The plaintiff had a jury verdict in the Superior Court in a tort action for libel. The case came before us on the defendant’s outline bill of exceptions and was decided on May 6, 1974. Stone v. Essex County Newspapers, Inc. 365 Mass. 246 (1974). Thereafter, on June 25, 1974, the Supreme Court of the United States decided Gertz v. Robert Welch, Inc. 418 U. S. 323 (1974). The plaintiff thereupon filed a petition, based on the holdings of the Supreme Court in the Gertz case, for a rehearing of the instant case. We granted the petition for rehearing, and have reconsidered the matter on new briefs filed by the parties.

The defendant claimed exceptions to the judge’s denial of the defendant’s motion for a directed verdict, and to certain of the judge’s instructions to the jury. We conclude that there was no error in the refusal to direct a verdict, but by reason of errors now apparent in the judge’s charge to the jury as considered in light of the holdings of the Gertz case, this case must be remanded to the Superior Court for a new trial.

*851In particular, as will be seen, we hold that a plaintiff who is not a public officer or a public figure may recover damages in an action for libel by proof of negligence in the publishing of the libel by the defendant, its agents or servants, even though the libel occurred in the reporting of an event of public or general concern. We further hold that a plaintiff who is a public officer or a public figure may in such an action recover only on proof of “actual malice” (wilful or reckless disregard of the truth in the publishing of the libel). Also, while it now appears that punitive damages may be constitutionally permissible in certain cases, we, placing primary emphasis on the necessity for protection of freedom of speech and the press, decline to adopt a rule allowing punitive damages. Rather, we affirm the principle that damages for defamation in this Commonwealth may be assessed only for actual injury and only on a compensatory basis, subject to searching judicial scrutiny at the trial and appellate levels. Finally, in any case where the plaintiff must show knowledge of falsity or reckless disregard of the truth, he must establish his proof, not merely by the fair preponderance of the evidence, but by “clear and convincing proof.”

We restate the facts as presented in the original hearing before this court. On November 4, 1969, Jeffrey C. Stone, the then twenty year old son of the plaintiff, appeared in District Court charged with being present where narcotic-drugs were illegally kept and with illegal possession of narcotics. A tablet alleged to be a “harmful drug” was introduced in evidence. The city marshal, Robert F. Jones, testified that the other defendants in the District Court case had indicated to him that the defendant Stone was the owner of the harmful drug.

The plaintiff from 1963 to 1972 served on the Newburyport Redevelopment Authority, owned a catering business, and was food service director for Newburyport schools.

*852Anthony Pearson, a reporter for the defendant’s newspaper, the Newburyport Daily News, was in court covering the proceedings. Pearson had been at work just four months as a reporter and had received only several hours of instruction in the work. Unaware that there was a reporter’s table near the witness stand, Pearson sat in the back of the court room. So positioned, he had trouble hearing some of the witnesses, including Jones.

Pearson interpreted Jones’s testimony to be that “Mr. Stone”1 was the owner of the “harmful drug,” and inferred that the title “Mister” was used to distinguish the father, who was in the court room, from the son.

That evening, Pearson wrote his story on the trial, translating the “Mr. Stone” of his notes to “John J. Stone,” which he had discovered the father’s name to be. He submitted it to William Coltin, the editor who ordinarily checked over and edited his copy. Coltin testified that he read it about midnight and was “surprised” at the information about the plaintiff (whom he had known for twenty years and whom he considered an “excellent citizen”), but accepted it as the testimony of a reliable public official under oath. He “may have” been surprised enough to question Pearson but did not see the reporter’s notes on the story; he very rarely went back to check a reporter’s notes. The article, which had been written for inclusion on November 5, 1969, the day following the trial, was crowded out and its publication postponed for twenty-four hours. During that time Coltin did not communicate any concern about the story to his superiors.

There also was evidence from which the jury could infer that police testimony was produced in the District Court proceeding to show that the substance in question *853was not a harmful drug or narcotic, and that Pearson’s notes and the news story did not include an account of that testimony.

The article was published on November 6, 1969. Shortly after it reached the public, the plaintiff called Coltin to complain of its inaccuracy. Coltin discussed the matter with John J. O’Neil, the managing editor, and then checked with Jones and discovered the plaintiff had had nothing to do with the case. O’Neil next consulted the editor and general manager of the paper, and then called the plaintiff and discussed on which page a retraction would be printed. O’Neil offered to get the plaintiffs approval of the retraction before printing it and they met the next morning for that purpose. The plaintiff “said it was fine but the damage had already been done.”

1. We turn first to a consideration of the instructions to the jury. The defendant’s preliminary argument, which is apposite to the directed verdict issue as well as to the instructions, is that the article did not charge the plaintiff with a crime as it only referred to his ownership of the drug, and the crime, if any, would have been in its sale or giving away. Even if we accept the defendant’s understanding of the criminal law as correct, this argument avails it nothing. While an imputation of crime is defamatory per se, Lynch v. Lyons, 303 Mass. 116, 118-119 (1939), the general test for libel is much broader: written words which would tend to hold the plaintiff up to scorn, hatred, ridicule or contempt, in the minds of any considerable and respectable segment in the community. Ingalls v. Hastings & Sons Publishing Co. 304 Mass. 31 (1939). The judge’s charge clearly and properly left these issues to the jury, who were instructed to consider damages only if they found the publication libellous, either by imputing commission of a crime or otherwise harming the plaintiff’s reputation.

2. The defendant excepted to the judge’s failure, in instructing on the issue of the common law privilege for *854reports of a judicial proceeding, to charge that the accuracy required to claim the privilege is substantial accuracy and does not require correctness in all particulars. Thompson v. Boston Publishing Co. 285 Mass. 344, 348-349 (1934). We find no error. The standard supplied by the trial judge, “fair and accurate report,” was at least as favorable to the defendant as it had any right to expect. We have previously held that accuracy is required “at least in regard to all material matters.” Sweet v. Post Publishing Co. 215 Mass. 450 (1913). Our statement in another mistaken identity case is almost directly applicable here: “A publication which identifies a person who had nothing to do with the proceedings as the one against whom the proceedings were directed can be neither fair nor accurate.” Whitcomb v. Hearst Corp. 329 Mass. 193, 199 (1952). The jury were wholly warranted in concluding that this privilege did not apply.

3. There was error in the judge’s instructions on the constitutional aspects of the case.2 The judge ruled that the standards enunciated in New York Times Co. v. Sullivan, 376 U. S. 254 (1964), were inapplicable. He charged the jury, in substance, that a verdict for the plaintiff was warranted on proof, without more, of publication by the defendant of a falsehood which was defamatory oil the plaintiff. Essentially, the judge thus instructed that the defendant newspaper could be found liable without fault. This, whether the plaintiff is a private person, or a public figure or public official, was error. ' Accordingly, there must be a new trial and the jury must be charged in accordance with the holdings of this opinion.

*855We turn now to the constitutional issues raised by the decision in Gertz v. Robert Welch, Inc. 418 U. S. 323 (1974), a decision that in large measure prompted our grant of this rehearing. We appreciate that we are dealing here with conflicting interests. On the one hand, the tort law of this Commonwealth has long recognized a right of redress to one who suffers injury to his reputation by the publishing of a defamatory falsehood. On the other hand, freedom of expression is guaranteed by the First Amendment to the United States Constitution as applicable to the States through the Fourteenth Amendment. In this case, involving as it does the right of the press to publish and disseminate news and the right of an individual to be free from a defamatory mark on his reputation, a balancing of interests is necessary.

“[T]o insure the ascertainment and publication of the truth about public affairs, it is essential that the First Amendment protect some erroneous publications as well as true ones.” St. Amant v. Thompson, 390 U. S. 727, 732 (1968). The suggestion has been made that the First Amendment provides absolute protection for the press, protection extending even to knowing publication of falsehoods, but this suggestion has been rejected and is not the law.3 As a result, the issue is under what circumstances the publication of certain statements is expression not protected by the First Amendment’s proscription of any law “abridging the freedom of speech, or of the press.” Thus in this case we must define as matter of State law the standards by which a publisher or broadcaster is to be adjudged liable for the publication of falsehoods, to wit, we must decide whether such publication need simply be proved negligent or whether a *856heavier burden such as publication with reckless or intentional disregard of the truth is required.

Limitations on the power of State courts to award damages in libel actions were established by certain decisions of the United States Supreme Court. New York Times Co. v. Sullivan, 376 U. S. 254, 283 (1964), held that public officials cannot recover damages against critics of their official conduct without proof of actual malice (i.e., reckless or wilful disregard of the truth) in the publishing. The rule in turn was extended to include public figures who are not public officials. Curtis Publishing Co. v. Butts, 388 U. S. 130, 155 (1967) (Harlan, J.).

Further extending the doctrine of the New York Times Co. and Butts cases, the Supreme Court, in a plurality opinion, decided that the actual malice standard applied not only to criticism of a public figure or of the official conduct of a public officer but also to the reporting of an event of public or general concern. Rosenbloom v. Metromedia, Inc. 403 U. S. 29, 44 (1971) (Brennan, J.). Thereafter, we set the standard in this Commonwealth for invoking the First Amendment protection for libellous material published without actual malice by holding, in accordance with the apparent requirements of the Metromedia case, that the relevant issue was not the status of the particular plaintiff involved, but rather the events which were the subject of the publication. Priestley v. Hastings & Sons Publishing Co. of Lynn, 360 Mass. 118, 123 (1971). See Twohig v. Boston Herald-Traveler Corp. 362 Mass. 807, 808-809 (1973).

Accordingly we held, on first deciding the instant case on May 6, 1974, that because the event which was reported was a matter of public interest, the New York Times Co. standard applied and the plaintiff was required to show that the libel was a wilful or reckless publication of a defamatory falsehood. We held, also, that in these circumstances involving a public prosecution, it was immaterial whether the plaintiff was a public *857person or a private person, since the matter was one of public interest under the Metromedia case. Cf. Cox Bdcst. Corp. v. Cohn, 420 U. S. 469 (1975).4 However, we remanded the case for a new trial because the judge had essentially charged that liability could be imposed without fault.

Thereafter, the case of Gertz v. Robert Welch, Inc. 418 U. S. 323 (1974), the case which impelled us to allow the plaintiffs petition for reconsideration of the instant case, was handed down. In an opinion written by Mr. Justice Powell for five members5 of the court, the Gertz case held, inter alla, that (1) the First Amendment protection afforded to defendants against defamation suits by public persons is not to be extended to defamation suits by private individuals even though the defamatory statements concern an issue of public or general interest, id. at 346, (2) so long as they do not impose liability without fault, the States may define for themselves the appropriate standard of liability for a publisher or *858broadcaster of defamatory falsehood injurious to a private individual, id. at 347, and (3) the States may not permit recovery of presumed or punitive damages, at least when liability is not based on a showing of knowledge of falsity or reckless disregard for the truth, id. at 349.

Accordingly, we hold that private persons, as distinguished from public officials and public figures, may recover compensation on proof of negligent publication of a defamatory falsehood. But see Walker v. Colorado Springs Sun, Inc. 188 Colo. 86 (1975). See also AAFCO Heating & Air Conditioning Co. v. Northwest Publications, Inc. Ind. App. (1974).a Publishers and broadcasters of defamatory falsehoods concerning public officers or figures must be protected by the New York Times Co. standard, but “injury to the reputation of private individuals requires that a different rule should obtain with respect to them.” Gertz v. Robert Welch, Inc., supra, at 343.6 The State’s interest here resides in the private individual’s right to the protection of his own good name, for this “reflects no more than our basic concept of the essential dignity and worth of every human being — a concept at the root of any decent system of ordered liberty.” Rosenblatt v. Baer, 383 U. S. 75, 92 (1966) (Stewart, J., concurring).

*859In considering whether private individuals should be governed by a different rule from that governing public persons, we find the Gertz case convincing in its reasoning that public officials and public figures usually enjoy greater access to the channels of effective communication and hence have a more realistic opportunity to counteract false statements than private individuals normally enjoy, and that public persons have thrust themselves forward, have invited attention and comment, and thus (unlike private persons) have voluntarily exposed themselves to increased risk of injury from defamatory falsehoods. Thus, private individuals are not only more vulnerable to injury than public officials and public figures — they are also more deserving of recovery. Gertz v. Robert Welch, Inc., supra, at 344.7

4. We turn now to a consideration of the principles which we hold shall control the assessment of damages in defamation actions in this Commonwealth. The rule defining the limits of damages, like the rules defining *860liability, must take cognizance of the need to reconcile the State interest in permitting defamation actions against the competing values of the First Amendment. The Gertz case so requires: “[T]he States have no substantial interest in securing for plaintiffs such as this petitioner, gratuitous awards of money damages far in excess of any actual injury.” Gertz v. Robert Welch, Inc., supra, at 349. However, the Gertz opinion also apparently provides that a State may constitutionally permit the award of punitive damages to a plaintiff who proves under the New. York Times Co. standard, wilful or reckless defamation by a publisher or broadcaster.. Id. at 349. Presumably in such aggravated cases, the damages may be permitted as private fines levied by civil juries to punish reprehensible conduct and to deter its future occurrence.

We reject the allowance of punitive damages in this Commonwealth in any defamation action, on any state of proof, whether based in negligence, or reckless or wilful conduct. We so hold in recognition that the possibility of excessive and unbridled jury verdicts, grounded on punitive assessments, may impermissibly chill the exercise of First Amendment rights by promoting apprehensive self-censorship.

We reaffirm the following as the controlling principles in this Commonwealth. In a case of defamation the plaintiff’s recovery is limited to actual damages, which are compensatory for the wrong done by the defendant. Ellis v. Brockton Publishing Co. 198 Mass. 538 (1908). Where specific harm is alleged to have resulted from the defendant’s tortious conduct, such harm may be pleaded and damages recovered. Muchnick v. Post Publishing Co. 332 Mass. 304 (1955). Cf. Lewis v. Vallis, 356 Mass. 662 (1970). Otherwise the plaintiff is limited to compensatory damages for actual injury, which include mental suffering, Chesley v. Tompson, 137 Mass. 136, 137 (1884); Pion v. Caron, 237 Mass. 107, 111 (1921), and harm to reputation. Ellis v. Brockton Publishing *861Co., supra. Punitive damages are prohibited, Ellis v. Brockton Publishing Co., supra, even on proof of actual malice. G. L. c. 231, § 93, as appearing in St. 1943, c. 360. In addition, the defendant may introduce evidence of a retraction in mitigation of damages, G. L. c. 231, § 94, as appearing in St. 1943, c. 361, but the degree of mitigation, if any, is for the jury. Whitcomb v. Hearst Corp. 329 Mass. 193 (1952).

The United States Supreme Court has recognized, at least by implication, the difficulties in instructing a jury on compensatory damages for such abstract elements as impairment of reputation and mental suffering. In the Gertz case, 418 U. S. at 349-350 (1974), it is stated: “We need not define ‘actual injury,’ as trial courts have wide experience in framing appropriate jury instructions in tort actions. Suffice it to say that actual injury is not limited to out-of-pocket loss. Indeed, the more customary types of actual harm inflicted by defamatory falsehood include impairment of reputation and standing in the community, personal humiliation, and mental anguish and suffering. Of course, juries must be limited by appropriate instructions, and all awards must be supported by competent evidence concerning the injury, although there need be no evidence which assigns an actual dollar value to the injury.”

Because of constitutional considerations, and the potential difficulties in assessing fair compensation in such cases, in our view both trial and appellate judges have a special duty of vigilance in charging juries and reviewing verdicts to see that damages are no more than compensatory. Thus, Chief Justice Shaw’s solution to the difficulty of measuring libel damages, that “the court will be slow to pronounce a verdict excessive,” Treanor v. Donahoe, 9 Cush. 228, 231 (1852), has limited applicability since First Amendment values are at stake. Cf. Curtis Publishing Co. v. Butts, 388 U. S. 130, 138 (1967) (Harlan, J.); Rosenbloom v. Metromedia, Inc. 403 U. S. 29, 40 (1971) (Brennan, J.).

*8625. Finally, we discuss several concepts that will be crucial in the retrial of this case, or in similar cases: concepts involving the nature of “public official,” “public figure,” and “actual malice.” “Negligence,” a more familiar concept, needs no such discussion. Of necessity, our discussion is broader than is required by the evidence at the prior trial of the instant case, because we have no way of knowing what the evidence will be on retrial, or what jury instructions will be required based on that evidence.

We turn first to the meaning of “public official.” When we first heard and decided this case, this term was unimportant in our consideration, because under the law then controlling the case clearly turned on the fact that the judicial proceeding which was reported was an event of public or general concern. Under the present law, the determination whether the plaintiff is a public official becomes a paramount issue in this case.

The determination of the plaintiffs status, whether public official or public figure or private person, “as is the case with questions of privilege generally, ... is for the trial judge in the first instance.” Rosenblatt v. Baer, 383 U. S. 75, 88 (1966). In Lewis v. Vallis, 356 Mass. 662, 668 (1970), where the point did not affect the decision, we interpreted this to mean that the question whether the plaintiff was a public figure “was a question of law for the court.”8 A full statement of the rule would seem to be that the question whether the plaintiff is a public official or a public figure is one for the court to answer whenever (a) all of the facts bearing thereon are uncontested or agreed by the parties, (b) the case is tried before a judge without a jury, or (c) all of the facts bearing thereon are specially found and reported by the *863jury by way of answers to special questions submitted to them; and that otherwise, in a case tried to a jury, it is a question for the jury to answer after instructions by the judge on the applicable law and on what facts must be found to constitute the plaintiff a public official or a public figure.

It is not crucial that the newspaper article did not refer to the plaintiff’s public capacity. It is clear that the defendant did not criticize the plaintiff’s official conduct per se, as in New York Times Co. v. Sullivan, 376 U. S. 254, 283 (1964), and Rosenblatt v. Baer, 383 U. S. 75, 87 (1966). Similarly unlike the plaintiffs in Monitor Patriot Co. v. Roy, 401 U. S. 265, 271 (1971), and Ocala Star-Banner Co. v. Damron, 401 U. S. 295, 299 (1971), the plaintiff was not at the time of the publication a candidate for public office. To a large extent his public position seems irrelevant to the defendant’s article. Yet that is not controlling, for “a charge of criminal conduct against an official or a candidate, no matter how remote in time or place, is always relevant to his fitness for office for purposes of applying the New York Times rule of knowing falsehood or reckless disregard of the truth.” Ocala Star-Banner Co. v. Damron, 401 U. S. 295, 300 (1971). Monitor Patriot Co. v. Roy, 401 U. S. 265 (1971).

The mere fact that a plaintiff was a government employee is also not determinative, for such employees in the lower ranks are clearly not public officials for purposes of the rule, but the designation of public official applies at least to government employees who have, or publicly appear to have, substantial responsibility for control of public affairs. Rosenblatt v. Baer, 383 U. S. at 85 (1966). New York Times Co. v. Sullivan, 376 U. S. at 283 (1964). In Rosenblatt v. Baer, supra, at 86, the court stated: “Where a position in government has such apparent importance that the public has an independent interest in the qualifications and performance of the person who holds it, beyond the general *864public interest in the qualifications and performance of all government employees, both elements we identified in New York Times are present and the New York Times malice standards apply.” With respect to this statement the court continued: “It is suggested that this test might apply to a night watchman accused of stealing state secrets. But a conclusion that the New York Times malice standards apply could not ■ be reached merely because a statement defamatory of some person in government employ catches the public’s interest; that conclusion would virtually disregard society’s interest in protecting reputation. The employee’s position must be one which would invite public scrutiny and discussion of the person holding it, entirely apart from the scrutiny and discussion occasioned by the particular charges in controversy.” Id. at 86-87 (fn. 13).

The New York Times Co. standard has been held to apply in cases involving an elected city commissioner of Montgomery, Alabama (New York Times Co. v. Sullivan); the mayor of a small city who was a candidate for county tax assessor (Ocala Star-Banner Co. v. Damron); a candidate for the United States Senate (Monitor Patriot Co. v. Roy); and a deputy chief of detectives for a large city police department (Time, Inc. v. Pape, 401 U. S. 279 [1971]). However, a lawyer was not held to be a public official, merely because he was an officer of the court. Gertz v. Robert Welch, Inc.

The plaintiff in this case was employed by the school department of the city of Newburyport; he was also a member of the Newburyport Redevelopment Authority. His position with the city’s school department was entitled “food service director.” His duties involved the purchasing of food and equipment for two school cafeterias, preparation of menus and hiring and firing of employees with the approval of the school committee. His salary, which was set by agreement with the school committee rather than by statute or ordinance, was in the amount of $150 or $160 a week in 1969, with $40 a *865week added during the football season for the handling of athletic funds.

In addition, the plaintiff was a member of the Newburyport Redevelopment Authority from 1963 to 1972. At the time of the publication of the alleged libel he was treasurer of the authority, bonded in the sum of $25,000 and in this position he was authorized to cosign checks, thus exercising a degree of management over $1,000,000 in authority funds during his term of office. It does not appear whether he was remunerated for these services, what his exact duties were and whether and to what extent he made policy decisions on public issues.

All of these facts must be considered at a new trial in light of the consideration that the purpose of the New York Times Co. privilege is to further “first, a strong interest in debate on public issues, and second, a strong interest in debate about those persons who are in a position significantly to influence the resolution of those issues.” Rosenblatt v. Baer, 383 U. S. 75, 85 (1966).

Since the inquiry into relevant facts such as remuneration, duties, and participation in decisions on public issues was limited at the prior trial, we express no opinion or prediction whether the plaintiff here was a public official at the time of the publication. We leave that ruling for development at a new trial.

Although we have before us for review the defendant’s exception to the denial of its motion for a directed verdict, this does not necessitate any further consideration of the public official issue. Since we have held, as shown later in this opinion, that the evidence warranted an inference of actual malice on the defendant’s part, the case was one for the jury’s consideration even if the plaintiff was ruled to be a public official.

6. If a ruling is required that the plaintiff was a public official, there would be no necessity for consideration whether he is also a public figure. Assuming that the “public figure” concept may be material, however, we turn to a discussion of its meaning. Whether a plaintiff *866is to be considered a “public figure” is also a matter not easily resolved by any sort of precise classification. In the Gertz case, the plaintiff was held not to be a public figure. This result was reached despite the facts that the plaintiff was a lawyer well known in legal circles, had published several books and articles on legal subjects, served as an officer of local civic groups and professional organizations and had long been active in community and professional affairs.

In the Gertz case, the court said: “That designation [public figure] may rest on either of two alternative bases. In some instances an individual may achieve such pervasive fame or notoriety that he becomes a public figure for all purposes and in all contexts. More commonly, an individual voluntarily injects himself or is drawn into a particular public controversy and thereby becomes a public figure for a limited range of issues. In either case such persons assume special prominence in the resolution of public questions. . . . We would not lightly assume that a citizen’s participation in community and professional affairs rendered him a public figure for all purposes. Absent clear evidence of general fame or notoriety in the community, and pervasive involvement in the affairs of society, an individual should not be deemed a public personality for all aspects of his life. It is preferable to reduce the public-figure question to a more meaningful context by looking to the nature and extent of an individual’s participation in the particular controversy giving rise to the defamation.” 418 U. S. at 351-352 (1974).

On the evidence presented at the trial of the instant case, it is our view that the plaintiff clearly was not shown to be a public figure. Although he was, aside from his official duties, an active participant in community affairs and a member of certain fraternal organizations, he had not “thrust himself into the vortex of . . . public issue [s], nor did he engage the public’s attention *867in an attempt to influence . . . [the] outcome [thereof].” Gertz v. Robert Welch, Inc., supra, 418 U. S. at 352 (1974).

7. We turn next to a consideration of the meaning of “actual malice.” This concept becomes relevant, of course, if the plaintiff is shown to be a public official or a public figure.

Mr. Justice Harlan’s plurality opinion in the earlier case of Curtis Publishing Co. v. Butts, 388 U. S. 130, 155 (1967), had allowed recovery, at least for a “public figure” who was not a public official, on a showing of “highly unreasonable conduct constituting an extreme departure from the standards of investigation and reporting ordinarily adhered to by responsible publishers.” This language implies an objective test and a requirement of something less than recklessness. This requirement fell short of actual malice and was never adopted by a majority of the court, however. Cf. Gertz v. Robert Welch, Inc. 471 F. 2d 801, 806, fn. 11 (7th Cir. 1972).

Actual malice is not necessarily proved in terms of ill will or hatred, but is proved rather by a showing that the defamatory falsehood was published with knowledge that it was false or reckless disregard of whether it was false. See New York Times Co. v. Sullivan, 376 U. S. 254, 279-280 (1964). “[Defeasance of the privilege is conditioned, not on mere negligence, but on reckless disregard for the truth.” Garrison v. Louisiana, 379 U. S. 64, 79 (1964).

“‘Reckless disregard,’ it is true, cannot be fully encompassed in one infallible definition. . . . [H]owever . . . [t]here must be sufficient evidence to permit the conclusion that the defendant in fact entertained serious doubts as to the truth of his publication. Publishing with such doubts shows reckless disregard for truth or falsity and demonstrates actual malice.” St. Amant v. Thompson, 390 U. S. 727, 730-731 (1968). Thus, the test is entirely a subjective one. That information was available *868which would cause a reasonably prudent man to entertain serious doubts is not sufficient. In order to negate the privilege, the jury must find that such doubts were in fact entertained by the defendant, or by the defendant’s servant or agent acting within the scope of his employment. The jury may, of course, reach this conclusion on the basis of an inference drawn from objective evidence, since it would perhaps be rare for a defendant in such a circumstance to admit to having had serious, unresolved doubts. But the jury must draw this conclusion in order to hold the defendant liable.

The constitutionally mandated standard of recklessness in this civil context thus requires actual malice,9 In this regard it is conceptually narrower than the Massachusetts common law concept of recklessness sufficient to impose criminal liability, e.g., in a homicide case. In the criminal law we have held that an objective standard of recklessness and thus proof of implied malice will suffice, and said of a defendant that although ‘“in fact he did not realize the grave danger, he cannot escape the imputation of wanton or reckless conduct in his dangerous act or omission, if an ordinary normal man under the same circumstances would have realized the gravity of the danger.’” Commonwealth v. Welansky, 316 Mass. 383, 398-399 (1944). Commonwealth v. Pierce, 138 Mass. 165 (1884). Cf. Am. Law. Inst., Model Penal Code, § 2.02 (2) (c) (Proposed Official Draft May 4, 1962).

Subsequent decisions, both in this court and in the United States Supreme Court, adopt and apply the subjective formulation in the St. Amant case rather than the objective standard of the criminal law. Twohig v. Boston Herald-Traveler Corp. 362 Mass. 807 (1973). Time, Inc. v. Pape, 401 U. S. 279 (1971).

*8698. In reviewing the denial of the defendant’s motion for a directed verdict we consider the evidence in light of the concept of actual malice. Thus we are faced with the issue whether the evidence adduced at trial constituted sufficient proof of the requisite state of mind for recklessness so as to warrant the jury in returning a verdict for the plaintiff. New York Times Co. v. Sullivan, supra. Despite evidence of what the jury could find to be gross carelessness on the part of Pearson, it does not appear that the evidence would have been sufficient to warrant a conclusion of recklessness on his part. Of course, we have no way of knowing what the evidence at a new trial may show as to his state of mind.

However, the evidence concerning the defendant’s news editor Coltin was sufficient to warrant submission of the case to the jury. Coltin allowed the story to be printed despite serious doubts as to its accuracy with respect to the plaintiff. Coltin admitted he was “surprised” by the report of the plaintiff’s involvement. He denied that this term was an understatement and stated that he accepted the reported testimony of the city marshal. Nevertheless, combining this admission with his testimony that he considered the plaintiff, whom he knew well, to be an “excellent citizen,” and the fact that the article was written by an inexperienced reporter, of whose minimal training Coltin was fully aware, a jury might draw the inference that the news editor had in fact entertained doubts as to the story’s accuracy. The detailed evidence of Coltin’s knowledge of the plaintiff’s reputation and character, all of which evidence could be found to be inconsistent with the nature of the crime charged, might well support such a finding.

Assuming a jury so found, the amount of time necessary and available for checking the accuracy of the story might be considered relevant to determine whether pushing aside or disregarding those doubts rose to the level of recklessness. In this regard, there was evidence of a delay of a full day in the publication of the story. *870Cf. Priestley v. Hastings & Sons Publishing Co. of Lynn, 360 Mass. 118, 123-125 (1971), and Curtis Publishing Co. v. Butts, 388 U. S. 130, 159 (1967). All told, there was sufficient evidence to warrant consideration by the jury. Therefore, it is appropriate for us to order a new trial of the case, rather than order that judgment should enter for the defendant.

9. In any case where the plaintiff is required to prove actual malice he must do so, not merely by the fair preponderance of the evidence, but by “clear and convincing proof.” See Gertz v. Robert Welch, Inc. 418 U. S. 323, 342 (1974); New York Times Co. v. Sullivan, 376 U. S. 254, 285-286 (1964); Restatement 2d: Torts (Tent, draft No. 21, 1975), § 580A, comment (f). It is an unusual development, indeed, that requires as matter of constitutional law that the jury must find facts according to such a standard, and must be charged accordingly; yet that is the apparent holding of the Supreme Court.10 In ruling on a defendant’s motion for a directed verdict in a defamation case, the judge certainly must apply the “clear and convincing proof” standard. That is to say, the judge must determine whether the jury would be warranted in concluding that malice was proved by clear and convincing evidence.11 Further, at least until the Supreme Court makes additional comment on the issue, if it does, the jury should be charged according to that standard.12

*871The New York Times and the Gertz cases offer no definition of the meaning of “clear and convincing proof,” to assist in formulating jury instructions. However, from other sources we find the phrase defined. Clear and convincing proof involves a degree of belief greater than the usually imposed burden of proof by a fair preponderance of the evidence, but less than the burden of proof beyond a reasonable doubt imposed in criminal cases. See Foley v. Coan, supra; Coghlin v. White, supra. It has been said that the proof must be “strong, positive and free from doubt” (Coghlin v. White, supra, at 55, quoting from Newell v. Homer, 120 Mass. 277, 280 [1876]), and “full, clear and decisive” (Kidder v. Greenman, supra, at 613, and cases cited). See generally, Wigmore, Evidence, § 2498 (3) (3d ed. 1940).

10. The defendants exceptions are sustained and the case is remanded to the Superior Court for a new trial on all issues.

So ordered.

(concurring in part and dissenting in part). I concur with all of the opinion in this case except part 9 thereof. Specifically, I disagree that as a matter of constitutional law (1) “[i]n any case where the plaintiff is required to prove actual malice he must do so, not merely by the fair preponderance of the evidence, but by ‘clear and convincing proof’” which “involves a degree of belief greater than the usually imposed burden of proof by a fair preponderance of the evidence, but less than the burden of proof beyond a reasonable doubt imposed in criminal cases,” (2) “the jury must find facts according to such a standard, and must be charged accordingly,” and (3) “ [i]n ruling on a defendant’s motion for a directed verdict in a defamation case, the judge certainly must apply the clear and convincing proof’ standard.”

I recognize that in New York Times Co. v. Sullivan, 376 U. S. 254, 285-286 (1964), the Supreme Court, through Mr. Justice Brennan, indicated that the proof of actual malice presented in that case would not constitutionally sustain the judgment because it lacked “convincing clarity.” I also recognize that in Rosenbloom v. Metromedia, Inc. 403 U. S. 29, 52 (1971), Mr. Justice Brennan, for three members of the court, transposed his New York Times Co. “convincing clarity” language into a requirement that actual malice be proved by “clear and convincing proof.” I further recognize that in Gertz v. Robert Welch, Inc. 418 U. S. 323, 342 (1974), a majority of the court interpreted the New York Times Co. standard as permitting a public official or public figure to “recover for injury to reputation only on clear and convincing proof that the defamatory falsehood was made with knowledge of its falsity or with reckless disregard for the truth.” It is doubtless possible to read these excerpts, as the court today has done, to require that juries be instructed to find for the defendant on the issue of malice unless they view the evidence on that issue as “clear and convincing” in the plaintiff’s favor. It is also possible to infer that this standard involves a degree of belief lying *873somewhere between that required to support a verdict in an ordinary civil case and that required to support a verdict of conviction in a criminal case. It may even be possible to interpret this language as requiring the trial judge in a jury case to direct a verdict1 for the defendant unless he is personally satisfied that the evidence establishing malice is clear and convincing. And perhaps the language can also be construed as demanding that members of appellate courts review trial transcripts and exhibits and thereafter reverse a denial of a motion for a directed verdict unless they are personally satisifed that the evidence establishing malice is clear and convincing. I, however, view this language as little more than confusing rhetoric — words which have been given no content by the United States Supreme Court. For reasons hopefully explained below, I would not give these words the content given them by the court today.

1. The Supreme Court has not defined “convincing clarity” or “clear and convincing.” Neither has it stated whether these undefined terms raise questions of law or of fact. The court today creates a definition for these phrases and seemingly rules that they raise questions of both law and fact. In considering these holdings, it is helpful to bear in mind that they are not explicitly mandated by the Supreme Court.

2. My feeling that the phrases “convincing clarity” and “clear and convincing” are essentially rhetorical is reinforced by the fact that in the Gertz case the court said: “Absent clear evidence of general fame or notoriety in the community, and pervasive involvement in the affairs of society, an individual should not be deemed a public personality for all aspects of his life” (emphasis supplied). *874418 U. S. at 352 (1974). It seems to me that if, as the court today holds, the plaintiff in a defamation action has a special burden of proof in regard to the issue of malice, then the defendant has a similar (although perhaps not identical) burden of proof on the issue whether the plaintiff is a public personality.

The question thus arises whether the burden of. adducing “clear” evidence on an issue is less than the burden of adducing “clear and convincing” evidence on an issue. If these phrases are not mere rhetoric, then each must set a precise, independent standard. If we hold, as the court apparently does, that one implication of the existence of such precise, independent standards is that juries must be instructed to find facts according to those standards, we raise the spectre of requiring trial judges in defamation cases to instruct juries as to four separate and distinct burdens of proof, falling variously on the plaintiff and defendant. That is to say that since the court today defines “clear and convincing” evidence as that which would satisfy a burden somewhere between those imposed by the ordinary preponderance of the evidence and reasonable doubt standards utilized in civil and criminal cases, and since “clear” evidence is presumably stronger than a preponderance of the evidence but not so strong as “clear and convincing” evidence, then a trial judge must instruct the jury as to the meaning of: (1) “a preponderance” of the evidence, by which most of the facts in issue must be found, (2) “beyond a reasonable doubt,” so that the standard can help define other terms, (3) “clear and convincing” evidence, by which malice must be proved, and (4) “clear” evidence, by which the public character of the plaintiffs personality must be proved. A juror listening to a judge instructing him to draw such fine distinctions in his levels of belief would likely agree with Mr. Bumble: “If the law supposes that, . . . the law is a ass, a idiot.” Dickens, Oliver Twist, ch. 51.

*875Moreover, it is by no means entirely clear that the problem of special burdens of proof is limited to the issues of malice and whether the plaintiff is a public personality. Circuit Judge Bell, specially concurring in Firestone v. Time, Inc. 460 F. 2d 712, 722-723 (5th Cir. 1972), cert. den. 409 U. S. 875 (1972), said: “The Supreme Court has not expressly added the requirement of clear and convincing proof of falsity to the plaintiff’s burden of proof. . . . Such a standard of proof seems implicit however, in the stated requirement in New York Times that plaintiff has the burden of showing by clear and convincing proof that publication was with knowledge of falsity or with reckless disregard as to falsity vel non. I conclude for the same constitutional reasons giving rise to this stringent proof requirement that the clear and convincing proof standard would also apply to proving that the statement was false in the first instance.” If I were to agree that a special burden of proof were properly applicable to any of the issues in a defamation case, I, like Circuit Judge Bell, would find it difficult to isolate one among many issues to which to apply such a special burden.

3. Even assuming that the “convincing clarity” and “clear and convincing” language in the New York Times Co., Rosenbloom, and Gertz cases does require some special instructions to the jury, I do not believe that this language requires that juries in this Commonwealth be instructed that they find certain facts with “a degree of belief greater than the usually imposed burden of proof by a fair preponderance of the evidence, but less than the burden of proof beyond a reasonable doubt imposed in criminal cases.” Admittedly, the quoted definition fairly defines a level of proof employed, as an outgrowth of ancient chancery practice, in many States on a few narrow issues, principally involving attempts to prove by oral evidence matters which ordinarily are required to be proved by written evidence. See McCormick, Evidence, § 340 (2d ed. 1972). It is admittedly also true, as *876pointed out in fn. 10 of the court’s opinion in this case, that some of our early cases indicated that we followed such a rule on similar issues. However, there are also indications in the latest of the cases cited by the court and in a subsequent case that phrases such as the ones involved here are substantially rhetorical, and too vague to admit of definition or to be employed in general use. In Kidder v. Greenman, 283 Mass. 601, 613-614 (1933), we said: “It is settled that where reformation of an instrument is sought on the ground of a mutual mistake the proof must be ‘full, clear, and decisive.’. . . Whether this rule requires proof ‘beyond a reasonable doubt’ as those words are used in the criminal law . . . or merely recognizes that, in determining the preponderance of evidence, a completed instrument is evidence of great weight of the intention of the parties thereto and such evidence is not readily overcome by paroi evidence of such intention, need not be considered, for the rule, however stated, is not applicable to the present case” (emphasis supplied). In Matter of Mayberry, 295 Mass. 155, 167 (1936), this court held “that cause for disbarment may be established by a fair preponderance of the evidence as in other civil causes and that proof beyond reasonable doubt as in criminal cases is not required.” We added: “Nor do we think that there is in this Commonwealth any rule of law establishing for such cases an intermediate standard of proof, such as that the evidence must be ‘clear and convincing’ or ‘not of doubtful character.’ Such midway expressions may have some place in emphasizing the care with which the trier of fact should approach the decision of an issue so important to the respondent as the loss of his profession. . . . But such terms are too vague to serve generally as a practical guide in the trial of cases.” Cf. Commonwealth v. Bell, 356 Mass. 724, 725 (1969).

In my view, we could accommodate our trial practice to the evidentiary requirements of New York Times Co., Rosenbloom, and Gertz simply by having judges instruct *877juries to find for the defendant in a defamation action unless the plaintiff proves every necessary element of his claim for relief by a preponderance of the evidence, but that on the issue of malice they should scrutinize the evidence with special care, not lightly inferring the presence of malice, and that they should find that the defendant acted with malice only if they find the evidence on that issue clear enough to convince them by a preponderance of the evidence. A similar instruction, running in favor of the plaintiff, could be given on whether the plaintiff is a public personality. Such instructions should be adequate to satisfy any constitutional requirements imposed by the Supreme Court without formally incorporating into our jurisprudence an unworkable standard of proof, a standard of proof which will be unintelligible to juries, and which we have heretofore generally eschewed.

4. The court’s opinion states that, “ [i]n ruling on a defendant’s motion for a directed verdict in a defamation case, the judge certainly must apply the ‘clear and convincing proof’ standard. That is to say, the judge must determine whether the jury would be warranted in concluding that malice was proved by clear and convincing evidence.” This statement, touching, as it does, on our historic allocation of functions between judge and jury, seems to me to raise a serious problem. Despite the court’s disclaimer of an intention to do so,2 it appears, inevitable that, by reason of the statement that the trial judge must make a preliminary determination whether the jury could find the evidence on the malice issue clear and convincing, the judge must to some degree evaluate thé weight and credibility of possibly conflicting and ambiguous evidence and draw his own inferences therefrom. I would prefer a statement to the effect that, in so far as malice is concerned, a case must go to the jury if there is any evidence from which malice could be *878inferred, and that it is for the jury alone to determine whether that evidence is clear and convincing. It is true that some Federal courts have adopted or advocated a practice in defamation cases being tried to juries of having judges grant directed verdicts in the face of conflicting or ambiguous evidence on the malice issue if personally unconvinced that the defendant in fact acted maliciously. I regard such a practice as improvident and would proscribe it in our courts unless the United States Supreme Court were unequivocally to require it as matter of Federal constitutional law. I do not believe that the Supreme Court has imposed such a requirement.

Perhaps the clearest statement of the practice I would not follow is contained in the concurring opinion of Circuit Judge Wright3 in Wasserman v. Time, Inc. 424 F. 2d 920, 922-923 (D. C. Cir. 1970), cert. den. 398 U. S. 940 (1970), setting out “what I conceive to be the proper procedure in handling the issue of actual malice .... In my judgment New York Times Co. v. Sullivan makes actual malice a constitutional issue to be decided in the first instance by the trial judge applying the Times test of actual knowledge or reckless disregard of the truth. . . . Unless the court finds, on the basis of pretrial affidavits, depositions or other documentary evidence, that the plaintiff can prove actual malice in the Times sense, it should grant summary judgment for the defendant. ... If the case survives the defendant’s summary judgment motion, the trial court at the close of the plaintiff’s case must decide whether actual malice has been shown with ‘convincing clarity.’ In making this judgment the court will judge the credibility of the witnesses and draw its own inferences from the evidence. *879If the trial is permitted to proceed, the court will be called upon again to make a judgment on the actual malice issue at the close of all of the evidence. If the motion for a directed verdict at this stage of the trial is denied, the actual malice issue, along with the other issues, is then submitted to the jury under the Times instruction without any indication from the court or counsel that the court has decided that the evidence shows actual malice with ‘convincing clarity.’ This two-step procedure in which both the trial judge and the jury must find actual malice before there can be judgment for the plaintiff provides the protection of the First Amendment freedom that Times sought to make secure in areas of public concern” (emphasis supplied). Circuit Judge Wright’s views were quoted with apparent approval by another Federal Court of Appeals in Bon Air Hotel, Inc. v. Time, Inc. 426 F. 2d 858, 864-865 (5th Cir. 1970).

This notion that the trial judge in a jury case, when ruling on a motion for a directed verdict, should “judge the credibility of the witnesses and draw . . . [his] own inferences from the evidence” is wholly foreign to our law. In upholding a denial of a motion for a directed verdict, we have said: “In deciding the correctness of this ruling upon the motion we need only consider evidence favorable to the plaintiff from whatever source it came, including evidence more favorable to her than that given by herself. If upon any reasonable view of the evidence there is found any combination of circumstances from which a rational inference may be drawn in favor of the plaintiff, then there was no error in the denial of the motion, even if there may be other and different circumstances disclosed in the evidence which, if accepted as true by the jury, would support a conclusion adverse to the plaintiff. The question presented by the motion was not the weight of the evidence but whether there was any evidence viewed in the light most favorable to the plaintiff that would support her cause of *880action.” Howes v. Kelman, 326 Mass. 696, 696-697 (1951). Identical or similar statements of the law are so numerous as to render any attempt at exhaustive citation pointless.

In light of our law on directed verdicts, I would adopt and apply the views expressed in Guam Fedn. of Teachers, Local 1581, of the Am. Fedn. of Teachers v. Ysrael, 492 F. 2d 438, 441 (9th Cir. 1974), cert. den. 419 U. S. 872 (1974). The court therein stated that “with respect, we are not persuaded by the second phase of Judge Wright’s analysis in Wasserman which suggests that in deciding these motions, the trial court should judge the credibility of witnesses and draw its own inferences from the evidence. We think that in a libel case, as in other cases, the party against whom a motion for summary judgment, a motion for a directed verdict, or a motion for a judgment notwithstanding the verdict is made is entitled to have the evidence viewed in the light most favorable to him and to all inferences that can properly be drawn in his favor by the trier of fact. We think, too, that in such cases it is not only not the duty of the judge, or of this court of appeal, to weigh the credibility of the evidence, or to draw inferences in favor of the moving party (except, of course, when no contrary inference can legitimately be drawn), but that neither the judge nor this court on appeal has the authority to weigh credibility or to choose among legitimate inferences in such cases. The standard against which the evidence must be examined is that of New York Times and its progeny. But the manner in which the evidence is to be examined in the light of that standard is the same as in all other cases in which it is claimed that a case should not go to the jury. If the evidence, so considered, measures up to the New York Times standard, the case is one for the jury, and it is error to grant a directed verdict, as the trial judge did in this case.”