concurring in part and dissenting in part.
I concur in that portion of the majority opinion that holds that the trial court did not err in requiring the submitted theory of liability of failure to treat to be limited to the modality of “TPN” as opposed to a broader scope of possible treatments for malnutrition. Appellant is correct in arguing that an instruction should not be required to submit evidentiary detail. Verdict directing instructions in particular, however, are dependent upon the evidence that is adduced at trial. The distinction between submitting ultimate facts but not evidentiary detail, as required by M.A.I., is often difficult to apply because the instruction must follow both the law and the evidence. In a medical negligence case, the testimony of the plaintiffs expert will *376largely shape the ultimate facts that must be submitted. If alternative treatments may, in the exercise of due care, be chosen by the physician, it would be improper to submit whether he was negligent only in not selecting one of the available modalities. Conversely, if the expert’s testimony established that only one treatment modality would satisfy the standard of care, although there might be some evidence of some other types of treatment, then it would be a roving commission to allow the jury to find the physician negligent for not using one of the other treatment modalities. The expert witness and the plaintiff’s lawyer are largely in control of this dilemma, and it must be resolved by looking at the evidence that is presented. If the expert believes that any one of several modalities would satisfy the standard of care, even though he personally might prefer one, then submission generally of all of the modalities of treatment would be proper. But this distinction between an expert’s personal preference and multiple choices proper under the standard of care must be made clear. This appellant failed to do and for that reason I concur in the majority opinion on this point. Despite protestations to the contrary on appeal, the evidence clearly reveals that appellant’s expert opined that, at the time in question, treatment by TPN was the only treatment that satisfied the standard of care.
I dissent, however, from that portion of the majority opinion that holds that the trial court did not err in refusing to submit appellant’s theory that the respondents were negligent in failing to diagnose appellant’s malnutrition. A party is generally entitled to submit any theory of recovery to the jury that falls within the scope of its pleadings, is supported by the evidence, and is in proper form. See Yoos v. Jewish Hosp. of St. Louis, 645 S.W.2d 177, 191 (Mo.App.1982). The record clearly reflects that appellant’s expert witness gave the opinion in proper form that respondents failed to diagnose malnutrition and that the failure fell below the standard of care. Appellant’s initial proffered verdict director disjunctively submitted two prongs of negligence: (1) failure to diagnose malnutrition, or (2) failure to treat malnutrition. The objections to the failure to treat theory have been extensively discussed in the majority opinion.
The trial court apparently began discussing instructions and the verdict director with counsel just after the plaintiffs evidence, ended. Such early communication and even earlier discussion is advisable. Free ranging discussions were conducted about possible theories of submission, the form of submission, and objections. Such informal instruction discussions are valuable in defining and even potentially remedying instructional problems. In this case, the “informal discussions” were done on the record and after various instructions were drafted and redrafted. The “formal instruction conference” was rather perfunctory, referring back largely to the more informal instruction conferences during the trial. As a result, the positions and objections are somewhat difficult to follow.
Nevertheless, it appears that there were two objections made to the appellant’s original verdict director.1 Respondents first objected to the submissibility of the case at all, contending that there was insufficient evidence to submit the case to the jury on any of the proffered issues. Secondly, they objected that it was improper to submit the alternative theories *377of recovery in the disjunctive because they claimed that the evidence did not support a disjunctive submission. The trial court also indicated some concern that the initial verdict director assumed a disputed fact, i.e., that appellant had malnutrition. Finally, the trial judge also indicated that it was her recollection that there was no testimony showing a causal relationship between failure to diagnose and appellant’s/decedent’s death.
Although it is the initial responsibility of a party to request instructions, Rule 70.02, it is ultimately the duty of the trial court to see that the jury is instructed and instructed correctly. M.A.I. 5th ed. “How to Use This Book” at p. LVII. Parties need to object to instructions before, generally, an appellate court will review an instruction for error. Daniels v. Bd. of Curators of Lincoln Univ., 51 S.W.3d 1, 10-11 (Mo.App.2001). Still, the trial judge may point out flaws in an instruction, to see that the jury is not misinstructed on the law. See Bench Book for Missouri Trial Judges, Civil Chapter 39 § 39.9 (2002). Where, as here, the trial judge raised two questions, herself, about the verdict director and, apparently, relied upon those concerns in making her ruling, it should be irrelevant that the respondents did not actually object to the instructions on those grounds.
The first concern expressed by the judge was whether the failure to diagnose and failure to treat submissions assumed a disputed fact. It is correct that it is improper to assume disputed facts in instructions. Young v. Kansas City Power & Light Co., 773 S.W.2d 120, 125 (Mo.App.1989). This writer does not believe the instruction was deficient in that respect. To find for the plaintiff, the jury would have to believe that respondents failed to diagnose malnutrition. There is no way that a jury could reach that finding without believing that the deceased had malnutrition. How can a physician fail to diagnose a condition that does not exist? Resolution of that disputed issue of whether the decedent suffered from malnutrition was embedded in the submission offered by appellant. Reed v. Sale Mem’l Hosp. & Clinic, 698 S.W.2d 931, 938 (Mo.App.1985) (instruction not erroneous when a finding of the essential element is necessarily implied from the other findings required). Acceptance of respondents’ novel reasoning would lead to the requirement that in a red light automobile case the submission also include whether the light was red as a separate finding before asking the jury if the defendant ran a red light.
Where evidence is conflicting, as it often is, a party is still entitled to submit its case under the theory and evidence that it has adduced. Highfill v. Brown, 340 S.W.2d 656, 661 (Mo. banc 1960). Accommodation of the other party’s conflicting evidence is not required, with the proviso that a disputed fact cannot be ignored. The disputed fact was not ignored here. Additionally, there were other ways for respondents to even highlight the issue such as the use of a converse under M.A.I. 33.01 (e.g., “Your verdict must be for defendant unless you believe that the plaintiff suffered from malnutrition”). Finally, if the failure to diagnose instruction suffered from this alleged infirmity then, likewise, the failure to treat malnutrition instruction drafted by the respondents and ultimately given by the court was incorrect for the same reason.
WAS THERE SUBSTANTIAL CAUSATION EVIDENCE TO SUPPORT SUBMISSION OF A FAILURE TO DIAGNOSE THEORY?
The majority opinion simply concludes that it was not error to submit appellant’s failure to diagnose theory because there *378was no substantial evidence to satisfy the requirement of a causal relationship between the alleged act of negligence (failure to diagnose malnutrition) and the death of Shirley Mast. In attempting to justify that conclusion, the majority quotes the following testimony from appellant’s expert witness:
Q: Doctor do you have an opinion within a reasonable degree of medical certainty whether the failure of Dr. Braverman to diagnose or treat the malnutrition of Shirley Mast caused her death?
A: I believe it did.
The majority unfortunately omits the immediately previous testimony of the expert as follows:
Q: Doctor ... I asked [sic] you also if you would have an opinion within a reasonable degree of medical certainty as to the cause of Shirley Mast’s death. Can you give us a cause of death?
A: I think she starved to death. There were factors that happened at the last. She had an infection of the line, she had liver necrosis, she had kidney failure, all of those things, but they were triggered by starvation, malnutrition....
Nevertheless, the majority finds the testimony on causation lacking in substantial probative value because the question asked the expert whether failure to diagnose OR failure to treat caused Mrs. Mast’s death. The reasoning used to reach this conclusion is tortured, ignores the common meaning of the words used, and is unsupported by any legal authority.
The majority seems to suggest that the word “or” means one but not the other. No reference is made to any dictionary, style manual, or other authority for that suggestion. In its customary meaning, however, “or” means “either.”2 The word “or” also usually, from a style analysis, includes “and.”3 Nevertheless, the fallacy of the majority’s view of the expert’s testimony is not dependent upon resolving, or even understanding, grammatical and stylistic rules.
Somehow, the majority makes a leap of logic that reinterprets the expert’s testimony to be that failure to diagnose was not a cause of death but that failure to treat was. If, by use of the word “or,” his testimony was not substantial evidence of causation on one theory, then how could it be so on the other? Yet the majority does not hold that by use of the word “or” that there was not sufficient causation evidence on either theory.
The majority then expands upon its fixation with the disjunctive testimony by stating “the problem with proceeding in this manner is that it is now unclear whether Dr. Schaefer believed that it was the failure to diagnose alone, failure to treat alone, or a combination of the two, which caused Mrs. Mast’s death.” (emphasis added). No citation is made to any legal authority that such proof is required where both alleged causes are the responsibility of the defendant.
There are, however, even more important difficulties with the majority analysis. The majority relies upon Super v. White, 18 S.W.3d 511 (Mo.App.2000) to support its conclusion that appellant’s expert testi*379mony on causation was not substantial evidence. In so doing it misconstrues both the holding in Super and ignores the actual testimony in this case. Although Super dealt with two possible causes of death, only one was allegedly the defendant doctor’s responsibility. The decedent had preexisting cirrhosis of the liver. The doctor was charged with negligently administering a treatment to the decedent for tuberculosis in the presence of chronic Hepatitis C. The plaintiffs expert testified (because he agreed that the TB treatment did not cause the cirrhosis) that he could not testify to a reasonable degree of medical certainty that the death was caused by the TB treatment as opposed to the cirrhosis. He testified that it was both possible and not possible that the alleged negligence of the defendant caused the death. Id. at 517. This court properly held both that testimony of causation must be to a reasonable degree of certainty and that where there are two or more possible causes the testimony must establish to a reasonable degree of certainty that the alleged negligence caused or contributed to cause the death. Id. at 516-17. See also Baker v. Guzon, 950 S.W.2d 635, 644, 646 (Mo.App.1997). This is a far cry from the situation in this case whether there was no evidence of a possible cause of death other than the defendants’ claimed negligence. Moreover, the appellant’s expert, as shown from the majority opinion’s own quotation from the record, testified, “to a reasonable degree of medical certainty” on the causation issue.
The majority next justifies its conclusion with the rationale that the causation testimony concerning the failure to diagnose theory was not as detailed as on the failure to treat theory. This rationale is disturbing as this writer is aware of no authority for a conclusion that one properly stated opinion of an expert lacks substantial evi-dentiary value because not as detailed as his testimony on another theory. The majority takes the position that Dr. Schae-fer’s testimony did not yield substantial evidence supporting the Plaintiffs theory of failure to diagnose because the doctor spent more time discussing the theory of failure to treat and rendered more detailed testimony concerning that theory. In this writer’s view, such comparisons are irrelevant under Missouri law. To determine whether evidence is substantial, the test is merely whether it is evidence from which the trier of fact could reasonably find the issue in harmony therewith. State v. Taylor, 445 S.W.2d 282, 284 (Mo.1969). “A trier of facts may find an issue in harmony with certain evidence when it is not inherently incredible, self-destructive, or completely impeached by contradictory evidence, and is such that reasonable minds might believe it.” State v. Charles, 537 S.W.2d 855, 857 (Mo.App.1976) (citing State v. Harris, 295 S.W.2d 94 (Mo.1956)). The doctor’s testimony upon the theory of failure to diagnose more than meets that threshold.
Finally, the majority justifies its conclusion by holding that, despite what the expert said, it was ultimately the failure to treat malnutrition that really caused her death. Such weighing of the evidence is inappropriate in determining whether a submissible case has been made. Moreover, a court has no business deciding which of a party’s legally supportable theories of recovery or defense is its best or real one. There was, in fact, expert testimony to support causation, and it was not proper to refuse appellant’s failure to diagnose theory on that basis.
The evidence in a disjunctive submission must support each negligence prong. Respondents make no argument on appeal that the failure to treat by use of TPN theory of recovery was not supported by the evidence. As discussed above, appel*380lant’s expert’s testimony clearly supports an alternative theory of recovery for failure to diagnose. The only possible question that could be raised is whether the denial of that submission constituted reversible error. I believe that it did, and the cause should be reversed and remanded for a new trial.
There are many cases stating the general rule that, when considering the propriety of an instruction, the evidence is to be viewed “in the light most favorable to the submission of the instruction, and a party is entitled to an instruction upon any theory supported by the evidence.” Vandergriff v. Mo. Pac. R.R., 769 S.W.2d 99, 104 (Mo. banc 1989) (emphasis added). This principle applies equally to plaintiffs and defendants. Entitlement reflects a legal right, and a trial judge commits an abuse of discretion by refusing an instruction that is supported by the evidence and in proper form. The choice of theories to be supported by the plaintiff in a verdict director and a defendant in an affirmative defense instruction is for the party to make. The trial judge has no role in the selection of theories if those theories are supported by the evidence and are correct statements of the law. Here, I believe that the appellant submitted and offered to submit an instruction that was proper in form on the failure to diagnose theory.
It seems obvious from the instruction conference, however, that the trial judge had no intention of ever submitting a theory of failure to diagnose malnutrition as a theory of recovery, regardless of the possible wording of such a submission. In this, she erred prejudicially to the rights of the appellant. In Williams v. Christian, 520 S.W.2d 139 (Mo.App.1974), the issue was whether the trial court erred in refusing plaintiff’s failure to keep a lookout instruction and giving instead a res ipsa instruction that the defendant contended was even more favorable to the plaintiff. This court said, “It is axiomatic that appellant was entitled to a verdict directing instruction predicated on respondent’s failure to keep a careful lookout, his theory of the case, if supported by the evidence.” Id. at 141 (emphasis added). After considering whether the instruction was factually supported and legally correct, Judge Sommer-ville stated, “Perforce, the trial court erred in refusing to give the tendered instruction.” Id. at 146. And, finally, with regard to the prejudice issue, the court said, “If more than lip service is to be given to the principle that a party is entitled to go to the jury on his theory of the case, if supported by the evidence, then the trial court’s error in refusing the requested lookout instruction cannot be sloughed off as merely harmless error.” Id. Separately, respondent Wuellner makes an argument that the instruction was improper because he did, in fact, diagnose malnutrition, albeit at an earlier time than in question with appellant’s expert. If that was his theory, he could have submitted his theory under M.A.I. 33.05. It was not a basis for refusing appellant’s instruction. See id. at 145-46.
For the reasons stated, I dissent and would reverse and remand the cause for a new trial.
. There were, of course, verdict directors against each respondent. Because they were identical, they are referred to in the singular for ease of understanding.
. "Or” is a coordinating conjunction introducing an alternative: specifically, introducing the second of two choices. Webster's Dictionary of the English Language Unabridged 1257 (1977).
. Bryan A. Garner, The Elements of Legal Style 103 (1991). For example, a statement that a person is incapable of reading or writing connotes the inability to do both tasks, not just one.