Taylor v. Commissioner of Mental Health

Annotate this Case

481 A.2d 139 (1984)

Henry A. TAYLOR, III v. COMMISSIONER OF MENTAL HEALTH AND MENTAL RETARDATION.

Supreme Judicial Court of Maine.

Argued March 6, 1984.

Decided August 14, 1984.

*141 Kelly, Remmel & Zimmerman, Graydon G. Stevens (orally) Portland, for plaintiff.

James E. Tierney, Atty. Gen., Wayne S. Moss (orally), James Erwin, Asst. Attys. Gen., Augusta, for defendant.

Before McKUSICK, C.J., and NICHOLS, ROBERTS, VIOLETTE and GLASSMAN, JJ.

McKUSICK, Chief Justice.

Henry A. Taylor, III, who has been committed to Augusta Mental Health Institute ("AMHI") since his 1979 acquittal by reason of insanity on two serious criminal charges, appeals to this court from the denial by the Superior Court (Kennebec County) of his petition for modified release treatment under 15 M.R.S.A. § 104-A (Supp.1983-1984). In absence of any legislative specification of the applicable standard of proof on such a petition, the Superior Courtfollowing the precedent of State v. Shackford, 262 A.2d 359 (Me.1970)required the petitioner to prove his eligibility for the proposed modified release treatment "beyond a reasonable doubt." On reexamination of the question in light of judicial and statutory developments elsewhere since Shackford, we conclude that the reasonable doubt standard is unduly stringent. We remand the case to the Superior Court for a rehearing at which petitioner to succeed must establish his eligibility for the proposed treatment by clear and convincing evidence.

I.

Henry Taylor was indicted for murder and reckless conduct in shooting to death one of his neighbors on November 17, 1978. On July 2, 1979, in a nonjury trial, Taylor was found not guilty by reason of insanity (hereinafter "BRI"). As directed by 15 M.R.S.A. § 103 (1980), the Superior Court automatically committed him to the custody of the Commissioner of Mental Health and Corrections.[1] The Commissioner placed him in the Augusta Mental Health Institute, where he has since been a patient.

Since his commitment five years ago Taylor and the AMHI staff have filed a series of petitions seeking court approval for progressively less restrictive treatment programs.[2] Judicial approval of some of *142 those petitions has given Taylor substantial freedom to work, leave the AMHI grounds, and travel with relatives. The present petition was filed with the Superior Court on April 15, 1983. Signed by both Taylor and Dr. Ulrich Jacobsohn, the clinical director of AMHI, the petition sought approval of a "modified release treatment" plan giving Mr. Taylor still more freedom during his continued treatment. Denominated a "petition for Release Treatment Program," it sought approval, under the predecessor to 15 M.R.S.A. § 104-A(2), of a plan allowing Taylor to live full-time off the AMHI grounds at any location in Maine except for Cumberland County. The plan required Taylor to report to AMHI once a week for out-patient evaluation and medication. The estimated duration of the release treatment program before further change in treatment status was six months. The petition alleged that this program would aid Taylor in his treatment and could be accomplished without a likelihood of injury to Taylor or others due to mental illness.

The Superior Court held a hearing on this petition on August 17, 1983. At the hearing, in which counsel for both Taylor and the State participated, the Superior Court heard testimony from Dr. Jacobsohn, Taylor, and Taylor's sister. By order dated August 24, 1983, the Superior Court denied the petition. On the basis of prior Law Court authority, the Superior Court required Taylor to prove his eligibility for the requested modified release treatment "beyond a reasonable doubt." The Superior Court was unpersuaded that Taylor had met that high standard of proof. Taylor appeals to this court, contending that the Superior Court should not have required him to prove his eligibility for release beyond a reasonable doubt.

II.

At the outset it is necessary to place Taylor's argument in the broader context of Maine's system for handling persons who assert the insanity defense. A determination that an accused is not guilty by reason of insanity requires two separate findings in the criminal trial. The State must carry its initial burden of proving beyond a reasonable doubt every element of the offense.[3]See 17-A M.R.S.A. §§ 32, 40 (1983). If the State is successful, the "defendant shall have the burden of proving, by a preponderance of the evidence, that he lacks criminal responsibility...." 17-A M.R.S.A. § 39(1) (1983). See State v. Burnham, 406 A.2d 889, 891-92 (Me.1979), appeal after remand, 427 A.2d 969 (Me. 1981). In Maine a defendant who gets a verdict of not guilty by reason of mental disease or defect has persuaded the factfinder by a preponderance of the evidence that "at the time of the criminal conduct, as a result of mental disease or defect, he either lacked substantial capacity to conform his conduct to the requirements of the law, or lacked substantial capacity to appreciate the wrongfulness of his conduct." 17-A M.R.S.A. § 39(1) (comparable to socalled "ALI test," see American Law Institute Model Penal Code § 4.01 (1962)).

A criminal defendant found not guilty by reason of insanity (hereinafter at times referred to as a "BRI acquittee") is subject to automatic commitment without a further hearing. 15 M.R.S.A. § 103 provides that "the court shall order such person committed to the custody of the Commissioner of Mental Health and [Mental Retardation] to be placed in an appropriate institution for the mentally ill or the mentally retarded for care and treatment."

Detailed statutory procedures govern the release and discharge of BRI acquittees committed pursuant to section 103. In addition to outright discharge of a BRI acquittee from the custody of the Commissioner of Mental Health, 15 M.R.S.A. *143 § 104-A[4] provides for his possible release from the mental institution in two forms: (1) release subject to conditions imposed by the Superior Court, such as out-patient treatment or supervision by the Division of Probation and Parole, § 104-A(1)(A); and *144 (2) modified release treatment, involving absence from the institution's grounds for not more than 14 days, under a treatment program proposed by the institution staff and not disapproved by the Superior Court, § 104-A(2).[5]

Proceedings for conditional release or discharge may be initiated either by the head of the institution where the BRI acquittee has been committed or by the acquittee, his spouse, or next of kin. Only the acquittee himself may petition the Superior Court for modified release treatment, supported by a detailed report from the institution's staff of the nature and expected results of the treatment program. The modified release program goes into effect automatically unless the court within 60 days orders a hearing on its own initiative or on request of the district attorney or Attorney General. Any release or discharge of a BRI acquittee may be approved by the Superior Court only if it finds that he "may be released or discharged without likelihood that he will cause injury to himself or to others due to mental disease or mental defect." § 104-A(1). Although section 104-A is far from being a model of drafting clarity, we read that statute, as did the court in the case at bar, to require that the statutory release eligibility test be satisfied on a petition for the approval of a program of modified release treatment, just as much as on a petition for conditional release or outright discharge. The interests of the public and the BRI acquittee in avoiding injury while the latter is away from the institution are the same whatever the basis for terminating his inpatient residency at the mental institution. That is not to say, however, that the evidentiary considerations bearing on the protection of the public and the BRI acquittee will be the same for any and all kinds of release and discharge.

A conspicuous lacuna exists in section 104-A. The legislature has not stated what standard of proof should be required by the Superior Court in determining whether a BRI acquittee satisfies the statutory release eligibility test.[6] Fourteen years ago we affirmed a Superior Court justice's requirement that as a condition for an acquittee's release he must establish "by evidence, convincing in its effect beyond a reasonable doubt that release could be effected without danger to the public within the foreseeable future." (Emphasis added) State v. Shackford, 262 A.2d 359, 365 (Me.1970). In Shackford, we recognized that the controlling statute[7] did not define the standard of proof, and that the court was thus left to define the appropriate standard of proof to be applied in the context of Maine's statutory scheme for the insanity defense and for institutionalizing BRI acquittees. Cf. Santosky v. Kramer, 455 U.S. 745, 755-56, 102 S. Ct. 1388, 1395-96, 71 L. Ed. 2d 599 (1982) ("the degree of proof required in a particular type of proceeding `is the kind of question which has traditionally been left to the judiciary to resolve'"). Finding compelling the reasoning of a 1960 decision of the United States Court of Appeals for the District of Columbia,[8] we declared:

*145 We think it reasonable to require that reasonable medical doubts and reasonable judicial doubts be resolved in favor of the public.

State v. Shackford, 262 A.2d at 366.

The next year, in Chase v. Kearns, 278 A.2d 132 (Me.1971), which upheld the constitutionality of the reasonable doubt standard of proof, we said of Shackford:

We based our conclusion that reasonable medical doubts must be resolved in favor of the public not on particular language of the statute but on the paramount necessity of protecting the special interest which the public acquired in the confinement and release of such a patient.

Id. at 137. In the 13 years that have elapsed since Chase,[9] we have not been called upon to reexamine the reasonable doubt rule, which our court in 1970 had itself developed as a proper weighing of the public interest against the BRI acquittee's claimed right to be set free. See State v. Shackford, 262 A.2d at 366.

III.

The time has come when we must take a fresh look at our court-made reasonable doubt rule. Statutory and decisional changes occurring outside Maine since Shackford make it now difficult, and perhaps impossible, to find another American jurisdiction that places a reasonable doubt burden of proof upon a BRI acquittee seeking release. Enlightened by those legal developments and by the substantial legal and medical literature in the field, we address the question whether in our best judgment the reasonable doubt rule strikes the proper balance between the protection of the public and the acquittee's interest in his freedom.

We start with the proposition that the court proceeding on a BRI acquittee's petition for release is not a part of the criminal case that resulted in his commitment; it rather is a type of original civil proceeding. See, e.g., Model Penal Code § 4.08(3) (1962); Hamann, The Confinement and Release of Persons Acquitted by Reason of Insanity, 4 Harv.J. on Legis. 55, 89. "[T]he reasonable doubt standard generally has been held inapplicable in civil cases, regardless of the nature of the issue involved." E. Cleary, McCormick's Handbook of the Law of Evidence § 341, at 802 (2d ed. 1972).

In Coffin v. United States, 156 U.S. 432, 453-56, 15 S. Ct. 394, 402-04, 39 L. Ed. 481 (1895), the Supreme Court traced the development of the reasonable doubt standard in criminal cases to early Greek and Roman times. This standard has been a fundamental part of our criminal law since the early days of the nation. See In re Winship, 397 U.S. 358, 361, 90 S. Ct. 1068, 1071, 25 L. Ed. 2d 368 (1970); see also Mullaney v. Wilbur, 421 U.S. 684, 701, 95 S. Ct. 1881, 1890, 44 L. Ed. 2d 508 (1975) (reasonable doubt standard is traditional element of criminal prosecutions); 17-A M.R.S.A. § 32 (1983). The insistence that the State prove each element of a crime beyond a reasonable doubt reflects a concern for the special nature of a criminal prosecution. The State may imprison anyone only on proof of guilt by the highest standard of proof known to the common law. A criminal conviction, with all its attendant consequences to individual freedom and reputation, is an awesome exercise of governmental power. Merely stating this principle suggests why the standard of proof imposed on the State in a criminal case should differ from the standard imposed on a person seeking his release from commitment to a mental institution. In the former case *146 the State, as the moving party, seeks to take away a person's liberty. In the latter case, the BRI acquittee, as the moving party, seeks to regain his liberty. Additionally, whereas we may impose on the State, with all of its resources, the highest standard of proof, we should not expect the same degree of exacting proof from a resident in a mental institution. For these reasons, "we should hesitate to apply [the reasonable doubt standard] too broadly or casually in noncriminal cases." Addington v. Texas, 441 U.S. 418, 428, 99 S. Ct. 1804, 1810, 60 L. Ed. 2d 323 (1979). In Allen v. Radack, 426 F. Supp. 1052, 1058 (D.S.D. 1977), the court rejected the reasonable doubt standard for release of BRI acquittees. The court there found, as do we, that the reasonable doubt standard is "peculiarly" necessary to safeguard individual liberty in criminal prosecutions and should not be imposed on persons seeking their release from commitment. See Hamann, 4 Harv.J. on Legis. at 89 (application of reasonable doubt standard to BRI acquittees unfair and inconsistent).

In the context of mental health evaluations, the reasonable doubt standard requires a degree of certainty in diagnosis and prediction that as a practical matter is seldom attainable. Under Shackford, the Superior Court has been required to find beyond a reasonable doubt that the BRI acquittee meets the statutory release eligibility test, namely, that he can be released "without likelihood that he will cause injury to himself or to others due to mental disease or mental defect." 15 M.R.S.A. § 104-A(2). By itself, the "without likelihood of injury" eligibility test requires no more than a showing of the probability of safe release. However, a demand that petitioner establish that probability to the exclusion of any reasonable doubt dramatically magnifies the quantum of required proof.

The ability of the State to establish the elements of a crime beyond a reasonable doubt does not give any assurance that this standard of proof is attainable in predicting the future conduct of a patient after release from a mental institution. The critical issues at a release hearing are inherently less capable of being determined with certainty than are the issues in most criminal trials. In a criminal case, "the basic issue is a straightforward factual questiondid the accused commit the act alleged?" Addington v. Texas, 441 U.S. at 429, 99 S. Ct. at 1811. Although there may be difficulties in collecting and presenting the evidence relevant to that question, once that evidence is before the factfinding tribunal it is possible to dispel any reasonable doubt as to whether the defendant committed a particular act at a particular place and time in the past. A release hearing, by contrast, requires the factfinder to predict future behavior. See State v. Krol, 68 N.J. 236, 260, 344 A.2d 289, 302 (1975). The future often eludes attempts at prediction, particularly where one attempts to foretell human conduct.

Resort to scientific evidence will not provide the certainty required by the reasonable doubt standard. Although a court may find psychiatric opinion useful in reaching its conclusions, psychiatry is "by no means an exact science." State v. Carter, 64 N.J. 382, 407, 316 A.2d 449, 463 (1974). Given the vagaries of human behavior, a psychiatrist could rarely, if ever, "swear under oath that there will be no relapse by the patient." Comment, Compulsory Commitment Following a Successful Insanity Defense, 56 Nw.U.L.Rev. 409, 453 (1961). After reviewing the literature on psychiatric predictions of dangerousness, Cocozza and Steadman concluded that such predictions cannot satisfy the degree of certainty required by the reasonable doubt standard. See Cocozza & Steadman, The Failure of Psychiatric Predictions of Dangerousness: Clear and Convincing Evidence, 29 Rutgers L.Rev. 1084, 1101 (1976). See also Addington v. Texas, 441 U.S. at 429, 99 S. Ct. at 1811 (uncertainty and fallibility of psychiatric diagnosis precludes proof of illness and dangerousness beyond a reasonable doubt).

*147 Testimony from other sources cannot sufficiently supplement the psychiatric evidence to generate a conclusion that dispels any reasonable doubt. One example of the difficulty in predicting future dangerousness is the range of possible environmental stimuli that the BRI acquittee may encounter after his release. Because a mental patient may be violent or criminogenic only in certain limited circumstances, an evaluation of the patient's eligibility for release "must take into account the likelihood that [he] will be exposed to such situations or come into contact with such individuals." Krol, 68 N.J. at 260, 344 A.2d at 302. The difficulty of predicting the likelihood, and consequences, of such encounters weighs heavily against the BRI acquittee who is "kept in strict custody and allowed no opportunity to demonstrate what his behavior would be if he were given his liberty." German & Singer, Punishing the Not Guilty: Hospitalization of Persons Acquitted by Reason of Insanity, 29 Rutgers L.Rev. 1011, 1064-65 (1976). A second factor hindering precise prediction is the necessity that the BRI acquittee prove a negative to secure his release. In contrast to a criminal prosecution where the State must prove the occurrence of a specific affirmative act in the past, in a release hearing the BRI acquittee must prove that certain events, potentially of a wide variety, will not occur following his future release. Id. at 1064. Proof of such a negative is inevitably a critical element in a BRI acquittee's attempt to secure his release, but the reasonable doubt standard overburdens an already demanding problem of proof.

Considering the inevitable difficulty of predicting human behavior through any sort of evidence, courts and commentators have widely denounced the reasonable doubt standard as demanding too much. See, e.g., Addington v. Texas, 441 U.S. at 429, 99 S. Ct. at 1811 ("serious question" whether standard can be met); Krol, 68 N.J. at 260, 344 A.2d at 302 (certainty of prediction "cannot reasonably be expected"); German & Singer, 29 Rutgers L.Rev. at 1064 (reasonable doubt standard "seems practically impossible to meet"); Hamann, 4 Harv.J. on Legis. at 89 (standard is "insuperable"); Comment, The Rights of the Person Acquitted by Reason of Insanity: Equal Protection and Due Process, 24 Me. L.Rev. 135, 145 (1972) (reasonable doubt standard makes release hearing "meaningless"); Comment, Commitment Following Acquittal by Reason of Insanity and the Equal Protection of the Laws, 116 U.Pa.L. Rev. 924, 938 (1968) (standard "borders on the impossible"). As a consequence of the near impossibility of making this showing beyond a reasonable doubt, whether the BRI acquittee will be released will be decided in practical fact by the State's attorney. Whenever, in his unreviewable discretion, the prosecutor chooses to oppose a release petition with vigor, the acquittee's chances of securing his release are virtually nil.

The reasonable doubt standard also works to the detriment of the judicial process by effectively precluding appellate review of a Superior Court finding that the BRI acquittee is ineligible for release. We review Superior Court factual determinations in section 104-A release hearings, as in other civil cases, only for findings that are "clearly erroneous." See M.R.Civ.P. 52(a). The "clearly erroneous" standard of review results in substantial deference to the trial court's factual findings in the ordinary civil case where the plaintiff has a burden of proving his case by a preponderance of the evidence. But that standard of review, as a practical matter, insulates from appellate review a lower court's denial of a release petition where that court need only have had a reasonable doubt of the petitioner's eligibility. It will be rarely, if ever, possible to conclude that "the justice acted irrationally in failing to be satisfied beyond a reasonable doubt that petitioner" qualified for release. See In re Fleming, 431 A.2d 616, 618 (Me.1981); cf. State v. Caouette, 446 A.2d 1120, 1124 (Me.1982) (appellate review of finding that reasonable doubt existed as to voluntariness of confession). Lower court determinations of issues such as future dangerousness *148 are inevitably somewhat resistant to effective appellate review, see Note, Commitment and Release of Persons Found Not Guilty By Reason of Insanity: A Georgia Perspective, 15 Ga.L.Rev. 1065, 1100 (1981), but application of the reasonable doubt standard largely precludes an appellate court from exercising its important function of correcting clear errors. See State v. Carter, 64 N.J. at 406, 316 A.2d at 462.

On examining the standards of proof used in other states for determining the eligibility of BRI acquittees for release from commitment, we find that Maine is unique in requiring the BRI acquittee to show his eligibility for release beyond a reasonable doubt. Although prior to 1970, when Shackford was decided, there were some jurisdictions that required the BRI acquittee to meet the reasonable doubt standard, or some rough equivalent, there is today no other jurisdictionso far as we can determinethat, either by statute or case law, unambiguously imposes the reasonable doubt standard on BRI acquittees. The Shackford decision, 262 A.2d at 366, relied heavily upon the use of the reasonable doubt standard in Ragsdale v. Overholser, 281 F.2d 943, 947 (D.C.Cir.1960). As Shackford recognized, Ragsdale had shortly before been overruled by the District of Columbia court, see Bolton v. Harris, 395 F.2d 642, 653 (D.C.Cir.1968), and by statute the rule in the District now is that the BRI acquittee need prove his eligibility for release only by a preponderance of the evidence. See Jones v. United States, 463 U.S. 354, ___, 103 S. Ct. 3043, 3045, 77 L. Ed. 2d 694, 700 (1983); D.C.Code Ann. § 24-301(d)(2)(B) (Supp.1981).[10] An earlier South Dakota decision imposing the reasonable doubt standard, see State ex rel. Barnes v. Behan, 80 S.D. 370, 124 N.W.2d 179 (1963), was invalidated on constitutional grounds by a federal district court, which substituted a clear-and-convincing-evidence standard.[11]See Allen v. Radack, 426 F. Supp. 1052. The decision of the Montana Supreme Court adopting the reasonable doubt standard, see State v. Taylor, 158 Mont. 323, 491 P.2d 877 (1971), cert. denied, 406 U.S. 978, 92 S. Ct. 2428, 32 L. Ed. 2d 677 (1972), was effectively overruled by a change in the Montana Code. See Mont.Code Ann. §§ 46-14-302(4), 46-14-303 (1983) (burden on BRI acquittee to prove eligibility for release by a preponderance of the evidence); see also State v. Olson, 181 Mont. 323, 593 P.2d 724 (1979) (citing and applying statutory preponderance standard). The state of Washington once permitted release of BRI acquittees only if the trier of fact, applying a standard comparable to reasonable doubt, found that the acquittee was sane, likely to remain sane, and safe if left at large. See State v. Blubaugh, 80 Wash. 2d 28, 36, 491 P.2d 646, 651 (1971). By statute, Washington now requires the petitioner to show his eligibility for discharge by "a preponderance of the evidence." Wash.Rev.Code § 10.77.200 (Supp.1984-1985); see State v. Kolocotronis, 34 Wash. App. 613, 663 P.2d 1360 (1983). The reasonable doubt standard, as adopted in Shackford, has thus become out of step with the legal developments in this field throughout the United States.

We therefore conclude that requiring a BRI acquittee such as Taylor to prove his *149 eligibility for release beyond a reasonable doubt contravenes important policy considerations, and we join the courts and legislatures of all other American jurisdictions in rejecting such a standard.[12] Our next task is to determine anew a proper standard of proof to fill the lacuna left in section 104-A by the legislature.

IV.

Having rejected the criminal reasonable doubt standard, we hold that BRI acquittees must prove their eligibility for release by "clear and convincing evidence." As between the two standards of proof in civil cases, proof by a preponderance of the evidence and proof by "clear and convincing evidence," we conclude that the former would not adequately address the important policy concerns involved in a BRI acquittee's petition for release.

The traditional requirement of proof by clear and convincing evidence, which developed in equity practice, has been applied to a wide variety of civil cases where for policy reasons the courts wish a higher than ordinary degree of certitude before making findings of fact. Wigmore gives us the following summary:

[A] stricter standard, in some such phrase as "clear and convincing proof," is commonly applied to measure the necessary persuasion for a charge of fraud, or of undue influence; for the existence and contents of a lost deed or will; for a parol gift or an agreement to bequeath or devise by will or to adopt; for mutual mistake sufficient to justify reformation of an instrument; for a parol or constructive trust; for an oral contract as a basis for specific performance; for impeaching a notary's certificate of acknowledgment; for prior anticipatory use of an invention; for an agreement to hold a deed absolute as a mortgage; and for sundry classes of cases in local practice.

9 Wigmore, Evidence § 2498, at 424-31 (Chadbourn rev. 1981) (italics in original omitted). For constitutional and other policy considerations, the United States Supreme Court has extended the requirement of clear and convincing evidence to many factual controversies beyond the equity cases that spawned the heightened civil proof requirement. The Court requires proof by clear and convincing evidence before diversion of water by one state from another,[13] before termination of parental rights,[14] before civil commitment of a mental patient,[15] before admission in a criminal case of an in-court identification as being not untainted by an inadmissible pretrial identification,[16] before a deportation, denaturalization, or expatriation under the immigration laws,[17] and before a public figure's *150 recovery of money damages for a defamatory statement made in reckless disregard of the truth.[18] In other jurisdictions, a form of proof more demanding than a mere preponderance is required to discipline an attorney,[19] to impeach an affidavit that swears to proper service of process,[20] to declare illegitimacy,[21] to prove arson as a defense to an insurance policy,[22] or to establish the character, whether a loan or a gift, of a money transfer to a decedent prior to his death.[23] In Maine, clear and convincing evidence has traditionally been required in the equity-type cases of the nature listed in the above-quoted Wigmore summary. See Horner v. Flynn, 334 A.2d 194 (Me.1975), and the cases there cited. In addition, the Law Courtfor policy reasons comparable to those motivating the United States Supreme Court and many other courtshas in recent years required some contested facts in civil cases to be proved by clear and convincing evidence. In enforcing the purposes of the Maine Human Rights Act, we concluded that an employer could rebut a prima facie case of hiring discrimination only by clear and convincing proof that the complaining applicant would for nondiscriminatory reasons not have been hired anyway. See Maine Human Rights Commission v. City of Auburn, 425 A.2d 990, 996 (Me.1981). Similarly, in holding for the first time that the Workers' Compensation Act applies to gradual mental injuries from ordinary work-related stress and strain, we required, as a protection against malingering, that the claimant show "by clear and convincing evidence that the trauma generated by the employment predominated [over other forces] in producing the resulting injury." Townsend v. Maine Bureau of Public Safety, 404 A.2d 1014, 1020 (Me.1979).

In sum, although the preponderance standard normally prevails in a civil case, appellate courts in a large number of categories of litigation have found compelling reasons for requiring a higher form of proof. The purpose of fixing a standard of proof for the trial courts is to "instruct the factfinder concerning the degree of confidence our society thinks he should have in the correctness of factual conclusions for a particular type of adjudication." In re Winship, 397 U.S. at 370, 90 S. Ct. at 1076 (Harlan, J., concurring). The answer to what "our society thinks" is not determined by a public opinion poll, but rather can come only from a reasoned balancing of all the interests, public and private, that are implicated in the particular factual determination. The contested fact question presented on a BRI acquittee's petition for release is: Can the petitioner be released without likelihood that he will cause injury to himself or to others due to mental disease or defect? Balancing public and private interests in the answer, we are convinced that proof merely that an affirmative answer is more probable than not does not adequately protect the public interest. We conclude that society appropriately should demand that to gain release, a BRI acquittee prove by clear and convincing evidence that he can be released with safety to himself and others.

The State has an obvious and substantial interest in seeing to it that the factfinder in resolving uncertain evidence err, if at all, on the side of caution. The State rightfully seeks to protect the public from the premature *151 release of dangerous individuals. It understandably acts from a belief that "[t]he fact that a person has been found, beyond a reasonable doubt, to have committed a criminal act certainly indicates dangerousness." Jones v. United States, 463 U.S. ___, ___, 103 S. Ct. 3043, 3049, 77 L. Ed. 2d at 705. That indication becomes weaker with the passage of time following the criminal act, but always remains in some degree as the compelling reason for the public interest in guarding against a repetition.[24] The New Jersey Supreme Court has explained that

[p]ublic safety is the primary concern in shielding the public from both criminals and those adjudicated insane. "[T]he aim of the law is to protect the innocent from injury by the sick as well as the bad." State v. Maik [60 N.J. 203, 213, 287 A.2d 715, 720 (1972)].

State v. Carter, 64 N.J. at 388, 316 A.2d at 452. See Note, Commitment and Release Standards and Procedures: Uniform Treatment for the Mentally Ill, 41 U.Chi. L.Rev. 825, 837 (1974). That governmental function has been described as an "overriding concern," Carter, 64 N.J. at 401, 316 A.2d at 459, and one of the "primary duties of the state," 56 Nw.U.L.Rev. at 458. The state also has a legitimate concern for the well-being of persons who suffer from mental illness. The premature release of a BRI acquittee could result in the acquittee's injuring himself and may deprive him of the opportunity to carry through a complete course of treatment. See Hamann, 4 Harv.J. on Legis. at 85, 87, 91. "One who is suffering from a debilitating mental illness and in need of treatment is neither wholly at liberty nor free of stigma." Addington v. Texas, 441 U.S. at 429, 99 S. Ct. at 1811. Of course, on the acquittee's side, "commitment for any purpose constitutes a significant deprivation of liberty that requires due process protection." Id. at 425, 99 S. Ct. at 1809 (civil commitment); see also Jones v. United States, 463 U.S. at ___, 103 S. Ct. at 3048, 77 L. Ed. 2d at 703 (commitment of BRI acquittees). His interest to be free of any restraints as soon as possible must be weighed in the balance against the public's right to be protected. The acquittee's best interests, however, do not lie entirely on the side of early release regardless of the correctness of the judicial prediction of his conduct. It does him little good to be prematurely released on a mistaken prediction that the mental disease or defect that resulted in his prior criminal act has been treated, and will continue to be treated after release, so as to prevent a recurrence. On balance, the public interest in assuring a correct decision on the eligibility question outweighs the acquittee's private interest to an extent sufficient to justify society in demanding that he prove the safety of his release by clear and convincing evidence. The decision on an acquittee's petition for release involves substantial public interests similar to those that have led other courts to require enhanced degrees of proof. See, e.g., Colorado v. New Mexico, ___ U.S. at ___, 104 S. Ct. at 2437 (stability of water rights); New York Times Co. v. Sullivan, 376 U.S. 254, 285-86, 84 S. Ct. 710, 728-29, 11 L. Ed. 2d 686 (1964) (newspaper's first amendment rights); Maine Human Rights Commission v. City of Auburn, 425 A.2d at 996 (prevention of employment discrimination).

The choice of standard of proof also serves to allocate the risk of error between the litigants. See Addington v. Texas, 441 U.S. at 423, 99 S. Ct. at 1808. In view of the very serious and irreversible consequences that flow from an erroneous prediction of an acquittee's harmlessness after *152 release, it is not inappropriate to put the risk of error on the acquittee, who is foreclosed from repetitioning only for six months. Particularly is this true in light of the high risk of error produced by the problems inherent in predicting the post-release conduct of a BRI acquittee. See State v. Cuvelier, 175 Conn. 100, 109 n. 5, 394 A.2d 185, 189 n. 5 (1978) (studies cited). In its report to the President's Commission on Mental Health, the Task Panel on Legal and Ethical Issues found that:

The research suggests that the validity of psychological predictions of dangerous behavior, at least in the sentencing and release situations we are considering, is extremely poor, so poor that one could oppose the use of such predictions on the strictly empirical grounds that mental health professionals are not competent to make such judgments.

Mental Health and Human Rights: Report of the Task Panel on Legal and Ethical Issues, 20 Ariz.L.Rev. 49, 129 (1978). See generally Cocozza & Steadman, 29 Rutgers L.Rev. at 1086-99.

In many of the clear-and-convincing-evidence cases, the courts require that stricter type of proof from the party who seeks to change a prior decision or to nullify its consequence. For example, in a hiring discrimination case, once the factfinder concludes that the applicant has proved a prima facie case of discrimination in not hiring the applicant, the employer can avoid the consequences of that finding only by proving by a heightened form of evidence that she would for other reasons have been rejected in any event. See Maine Human Rights Commission v. City of Auburn; cf. State v. Commeau, 409 A.2d 247 (Me.1979) (avoidance of tainting consequence on in-court identification of pretrial identification that has been found inadmissible). Under Maine's system for handling the insanity defense, a BRI acquittee has proved by a preponderance (at the criminal trial in which the State first has proved that he committed a crime as charged) that he was suffering from mental disease or defect at the time of the crime. On the basis of that proof by the defendant-acquittee himself, he is absolved of responsibility for his crime and he is automatically committed to a mental institution. That is the status quo that the acquittee on his petition for release seeks to reverse. He should not be able to upset that status quo unless he can prove his future non-dangerousness by a standard higher than the mere preponderance that sent him to a hospital rather than a prison. The time reference of the issue in the release hearing is different from that in the criminal proceeding, but that fact will merely aid the acquittee in carrying his enhanced burden on his release petition.

These considerations lead us to agree with those other jurisdictions, which include Alaska,[25] Illinois,[26] and Missouri,[27] that release BRI acquittees only upon proof by clear and convincing evidence that their mental illness no longer makes them dangerous to themselves or others.

V.

We recognize that the "clear and convincing evidence" standard of proof has in Maine law acquired two different meanings. See City of Auburn, 425 A.2d at 996 n. 3. First, viewed as an intermediate standard, clear and convincing proof is such as leads the trier of fact to find that the existence of the contested fact is highly probable, rather than merely more probable than not. See Colorado v. New Mexico, ___ U.S. at ___, 104 S. Ct. at 2437; McCormick's Handbook, supra, § 340, at 796. *153 The other definition of the "clear and convincing evidence" standard was laid out by us in Horner v. Flynn, 334 A.2d 194 (Me. 1975). We there held that the "clear and convincing evidence" required to establish fraud does not identify a level of proof higher than proof by a preponderance, but rather denotes the better quality of evidence that is required to satisfy the preponderance standard in that and some other special cases. In the cases since Horner where this court has extended the use of a "clear and convincing evidence" standard to new applications, we have, with one possible exception, adopted that phrase to refer to an intermediate standard of proof. See City of Auburn, 425 A.2d at 996 n. 3; State v. Commeau, 409 A.2d at 248; State v. Cefalo, 396 A.2d 233, 238 n. 11 (Me. 1979). But see Townsend v. Maine Bureau of Public Safety, 404 A.2d at 1020 n. 5.

For the reasons expressed below we conclude that the Horner definition of "clear and convincing evidence" should be abandoned. We adopt the first definition, by which the party with the burden of persuasion may prevail only if he can "place in the ultimate factfinder an abiding conviction that the truth of [his] factual contentions are `highly probable'." Colorado v. New Mexico, ___ U.S. at ___, 104 S. Ct. at 2438.

As a practical matter, the Horner definition of "clear and convincing evidence" removes the higher standard of proof aspect of the lower court's factual findings from appellate review. Under Horner the question whether the evidence "which by its nature is capable of inducing belief does in fact induce belief is the responsibility of the factfinder to determine." Horner, 334 A.2d at 200 (emphasis in original). In Horner itself, the Law Court upheld a finding of fraud even though the trial court had given the jury, albeit without objection, a mere preponderance instruction without any of the "clear and convincing evidence" qualifications. Id. at 203. In effect, the appellate court reviews a finding in favor of the moving party under Horner just as if the moving party needed only to establish his allegations by a preponderance. Believing, as we do, that the policies that motivated the imposition of the "clear and convincing evidence" standard apply with equal force at both the factfinding and appellate stages, we prefer a definition of "clear and convincing evidence" that allows meaningful appellate review of the lower court's findings. Under the intermediate standard of proof we can address the question whether the factfinder could reasonably have been persuaded that the required factual finding was or was not proved to be highly probable.

The application of the Horner definition at the factfinding level creates needless difficulties. Under Horner the factfinder, often a jury, is required to make separate, confusing and potentially inconsistent determinations. After the jury has heard all of the evidence, it is asked to decide whether the moving party has offered evidence with a "high belief-inducing capability," i.e., evidence of a type that may prove a factual conclusion to be highly probable. Any factfinder may be hard pressed to separate the question whether the moving party's evidence has a "high belief-inducing capability" from the question whether the evidence, as a whole, is persuasive of the fact in dispute. As one commentator has argued:

By requiring that the jury evaluate the burdened party's presentation to determine whether it has a "high belief-inducing capability," however, Horner seems to require a preliminary assessment that the asserted facts are highly probable. Any effort to convey to a jury in a single set of instructions the concepts of "preponderance" or greater weight of the evidence and "high belief-inducing capability," with or without acknowledging this contradiction, is likely to result in confusion. [footnote omitted]

Note, Horner v. Flynn: A Preponderance of Clear and Convincing Evidence, 28 Me. L.Rev. 240, 248 (1976). In theory, Horner *154 requires the same two-step assessment of the evidence by a judge sitting as the trier of facts. But, for a trial judge also, the Horner double process is difficult at best to carry out, and at worst is likely to water down the significance of the heightened standard of proof required to protect important public interests.

As we have explained, a standard of proof serves to allocate the risk of error and to instruct the factfinder as to the degree of confidence society expects for a particular decision. To effectuate those purposes a standard of proof should operate to set the degree to which the factfinder must be persuaded of a particular factual conclusion. Where, as here, an important public interest and the desire to preserve prior judicial orders and adjudications lead us to employ the "clear and convincing evidence" standard for the release of BRI acquittees, the lower court must find the required factual conclusion to be "highly probable." Under the Horner approach the factfinder need only be persuaded that the factual conclusion in dispute is more probable than not. The additional requirement in Horner that the conclusion be supported by high quality evidence cannot adequately satisfy the objectives of the "clear and convincing evidence" standard. A "high quality evidence" requirement does not serve to allocate the risk of error and serves only indirectly to instruct the factfinder of the degree of confidence expected for a certain result. For example, there are many instances in which the evidence on both sides might be deemed of "high quality." In such instances, Horner permits the party bearing the burden of proof to prevail despite having only a bare preponderance of the evidence. Although the introduction of high quality evidence may well be an important element in meeting the intermediate standard of proof, that alone would not suffice. The factfinder must be persuaded, on the basis of all the evidence, that the moving party has proved his factual allegations to be true to a high probability. That degree of confidence effectuates the policy purposes for which we have, in this case and others, adopted the "clear and convincing evidence" standard.

Finally, Horner is out of step with the law of the rest of the country. The federal courts and virtually all other states treat "clear and convincing evidence" as an intermediate standard of proof lying between the preponderance and the reasonable doubt standards. See 28 Me.L.Rev. at 242-45. If we found some wisdom in Horner not recognized elsewhere or if Horner served some special Maine need, we would not follow the near-universal authority elsewhere. Finding no such considerations, however, we rejoin the mainstream of legal jurisprudence on this issue.

For these theoretical and practical concerns, we overrule Horner v. Flynn, 334 A.2d 194 (Me.1975).[28] We define "clear and convincing evidence" as that which establishes a factual conclusion to be "highly probable."

VI.

At this time we elect not to decide finally any question of the constitutionality of requiring clear and convincing proof before a BRI acquittee may be released. On his appeal Taylor has attacked the reasonable doubt standard of Shackford as violative of both due process and equal protection; but the parties have not addressed in any focused fashion the constitutionality of the reduced level of proof that we today announce for release proceedings. It is sufficient for present purposes to note that the clear-and-convincing-evidence standard has been held constitutional in other jurisdictions. See In re King, 114 Ill.App.3d 346, 70 Ill.Dec. 9, 448 N.E.2d 887 (1983); Allen *155 v. Radack, 426 F. Supp. at 1059 ("the state of South Dakota is free to adopt for future cases any standard it wishes, short of `beyond a reasonable doubt'"). When and if the constitutional issue arises on a future appeal by Taylor or any other BRI acquittee, we will be able to decide the question in a fully developed factual context and with the aid of focused argument by counsel. See United States v. Fruehauf, 365 U.S. 146, 157, 81 S. Ct. 547, 553, 5 L. Ed. 2d 746 (1961); UV Industries, Inc. v. Posner, 466 F. Supp. 1251, 1258 (D.Me.1979).

VII.

On remand the Superior Court should open up the hearing on Taylor's petition for the receipt of up-to-date information on his mental condition and the likely course of his future recovery. Only if Taylor successfully carries his burden of showing his eligibility by clear and convincing proof should the Superior Court approve his proposed modified release treatment. He may be released on that program only if he can "place in the [Superior Court justice as] the ultimate factfinder an abiding conviction that the truth of [his] factual contentions are `highly probable'." Colorado v. New Mexico, ___ U.S. at ___, 104 S. Ct. at 2438.

The entry is:

Judgment of the Superior Court vacated.

Remanded for further proceedings consistent with the opinion herein.

All concurring.

NOTES

[1] By P.L.1981, ch. 493, the Department of Mental Health and Corrections was split into two departments, the Department of Mental Health and Mental Retardation and the Department of Corrections. The responsibility for the operation of the state mental health institutes lies with the Commissioner of Mental Health and Mental Retardation. See 34 M.R.S.A. § 1 (Supp. 1983-1984). References in 15 M.R.S.A. §§ 103, 104-A to the "Commissioner of Mental Health and Corrections" have been amended to read and mean the "Commissioner of Mental Health and Retardation." P.L.1981, ch. 493, § 2.

[2] Taylor appealed to the Law Court from the denial of one of those petitions but the Law Court dismissed the appeal as moot. See Taylor v. Comm'r of Mental Health and Corrections, 431 A.2d 1304 (Me.1981).

[3] Maine is said to be one of only two states (the other being Louisiana) that require the criminal act to be proved beyond a reasonable doubt before the insanity issue is considered. See S. Brakel et al., The Mentally Disabled and the Law ch. 12, pt. III(D)(2) (3d ed., in preparation, American Bar Foundation 1984).

[4] 15 M.R.S.A. § 104-A provides in pertinent part:

1. Release and discharge. The head of the institution in which a person is placed under section 103 shall, annually, forward to the Commissioner of Mental Health and Corrections a report containing the opinion of a staff psychiatrist as to the mental condition of that person, stating specifically whether he may be released or discharged without likelihood that he will cause injury to himself or to others due to mental disease or mental defect. The report shall also contain a brief statement of the reasons for the opinion. The commissioner shall forthwith file the report in the Superior Court for the county in which the person is hospitalized. The court shall review each report and, if it is made to appear by the report that any person may be ready for release or discharge, the court shall set a date for and hold a hearing on the issue of the person's readiness for release or discharge. At the hearing, the court shall receive the testimony of at least one psychiatrist who has observed or treated that person and any other relevant testimony. If, after hearing, the court finds that the person may be released or discharged without likelihood that he will cause injury to himself or to others due to mental disease or mental defect, the court shall order, as applicable:

A. Release from the institution, subject to conditions deemed appropriate by the court which conditions:

(1) May include, but are not limited to, outpatient treatment;

(2) Continue until terminated by the court;

(3) Are subject to annual review by the court; and

(4) May include supervision by the State Division of Probation and Parole for one year, which period may be extended for an additional year by the court upon review after the expiration of the first year; or

B. Discharge from the custody of the Commissioner of Mental Health and Corrections.

2. Modified release treatment. Any individual hospitalized pursuant to section 103 may petition the Superior Court for the county in which that person is hospitalized for a release treatment program allowing the individual to be off institutional grounds for a period of time, not to exceed 14 days at any one time. The petition shall contain a report from the institutional staff including at least one psychiatrist, and the report shall define the patient's present condition; the planned treatment program involving absence from the institution; the duration of the absence from the institution; the amount of supervision during the absence; the expectation of results from the program change and the estimated duration of the treatment program before further change. This petition shall be forwarded to the court no later than 60 days prior to the beginning of the modified treatment program. If the court considers that the individual being off the grounds as described in the treatment plan is inappropriate, it shall notify the hospital that the plan is not approved and shall schedule a hearing on the matter. The clerk of courts upon receipt of the proposed treatment program shall give notice thereof by mailing a copy to the district attorney and Attorney General, who may file objections and request hearing on the matter. If the court does not respond within 60 days to the proposed treatment plan and no objections and request for hearing are filed by the district attorney or Attorney General, it may then be put into effect by the administrator of the hospital on the assumption that the court approved the treatment plan.

The term "release" as used in this section means termination of institutional inpatient residency and return to permanent residency in the community.

A report shall be forwarded and filed and hearings shall be held in accordance with the first paragraph of subsection 1 without unnecessary delay when, at any time, it is the opinion of a staff psychiatrist that a patient hospitalized under section 103 may be released or discharged without likelihood that he will cause injury to himself or to others due to mental disease or mental defect.

A person hospitalized under section 103, or his spouse or next of kin, may petition the Superior Court for the county in which that person is hospitalized for a hearing under this section. Upon receiving the petition, the court shall request and be furnished by the Commissioner of Mental Health and Corrections a report on the mental condition of that person, as described in the first paragraph of subsection 1. A hearing shall be held on each petition, and release or discharge, if ordered, shall be in accordance with the first paragraph of subsection 1. If release or discharge is not ordered, a petition shall not be filed again for the release or discharge of that person for 6 months. Any person released under this section or his spouse or next of kin may at any time after 6 months from the release petition the Superior Court for the county in which he was hospitalized for his discharge under this section. If discharge is not ordered, a petition for discharge may not be filed again for 6 months.

[5] A hearing on a proposed program for modified release treatment may be scheduled on the Superior Court's own initiative or on request by either the district attorney or the Attorney General. 15 M.R.S.A. § 104-A(2).

[6] Section 104-A also lacks an express declaration as to who bears the burden of proof in a proceeding to determine a BRI acquittee's eligibility for release. The implication seems clear, however, that the burden of proof (both as to initial production and ultimate persuasion) restsas in any other civil proceedingupon the moving party, namely, the acquittee, whether he is the petitioner or only the person on whose behalf the proceeding is undertaken. See State v. Carter, 64 N.J. 382, 407, 316 A.2d 449, 463 (1974); E. Cleary, McCormick's Handbook of the Law of Evidence § 337 (2d ed. 1972); Hamann, The Confinement and Release of Persons Acquitted by Reason of Insanity, 4 Harv.J. on Legis. 55, 89 (1966).

[7] 15 M.R.S.A. § 104 (Supp.1967), the forerunner of our present section 104-A.

[8] Ragsdale v. Overholser, 281 F.2d 943 (D.C.Cir. 1960). As the Law Court noted in State v. Shackford, 262 A.2d at 366, Ragsdale had already been expressly overruled in the same circuit by Bolton v. Harris, 395 F.2d 642 (D.C.Cir. 1968).

[9] See, e.g., Taylor v. Comm'r of Mental Health and Corrections, n. 1 above (appeal mooted by subsequent modified release treatment); In re Fleming, 431 A.2d 616 (Me.1981) (appeal on sufficiency of the evidence only); Buzzell v. Comm'r of Mental Health and Corrections, 423 A.2d 246, 246 (Me.1980) (memorandum of decision) ("Regardless of the level of proof required, we find no error in the Court's conclusion").

[10] Like Maine, the District of Columbia makes the insanity defense an affirmative plea that must be proved by the accused by a preponderance of the evidence. D.C.Code Ann. § 24-301(j) (Supp.1981).

[11] At present, by statute in South Dakota, a person committed as guilty but mentally ill has all of the rights to review and rehearings given to persons under civil commitment. See S.D. Comp.Laws Ann. § 23A-26-12 (Supp.1983). Persons under civil commitment have a right to periodic rehearing, which rehearings are the same as an initial civil commitment proceeding. See S.D.Comp.Laws Ann. § 27A-12-17 (Supp. 1983). A court may order civil commitment if it "finds by clear and convincing evidence that the proposed patient is mentally ill and in need of treatment ...." S.D.Comp. Laws Ann. § 27-9-18 (1976). It appears, therefore, that BRI acquittees in South Dakota are entitled to annual reviews of their status, at which hearings the state must prove the need for continued commitment.

[12] This conclusion is not contradicted by our cases holding that the presumption of legitimacy of any child born in wedlock is so strong that the contrary must be proved beyond a reasonable doubt. Buzzell v. Buzzell, 235 A.2d 828 (Me.1967); Ventresco v. Bushey, 159 Me. 241, 191 A.2d 104 (1963). The strength of the presumption, even though rebuttable, coupled with the advancing technology of blood groupings, have led many jurisdictions to use the reasonable doubt standard for proving the past fact of illegitimacy. See, e.g., Simmons v. Simmons, 479 S.W.2d 585 (Ky.1972) (decided under Uniform Act on Paternity).

[13] Colorado v. New Mexico, ___ U.S. ___, ___, 104 S. Ct. 2433, 2437, 81 L. Ed. 2d 247 (1984).

[14] Santosky v. Kramer, 455 U.S. 745, 769-70, 102 S. Ct. 1388, 1402-03, 71 L. Ed. 2d 599 (1982). See also 22 M.R.S.A. § 4055 (Supp.1983-1984), enacted by P.L. 1979, ch. 733, § 18 (termination of parental rights).

[15] Addington v. Texas, 441 U.S. 418, 431-33, 99 S. Ct. 1804, 1812-13, 60 L. Ed. 2d 323 (1979).

[16] United States v. Wade, 388 U.S. 218, 240, 87 S. Ct. 1926, 1939, 18 L. Ed. 2d 1149 (1967). For application of the same principle in Maine, see State v. Commeau, 409 A.2d 247 (Me.1979); State v. Cefalo, 396 A.2d 233, 238 (Me.1979).

[17] Woodby v. Immigration and Naturalization Service, 385 U.S. 276, 285, 87 S. Ct. 483, 487, 17 L. Ed. 2d 362 (1966) (deportation); Nowak v. United States, 356 U.S. 660, 668, 78 S. Ct. 955, 960, 2 L. Ed. 2d 1048 (1958) (denaturalization); Nishikawa v. Dulles, 356 U.S. 129, 133, 78 S. Ct. 612, 615, 2 L. Ed. 2d 659 (1958) (expatriation).

[18] Gertz v. Robert Welch, Inc., 418 U.S. 323, 332, 94 S. Ct. 2997, 3003, 41 L. Ed. 2d 789 (1974); New York Times Co. v. Sullivan, 376 U.S. 254, 285-86, 84 S. Ct. 710, 728-29, 11 L. Ed. 2d 686 (1964).

[19] Louisiana State Bar Ass'n v. Edwins, 329 So. 2d 437, 441-42 (La.1976); In re Conduct of Chambers, 292 Or. 670, 673, 642 P.2d 286, 288 (1982).

[20] Occidental Life Ins. Co. v. Marsh, 5 Ariz.App. 74, 75-76, 423 P.2d 150 (1967).

[21] Shepherd v. Shepherd, 81 Mich.App. 465, 265 N.W.2d 374, 376 (1978) ("clear and convincing proof"). See also note 12 above.

[22] American Universal Insurance Co. v. Falzone, 644 F.2d 65, 67 (1st Cir.1981) (diversity case applying Maine law).

[23] DeBlanco v. Dooley, 164 N.J.Super. 155, 158, 395 A.2d 909, 911 (1978).

[24] Along with a consideration of the amount of time that has passed since the acquittee committed the criminal act or any other overt act injuring others, the court considering the acquittee's petition for release should assess "the possibility that acquittee's confinement may have prevented the commission of other overt acts indicative of a substantial threat of serious bodily harm to others." Standard 7-7.8(b), American Bar Association Criminal Justice Health Standards (adopted at Chicago, Ill., August 7, 1984). See Pitts v. State, 151 Ga.App. 691, 261 S.E.2d 435, 438 (1979); People v. Howell, 196 Colo. 408, 586 P.2d 27, 30-31 (1978).

[25] See Alaska Stat. § 12.47.090(c)(e) (Supp. 1983) ("clear and convincing evidence").

[26] See Ill.Ann.Stat. ch. 38, § 1005-2-4(g) (Smith-Hurd Supp.1985-1985); In re Risner, 110 Ill.App.3d 368, 372, 66 Ill.Dec. 115, 119, 442 N.E.2d 541, 545 (1982) ("clear and convincing evidence").

[27] See State v. Pedersen, 651 S.W.2d 639, 641 (Mo.App.1983) (BRI acquittee released only if he proves "clearly and obviously" that he "does not have and in the reasonable future is not likely to have a mental disease or defect rendering him dangerous").

[28] We simultaneously overrule the adoption of the Horner quality-of-evidence meaning of the phrase "clear and convincing evidence" used in 22 M.R.S.A. § 4055 (Supp.1983-1984) for the required proof of the grounds for termination of parental rights. See In re Shannon R., 461 A.2d XXX-XXX-XX (Me.1983); see also In re Merton R., 466 A.2d 1268, 1269 (Me.1983); In re Edmund M., 475 A.2d 1143 (Me.1984).