****************************************************** The ‘‘officially released’’ date that appears near the beginning of each opinion is the date the opinion will be published in the Connecticut Law Journal or the date it was released as a slip opinion. The operative date for the beginning of all time periods for filing postopinion motions and petitions for certification is the ‘‘officially released’’ date appearing in the opinion. In no event will any such motions be accepted before the ‘‘officially released’’ date. All opinions are subject to modification and technical correction prior to official publication in the Connecti- cut Reports and Connecticut Appellate Reports. In the event of discrepancies between the electronic version of an opinion and the print version appearing in the Connecticut Law Journal and subsequently in the Con- necticut Reports or Connecticut Appellate Reports, the latest print version is to be considered authoritative. The syllabus and procedural history accompanying the opinion as it appears on the Commission on Official Legal Publications Electronic Bulletin Board Service and in the Connecticut Law Journal and bound volumes of official reports are copyrighted by the Secretary of the State, State of Connecticut, and may not be repro- duced and distributed without the express written per- mission of the Commission on Official Legal Publications, Judicial Branch, State of Connecticut. ****************************************************** STATE OF CONNECTICUT v. A. M.* (SC 19497) Rogers, C. J., and Palmer, Zarella, Eveleigh, McDonald, Espinosa and Robinson, Js. Argued September 16—officially released December 23, 2016**
Jennifer F. Miller, deputy assistant state’s attorney, with whom, on the brief, were Stephen J. Sedensky III, state’s attorney, and Colleen P. Zingaro, assistant state’s attorney, for the appellant (state). Bethany L. Phillips, for the appellee (defendant). Opinion
EVELEIGH J. In this certified appeal, we consider whether the state deprived the defendant, A. M., of his fifth amendment right to remain silent when the prosecutor twice noted during closing arguments that the defendant had not testified in his own defense. After a trial, a jury found the defendant guilty of multiple offenses, including sexual assault in the first degree, attempt to commit sexual assault in the first degree, and risk of injury to a child.1 The trial court rendered judgment in accordance with the jury’s verdict, and the defendant appealed to the Appellate Court. In his appeal, the defendant claimed, among other things, that the prosecutor’s comments during closing argument were improper because they infringed on his fifth amendment right to remain silent, depriving him of a fair trial. The Appellate Court agreed and reversed the judgment and remanded the case for a new trial. State v. A. M., 156 Conn. App. 138, 156, 111 A.3d 974 (2015). We agree that the prosecutor’s comments were improper and we find that the state has failed in its burden of proof to show that the comments were harm- less beyond a reasonable doubt. Therefore, we affirm the judgment of the Appellate Court. The jury reasonably could have found the following facts. The defendant began dating the victim’s mother in 2003 and moved in with the mother and her three children later that same year. In August, 2009, the vic- tim, who was approximately ten years old at the time, told her mother’s cousin that the defendant had squeezed her buttocks while she was washing dishes. The cousin alerted the victim’s mother who then removed the defendant from the home. Approximately two weeks later, the victim’s mother allowed the defen- dant back into the home. She did not call the police or alert the victim’s father to the allegations. Seven months after the victim first disclosed her allegations of abuse, a member of the victim’s extended family went to the victim’s father and spoke to him about the allegations. Soon thereafter, the victim’s father questioned the vic- tim about the allegations, which she confirmed as true. Later that same day, he filed a report with Danbury Police Department. After receiving the report, the police commenced an investigation. Officers interviewed the victim, who con- firmed that the defendant had been touching her inap- propriately. Police officers also interviewed the defendant, who gave oral and written statements to police discussing the allegations and denying that he had any inappropriate contact with the victim. At the request of the police, the victim underwent a forensic interview, which was video recorded. In the interview, the victim described numerous incidents in which the defendant sexually abused her. She stated that several times when she was washing dishes, the defendant came up behind her and ‘‘squeezed,’’ ‘‘smacked,’’ and otherwise touched her buttocks while telling her that she was pretty. She also stated that on other occasions, the defendant squeezed, touched, and tried to lick her breasts. The victim recalled another occasion during which the defendant carried the victim into his room, told her he loved her and pulled down her pants. He then pulled open her legs and licked her genitals. Finally, the victim stated that the defendant once pinned her in the corner of a room, made her lay on the floor, pulled down her pants, and placed his penis on her buttocks. The defendant then attempted to penetrate the victim’s anus with his penis, but was unable to do so. The victim initially told the forensic interviewer that these events occurred after August, 2009, when the defendant had returned after being kicked out of the home, but later, she clarified that these incidents occurred prior to August, 2009. The victim also underwent a physical examination at the request of the police. The examination did not reveal any physical evidence of abuse. The examining physi- cian later testified at trial, however, that the lack of physical evidence did not prove the absence of abuse based on the victim’s description of how the abuse occurred. On the basis of the information gathered during their investigation, the police obtained a warrant for the defendant’s arrest, and the state later charged the defen- dant with numerous crimes relating to sexual assault and risk of injury to a child. At trial, the victim testified about the incidents in which the defendant grabbed her buttocks. At one point during her testimony, the victim stated that the defen- dant had squeezed her buttocks six or seven times. Subsequently, she testified that he had done so only two or three times. When the prosecutor questioned the victim about the other instances of abuse—the defendant licking her genitals and attempting to pene- trate her—the victim would not discuss them, saying, ‘‘I don’t want to talk about this. Can’t.’’ After the victim refused to discuss these incidents, the trial court admit- ted into evidence the video recording of the victim’s forensic interview, which the jury was able to consider as substantive evidence of guilt pursuant to State v. Whelan, 200 Conn. 743, 753, 513 A.2d 86, cert. denied, 479 U.S. 994, 107 S. Ct. 597, 93 L. Ed. 2d 598 (1986). The portion of the video recording played at trial included the victim’s descriptions of the incidents that she was unwilling to testify about at trial. Defense coun- sel was able to extensively cross-examine the victim about the allegations, including those made during the forensic interview and depicted in the video recording. The state presented expert testimony to explain why the victim’s recollection of when the sexual abuse occurred might have varied. The state’s expert, Larry M. Rosenberg, a clinical psychologist, testified that chil- dren view and experience time differently than adults. Rosenberg also testified that the effect of trauma on memory is commonly to make it less accurate. Noting that the victim was only approximately ten years old at the time of the assaults, Rosenberg testified that children between the ages of eight and ten years old were not as capable of the abstract reasoning and enhanced judgment possessed by adolescents, and though they could be more suggestible, they were less likely to make false reports than adolescents. In his defense, the defendant presented witnesses to challenge the victim’s credibility. The defendant primar- ily relied upon the testimony of the victim’s mother, who testified regarding the defendant’s positive character traits and her disbelief that he would have committed these crimes—testimony that the state impeached with evidence of the defendant’s history of violence toward her and other women. The defendant also presented his own expert who testified, contrary to the state’s expert, that a ten year old child’s memory skills are the same as an adult’s, and that such a child would remember traumatic events as clearly as other memo- ries, if not better. The defendant did not testify at trial. After the close of evidence, the state and defense counsel gave closing arguments to the jury. During the state’s rebuttal argument, the prosecutor twice remarked to the jury that the defendant had not tes- tified. The prosecutor made the first remark while she reviewed the contents of the written statements that the defendant gave to the police, which had been admit- ted into evidence during trial. After summarizing por- tions of the testimony from the victim’s mother, the prosecutor turned her attention to the defendant’s state- ments to the police, telling the jury: ‘‘This is the other thing. Counsel did not present his client to testify. That’s their right guaranteed by the constitution if any of us were accused. But there is evidence as to things [the defendant] said. His sworn statement. Also, testimony by a couple of police officers as to what he said to them, and that’s before you.’’ The prosecutor then discussed portions of the defendant’s statements to the police, before returning to discuss additional portions of the testimony of the victim’s mother and arguing that her testimony lacked credibility. Thereafter, the prosecutor again turned her attention to the defendant’s statements to the police, this time to argue that they also lacked credibility. Specifically, the prosecutor stated: ‘‘You’ve got to look at the credi- bility of the defendant as well. I mean, he didn’t testify. Again, that’s his right, but there are some statements that are contained in the evidence. One [of] which that I’ve just referred to was his sworn statement as well as some statements by the police that he made the night he was arrested. Statements like, ‘I’ve never seen her naked.’ Three year olds in the house? We’re all parents. Come on, never? Never. Never touched or tickled her ever. Really? Does that make sense?’’ The prosecutor then went on to point out inconsistencies within the defendant’s own statements and between the defen- dant’s statements and some of the testimony given at trial. Defense counsel did not object to either of the prose- cutor’s comments about the defendant’s failure to tes- tify, either during or after the state’s argument. After the state’s rebuttal argument, counsel did, however, make an unrelated objection about a different aspect of the prosecutor’s rebuttal argument. After resolving the unrelated objection, the court then instructed the jury on the law that applied to the case. Although the defendant had not objected to the prosecu- tor’s comments, the court’s charge nevertheless included the following instruction about the defendant’s decision not to testify, as required by General Statutes § 54-84 (b): ‘‘Now, the defendant has elected not to testify in this particular case. An accused has the right and the option to testify or not to testify at his or her own trial, and is under no obligation to testify. He has a constitutional right that is protected by the [United States constitution] as well as the Connecticut [consti- tution] not to testify, and you may draw no unfavorable inferences from the defendant’s choice not to testify in this particular instance.’’ After receiving the court’s instructions, the jury retired to deliberate and ultimately returned a verdict finding the defendant guilty of all but one charge. The trial court thereafter rendered judgment in accordance with the jury’s verdict, issuing a total effective sentence of twenty years of imprisonment, execution suspended after twelve years, followed by twenty years of proba- tion with special conditions. The defendant appealed from the judgment of the trial court to the Appellate Court, claiming, among other things, that the prosecutor’s references to the defen- dant’s decision not to testify were improper and deprived him of his right to a fair trial.2 The Appellate Court agreed with the defendant and reversed the trial court’s judgment. State v. A. M., supra, 156 Conn. App. 148–49, 156. We granted the state’s petition for certifica- tion on the following issue: ‘‘Did the Appellate Court properly determine that certain comments made by the prosecutor amounted to prosecutorial improprieties and that such improprieties deprived the defendant of his due process rights to a fair trial?’’ State v. A. M., 317 Conn. 910, 116 A.3d 309 (2015). On appeal to this court, the state claims that the Appellate Court improp- erly concluded that the prosecutor’s comments were improper. The state also contends that even if the com- ments were improper, they were harmless beyond a reasonable doubt. As a threshold matter, we must revise the certified question and clarify the proper standard of review. The Appellate Court treated the defendant’s claim as impli- cating his general due process right to a fair trial rather than as a claim arising under the fifth amendment right to remain silent. The Appellate Court thus decided the defendant’s appeal using the due process standard for prosecutorial misconduct claims from State v. Wil- liams, 204 Conn. 523, 540, 529 A.2d 653 (1987), to con- clude that the prosecutor’s comments in the present case deprived the defendant of his due process right to a fair trial. State v. A. M., supra, 156 Conn. App. 156. We have previously clarified, however, that the Williams standard applies only when a defendant claims that a prosecutor’s conduct did not infringe on a specific constitutional right, but nevertheless deprived the defendant of his general due process right to a fair trial. State v. Payne, 303 Conn. 538, 562–63, 34 A.3d 370 (2012). For example, in Williams, the defen- dant argued that the prosecutor’s comments deprived him of a fair trial because the prosecutor unfairly expressed his personal opinions about a witness’ credi- bility and appealed to the passions and emotions of the jurors. State v. Williams, supra, 540–41. There was no claim in Williams that the prosecutor’s comments infringed an enumerated constitutional right, and, there- fore, the court applied the general due process standard for prosecutorial impropriety claims. Id., 539–40. Under the Williams general due process standard, the defen- dant has the burden to show both that the prosecutor’s conduct was improper and that it caused prejudice to his defense. Id.; see also State v. Payne, supra, 562–63. We held in Payne, however, that a different standard applies when the defendant claims that the prosecu- torial impropriety ‘‘infringed a specifically enumerated constitutional right, such as the fifth amendment right to remain silent . . . .’’ (Emphasis added.) State v. Payne, supra, 303 Conn. 562. If the defendant raises this type of claim, the defendant initially has the burden to establish that a constitutional right was violated. Id. If the defendant establishes the violation, however, the burden shifts to the state to prove that the violation was harmless beyond a reasonable doubt.3 Id. Because the defendant’s claim in the present case implicated his fifth amendment right to remain silent, we apply the harmless error standard as called for in Payne rather than the general due process standard from Williams. We therefore restate the certified ques- tion to ask: ‘‘Did the Appellate Court properly determine that certain comments made by the prosecutor violated the defendant’s right to remain silent and, if so, was the violation harmless beyond a reasonable doubt?’’ Turning to the merits of the present case, we first consider whether the prosecutor’s comments violated the defendant’s fifth amendment right to remain silent.4 The fifth amendment prohibits the state from forcing the defendant to be a witness against himself, and the United States Supreme Court has concluded that this protection also prohibits prosecutors from commenting at trial on the defendant’s decision not to testify. Griffin v. California, 380 U.S. 609, 615, 85 S. Ct. 1229, 14 L. Ed. 2d 106 (1965); see also State v. Parrott, 262 Conn. 276, 292, 811 A.2d 705 (2003) (‘‘[i]t is well settled that comment by the prosecuting attorney . . . on the defendant’s failure to testify is prohibited by the fifth amendment to the United States constitution’’ [internal quotation marks omitted]). In Griffin, the court rea- soned that allowing a prosecutor to comment on the defendant’s refusal to testify would be equivalent to imposing a penalty for exercising his constitutional right to remain silent.5 Griffin v. California, supra, 614. In the present case, the prosecutor’s two statements at issue clearly violated the defendant’s fifth amend- ment right to remain silent. The prosecutor’s remarks during closing arguments directly and unambiguously called the jury’s attention to the defendant’s decision not to testify. This is precisely the type of comment proscribed by Griffin.6 The state argues, however, that despite the prosecu- tor’s direct reference to the defendant’s failure to tes- tify, the comments nevertheless did not violate the protections of the fifth amendment. According to the state, we must apply what is known as the ‘‘ ‘naturally and necessarily’ ’’ test to determine whether a prosecu- tor’s comments violate the fifth amendment. That test asks whether ‘‘the language used [by the prosecutor was] manifestly intended to be, or was . . . of such a character that the jury would naturally and necessarily take it to be a comment on the failure of the accused to testify.’’ (Emphasis added; internal quotation marks omitted.) State v. Haase, 243 Conn. 324, 333, 702 A.2d 1187 (1997), cert. denied, 523 U.S. 1111, 118 S. Ct. 1685, 140 L. Ed. 2d 822 (1998). The state argues that a prosceu- tor’s statements violate the fifth amendment only if they ‘‘ ‘naturally and necessarily’ ’’ imply to the jury that it should draw an adverse inference and infer guilt from the defendant’s silence. According to the state, the pros- ecutor’s comments in the present case did not violate the fifth amendment because the comments did not naturally and necessarily ask the jury to infer the defen- dant’s guilt from the defendant’s silence, but were made to remind the jury that although the defendant had not testified, the jury still had evidence of his statements before it that it could consider. We disagree. First, the state’s reliance on the ‘‘naturally and neces- sarily’’ test is misplaced. The ‘‘naturally and necessarily’’ standard applies only when it is unclear whether the prosecutor’s comments at issue referred to the defen- dant’s failure to testify, not when the prosecutor makes comments that explicitly refer to the defendant’s silence. See, e.g., United States v. Tanner, 628 F.3d 890, 899 (7th Cir. 2010) (‘‘[a]bsent [a] direct comment, [t]he right against self-incrimination is violated only when . . . the remark was of such a character that the jury would naturally and necessarily take it to be a comment on the defendant’s silence’’ [internal quotation marks omitted]); Lent v. Wells, 861 F.2d 972, 975 (6th Cir. 1988) (noting that direct reference to failure to testify clearly violates fifth amendment, while indirect remark is pro- hibited only if it ‘‘ ‘naturally and necessarily’ ’’ would be understood as reference to defendant’s silence); see also State v. Ruffin, 316 Conn. 20, 31, 110 A.3d 1225 (2015) (applying ‘‘ ‘naturally and necessarily’ ’’ test to determine if prosecutor’s comments indirectly referred to defendant’s failure to testify); State v. Colon, 70 Conn. App. 707, 713–14, 799 A.2d 317 (noting that even indirect remark about defendant’s failure to testify can violate fifth amendment if it ‘‘naturally and necessarily’’ would be understood as reference to defendant’s silence at trial), cert. denied, 261 Conn. 933, 806 A.2d 1067 (2002). For example, in State v. Ruffin, supra, 316 Conn. 30, the prosecutor stated, during closing argument, that no witness had controverted the victim’s testimony, and the defendant claimed that this statement was a com- ment on his failure to testify. Because the prosecutor had not directly mentioned the defendant’s failure to testify, we used the ‘‘ ‘naturally and necessarily’ ’’ stan- dard to determine whether the jury would nevertheless have understood the prosecutor’s statement as referring to the defendant’s failure to testify. Id., 29. Using this standard, we concluded that the most reasonable inter- pretation of the comment was that it referred to the credibility of the victim, and thus did not ‘‘naturally and necessarily’’ refer to the defendant’s failure to testify. Id., 31. As a result, we concluded that the comments did not refer to the defendant’s failure to testify and were therefore not improper. Id., 32. In the present case, however, there is no question that the prosecutor referred directly to the defendant’s failure to testify. There is, therefore, no need to apply the ‘‘naturally and necessarily’’ test to determine whether the comments violated the fifth amendment. Second, to violate the fifth amendment, the prosecu- tor’s comments need not imply that the jury should draw an adverse inference from the defendant’s silence. It is enough that the comment by an adverse party calls the jury’s attention to the defendant’s silence because any such comment ‘‘heighten[s] the jury’s awareness of the defendant’s silence, namely, his failure to answer to the state’s charges . . . [and] is improper.’’ State v. Smalls, 78 Conn. App. 535, 543, 827 A.2d 784, cert. denied, 266 Conn. 931, 837 A.2d 806 (2003). When the defendant chooses not to testify, he takes the risk that the jury will view his silence with skepticism—a prose- cutor’s explicit reminders to the jury of the defendant’s decision serves only to heighten this risk, burdening the defendant’s constitutional right to remain silent. Id.; see also Griffin v. California, supra, 380 U.S. 614. Although the extent to which the prosecutor’s com- ments implied that the jury should hold the defendant’s silence against him may relate to the harmfulness of the violation, it does not impact our conclusion that a violation occurred in the first place. Because the defendant here has established that the prosecutor violated his fifth amendment rights by directly referencing his failure to testify, we next must determine whether the state has proven beyond a rea- sonable doubt that the violation was harmless. ‘‘[T]here may be some constitutional errors which in the setting of a particular case are so unimportant and insignificant that they may, consistent with the [f]ederal [c]onstitu- tion, be deemed harmless, not requiring the automatic reversal of the conviction.’’ Chapman v. California, 386 U.S. 18, 22, 87 S. Ct. 824, 17 L. Ed. 2d 705 (1967); see also United States v. Hasting, 461 U.S. 499, 509, 103 S. Ct. 1974, 76 L. Ed. 2d 96 (1983) (‘‘[i]t is the duty of a reviewing court to consider the trial record as a whole and to ignore errors that are harmless, including most constitutional violations’’). The state has the bur- den to prove that this error was harmless beyond a reasonable doubt. State v. Payne, supra, 303 Conn. 563 (‘‘if the defendant raises a claim that the prosecutorial improprieties infringed a specifically enumerated con- stitutional right, such as the fifth amendment right to remain silent . . . and the defendant meets his burden of establishing the constitutional violation, the burden is then on the state to prove that the impropriety was harmless beyond a reasonable doubt’’). The focus of our harmless error inquiry is on whether the state has demonstrated that the otherwise improper comments did not influence the outcome of the trial. United States v. Hasting, supra, 510–11; Chapman v. California, supra, 24. Applying this test to the present case, we are con- vinced that the state has not met its burden of showing that the comments were harmless beyond a reasonable doubt. Although we are not required to do a complete Williams analysis due to the nature of the right infringed, we find the Williams factors to be a useful guide in our evaluation of whether the state has met its burden of proof to show harmlessness beyond a reasonable doubt.7 ‘‘We are mindful throughout this inquiry, however, of the unique responsibilities of the prosecutor in our judicial system. A prosecutor is not only an officer of the court, like every other attorney, but is also a high public officer, representing the people of the [s]tate, who seek impartial justice for the guilty as much as for the innocent. . . . By reason of his [or her] office, [the prosecutor] usually exercises great influence upon jurors. [The prosecutor’s] conduct and language in the trial of cases in which human life or liberty are at stake should be forceful, but fair, because he [or she] represents the public interest, which demands no victim and asks no conviction through the aid of passion, prejudice or resentment. If the accused be guilty, he [or she] should [nonetheless] be convicted only after a fair trial, conducted strictly according to the sound and [well established] rules which the laws prescribe. While the privilege of counsel in addressing the jury should not be too closely narrowed or unduly hampered, it must never be used as a license to state, or to comment upon, or to suggest an inference from, facts not in evidence, or to present matters which the jury [has] no right to consider.’’ (Internal quotation marks omitted.) State v. Singh, 259 Conn. 693, 701–702, 793 A.2d 226 (2002). With that proviso in mind, we now turn to the Williams factors for guidance. First, as we noted previously, there is no question that the comments of the prosecutor were not invited by either the argument or conduct of defense counsel. Second, the remarks of the prosecutor were particularly severe. The state argues that the prosecutor’s remarks were not of such a character that the jury would ‘‘natu- rally and necessarily’’ have taken them to be an adverse comment on the defendant’s failure to testify. It argues that their context diluted any impropriety and demon- strated a purpose unrelated to the impropriety, namely, to focus on the statements given by the defendant to the police. We disagree. The statements of the prosecu- tor were particularly severe in this case because the prosecutor, as an officer of the court, violated a state statute, § 54-84 (a), by commenting on the fact that the defendant did not testify, in open court, and did it on two occasions. The prosecutor then went on to tell the jurors that they could evaluate the credibility of the defendant through the statements he gave to the police, yet the only charge that the judge gave regarding credi- bility related to witnesses who testified.8 In spite of the prosecutor’s comments asking the jurors to judge the defendant’s credibility by the statements he gave to the police, the jury was instructed by the judge that credibility related to witnesses who testified. Thus, the prosecutor, through her statements on credibility, was reminding the jurors again that the defendant did not testify. The state argues that the fact that the defense counsel did not object to the prosecutor’s comments, either during the prosecutor’s argument or afterward, weighs heavily in any analysis of the harmfulness of the statements. We acknowledge that the failure to object on the part of defense counsel is certainly a factor to consider. We note that, in nearly all cases where defense counsel fails to object to and request a specific curative instruction in response to a prosecu- torial impropriety, especially an impropriety that we do not consider to be particularly egregious, and the court’s general jury instruction addresses that impropri- ety, we have held that the court’s general instruction cures the impropriety. See, e.g., State v. Jones, 320 Conn. 22, 38–39, 128 A.3d 431 (2015); State v. Luster, 279 Conn. 414, 446, 902 A.2d 636 (2006); State v. War- holic, 278 Conn. 354, 402, 897 A.2d 569 (2006); State v. Stevenson, 269 Conn. 563, 597–98, 849 A.2d 626 (2004). In fact, we have noted that, ‘‘defense counsel’s failure to object to the prosecutor’s argument when it was made suggests that defense counsel did not believe that it was unfair in light of the record of the case at the time.’’ (Internal quotation marks omitted.) State v. Ceballos, 266 Conn. 364, 414, 832 A.2d 14 (2003). We have recognized, nevertheless, that defense counsel may elect not to object ‘‘for tactical reasons, namely, because he or she does not want to draw the jury’s attention to it or because he or she wants to later refute that argument.’’ (Internal quotation marks omitted.) Id.9 If, however, defense counsel fails to object to particu- larly egregious or pervasive misconduct, and the gen- eral jury instructions do not specifically address the prosecutor’s misconduct, we have held that the general jury instructions were insufficient to cure the miscon- duct, in spite of defense counsel’s failure to object. We have stated that prosecutorial misconduct can be ‘‘sufficiently egregious to overcome the suggestion that defense counsel did not think it was unfair at the time.’’ Id., 415; id. (concluding that prosecutorial statements, relating to comments on veracity of other witnesses and inflammatory religious comment, were severe despite defense counsel’s failure to object). Further, our case law has recognized that ‘‘the defendant’s failure to object at trial to each of the occurrences that he now raises as instances of prosecutorial impropriety, though relevant to our inquiry, is not fatal to review of his claims.’’ (Internal quotation marks omitted.) State v. Taft, 306 Conn. 749, 762, 51 A.3d 988 (2012). In the present case, we already have decided that the prosecu- tor’s comments were particularly egregious. Therefore, we follow cases such as State v. Angel T., 292 Conn. 262, 291, 973 A.2d 1207 (2009), and State v. Maguire, 310 Conn. 535, 560–61, 78 A.3d 828 (2013), wherein we have held that in instances of particularly grievous improprieties the failure to object is overcome by a general instruction that did not fully address the impro- priety. In Angel T., for instance, we concluded that the general instructions were insufficient to cure miscon- duct that had occurred repeatedly throughout the trial because the general jury instructions did not directly address all of the misconduct. State v. Angel T., supra, 291. We concluded that the prosecutor had improperly suggested that the defendant had sought the aid of counsel prior to being charged with a crime because he had something to hide. Id., 290. The prosecutor repeatedly elicited improper evidence about the defen- dant retaining counsel, first through direct examination of a police officer and cross-examination of the defen- dant, and then by discussing that testimony at length during opening and rebuttal summations. Id. The jury instructions addressed one aspect of the impropriety— emphasizing that attorneys’ arguments were not evi- dence—but the instructions did not specifically address the prosecutor’s suggestion that the defendant’s choice to retain counsel indicated that he had something to hide. We found that although the general jury instruc- tions might have mitigated some of the harm from the improprieties, they did not address all of the improprie- ties, and, therefore, were not sufficiently curative. Id., 291 and n.22. State v. Maguire, supra, 541, involved a mother who hired the defendant to help her care for her four chil- dren. Subsequently, the defendant was charged with two counts of risk of injury to a child and one count each of the crimes of sexual assault in the fourth degree and attempt to commit sexual assault in the first degree involving one of the children. Id., 544. Maguire also involved a forensic interview that the victim had with a multidisciplinary team of mental health and law enforcement employees, ‘‘all of whom work[ed] collab- oratively to investigate and treat cases of reported sex- ual abuse.’’ Id., 543. In Maguire, the defendant claimed that ‘‘the prosecutor engaged in impropriety by repeat- edly asserting, during rebuttal closing argument, that both the defendant and defense counsel were asking the jury to ‘condone child abuse’ and by claiming, inter alia, that defense counsel had ‘lied’ to the victim during cross-examination, that the defendant’s testimony was ‘coached’ and that the defense strategy was a game of ‘smoke and mirrors . . . .’ ’’ Id., 545. The defendant also contended that the prosecution engaged in impro- priety during defense counsel’s cross-examination of the director of the multidisciplinary team. Id. Specifi- cally, on appeal, the defendant asserted that the prose- cutor suggested, contrary to the assertions of defense counsel that the redacted portions of the transcript of the forensic interview revealed that the director did, in fact, question the victim about inconsistent statements that the victim made as to whether she had actually seen the defendant’s penis. The defendant maintained that, ‘‘because these improprieties bore directly on the central issue in the case, namely, the defendant’s credi- bility, they very easily could have tipped the balance in a case that turned entirely on whom the jury found to be more credible—the defendant or the victim.’’ Id. We concluded in Maguire that prosecutorial improprie- ties occurred and then examined whether the impropri- eties deprived the defendant of a fair trial. Id., 560–62. It is of note that, in Maguire, defense counsel objected to only one of the many improprieties (addressing the redacted portions of the interview). Id., 561. We con- cluded that the failure of defense counsel to object was the only factor that weighed in favor of the state. Id. In examining the jury charge in that case, we concluded as follows: ‘‘Although the trial court charged the jury in its final instructions that the law prohibits the prose- cutor from offering her personal opinion as to the credi- bility of any witness, the trial court gave no curative instructions at the time of the improper remarks, which, in our view, were relatively severe because they demeaned the integrity of defense counsel by improp- erly calling into question his veracity and denigrated the defense theory of the case by mischaracterizing it.’’ Id. While it may be argued that the remarks in both Angel T. and Maguire may be explained by the overexuber- ance of counsel in the ‘‘heat of battle,’’ there can be no such explanation for the statements of the prosecutor in the present case that are specifically forbidden by state statute and which the prosecutor is presumed to have known prior to the trial. Therefore, we conclude that the egregiousness of the remarks in this case were equal to, if not greater than, the remarks in both Angel T. and Maguire. Any curative instruction should have been immediately given by the trial judge, and the gen- eral jury instruction, as we discuss hereinafter, was insufficient to cure the impropriety. The prosecutor, an officer of the court, twice violated a state statute prohibiting her from commenting on the defendant’s failure to testify. Therefore, we hold that the remarks were egregious and were not cured by the general jury instructions. Third, we turn to the frequency of the prosceutor’s remarks. Although it may be argued that the prosecutor only mentioned the defendant’s failure to testify on two occasions, in our opinion, the violation of a specific state statute forbidding the comments means that two occasions of such violations is two too many. Indeed, we could not condone even one violation of such an extreme nature. Therefore, we conclude that, given the context of the remarks, they were frequent. Fourth, we turn to the centrality of the impropriety to the critical issues in the case. It is not contested that the prosecutorial impropriety in the present case went directly to the central issue, namely, credibility. The central issue for the jury was whether it would believe the allegations of the victim, particularly when there was no physical evidence. As this court previously has recognized, ‘‘[b]ecause there was no independent physi- cal evidence of the assault and no other witnesses to corroborate [the victim’s] testimony, her credibility was crucial to successful prosecution of the case.’’ State v. Ritrovato, 280 Conn. 36, 57, 905 A.2d 1079 (2006); see also State v. Iban C., 275 Conn. 624, 641–42, 881 A.2d 1005 (2005) (when there was no physical or medical evidence of abuse and no eyewitness testimony other than victim’s testimony, victim’s credibility was central to state’s case). As we stated in State v. Maguire, supra, 310 Conn. 561–62, ‘‘[i]n addition, each instance of prose- cutorial impropriety bore directly on the central issue in the case, namely, the credibility of the defendant and his theory of defense. As the defendant maintains, because the state’s case rested entirely on the victim’s credibility, any improper remarks by the prosecutor that tended to bolster [the victim’s] credibility, or to diminish that of the defendant, may very well have had a substantial impact on the verdict.’’ See also State v. Alexander, 254 Conn. 290, 308, 755 A.2d 868 (2000) (‘‘[t]he improper comments directly addressed the criti- cal issue in this case, the credibility of the victim and the defendant’’ [internal quotation marks omitted]). Furthermore, this court has recognized that ‘‘[t]he jury might have doubted the uncorroborated testimony offered [to establish] the defendant’s guilt . . . but [found] him [guilty] because he did not take the stand to deny his guilt.’’ State v. Dudla, 190 Conn. 1, 7, 458 A.2d 682 (1983); see also State v. Ruocco, 322 Conn. 796, 806, 144 A.3d 354 (2016) (‘‘given the discrepancies in [uncorroborated testimony by a witness] and the fact that there was only one eyewitness . . . we cannot discount the possibility that the jury might have doubted the [testimony of other witnesses] but found the defen- dant guilty because of his failure to testify’’). Similarly, in the present case, the prosecutorial improprieties focused the jury’s attention on the defendant’s failure to take the stand to deny his guilt and, therefore, improperly bolstered the testimony of the victim. The credibility of the defendant and the victim were particu- larly central to the present case because there was no physical evidence, no eyewitness testimony and the victim’s testimony was inconsistent regarding the inci- dences of abuse. Therefore, we conclude, as we did in State v. Maguire, supra, 561–62, that because, in the present case, the state’s charges rested entirely on the credibility of the victim, the comments of the prosecu- tor tended to diminish the credibility of the defendant and to bolster the victim’s credibility. Fifth, we examine the strength of the state’s case. This court repeatedly has recognized that ‘‘a sexual assault case lacking physical evidence is not particu- larly strong, especially when the victim is a minor.’’ State v. Ritrovato, supra, 280 Conn. 57; see also State v. Ceballos, supra, 266 Conn. 416 (‘‘a child sexual abuse case lacking conclusive physical evidence, when the prosecution’s case rests on the credibility of the victim, is not particularly strong’’ [internal quotation marks omitted]). Furthermore, as this court explained in State v. Maguire, supra, 310 Conn. 561, ‘‘the state’s case was not particularly strong. Significantly, there was no phys- ical evidence of abuse, and there was no eyewitness testimony other than that of the victim, whose testi- mony at times was both equivocal and vague.’’ In fact, this court has previously determined that when there is no physical evidence to support the allegations of sexual assault, the significance of the prosecutorial impropriety increased considerably. ‘‘[A]lthough there exists evidence in the present case supporting the defendant’s conviction despite the lack of direct physi- cal evidence linking the defendant to the sexual assault of the victim—namely, the victim’s testimony and the testimony of her mother and cousin—without indepen- dent physical evidence to prove that the defendant sexu- ally assaulted [the victim], or even that [the victim] had been sexually assaulted at all, the significance of the [prosecutor’s] improper conduct increases consider- ably.’’ State v. Angel T., supra, 292 Conn. 293. Similar to the circumstances in Angel T., in the present case, there was no physical evidence linking the defendant to the sexual assault. There were also inconsistent state- ments made by the victim, and, in fact, the victim refused to answer certain questions during her direct testimony. The only other evidence on point concerned the defendant’s statements to the police wherein he denied that any of the incidents occurred or if some- thing did happen, it was an accidental brushing. Consid- ering the totality of the circumstances and the evidence adduced at trial, we conclude that the state’s case was not particularly strong. The last Williams factor involves the strength of the curative measures taken by the court. The state argues that the court’s charge to the jury that it could not consider the defendant’s failure to testify in its delibera- tions cured any defect in the prosecutor’s argument. We are not persuaded. As we noted previously, no curative instruction was given at the time the prosecutor uttered the remarks. In State v. Maguire, supra, 310 Conn. 561, we held that the general instructions to the jury were not enough when ‘‘[a]lthough the trial court charged the jury in its final instructions that the law prohibits the prosecutor from offering her personal opinion as to the credibility of any witness, the trial court gave no curative instructions at the time of the improper remarks, which, in our view, were relatively severe . . . .’’ Similarly, in the present case, the prosecutor’s remarks not only were severe but also constituted both a statutory and constitutional violation, therefore, the judge’s charge, which did not reference the specific remarks but was general in nature, was ineffectual and did not cure the impropriety of the prosecutor’s remarks. The remarks were so egregious that, in the absence of a strong curative instruction at the time when the remarks were made, no general charge would have sufficed. Therefore, we conclude that the curative remarks made in the jury charge in the form of a general statement based on the accused’s failure to testify, was insufficient in this case. In our analysis regarding whether the state has proven that the remarks were harmless beyond a rea- sonable doubt, we acknowledge that the jury found the defendant guilty of eight of the nine counts with which he was charged. Our case law does suggest that in the absence of evidence that the jury disregarded any of the court’s instructions, we presume that the jury fol- lowed the instructions. State v. Ancona, 256 Conn. 214, 219, 772 A.2d 571 (2001). It may be argued that the fact that the jury acquitted the defendant on one of the charges supports a conclusion that the jury was able to analyze all of the charges and weigh the evidence against the defendant, rather than totally condemn him for his failure to testify. We note, however, that a similar argument was rejected in State v. Maguire, supra, 310 Conn. 562–63 n.13, wherein the defendant was found not guilty on two counts. Although in our analysis we do credit the state for this conclusion, on closer exami- nation it is not as persuasive as it initially may appear. The defendant was charged with a total of nine counts, eight of which related to various types of sexual assault and risk of injury charges for his alleged conduct relat- ing to the victim. The ninth charge did not. The ninth charge alleged, in relevant part, that the defendant ‘‘wil- fully and unlawfully caused and permitted a child under the age of sixteen to be placed in a situation that the morals of said child were likely to be impaired, to wit: witnessing [the defendant] being violent towards the child’s mother . . . .’’ As the prosecutor stated in his closing argument: ‘‘Now, count nine is the one count that’s a little bit different. That is a risk of injury under a—it’s the same statute number, which is § 53-21 (a) (1), but it has to do with the child witnessing domestic violence and—not domestic violence, but violent actions by the defendant in the home.’’ The prosecu- torial impropriety in the present case improperly impugned the defendant’s credibility for not testifying and bolstered the victim’s credibility as it related to the central issue in the case—whether the sexual assaults had taken place. Indeed, there were no statements made by the defendant to the police regarding the ninth count. Therefore, the prosecutor’s improper comments would not have affected the jury’s deliberation on that count. Accordingly, the fact that the defendant was acquitted on an unrelated charge is not relevant to the consider- ation of whether the constitutional and statutory viola- tion harmed the defendant on the charges in which the jury was required to pit his testimony, through his statements to the police, against the testimony of the victim. Therefore, we conclude that all of the Williams fac- tors weigh in favor of the defendant. Similar to the situation in State v. Maguire, supra, 310 Conn. 561, the only factor in the Williams analysis favorable to the state is defense counsel’s failure to object to the state- ments. This failure, however, is overcome by the egre- giousness of the statements. Furthermore, we have credited the state with the fact that the jury found the defendant guilty of eight out of the nine charges. How- ever, due to the difference between the first eight charges on which the defendant was found guilty, and the ninth charge on which he was acquitted, we do not find this fact to be either strongly in favor of the state or dispositive of the case and we arrive at a similar conclusion as the court did in Maguire, when the defen- dant was found not guilty on two charges. Id., 562–63 n.13. Therefore, we are convinced that the state has failed to meet its burden of proof that the statements made by the prosecutor were harmless beyond a reason- able doubt. The judgment of the Appellate Court is affirmed. In this opinion ROGERS, C. J., and PALMER, McDON- ALD and ROBINSON, Js., concurred. * In accordance with our policy of protecting the privacy interests of the victims of sexual assault and the crime of risk of injury to a child, we decline to identify the victim or others through whom the victim’s identity may be ascertained. See General Statutes (Supp. 2016) § 54-86e. ** December 23, 2016, the date that this decision was released as a slip opinion, is the operative date for all substantive and procedural purposes. 1 More specifically, the defendant was found guilty of one count of attempt to commit sexual assault in the first degree in violation of General Statutes §§ 53a-49 (a) (2) and 53a-70 (a) (2), three counts of risk of injury to a child in violation of General Statutes (Rev. to 2009) § 53-21 (a) (2), two counts of risk of injury to a child in violation of General Statutes (Rev. to 2009) § 53-21 (a) (1), and one count each of sexual assault in the first degree in violation of General Statutes § 53a-70 (a) (2), and sexual assault in the fourth degree in violation of General Statutes § 53a-73a. 2 Although the defendant did not preserve his prosecutorial impropriety claim by objecting at trial, such claims are nevertheless reviewable on appeal. See State v. Payne, 303 Conn. 538, 560, 34 A.3d 370 (2012). 3 We acknowledge that in State v. Ruffin, 316 Conn. 20, 27–29, 110 A.3d 1225 (2015), a case decided after Payne, we used the general due process standard when considering a prosecutorial misconduct claim that implicated the fifth amendment right to remain silent. In light of our earlier decision in Payne, however, we now clarify that under Payne, our analytical approach in Ruffin was not correct. We note, however, that this oversight did not impact the outcome of that appeal because we found that the prosecutor’s statements were not improper, and thus it did not matter which party had the burden to prove harm. Id., 32. 4 The fifth amendment to the federal constitution, although addressed to the federal government, also applies to the states under the due process clause of the fourteenth amendment. State v. Ruffin, 316 Conn. 20, 29 n.4, 110 A.3d 1225 (2015). 5 Section 54-84 (a) also prohibits a prosecutor from commenting on the ‘‘neglect or refusal of an accused party to testify . . . .’’ In the past, however, we have treated the protections of the statute as being synonymous with those of the fifth amendment. See, e.g., State v. Parrott, supra, 262 Conn. 293; State v. Haase, 243 Conn. 324, 332–33, 702 A.2d 1187 (1997), cert. denied, 523 U.S. 1111, 118 S. Ct. 1685, 140 L. Ed. 2d 822 (1998). To the extent that the defendant argues that the prosecutor’s conduct violated § 54-84, the defendant has not argued that the statute provides greater protection than the fifth amendment, and, therefore, we do not separately analyze the defen- dant’s claims of impropriety under the statute, except insofar as we deter- mine in our analysis that the prosecutor’s comments, in view of § 54-84, were particularly egregious. 6 We note that the United States Supreme Court has held that a direct comment by a prosecutor on the defendant’s silence will not violate the fifth amendment when a ‘‘prosecutor’s reference to the defendant’s opportunity to testify is a fair response to a claim made by [a] defendant or his counsel . . . .’’ United States v. Robinson, 485 U.S. 25, 32, 108 S. Ct. 864, 99 L. Ed. 2d 23 (1988). There is no claim in the present case that the defendant invited the challenged statements. 7 ‘‘In determining whether prosecutorial misconduct was so serious as to amount to a denial of due process, this court, in conformity with courts in other jurisdictions, has focused on several factors. Among them are the extent to which the misconduct was invited by defense conduct or argument . . . the severity of the misconduct . . . the frequency of the misconduct . . . the centrality of the misconduct to the critical issues in the case . . . the strength of the curative measures adopted . . . and the strength of the state’s case.’’ (Citations omitted.) State v. Williams, supra, 204 Conn. 540. 8 The trial court instructed the jury as follows: ‘‘I now want to discuss the subject of credibility, by which I mean the believability of testimony. You have observed the witnesses. The credibility or the believability of the witnesses and the weight to be given to their testimony are matters entirely within your hands. It is for you alone to determine their credibility. Whether or not you find a fact proven is not to be determined by the number of witnesses testifying for or against it. It is the quality and not the quantity of testimony which should be controlling. As I previously indicated, a single witness’ testimony may be sufficient to justify a conviction, provided, of course, that such testimony satisfies you beyond a reasonable doubt that the state has sustained its burden of proving each and every element of the given charge or charges beyond a reasonable doubt.’’ 9 We note in this case that, since the improprieties occurred during rebuttal argument, defense counsel would not have had the opportunity to rebut the arguments. Further, after closing argument, although defense counsel objected to certain portions of the prosecutor’s argument, he did not object to the statements regarding the failure of the defendant to testify.