and Local Rule 35), the suggestion for Rehearing En Banc is DENIED.

E. GRADY JOLLY, Circuit Judge, with whom GEE, GARWOOD and JONES, Circuit Judges, join dissenting from denial of petition for rehearing en banc:

... Words strain,

Crack and sometimes break, under the burden,

Under the tension, slip, slide, perish,

Decay with imprecision, will not stay in place,

Will not stay still ...

T.S. Eliot, Burnt Norton, Part V.

Because I believe that Trevino has effectively rewritten the Supreme Court's test for government contractor immunity given in Boyle, I respectfully dissent from the denial of en banc review. Further, I believe that en banc review is necessary because Trevino is at odds with, if not in conflict with, this circuit's almost contemporaneous opinion in Smith v. Xerox, 866 F.2d 135 (5th Cir.1989), and that we have a responsibility to clarify ourselves. Instead, we leave the bench and bar to divide and argue over whether they can be reconciled, and if not, which will be followed as the law of the circuit.

* The United States Supreme Court, only less than a year ago specifically chose to use the word "approve" when it held that "Liability for design defects in military equipment cannot be imposed, pursuant to state law, when (1) the United States approved reasonably precise specifications." Boyle v. United Technologies, --- U.S. ----, 108 S. Ct. 2510, 2518, 101 L. Ed. 2d 442 (1980). In doing so, the Court selected a less demanding word than any of the synonyms it might have chosen:

Approve the most widely applicable, may indicate varying degress of admiration. Endorse (or indorse ), stronger than approve, implies expression of support, often by public statement. Sanction adds authorization, usually official, to approval. Certify and accredit imply official endorsement gained upon conforming to set standards. Ratify refers to making legal by formal official approval....

The American Heritage Dictionary (Second College Edition 1982). Yet the panel in Trevino finds that the term "approve" as used in Boyle actually means to establish reasonably precise specifications, id. at 1479; it actually means to "choose" a design feature, id. at 1480; it actually means to exercise judgment or policy choice, id. at 1484; it actually means to substantively review and evaluate, id. at 1487 n. 12; in short, it actually means to exercise a discretionary function within the strict meaning of the FTCA. We learn further in Trevino that one cannot "approve" a design by full awareness and acceptance of that design, defect and all, for thirteen years, id.; and finally, we learn that "approve" is a term that cannot be applied to work done by someone else, id. at 1486 ("If the government delegates its design discretion to the contractor and allows the contractor to develop the design, the government contractor defense does not apply....").

Indeed, the panel, in departing from any plain meaning of the word, dolorously observes that "the Supreme Court's use of the terms 'approve' and 'considered' is unfortunate," id. at 1481, n. 7. I would doubt, however, that the author of Boyle thinks it quite such a mishap. Mr. Justice Scalia was plainly aware of the precise requirements of exercising a discretionary function. He was aware that the "selection" of the appropriate design for military equipment is a discretionary function, 108 S. Ct. at 2517; and he was aware, as Justice Brennan reminded him, that the use of the word "approval" might be construed so broadly as to include rubberstamping, id. 108 S.Ct. at 2519. Yet, when he concluded the analysis for the majority, and defined the specific test to be applied, he chose the word "approve." He did not hold that the government must "select" the design, or "choose" the design, or "review and evaluate" the design, or "exercise its discretion," or any of the other words the panel seeks to substitute for "approve." All of the words that the panel would force on Mr. Justice Scalia, it should be added, were immediately before him as he defined the test, and he chose none of them.

Mr. Justice Scalia specifically explained that government "approval" assures that the design is within the area where the discretionary function policy would not be frustrated, because it assures that the design feature in question was considered by a government officer and not "merely by the contractor itself." Boyle, 108 S. Ct. at 2518. This explanation appears immediately after the use of the word "approved" and gives it its meaning. That government involvement be "within the area" does not mean that strict compliance with all of the elements of the discretionary function exception is required. Indeed, the Court indicated that the discretionary function exception only "suggests the outlines of 'significant conflict' between federal interests and state law in the context of government procurement." 108 S. Ct. at 2517.

I agree with the panel that "approve" is not synonymous with "rubber stamp." It does seem to me, however, that when the government furnishes 339 pages of specifications and involves a review team with the plans, "rubber stamp" is not an apt description of the government's involvement. Generally speaking, I would think that in this case if the government considered the plans and affirmatively assented to them, it has "approved"; or if the government, in the panel's words, was "charged with the knowledge of the defect" and accepted the design, especially for a period of years with no complaint, it has "approved" contract specifications within the meaning of Boyle. More specifically, I would suggest that an appropriate inquiry under the first Boyle prong is to ask: was the challenged feature disclosed with reasonable precision in the accepted design documents, and was the design accepted on behalf of the government by an officer who had, and knew he or she had, the duty and authority to decline to accept the challenged design feature on safety grounds if he or she deemed it appropriate to do so? I believe this test would be truer to the spirit and deliberately chosen language of Boyle.

I also dissent from the denial of rehearing en banc in this case because the panel opinion cannot comfortably coexist with our circuit's holding in Smith v. Xerox, 866 F.2d 135 (5th Cir.1989). Both cases were decided on the same day.

Both Smith and Trevino construe Boyle 's first prong of the government contractor defense: that the government have "approved reasonably precise specifications." The cases have different focuses, however. At issue in Smith was whether Xerox, the government contractor, was improperly granted summary judgment on its government contractor defense because it had not been able to produce its original specifications. In Smith, we looked at the language "reasonably precise," catalogued the information Xerox had produced, and found that from the evidence, the specifications that had been produced satisfied the "reasonably precise" requirements of Boyle. Because those specifications had been approved by the government, we found that Xerox had satisfied the first prong of the Boyle test.

It is true that superficially the holdings of Smith and Trevino do not appear to be inconsistent. In applying the first prong of the Boyle test and asking whether the government "approved reasonably precise specifications," we focused in Smith upon the requirement that the specifications be "reasonably precise," not that they be "approved," which is the focus of the Trevino opinion. If, however, Trevino had been applicable law, our analysis and holding in Smith would surely have been different, since the appeal was from the grant of summary judgment and there was little or no evidence in the record on how extensive the government's "substantive review and evaluation," i.e., approval, had been. Thus, although the opinions may be able to coexist, they can do so only because they were simultaneous.

Second, to the limited extent that Smith referred to "approval," we quoted the Ninth Circuit in McKay v. Rockwell International Corp., 704 F.2d 444 (9th Cir.1983) to the effect that approval meant "examining and agreeing" to a detailed description of the workings of the system. Id. at 138. Further, we referred to the Fourth Circuit's opinion in Boyle, and found that the Navy had "approved" reasonably precise specifications, where the contractor and Navy worked together to prepare detailed specifications, engaging in back and forth discussions, and where the Navy reviewed the mock-up and approved the design (all present in Trevino to some extent). Id. at 138.

Further, we held in Smith that the government had satisfied its burden on the first prong of the Boyle test, when the government "reviewed and approved Xerox final drawings and specifications." Id. at 138.

It is thus suggested in Smith that government review of the contractor's final precise specifications, followed by assent, is all that is required for contractor immunity, which is a considerably less stringent requirement than imposed by the Trevino panel. Thus, I would suggest that there is significant potential for conflict between Smith and Trevino. In any event, conflict or not, analytical comparison yields confusion. The bench and bar have a right to a clearer statement from us as to what Boyle means and how it should be applied.

Finally, I would observe that the Trevino panel's microscopic focus on the word "approval," unwarranted by Boyle, allows that term to eclipse the significant remainder of the phrase--"of reasonably precise specifications." The phrase "approved reasonably precise specifications" must be interpreted as a whole. It is the approval "of reasonably precise specifications" that assures that the government retains veto power over all design details and assures that "the suit is within the area where the policy of 'discretionary function' would not be frustrated." Boyle, 108 S. Ct. at 2518. If Trevino is sincere when it states that "[t]he requirement that specifications be precise means that the discretion over significant details and all critical design choices will be exercised by the government; if the government approved imprecise or general guidelines, then discretion over important design choices would be left to the government contractor," Trevino, at 1418, then it is not clear why the submission and simple approval of reasonably precise specifications cannot serve as a proxy for government involvement in the design decisions.

In sum, I do not think that it is necessary to overdefine the word "approved" in order to make certain in these cases that the government has sufficiently exercised its discretion and judgment over the design at issue. With great respect for those whose views seem to differ, because I think that this case is enbancworthy, I respectfully dissent from the failure of the court to grant en banc review.