HECHT, Justice, joined by PHILLIPS, Chief Justice, dissenting.

Maritime Overseas Corporation seeks a new trial because, while Richard Ellis was undeniably injured by his exposure to diazinon, the scientific evidence does not support the conclusion that he suffers from permanent neurotoxicity, and thus the $8,576,000 awarded him in damages is excessive. The Court holds that it could not order a new trial even if it agreed with Maritime Overseas' contention, completely ignoring its decision to grant a new trial in indistinguishable circumstances just one year ago in Texarkana Memorial Hospital, Inc. v. Murdock, 946 S.W.2d 836 (Tex. 1997). The Court also holds that Maritime Overseas failed to preserve its complaint for appeal because it did not object to Ellis's evidence at trial, even though Maritime Overseas' position has always been — in its opening statement, its extensive examination of the expert witnesses, its closing argument, its motion for new trial, and on appeal — that no reliable scientific evidence shows that diazinon can cause long-term neurotoxicity. As Ellis's attorney told the jury in his opening statement, Maritime Overseas' "position is that this chemical just cannot cause an injury to a worker's nervous system." Maritime Overseas' position has never been in doubt.

Not one case the Court cites so much as hints that a party in Maritime Overseas' circumstances has failed to preserve error, and one of those cases, Sumitomo Bank v. Product Promotions, Inc., 717 F.2d 215, 218 (5th Cir. 1983), actually suggests that Maritime Overseas has preserved its position. The Court refuses to acknowledge, much less reconcile, its own numerous precedents that require reversal of a judgment based on non-probative evidence, even though the evidence was admitted without objection. The Court appears to think that if it ignores these cases they will somehow go away. The Court steadfastly evades the one and only issue over which these parties have fought since the day this litigation began — whether there is reliable evidence that Ellis suffers from neurotoxicity. I would decide this issue; therefore I dissent.

I

It is undisputed that Ellis suffered some injury from his exposure to diazinon and should recover some damages, but it is equally undisputed that if he did not suffer long-term neurotoxicity, his damages are nowhere near $8,576,000. The court of appeals, in determining the factual sufficiency of the evidence, considered expert testimony that Ellis not only was injured but that he suffers from neurotoxicity. Maritime Overseas argues that evidence offered in support of Ellis's long-term injury claims is unreliable and therefore no evidence at all. Thus, Maritime Overseas contends that the court of appeals erred in considering such testimony in its factual sufficiency review. The Court correctly summarizes Maritime Overseas' argument: "In essence, Maritime would have this Court conduct a no evidence review of the evidence about delayed neurotoxicity within the Court's review of whether the court of appeals properly reviewed the factual sufficiency of the evidence." Ante at 412. Then the Court says: "We decline to do so." Id.

But the Court did not "decline to do so" last year in Texarkana Memorial Hospital, Inc. v. Murdock, 946 S.W.2d 836 (Tex. 1997). Murdock sued the Texarkana Memorial Hospital for negligence in delivering her daughter. The child was born with severe congenital defects and died about a year later. Murdock claimed that she was entitled to damages equal to all of the child's medical expenses, but the Hospital argued that Murdock could recover only for those expenses caused by its negligence, excluding expenses for treatment necessitated by the child's congenital defects. The district court awarded Murdock the total expenses, and the court of appeals affirmed, holding that legally and factually sufficient evidence supported the conclusion that all the medical expenses were caused by the Hospital's negligence. Texarkana Memorial Hosp., Inc. v. Murdock, 903 S.W.2d 868, 877-880 (Tex.App.-Texarkana 1995), rev'd, 946 S.W.2d 836 (Tex. 1997).In this Court, the Hospital argued that there was "no evidence of a direct causal link between the amount of medical expenses awarded and any injuries caused by [the Hospital's] negligence." Murdock, 946 S.W.2d at 837. We agreed and reversed the award, explaining:

[W]hile [there] is some evidence of damage caused by [the Hospital's] negligence, a plaintiff may recover only for those injuries caused by the event made the basis of suit. Morgan v. Compugraphic Corp., 675 S.W.2d 729, 732 (Tex. 1984). The case before us is analogous to other cases where a suit for medical expenses involved another injury or pre-existing condition. . . . We . . . hold that a plaintiff should recover only for medical expenses specifically shown to result from treatment made necessary by the negligent acts or omissions of the defendant, where such a differentiation is possible.

Id. at 839-840 (citation omitted). Although the Hospital couched its complaint in no-evidence terms, for which the remedy is ordinarily rendition of judgment, we concluded that "[b]ecause Murdock . . . presented legally sufficient evidence that some of the medical expenses resulted from [the Hospital's negligence], [she] should be afforded an opportunity to develop this evidence further." Id. at 841. Thus, we remanded the case for a new trial. In support of this conclusion we cited Stewart Title Guaranty Co. v. Sterling, 822 S.W.2d 1, 10-12 (Tex. 1991), in which we remanded a case for a new trial on attorney fees because the evidence supported an award of some fees for some claims, even though fees could not be awarded on all claims.

The present case is indistinguishable from Murdock. There, as here, the argument was that while some evidence showed some damages, no evidence supported all the damages awarded. Although the Hospital complained of the legal sufficiency of the evidence, it in effect challenged the court of appeals' factual sufficiency review for considering non-probative evidence, and we treated the complaint as being directed to that review, remanding for a new trial rather than rendering judgment for the Hospital. Maritime Overseas' application for writ of error states: "There is no evidence that diazinon causes delayed neurotoxicity and thus insufficient evidence that Ellis suffered $8,576,000 in actual damages." The arguments in the two cases, while phrased differently, are indistinguishable in import and effect.The arguments and the relief sought are the same in both.

Why isn't Murdock controlling or at least instructive? The Court refuses to answer, refuses even to cite Murdock. The argument that there is some significance in the Hospital's no-evidence challenge and Maritime Overseas' insufficient-evidence challenge is too weak even for the Court to employ. If anything, Maritime Overseas' contention that the evidence of damages is insufficient because there is no evidence of some damages awarded is more straightforward than the Hospital's contention that there was no evidence of the damages awarded because there was some evidence of only lesser damages. But in fact, both arguments come out at the same place, in substance — some but not all of the damages are supported by the evidence — and in result — a new trial excluding the unsupported claims. Maritime Overseas' first point of error in this Court asserts: "The court of appeals erred in failing to examine whether any well-founded scientific methodology supports the award of . . . actual damages." Even if Maritime Overseas could be faulted for misphrasing its point of error, that mistake cannot dictate the result in the case.

A point of error "is sufficient if it directs the attention of the appellate court to the error about which complaint is made." Courts are to construe rules on briefing liberally. An appellate court should consider the parties' arguments supporting each point of error and not merely the wording of the points.

Anderson v. Gilbert, 897 S.W.2d 783, 784 (Tex. 1995) (per curiam) (citations omitted). Maritime Overseas' argument in its application for writ of error is crystal clear:

In this case, Ellis offered no epidemiological study, no peer-reviewed theory, nor any evidence of general scientific acceptance to support the conclusion of his experts that his exposure to diazinon caused delayed neurotoxicity. The premise upon which his experts' conclusion was based — that because some organophosphates can cause delayed neurotoxicity, diazinon therefore must cause delayed neurotoxicity — is false logic, as pointed out by Justice Robertson's concurring and dissenting opinion, because some organophosphates do not cause delayed neurotoxicity.

To make the matter even clearer, Maritime Overseas summarizes its position thusly: "There is no evidence that diazinon causes delayed neurotoxicity and thus insufficient evidence that Ellis suffered $8,576,000 in actual damages."

The result in Murdock was correct, and the same analysis should be applied in this case. A party must have a means of contesting the amount of damages when there is evidence for some claims but not all of them. Following Murdock, Maritime Overseas is entitled to a new trial if its evidentiary complaint has been preserved and has merit. The Court holds that Maritime Overseas' complaint was not preserved and does not reach the merits.

II

As early as 1912, and as recently as last year, this Court has held that a party may complain after verdict and on appeal that evidence admitted without objection is neither legally nor factually sufficient to support the verdict. The Court ignores a solid line of cases establishing this principle with respect to all kinds of evidence, including scientific testimony. There is no authority for the Court's holding that "[t]o preserve a complaint that scientific evidence is unreliable and thus, no evidence, a party must object to the evidence before trial or when the evidence is offered." Ante at 409. The notion that a party must as a matter of course object to evidence before trial is a complete stranger to our procedure. Despite this lack of authority, it seems clear that parties should be required to contest the reliability of scientific testimony in some way prior to the verdict in most instances. However, Maritime Overseas did so in this case.

A

As a rule, a contention that evidence is insufficient to support a judgment need not be raised before the verdict. Rule 279, TEX.R. CIV. P., states: "A claim that the evidence was legally or factually insufficient to warrant the submission of any question may be made for the first time after verdict, regardless of whether the submission of such question was requested by the complainant." Prior to the verdict, a party may, but is not required to, raise the complete absence of evidence on a point. This differs from federal procedure, which requires that a motion for judgment as a matter of law be made before the case is submitted to the jury "to assure the responding party an opportunity to cure any deficiency in that party's proof that may have been overlooked until called to the party's attention". Fed.R.CIV. P. 50(a)(2) advisory committee's note. Texas procedure does not afford parties the same protection. Thus, for example, a defendant sued for reasonable and necessary expenses can wait until after the verdict to point out that the plaintiff never offered evidence that the expenses claimed were reasonable. See McCreless Properties, Ltd. v. F. W. Woolworth Co., 533 S.W.2d 863, 868 (Tex.Civ.App.-San Antonio 1976, writ ref'd n.r.e.); Holt v. Purviance, 347 S.W.2d 321,324-325 (Tex.Civ.App.-Dallas 1961, writ ref'd n.r.e.). A defendant sued for attorney fees may wait until after the verdict to assert that no evidence of the required presentment of the claim was offered. See Jim Howe Homes, Inc. v. Rogers, 818 S.W.2d 901, 905 (Tex.App.-Austin 1991, no writ). A pre-verdict objection to the factual insufficiency of the evidence cannot preclude submission to the jury of pleaded claims, Brown v. Goldstein, 685 S.W.2d 640, 641 (Tex. 1985), and thus has essentially no effect.

Even if evidence is admitted without objection, it may be insufficient to support a judgment. This Court held eighty-six years ago that "incompetent testimony can never form the basis of a finding of facts in an appellate court, notwithstanding its presence in the record without objection." Henry v. Phillips, 105 Tex. 459, 151 S.W. 533, 538 (Tex. 1912). In that case, testimony admitted without objection was held to be no evidence on appeal because it was hearsay. Id. at 537. The Court repeatedly treated hearsay as no evidence even if it was not objected to, until Rule 802 of the Texas Rules of Civil Evidence was adopted in 1983. Zobel v. Slim, 576 S.W.2d 362, 369 (Tex. 1978); Cooper Petroleum Co. v. LaGloria Oil Gas Co., 436 S.W.2d 889, 891 (Tex. 1969); Aetna Ins. Co. v. Klein, 160 Tex. 61, 325 S.W.2d 376, 379 (Tex. 1959); City of Mission v. Popplewell, 156 Tex. 269, 294 S.W.2d 712, 717 (Tex. 1956); Texas Co. v. Lee, 138 Tex. 167, 157 S.W.2d 628, 631 (Tex. 1941). But the principle in Henry has been applied to evidence other than hearsay.

In Casualty Underwriters v. Rhone, 134 Tex. 50, 132 S.W.2d 97 (Tex. 1939), Rhone sought compensation for injuries sustained while working on a construction site. The dispute centered on whether at the time of his injuries he was employed by the general contractor, Beaumont Development Corporation, or a subcontractor, McDaniel. The jury found that Rhone was employed by the general contractor, but the court of civil appeals reversed, holding as a matter of law that Rhone was employed by the subcontractor. We affirmed the court of civil appeals, holding that testimony by Rhone and McDaniel contrary to its conclusion, though not objected to, was no evidence.

The only testimony in the record which would in the least tend to support the conclusion that Rhone was working for the Beaumont Development Corporation was given by Rhone and McDaniel, each of whom testified that, at the time of the injury, Rhone was working for it. Those statements did not amount to any evidence at all. They were but bare conclusions and therefore incompetent, and the fact that they were admitted without objection adds nothing to their probative force.

Id. at 99.

The Court followed Rhone in Dallas Railway Terminal Company v. Gossett, 156 Tex. 252, 294 S.W.2d 377 (Tex. 1956). In that case, a bus passenger, Gossett, recovered damages for injuries she sustained when the bus struck a car. The bus company, Dallas Railway, impleaded the driver of the car, Sample, contending that her negligence in driving the wrong way on a one-way street caused the accident. The jury failed to find Sample negligent. On appeal, Dallas Railway argued that the evidence established Sample's negligence because it was undisputed that she was driving the wrong way on a one-way street. The bus driver, Gossett, Sample, and an accident investigator all testified that they believed traffic on the street was one-way, but no evidence was offered showing that traffic was legally restricted. The Court held that the witnesses' testimony did not establish that the street was one-way, explaining: "It is well settled that the naked and unsupported opinion or conclusion of a witness does not constitute evidence of probative force and will not support a jury finding even when admitted without objection." Id. at 380-381.

Two cases cited by Gossett with approval apply the same principle in other settings. In one, Webb v. Reynolds, 207 S.W. 914 (Tex. Comm'n App. 1919, judgm't adopted), the court held that a plaintiff's testimony that he owned a promissory note was no evidence to support his claim because the statement "was a bare conclusion or opinion of the witness without any basis of fact". Id. at 916 Plaintiff's own pleadings asserted that the note was owned by an estate. Id. The court added: "The fact that [the testimony] was not objected to could add nothing to its probative force." Id. In the other, Perren v. Baker Hotel, 228 S.W.2d 311 (Tex.Civ.App.-Waco 1950, no writ), the court held that a wife's testimony that her husband had agreed to rent hotel rooms "was nothing more than a bare conclusion on the part of the witness concerning a question of law and such testimony had no probative force, even though it had been admitted without any objection." Id. at 317.

In Robertson Tank Lines, Inc. v. Van Cleave, 468 S.W.2d 354 (Tex. 1971), this Court held that a plaintiff's testimony that he was acting in the course and scope of his employment at the time he was injured was no evidence to support a finding to that effect. Even though the testimony was admitted without objection, it was attacked in cross-examination. The Court stated:

This court has approved the holding that testimony of an employee (driver) that he was acting within the course of his employment at the time of an accident is not admissible.If such testimony is admitted, with or without objection, it has been held to be incompetent and without probative force.It will not support a verdict or a finding of fact by a court.

Id. at 360 (citations omitted).

In Schaefer v. Texas Employers' Insurance Association, 612 S.W.2d 199 (Tex. 1980), Schaefer claimed compensation benefits, alleging that he suffered from an occupational disease, atypical tuberculosis. The carrier disputed that Schaefer contracted his disease at work. His treating physician, Dr. Anderson, testified "that in his opinion, based on reasonable medical probability, Schaefer's disease resulted from his employment." Id. at 202. The defendant attacked Dr. Anderson's opinion on cross-examination but did not object to its admission. The jury found for Schaefer, but the court of civil appeals reversed and rendered judgment for the carrier. This Court affirmed, refusing to take Dr. Anderson's opinion at face value and looking instead to the basis for it. The Court explained:

The basis for [Dr. Anderson's] opinion is that persons engaged in "dirty" occupations, such as farmers, tend to have a greater exposure to the bacteria; that Schaefer frequently worked in soil contaminated by bird droppings; that Schaefer suffers from one of the serotypes of m. intracellularis; and, therefore, he has an occupational disease. Notwithstanding Dr. Anderson's opinion, there is a crucial deficiency in the proof of causation. The evidence fails to establish that any bacteria was present in the soil where Schaefer worked.

Id. at 203. After quoting extensively from Dr. Anderson's testimony, the Court continued that his opinion was no evidence of the cause of Schaefer's disease because it lacked any real basis:

Dr. Anderson assumes that Schaefer is infected with an avian serotype m. intracellularis pathogenic to fowl. He further assumes that this serotype was present in bird droppings where Schaefer worked. It is admitted that the particular strain of m. intracellularis from which Bobby Schaefer suffers has not been identified. It is also admitted that the manner in which the disease was transmitted to Schaefer is unknown. It is further admitted that there is no evidence that the bacteria is present in the soil where Schaefer worked, or even in Nueces County.

We have reviewed the substance of Dr. Anderson's testimony in its entirety and we find that it does no more than suggest a possibility as to how or when Schaefer was exposed to or contracted the disease. We hold that his opinion is not based upon reasonable medical probability but relies on mere possibility, speculation, and surmise. We hold there is no evidence that the disease suffered by Bobby Schaefer is an occupational disease "arising out of and in the course of employment." The fact that proof of causation is difficult does not provide a plaintiff with an excuse to avoid introducing some evidence of causation. To ignore the substance of Dr. Anderson's testimony and accept his opinion as "some" evidence simply because he used the magic words "reasonable probability" effectively removes this Court's jurisdiction over any case requiring expert opinion testimony. Under such view, so long as an expert states the words "reasonable probability," in giving his opinion, there would be some evidence.The question would then be solely one of sufficiency of the evidence over which this Court has no jurisdiction.

Id. at 204-205 (citations omitted).

We reaffirmed Schaefer in Burroughs Wellcome Company v. Crye, 907 S.W.2d 497 (Tex. 1995). In that case plaintiff Crye's treating physician, Dr. Blesius, testified without objection that Polysporin sprayed on Crye's foot caused frostbite. The jury found for Crye, and the court of appeals affirmed, concluding that the evidence was factually and legally sufficient to support the verdict. Burroughs Wellcome Co. v. Crye, 912 S.W.2d 251, 259 (Tex.App.-El Paso 1994), rev'd, 907 S.W.2d 497 (Tex. 1995). We reversed, despite the admission of Dr. Blesius' testimony without objection, because his opinion had no factual basis. We stated:

We hold that Dr. Blesius' testimony constitutes no evidence that Polysporin spray caused Crye to sustain a frostbite injury. When an expert's opinion is based on assumed facts that vary materially from the actual, undisputed facts, the opinion is without probative value and cannot support a verdict or judgment. See Schaefer v. Texas Employers' Ins. Ass'n., 612 S.W.2d 199, 202-05 (Tex. 1980) (reviewing substance of medical expert's testimony and holding that this testimony constitutes no evidence of causation, as it is based on assumptions, possibility, speculation, and surmise).

Id. at 499-500 (citation omitted).

Just last year in Merrell Dow Pharmaceuticals, Inc. v. Havner, 953 S.W.2d 706 (Tex. 1997), we reiterated that "an expert's bare opinion will not suffice" to provide evidence of causation of an injury; "[t]he substance of the testimony must be considered." Id. at 711. Merrell Dow asserted in the trial court that scientific evidence of any causal connection between the use of Bendectin and birth defects was unreliable, and it "objected to the admission of some, but not all, of this evidence." Id. at 709. We held that the expert testimony, even that admitted without objection, was no evidence to support a judgment for Havner because the testimony showed that there was no basis for the experts' opinions. We said: "When the expert `br[ings] to court little more than his credentials and a subjective opinion,' this is not evidence that would support a judgment." Id. at 712 (citation omitted). We added:

Justice Gonzalez, in writing for the Court, gave rather colorful examples of unreliable scientific evidence in E.I. du Pont de Nemours Co. v. Robinson, 923 S.W.2d 549, 558 (Tex. 1995), when he said that even an expert with a degree should not be able to testify that the world is flat, that the moon is made of green cheese, or that the Earth is the center of the solar system. If for some reason such testimony were admitted in a trial without objection, would a reviewing court be obliged to accept it as some evidence? The answer is no. In concluding that this testimony is scientifically unreliable and therefore no evidence, however, a court necessarily looks beyond what the expert said. Reliability is determined by looking at numerous factors including those set forth in Robinson and [ Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 113 S.Ct. 2786, 125 L.Ed.2d 469 (1993)]. The testimony of an expert is generally opinion testimony. Whether it rises to the level of evidence is determined under our rules of evidence, including Rule 702, which requires courts to determine if the opinion testimony will assist the jury in deciding a fact issue. While Rule 702 deals with the admissibility of evidence, it offers substantive guidelines in determining if the expert testimony is some evidence of probative value.

Id.(emphasis in original).

Within the past few months we denied the application for writ of error in Williams v. Gaines, 943 S.W.2d 185 (Tex.App.-Amarillo 1997, writ denied).In that case, Gaines sued Williams for removing her as president of a corporation in which he was sole shareholder and terminating her employment with the corporation. The jury found that Williams breached his agreement with Gaines and that her damages included $92,500 as the value of the stock as of a specific date that Williams promised Gaines but did not convey. The court of appeals reversed the judgment for Gaines and remanded the case for a new trial, holding that there was no evidence to support the jury's damages finding. Gaines and an expert witness had testified without objection to the value of the stock based solely on data after the date at issue. The court concluded: "Because the data relied upon by Ms. Gaines to support the jury's award is based on subsequent data, there was no probative evidence of the fair market value of one-half of the [corporation's] stock on [the specified date]". Id. at 193. The court explained: "Opinion evidence based on conjecture or speculation lacks probative value. Incompetent evidence, even if not objected to at trial, may not be considered as probative in determining the legal and factual sufficiency of the evidence." Id. (citation omitted).

To summarize, bare conclusions and assertions unsupported by facts of record, expert opinions based on facts merely assumed and not proved, or facts different from those proved, and scientific testimony without any reliable basis, even if admitted without objection, are no evidence to support a finding of fact. An expert's opinion that disease was contracted through working conditions, or that a spray caused frostbite, or that a medication caused birth defects, even if admitted without objection, is not probative evidence if the testimony shows that the opinion lacks any substantial basis. This is not to say that the deficiency in the evidence need not be pointed out in any way before the verdict, but only that it can be done by cross-examination and means other than objections.

B

The Court holds: "To preserve a complaint that scientific evidence is unreliable and thus, no evidence, a party must object to the evidence before trial or when the evidence is offered." Ante at 417. Whatever the Court means by objecting to evidence before trial, the four cases the Court cites as authority do not support this holding. The first case, Robinson, does not consider the issue. In that case, the subject evidence was objected to and excluded by the trial court. Whether any objection was necessary was never addressed by this Court. In the second case, Havner, we stated quite plainly that objection was made to the admission of "some, but not all" of the evidence at issue. "[T]he question of scientific reliability was raised repeatedly", but not consistently by objection. Havner, 953 S.W.2d at 709.

The other two cases, Marbled Murrelet v. Babbitt, 83 F.3d 1060 (9th Cir. 1996), cert. denied, ___ U.S ___, 117 S.Ct. 942, 136 L.Ed.2d 831 (1997), and Sumitomo Bank v. Product Promotions, Inc., 717 F.2d 215 (5th Cir. 1983), the Court cites for the proposition that "[w]ithout requiring a timely objection to the reliability of the scientific evidence, the offering party is not given an opportunity to cure any defect that may exist, and will be subject to trial and appeal by ambush." Ante at 409. There are two flaws in the Court's reliance on these cases. First, as noted earlier, Texas procedure allows the sufficiency of the evidence to be challenged for the first time after verdict, whereas federal procedure does not. Thus, Texas procedure allows for some ambush that federal procedure precludes. Second, Sumitomo Bank holds only that in determining whether there is no evidence to support a finding such that judgment should be rendered notwithstanding the verdict, evidence ruled admissible cannot be excluded from consideration. See also Schudel v. General Elec. Co., 120 F.3d 991, 995 (9th Cir. 1997) ("when deciding a motion for JNOV, a trial court may not ignore evidence erroneously admitted at trial,[because] excluding evidence after the verdict is unfair to a party who may have relied on the determination that the evidence was admissible.").While this reasoning applies in deciding whether to render judgment notwithstanding the verdict, it does not apply in deciding whether to grant a new trial. As the court explained in Sumitomo Bank:

The trial judge erred in retroactively striking the summary exhibits and then gauging the jury's performance on the fictive basis that the summary evidence was not before it. Although acceptable in the context of a motion for new trial, see Montgomery Ward Co. v. Duncan, 311 U.S. 243, 61 S.Ct. 189, 85 L.Ed 147 (1940), this methodology is not appropriate in connection with a motion for judgment n.o.v.

717 F.2d at 218 (emphasis added). As the court noted, the Supreme Court explained the difference between motions for judgment n.o.v. and motions for new trial in Montgomery Ward:

Each motion, as the rule recognizes, has its own office. The motion for judgment cannot be granted unless, as matter of law, the opponent of the movant failed to make a case and, therefore, a verdict in movant's favor should have been directed. The motion for a new trial may invoke the discretion of the court in so far as it is bottomed on the claim that the verdict is against the weight of the evidence, that the damages are excessive, or that, for other reasons, the trial was not fair to the party moving; and may raise questions of law arising out of alleged substantial errors in admission or rejection of evidence or instructions to the jury.

311 U.S. at 251, 61 S.Ct. 189 (emphasis added).

Maritime Overseas contends here that it is entitled to a new trial, not that judgment should be rendered in its favor. Thus, the Court's reasoning, and the cases it cites, are inapposite. Our rules of procedure do not require a party to assert before the verdict that the evidence is insufficient to support a verdict. The factual sufficiency of the evidence may always be attacked post-verdict, even if no objection was made to its admissibility. Indeed, as the Supreme Court observed, one consideration in deciding whether to grant a new trial is whether there were substantial errors in the admission or rejection of evidence. As already demonstrated, our own precedents permit evidence to be rejected post-verdict as non-probative in at least some instances, even if it was admitted without objection.

C

The Court holds that the reliability of scientific evidence must be objected to before trial or when the evidence is offered. How one objects to evidence before trial is not entirely clear. The Court mentions Merrell Dow's motion for summary judgment and motion in limine in Havner, suggesting that these are ways in which scientific evidence can be challenged. As already noted, the Court states that "Merrell Dow objected to the admission of the Havners' scientific evidence", ante at 411, but this is only partly true. Merrell Dow only objected to some of the Havners' evidence. Had Merrell Dow been foreclosed from attacking the reliability of evidence to which it did not object, there would have been evidence to support the verdict. Thus, the Court's holding that no evidence supported the verdict was despite the absence of objections.

The Court states that Havner "emphasized that the offering party should be allowed the opportunity to `pass muster' under a trial court Robinson challenge — `to present the best evidence available' — before an appellate court considers whether legally sufficient evidence supports a judgment." Ante at 411. What Havner actually said was:

In sum, we emphasize that courts must make a determination of reliability from all the evidence. Courts should allow a party, plaintiff or defendant, to present the best available evidence, assuming it passes muster under Robinson, and only then should a court determine from a totality of the evidence, considering all factors affecting the reliability of particular studies, whether there is legally sufficient evidence to support a judgment.

953 S.W.2d at 720. The point was, as we said, that the reliability of scientific evidence must be determined from a review of all the evidence, not simply the evidence of one party or the other. Only by alchemy can this passage be turned into a requirement that evidence be objected to before its reliability can be determined.

The Court does not explain the holding in Schaefer and other cases cited above, where evidence was held to be non-probative even though it had been admitted without objection. Instead, the Court refers vaguely to a pretrial " Daubert/Robinson-type hearing." Ante at 411. The Court does not explain what kind of hearing this is, how it is invoked, when it is to be conducted relative to the commencement of trial, and whether it is required.

Our precedents seem to teach that parties should not be permitted to attack evidence for the first time after the verdict unless it is plainly without probative value — such as an opinion based on the moon's being made of green cheese, or a mere assertion that a person is another's employee, or that a person was injured in the course of work, or that a person made an agreement. In most situations, however, if the probative value of evidence is to be in question, then ordinarily the issue must be raised before the verdict. This prevents the ambush that concerns the Court and puts both parties and the trial court on notice of the contentions in the case. But it hardly makes sense to require a specific objection to each line of scientific opinion testimony when a party's stated, clear position is that the opinion is baseless. In Schaefer, for example, the carrier's position was plain from its cross-examination of the claimant's physician: his opinion that the claimant contracted atypical tuberculosis at work had no basis in fact. Likewise, in Havner, there could be no mistake that Merrell Dow's position throughout, as in all the other Bendectin cases previously tried, was that there was no reliable evidence that Bendectin caused birth defects.

In the case before us, there was never any doubt about Maritime Overseas' position. In his opening statement, Ellis's attorney told the jury:

The attorney representing the company told you yesterday that — well, their position is that this chemical just cannot cause an injury to a worker's nervous system. That's just not true. In fact, you'll hear evidence from the witnesses that it can cause an injury if it is — if the exposure is sufficiently great and if the exposure is on the order of what this man was exposed to.

Maritime Overseas' counsel responded in his opening statement:

[W]e think the medical evidence will show that the effects of diazinon are not long-term but, indeed, are confined within a specific period of time. Certainly no more than months.

And the evidence will show, and we'll bring in a toxicologist and a neuropsychologist who will testify that there is no relationship between the current situation exhibited by Mr. Ellis in the exposure to diazinon on the ship in 1982.

The dispute over this issue pervaded the examination and cross-examination of the eight expert witnesses. The focus of all the testimony was not on Ellis's initial poisoning from his exposure to diazinon, but whether he suffered any long-term injury. The possibility that diazinon causes neurotoxicity was thoroughly explored, and Maritime Overseas established that no studies or other evidence exist to support the opinions of Ellis's experts that he suffered from neurotoxicity caused by exposure to diazinon.

In summation, Ellis's counsel again addressed the issue:

I acknowledge that the difficulty I have labored under is that you cannot show clearly a damage to the central nervous system. Nobody can, but that doesn't mean you don't have a right to be treated fairly when you have it.

Maritime Overseas' counsel stressed in summation:

There wasn't a single article out of all the articles that we all went over bit by bit, line by line. Not a single one ever says that diazinon causes these sort of effects [ i.e., [neurotoxicity]. Not one.

* * *

There's an article and it's Defendant's Exhibit No. 3. I want you to look on page 149 of that article, in particular. It's an article written by Al Johnson together with Dr. Lassetor and two other people. And one of the conclusions of that article is that pesticides — some pesticides have neurotoxic effects, yes.It doesn't mention diazinon. . . . And the reason is because all organophosphates are different. Some are nerve gas, some kill people, some are insecticide. There's not a single article anywhere that says diazinon causes these effects.

* * *

We have never taken the position that Mr. Ellis did not have acute symptoms due to exposure of the diazinon. Where the case differs and where we differ from the plaintiff is whether Mr. Ellis's current complaints are a result of the exposure to diazinon. Does he have long-term, delayed neurotoxicity as a result to the exposure to the diazinon. That's the key issue in this case. All these other issues that you have to answer, especially the ones relating to damages, to medical expenses, to loss of wages, it all falls from that decision that you have to make.

* * *

We have had article after article referred to, that have all been discussed, organophosphate poisoning and the effects of organophosphate poisoning. We've tried to show — and I've been accused of nitpicking for doing it — that each article relied on . . . doesn't support a determination that exposure to diazinon does cause long-term delayed neurotoxicity, period. It didn't support it. And what the plaintiff has tried to do is say the literature talks about organophosphate exposure, diazinon is an organophosphate, therefore this has got to be it. . . .

(Emphasis added.)

The Court states that to determine now whether Maritime Overseas' scientific evidence was unreliable "would base appellate review on a record that was not made." Ante at 411. That simply is not true. Maritime Overseas did not ambush Ellis on the substance of the expert testimony. The record shows that it was, in counsel's words, "the key issue" in the case. The parties purported to present all available evidence on the issue whether diazinon could cause neurotoxicity. This is not a case where a party could have offered more or different scientific evidence had it known that its opponent objected to the evidence as unreliable. Maritime Overseas reasserted its contentions in its motion for new trial and on appeal. There can be no question that Maritime Overseas challenged the reliability of Ellis's scientific evidence.

D

The Court does not attempt to argue that Ellis's evidence had any probative value. It holds that even if the evidence had no probative value, it must be considered some evidence to support the judgment on appeal if it was not objected to. This holding is squarely contrary to Schaefer, Crye, Havner, and the other cases I have cited. The Court has two responses.

First, the Court says that to allow an argument that scientific evidence admitted without objection was nevertheless unreliable and non-probative would "take away the trial court's gatekeeping function" and thus would:

usurp the orderly and efficient disposition of appeals, deprive the proffering party of an opportunity to cure any defects in its evidence that the objecting party might pose, and in some cases, place appellate courts in the undesirable position of making decisions about evidentiary reliability absent a fully developed record.

Ante at 412. Of course, none of these evil effects is present in this case. Ellis not only understood Maritime Overseas' position and had every opportunity to cure the defects in his evidence, he and Maritime Overseas purported to offer all the evidence in existence on whether diazinon can cause neurotoxicity. There can be no question in this case that the record was fully developed. To say that a review of the sufficiency of evidence admitted without objection deprives the trial court of its gatekeeping function is to say that Schaefer, Crye, and Havner were wrongly decided. In Schaefer, for example, a physician testified, just as in the present case, that the plaintiff's injury was caused by a particular agent. Defendant did not object to this testimony. Still, this Court held that the evidence had no probative value because there was nothing in the record to indicate that the injury could have occurred as the witness testified. The witness's mere opinion was not enough to support a judgment. The same situation is present in this case, except that here the parties clearly made every effort to produce all available evidence, whereas that is not at all clear in Schaefer.

Second, the Court says that the cases I have cited — it refers to none of them by name — are distinguishable because "those cases involve no evidence challenges where, on the face of the record, the evidence lacked probative value. . . . In contrast, by its own admission, Maritime is not making a no evidence complaint." Ante at 412. I have already explained that Maritime Overseas' complaint is really that there is no evidence of some damages, and that the Court's effort to categorize Maritime Overseas' position more rigidly is unfair to the arguments made in its briefs. But assume that all the cases I have cited involved no-evidence challenges and that this case does not. What possible difference can that make to the Court? Why is the necessity of objection to the evidence less important when the appellate complaint is no evidence? As the Court's own authority, Sumitomo Bank, points out, the necessity of objection is more important when the complaint is that there is no evidence to support a judgment and therefore judgment should be rendered in the complainant's favor. When the request is only for a new trial, a reassessment of evidence admitted without objection is "acceptable". Sumitomo Bank, 717 F.2d at 218. Moreover, the trial court's gatekeeping function which the Court argues must be preserved is "take[n] away", ante at 412, just as effectively in a no-evidence appeal.

The Court's attempts to distinguish Havner, Crye, Schaefer, and the long line of cases that precedes them are flawed.

E

The use of scientific evidence at trial poses unique problems. Sometimes, as in Havner, the entire body of evidence is unreliable from a scientific viewpoint. At other times, as in Crye and Schaefer, the evidence is unreliable because it is based on assumptions that cannot be demonstrated. In still other cases, like this one, the evidence is unreliable only as it pertains to a part of the claims. For the most part, I agree with the Court that the issue of the reliability of scientific evidence should be raised in the trial court. The exception is when the evidence is plainly lacking in probative value — the moon is made of green cheese. But it is not at all clear what procedures should be used to raise reliability challenges. The Court refers to motions in limine, although as a general rule rulings on such motions do not preserve error. The Court also refers to summary judgments, although this procedure may not work well when testimony is important to illuminate the issue. The Court insists that there be an objection, but Havner shows the difficulty of objecting to an entire case. Moreover, once the issue has been identified, why should further objection be necessary?

For over two years, the Supreme Court Advisory Committee, which advises the Court on all rules of procedure, and the State Bar of Texas Committee on the Administration of Rules of Evidence, which monitors the operation of the Rules of Evidence, have tried to fashion rules governing the timing and manner of objections to scientific evidence. The seventy-plus members of these highly respected committees have broad experience and expertise in procedural and evidentiary matters. Last fall the Advisory Committee, after considering the work of the State Bar Committee, concluded that the problem of how and when to object to scientific evidence is complex and involves many difficult considerations. The Advisory Committee recommended to this Court that any rules await a development of the issues in appellate opinions carefully analyzing the various concerns. That counsel seemed sound at the time, but today's confusing opinion makes the alternative of a rules solution far more appealing.

In simply mandating an objection before or during trial, the Court appears oblivious to the considerations its advisory committees believed to be complex and difficult. The Court's analysis is really confined to a single thought: parties should not be "ambushed". That relatively innocuous proposition simply cannot support the addition to our procedural jurisprudence of a vague and universal duty to object to scientific evidence before or during trial.

III

Maritime Overseas' challenge to Ellis's scientific evidence is valid. Although Ellis's experts testified that Ellis's exposure to diazinon caused neurotoxicity, there was no basis for their opinions in any scientific literature or experimentation. The experts reviewed all the literature regarding neurotoxicity from exposure to pesticides in general and organophosphates in particular; none was omitted. Nowhere in the literature is there any demonstration that diazinon causes neurotoxicity.

Ellis's position is that diazinon is an organophosphate, some organophosphates cause neurotoxicity (although some do not), and therefore diazinon causes neurotoxicity. The logical fallacy in this syllogism is apparent. The record establishes that no scientific evidence exists for concluding that diazinon is among the organophosphates that causes neurotoxicity or among those that do not. There is simply no way to tell.

In Havner, plaintiffs offered extensive epidemiological evidence showing a relationship between Bendectin and birth defects, but the relationship was never shown to be statistically significant. We held that that was no evidence to support a finding that Bendectin causes birth defects. The evidence in the present case is even weaker than the evidence in Havner. Here there is no evidence at all, other than Ellis's experts' bare opinions, showing a relationship between diazinon exposure and neurotoxicity. Moreover, all physical medical evidence electroencephalograms, peripheral nerve tests, an MRI, and a CAT scan — have shown Ellis to be in normal health, aside from problems relating to obesity, high blood pressure, smoking, and alcohol dependency. Under our precedents, the experts' unsupported opinions cannot provide a basis for a judgment against Maritime Overseas.

* * * * *

Because there is no basis for Ellis's experts' opinions that his exposure to diazinon caused him to suffer from neurotoxicity, those opinions were not probative evidence and should not have been considered by the court of appeals in assessing the factual sufficiency of the evidence of causation of Ellis's damages. Accordingly, I would reverse the court of appeals' judgment and remand the case to that court to redetermine the factual sufficiency of the evidence.