This is a review and cross-review of an unpublished court of appeals decision dated December 28, 1981. The appeal to the court of appeals was from a judgment of the circuit court for Portage county, James H. Levi, Circuit Judge. The defendant, Rita A. Felton, was convicted of the second degree murder of her husband, Robert Felton, in violation of sec. 940.02 (1), Stats.
The appeal affirmed the judgment of guilty, but directed that the cause be remanded for the purpose of allowing the defendant to interpose the defense of not guilty by reason of mental disease or defect. The state petitioned this court for review in respect to the remand. The defendant petitioned for a cross-review, asking for a new trial on all issues because of the ineffectiveness of trial counsel. Both petitions were granted by this court. We reverse the conviction and remand for a new trial.
On January 2, 1979, Rita Felton shot her husband while he lay sleeping. She was charged with first degree murder. On July 5,1979, her attorney filed a motion to amend her plea of not guilty to not guilty by reason *488of mental disease or defect. On the day the trial began, the plea of not guilty by reason of mental disease or defect was withdrawn. After a lengthy trial, the jury found Rita Felton guilty of murder in the second degree. Her defense was that she was a “battered” spouse who acted in self-defense. After the close of testimony, the trial court instructed the jury on first degree murder, second degree murder, manslaughter (imperfect self-defense) — sec. 940.05(2), Stats., and on the privilege of self-defense. There was no request for instruction on heat-of-passion manslaughter under sec. 940.05(1).
Defendant’s appellate counsel filed a motion for new trial, which was denied by the trial court after an evi-dentiary hearing. The appeal to the court of appeals was from the conviction and the denial of the motion for a new trial.
It was contended on appeal that trial counsel was ineffective because he improvidently and without sufficient investigation withdrew the defense that the defendant should be exonerated of responsibility because of mental disease or defect. It was also alleged that counsel was ineffective when he failed, because he was unaware of the law, to consider as a defense the lesser crime of manslaughter — heat-of-passion, as provided in sec. 940.05 (1), Stats.
It was also contended on the appeal that the record demonstrated erroneous and prejudicial limitations by the trial judge in counsel’s questioning of witnesses in respect to acts of violence by the victim, Robert Felton. It is also urged that the admission of a statement elicited from Rita Felton in violation of Miranda was prejudicial. The court of appeals found the admission of the statement erroneous but not prejudicial; and in respect to the limitation on witnesses, it concluded that the evidence which would have been elicited in respect to *489Robert Felton’s acts of violence was cumulative, and, hence, the exclusion was not prej udicial.
Counsel also argues that the “bridging” instruction which was given in this case and which is applicable when the evidence arguably permits a finding of first or second degree murder or manslaughter, coerces the jury into returning a murder, rather than a manslaughter, verdict.
On this review we hold that trial counsel was ineffective, and accordingly the verdict and conviction must be set aside and a new trial ordered. We reverse that portion of the decision of the court of appeals that sustained the conviction adjudging guilt.
Because the defense proffered by Rita Felton was that she was a battered spouse and because she recounted that her maltreatment had persisted throughout almost the entire course of her twenty-three-year marriage with Robert, the facts developed at trial are extensive. We recount some of them, because they are essential to an understanding of her defense and to her defense which ought to have been undertaken or ought to have been considered had counsel been effective.
Rita testified that Robert commenced beating her after they had been married for about six months. He beat her during the course of her six pregnancies and on one occasion she miscarried. On another he beat her, held her down, and threatened to burn her with a blowtorch. On occasions she would wake up and find that Robert was choking her. One time he struck her with his fist so hard that her upper denture broke and it cut through her lip. On various occasions he threatened to kill her. She was forced to commit sexual acts which she considered degrading. Several times the police were called to quell these acts of domestic violence. The police were called either by the older children or by neighbors who heard the disturbances. She testified that she was fre*490quently bruised, and on one occasion she had several broken ribs.
In 1977, after Robert had severely beaten her, the police department removed Rita and the children from the house, and she hired a lawyer and filed for divorce. The Feltons were separated for about ten months. She testified that they got back together again because Robert was attempting to be nice to her and because she was having financial problems. Robert also persuaded her that the children would have fewer disciplinary and school problems if he were around to take care of them.
Rita testified that she had received counseling from a clergyman and was told that she should be a better wife. She stated that she felt guilty about the family problems and had the feeling that if she were a better wife Robert would not conduct himself in the way he did and would not beat her.
Apparently the relations between Robert and Rita and the rest of the family were comparatively tranquil during part of 1978; but, according to Rita’s testimony and that of the children, around September of 1978 Robert’s violence commenced to escalate. He struck the children and choked Greg, the seventeen-year-old son. During November and December, according to Rita, the situation got even worse. She said that Robert would wake up in the middle of the night and without provocation would commence to beat her.
She said that, about two or three weeks prior to the shooting, Robert beat her because their fifteen-year-old, Rhonda, was doing poorly in school. Rita testified that she attempted to commit suicide that night by taking a whole bottle of Librium. Greg, the seventeen-year-old, testified that his father had struck him a number of times, sometimes with his fist, and that in November and December, there were more fights and pummelings than usual.
*491Rhonda testified her father beat her frequently and that he commenced beating her when she was twelve or thirteen years old. At the end of November, she said her father was angry more often than usual and hit her almost every day.
There was testimony that on an earlier occasion, without apparent provocation, Robert got a gun and shot a visitor to the house in the leg.
The trial court refused to allow other non-family witnesses to testify about specific acts of violence by Robert against Rita or the children. Two neighbors, however, were allowed to testify to Robert Felton’s reputation. One said he had the reputation for getting drunk and abusing his wife and his family. The other said his reputation was that he was mean and beat up his children.
The shooting occurred on the evening of January 2, 1979. That day was the ninth birthday of the two youngest children, Kim and Kelly. Rita testified that she and Robert went to the bank that morning to cash her AFDC check. The record showed that, because of advanced emphysema, Robert was totally disabled from performing his work as a concrete worker. He received a disability payment from the Social Security Administration; and Rita was the recipient of some AFDC assistance.
After cashing the check, they both went to a bar and drank beer. Rita went to a bakery and ordered birthday cakes for the twins and then stopped at a bar. When Robert came into the bar, Rita said he looked furious and that his tone of voice was “scary.” After they returned home, the altercation continued, and Robert continued to shout at her because the snowmobile suits they had bought for the twins were too large. Rita said that, because she was scared, she asked Charles, her nineteen-year-old son, to stay. Charles, however, left the house to return to his own living quarters.
*492A fight then ensued between the deceased and Rhonda, the fifteen-year-old. According to Rita’s testimony, after the fight Rhonda’s lip was bleeding.
Greg, the seventeen-year-old, testified that, when his parents returned, it was obvious to him that both of them had been drinking. He also stated that his father and Rhonda got into a fight, which he broke up. Shortly thereafter, he left the house. Rhonda also testified that both of her parents were drunk and that her father was angry with her mother, and later her father came after Rhonda, hit her with his fist, and split her lip, and that Greg then came in and pulled the father off of Rhonda. She left the house with Greg.
Rita testified that she was frightened after Greg and Rhonda left the house and that Robert started yelling at her because the children were “brats.” Robert followed her into the kitchen and took a swing at her, but at that point the nine-year-old twins grabbed Robert and pushed him away and into the livingroom.
Rita stated that at about 8 or 8:30 p.m., she went into her bedroom. She testified that she was scared and shaking. She testified, “Everything was — everything was just — it just looked — it was like a powder keg, like something’s gonna blow.” She thinks she slept for a little while.
Then she got up, went to the kitchen, and looked in on Robert, who was sleeping on the couch. She stated that she went back and sat on her bed and was scared because Robert often woke up in the middle of the night and on some occasions when he woke up he would beat her. She was afraid that if he woke up he would kill her. She said that she never felt the way she felt that night — it was more than just being scared.
She testified that she remembered sitting on the bed holding Robert’s .22 caliber single shot rifle. She stated that she debated what to do, and she testified that she *493knew there was no other way out without Robert killing her or the children. She said that she knew that she would be arrested but at least the children would be taken care of and that the twins would be placed in a foster home so they could have a real family life.
On cross-examination, Rita said that she knew then that she had to kill him and that she knew that it was wrong. She stated that she did not want to kill him but that she had to. She said she does not remember firing the shot, but she remembers standing by the couch with the gun against her shoulder and she remembers feeling the trigger. The next thing she said that she remembered was lying down on her own bed. She said she was still scared because she thought Robert would come and kill her. She heard footsteps in the hallway, but they stopped; and she testified that then she heard Robert snoring. The next thing she remembered was seeing Greg and Rhonda. She testified that, at that time, she was in a daze. She remembers that she called the police and remembers seeing the police come, but she does not remember making any statements to them. On re-direct examination, she testified that she intended to kill Robert.
Greg testified that he and Rhonda returned to the house at about 9 or 10 o’clock. He said that his father was yelling at his mother, and then Greg watched television for a while and then went to bed. He said that at this time his mother was in her bedroom and that he talked to her there. Greg stated that he thought he went to bed at about 10:30 p.m. He later heard a thud, woke up, and found his mother in the kitchen. She told him she “shot your dad.” He told her to call an ambulance, but she followed him into his bedroom. Greg said Rhonda got up at about that time and she also told Rita to call an ambulance. He said his mother was in a daze and was weeping.
*494Rhonda testified that, after she had gone to bed, she heard a noise and when she looked at her father he was making a noise like he was snoring. She then again looked at her father and, upon seeing the injury, told her mother to call an ambulance. When Rita did not, Rhonda called a neighbor, who came over and told Rita to call the police. Rita picked up the phone and then put it down and said, “I did that for you kids." The neighbor testified that when he came over and he asked Rita if she was going to call the police, she said “no” — she wanted Robert to die. The neighbor testified that Rita said she was sick of being beat up all the time and that she shot him for the twins’ birthday so they would not have to live with him. After Rita made the call, the police arrived and took her into custody.
Officer Kraklow, who came to the house, testified that Rita Felton said she pointed the gun between her husband’s eyes and pulled the trigger but he would not stop breathing. He testified that Rita was calm and that she told him she had been living like this for twenty-three years and was glad that it was over.
After Rita was taken to the police station, she was given the Miranda warnings. She said she would not make a statement. Nevertheless, the officer asked her how long she had contemplated this, and she said, “A long time." Another officer testified that, when she was washing her hands to be fingerprinted, Rita said that, “I should have finished the job right away.”
The principal defense witness was Dr. Joy Kenworthy, a clinical psychologist. She recounted that she had heard Rita’s in-court testimony and said that she had spent over twenty-five hours in consultation with her. The essence of her testimony was that Rita’s experience was typical of a battered spouse and that the evidentiary picture of Robert was typical of an assailant.
*495She explained the mutual dependency of persons in such a relationship — that an assailant such as Robert developed a dependency on the victim and that he felt he couid not survive without her. On the other hand, over the years, one in the position of Rita had guilt feelings and became detached from reality in order to tolerate the unrelenting tension and fear. This, the witness said, explained Rita’s calmness on the night of the shooting. Dr. Kenworthy stated that Rita was in a state that anyone would be in under similar circumstances. Dr. Ken-worthy acknowledged she did not know the legal standard for non-responsibility for one’s acts. She did state, however, that Rita functioned automatically to survive, and that her aggressive response was not in fact consciously thought out. On cross-examination, Dr. Kenworthy stated that Rita was not aware of her real reasons for her actions and the explanations given to the police officers were inaccurate.
Dr. Frederick Fosdal, a psychiatrist, appeared as a witness for the state. He testified that the killing stemmed from the very unusual and pathological marriage and that the shooting was the culmination of years of marital discord.
The jury returned a verdict of second-degree murder. Subsequently, appellate counsel brought a postconviction motion under sec. 974.06, Stats., based primarily upon the allegation that trial counsel was ineffective. An evi-dentiary hearing was held, at which the trial counsel testified.
The trial attorney acknowledged that he was not well versed in the criminal law. He stated that, although he had practiced law for over three years, it had not been in Wisconsin and that he had never handled an entire felony case. This case became his responsibility on the first day he actually assumed the duties as a public defender. He acknowledged that he was probably incompetent to han-*496die a case of this magnitude. He stated that he attempted to get experienced private counsel appointed, but his superiors encouraged him to undertake the defense. The administrators of the public defender’s program provided him with the assistance of a more experienced attorney; but according to the trial attorney, he had only limited opportunity to consult with that attorney until just before the time of the trial.
He testified that his theory of defense was self-defense based upon the fact that he perceived, correctly as the record showed, Rita Felton as a battered spouse. He acknowledged that he was ignorant of the possible defense afforded by the Wisconsin law under sec. 940.05 (1), Stats., manslaughter — heat-of-passion.
He also admitted that in his focus upon self-defense he never discussed the problems of mental disease or defect with Dr. Kenworthy. He was aware of the fact that Dr. Kenworthy did not hold herself out as a forensic psychologist and that it would be his responsibility as defense counsel to explain legal consequences to the psychologist, but he never did so.
Counsel admitted that he entered the plea of not guilty by reason of mental disease or defect without discussing it with his client and he withdrew that plea without any discussion with Rita of the legal consequences or the effect of the withdrawal.
There was also evidence that surfaced at this hearing that the trial attorney had some knowledge that Robert Felton had molested the daughters. Whether the trial attorney was fully aware of these molestations is unclear from his testimony, but Rita Felton said she had informed him of the molestations shortly before trial. One of the molestations allegedly occurred on the day of the shooting.
Dr. Kenworthy at the postconviction hearing stated that she did have expert knowledge of the problems of *497battered women, but she did not have a knowledge of the problems involved in the criminal law. She said that, when she testified, she did not have an understanding of the heat-of-passion defense or of nonresponsibility by reason of mental disease or defect.
She testified that now that she knows these standards she would be able to testify that the battering was the provocation which caused Rita to be deaf to the voice of reason and to act from that impelling force rather than from wickedness, cruelty, or recklessness. The implication of her testimony was that she would have been able to so testify at trial if the attorney had properly prepared her as a witness. Significantly, she also testified that any other person in that battering relationship would experience the same emotions and have the reaction that Rita did.
She also testified that, now that she understands the plea of not guilty by reason of mental disease or defect, she could testify that Rita had a severe personality disorder and did not have the substantial capacity to appreciate the wrongfulness of her acts and was unable to conform her conduct to the requirements of the law.
Dr. Fosdal, the state’s psychiatrist, in an affidavit stated that, had Felton’s trial attorney asked him he would have recommended a plea of not guilty by reason of insanity, because the determination of mental non-responsibility is for the jury. He also stated that the jury “might” have accepted the heat-of-passion defense. He did acknowledge telling the lawyer that Rita Felton, in his opinion, did not have a particular mental disease or defect at the time of the offense.
Following the evidentiary hearing, the trial judge denied the motion for a new trial. He made no formal findings of fact and entered no conclusions of law, but at the close of the hearing, he addressed himself to the *498principal matter raised — the ineffectiveness of counsel. His oral holding from the bench failed to refer to any of the facts adduced during the post-trial hearing in respect to ineffectiveness of counsel. Rather, he stated, referring to the lead attorney and to the attorney who furnished assistance at trial:
“[I]t is the Court’s recollection that while they may not have presented a perfect defense, the Court is aware or was made aware of the fact that both of these attorneys had previous experience and that they handled themselves well in the Courtroom and they adequately represented the defendant and participated in her defense in an adequate manner. . . . [T]here may have been some shortcomings in the matters handled during the trial, but very often that is a matter of trial strategy, and in this case the defense relied on the battered wife syndrome defense, and results of the jury verdict indicate that they were successful and presented a defense by which the State . . . did not receive [a first degree murder conviction] .... [T]he Court finds that the defendant was adequately represented by her attorneys during the course of the trial, and that the defenses they put forth were a matter of choice and of trial strategy, and not grounds for a new trial.”
The court also listed the seven other grounds which had been urged in the motion for a new trial. The court denied relief on these grounds. It merely restated its original rulings without any discussion whatsoever of either the evidence at trial or that produced at the hearing.
The motion for a new trial was coupled with a motion for reduction of sentence; and in respect to that motion, the judge gave a detailed statement of the facts adduced at the hearing and concluded that the original sentence was too severe. The judge found that Rita Felton was making rehabilitative progress and, accordingly, reduced the original sentence from sixteen years to ten years.
The basic question on this review, as on the appeal, is the one to which the trial court initially addressed its *499attention on the motion for new trial — whether counsel for Rita Felton had been ineffective.
The United States Constitution and the Constitution of the State of W isconsin guarantee the right to counsel. The right to counsel is more than the right to nominal representation. Representation must be effective. Cuyler v. Sullivan, 446 U.S. 335 (1980); State v. Koller, 87 Wis. 2d 253, 274 N.W.2d 651 (1979); State v. Harper, 57 Wis. 2d 543, 205 N.W.2d 1 (1973).
Whether or not an attorney is experienced is not the criterion for determining whether counsel was effective in a particular case, and the fact that an attorney is ineffective in a particular case is not a judgment on the general competency of that lawyer. It is merely a determination that a particular defendant was not appropriately protected in a particular case. As Judge Bazelon has written:
“Ineffectiveness is neither a judgment of the motives or abilities of lawyers nor an inquiry into culpability. The concern is simply whether the adversary system has functioned properly: the question is not whether the defendant received the assistance of effective counsel but whether he received the effective assistance of counsel. In applying this standard, judges should recognize that all lawyers will be ineffective some of the time; the task is too difficult and the human animal too fallible to expect otherwise.” Bazelon, The Realities of Gideon and Argersinger, 64 Georgetown Law J. 811, 822-23 (1976).
Accordingly, the trial judge’s conclusion that the lead lawyer at the trial and upon whom we are obliged to focus our attention, “had previous experience,” is irrelevant in respect to whether Rita Felton had effective assistance of counsel in this case. It is a fact, however, that the attorney was new to the practice of law in Wisconsin, had little criminal experience, and had never before tried a felony case. This, however, does not establish ineffective assistance of counsel. As the United *500States Court of Appeals for this circuit said in U.S. ex rel. Williams v. Twomey, 510 F.2d 634, 639 (7th Cir. 1975):
“[E]very lawyer must begin his career without experience. His first case is not inevitably so ill-prepared or poorly presented as to justify a finding of his incompetence.”
Thus, even inexperienced counsel may provide representation which is equal to that that would be given by an ordinarily prudent lawyer skilled and versed in the criminal law and, conversely, on occasions even experienced counsel may be ineffective. Accordingly, we judge not the attorney’s general competency, but rather we focus upon whether the representation given in this case to Rita Felton was effective. Nevertheless, from an examination of the record, we are fully aware of the relative inexperience of the trial attorney and the fact that he felt he was obliged by his employment as a public defender to assume representation in this case although he had serious doubts of his ability to provide optimum representation.
The case was a complex and lengthy one, raising numerous issues, only a few of which we consider in this opinion. This case would have provided a substantial challenge to even the most experienced lawyer. Many of the challenges posed by the case were handled well by counsel. Nevertheless, we are obliged to judge effectiveness of counsel, not on the basis of whether a lawyer did his best in light of his experience, but rather on the basis of the standards for “effective” counsel that have been established by this court. This court adopted the modern standard for judging effective assistance of counsel in State v. Harper, supra, wherein we said:
“Effective representation is not to be equated, as some accused believe, with a not-guilty verdict. But the representation must be equal to that which the ordinarily *501prudent lawyer, skilled and versed in criminal law, would give to clients who had privately retained his services.” P. 557.
In Harper we specifically rejected the test previously accepted in this state and in numerous other jurisdictions that a new trial would not be granted upon the grounds of ineffective counsel unless counsel’s performance was “so inadequate and of such low competency as to amount to no representation” or “so inadequate as to amount to no counsel at all and to reduce the trial to a sham and a mockery of justice.” Harper emphasized the duty of a lawyer to investigate adequately the circumstances of the case and to explore all avenues which could lead to facts that are relevant to either guilt or innocence, and it specifically adopted sec. 4.1 of the American Bar Association Standards Relating to The Prosecution Function and The Defense Function.1 Harper pointed out that courtroom performance and demeanor were not the sole criteria whether a proper antecedent investigation had been made. It emphasized that, in determining whether or not counsel had been effective, it is not appropriate by hindsight to insist on what would have been an ideal defense — but rather that a defendant is entitled to a defense which, under all the facts, would have afforded reasonably effective representation.
Since Harper we have reiterated that trial counsel and the defendant may, on the basis of a considered judgment, select a particular defense from among the alterna*502tive defenses that are available. Weatherall v. State, 73 Wis. 2d 22, 242 N.W.2d 220 (1976), cert den 429 U.S. 923 (1976); Lee v. State, 65 Wis. 2d 648, 223 N.W.2d 455 (1974). The defense selected need not be the one that by hindsight looks best to us.
This court has often stated that it disapproves of postconviction counsel second-guessing the trial counsel’s considered selection of trial tactics or the exercise of a professional judgment in the face of alternatives that have been weighed by trial counsel. Weatherall, supra, p. 30; Kain v. State, 48 Wis. 2d 212, 222, 179 N.W.2d 777 (1970).
Nevertheless, it is apparent that the prudent-lawyer standard adopted in Harper implies that there be conduct that is more than just acting upon a whim. It implies deliberateness, caution, and circumspection. It is substantially the equivalent of the exercise of discretion; and, accordingly, it must be based upon a knowledge of all facts and all the law that may be available. The decision must evince reasonableness under the circumstances.
Consistent with the express language of Harper, a prudent lawyer must be “skilled and versed” in the criminal law. The prudent-lawyer standard requires that strategic or tactical decisions must be based upon rationality founded on the facts and the law. If tactical or strategic decisions are made on such a basis, this court will not find that those decisions constitute ineffective assistance of counsel, even though by hindsight we are able to conclude that an inappropriate decision was made or that a more appropriate decision could have been made. Thus, when we look to a lawyer’s conduct and measure it against this court’s standard to determine effectiveness, we cannot ratify a lawyer’s decision merely by labeling it, as did the trial court, “a matter of choice and of trial strategy.” We must consider the law and the facts as they existed when trial counsel’s conduct *503occurred. Trial counsel’s decisions must be based upon facts and law upon which an ordinarily prudent lawyer would have then relied. We will in fact second-guess a lawyer if the initial guess is one that demonstrates an irrational trial tactic or if it is the exercise of professional authority based upon caprice rather than upon judgment.
An essential ingredient which must go into the determination of whether assistance of counsel is ineffective is prejudice to the defendant. As stated in the concurring opinion in State v. Fencl, 109 Wis. 2d 224, 241, 325 N.W. 2d 703 (1982):
“If the failure could have had no adverse effect on the defendant, the representation would not have been any more effective had that failure not occurred.”
See also, Weatherall, supra, p. 32.
The omission or commission sufficient to make a failure of counsel prejudicial has not been defined with clarity or consistency by this court, or for that matter by courts in other jurisdictions. Nor do we have occasion in this opinion to explore in detail the boundaries of what failures of counsel are prejudicial. It is appropriate to continue as we have heretofore, that is, to make such determinations of prejudice on a case-by-case basis. See, Fend, supra, Abrahamson, J. concurrence p. 256.
The question in this case is whether there was ineffective counsel where the lawyer failed to inform himself of a defense provided for in the statutes, and where he failed to adequately investigate the facts in respect to a potential defense, when the record indicates that, had these failures not occurred, a judge, after a proper evaluation of the record, could be impelled to instruct the jury in respect to these defenses and a jury could return a verdict based on one of these defenses.
*504In the instant case the principal grounds for a claim of ineffective counsel are the admitted ignorance of trial counsel of the statutes authorizing the heat-of-passion manslaughter defense and the failure to give due consideration to the defense of not guilty by reason of mental disease or defect and to make any meaningful investigation of the facts in respect to that defense. The record clearly indicates that Rita Felton was deprived by the conduct of her counsel in this particular case of the benefit of these two crucial defenses. She was prejudiced in her defense, and counsel was ineffective. We accordingly reverse and remand for a new trial.
The court of appeals raised the question of the proper standard of review in this case where the issue is the ineffectiveness of counsel. The first step in the analysis of a claim of ineffective assistance of counsel is to determine what the attorney did or did not do and the basis for the challenged conduct. State v. Fencl, 109 Wis. 2d 224, 247, 325 N.W.2d 703 (1982) (Abrahamson, J., concurring). These are factual determinations. A trial court’s findings of fact will ordinarily be upheld unless they are against the great weight and clear preponderance of the evidence. Wurtz v. Fleischman, 97 Wis. 2d 100, 107, 293 N.W.2d 155 (1980). This is the type of finding to which this court referred in applying the great weight and clear preponderance of the evidence standard in State v. Rock, 92 Wis. 2d 554, 556, 285 N.W. 2d 739 (1979).
In this case, the trial court did not make any findings of fact. However, the facts and the inferences to be drawn therefrom are undisputed. This court is not bound by a determination of the trial court which is based on undisputed facts, for, under those circumstances, only a question of law is presented. Compton v. *505Shopko Stores, Inc., 93 Wis. 2d 613, 616, 287 N.W.2d 720 (1980). The ultimate conclusion of whether the attorney’s conduct resulted in a violation of the right to effective assistance of counsel is a question of law. Only a question of law is presented in this case, so deference to the trial court’s determinations is not appropriate.
One of the major aspects of the claim of ineffective counsel was the failure of the attorney to inform himself on the possible defense of heat-of-passion manslaughter. At the postconviction hearing, the trial attorney admitted that he was unaware of sec. 940.05(1), Stats., which provides:
“Whoever causes the death of another human being under any of the following circumstances is guilty of a Class C felony:
“(1) Without intent to kill and while in the heat of passion . . . .”
The trial counsel focused entirely upon self-defense as it applied to a battered spouse, as Rita Felton clearly had been over the course of the years. Because of his preoccupation with this defense, counsel failed to explore the statutes further, and he never discovered the heat-of-passion provision. Accordingly, he never was in a position even to consider whether, in light of the facts, heat of passion was an appropriate defense; and he never explored the circumstances to determine what evidence existed that would support a finding that Rita Felton acted in the heat of passion. That defense was unknown to him, and he asked for no instructions on heat-of-passion manslaughter.
The standards adopted in Harper, supra, mandate that, to be effective, counsel must act in a manner that demonstrates that he is versed in the criminal law. The failure to be informed of this defense in the circumstances of this case constitutes a glaring deficiency in trial counsel’s knowledge of the law. Without that knowl*506edge, it was impossible for him to weigh alternatives and to make a reasoned decision consistent with the standard of performance expected of a prudent lawyer.
In Harper, we adopted the defense function standard 4.1 mandating the duty of a trial lawyer to fully investigate the facts of the case. The failure of counsel here to inform himself of the statutory defense of heat-of-passion manslaughter violates A.B.A. Defense Function Standard 5.1 (a). That section provides:
“5.1 Advising the defendant.
“(a) After informing himself fully on the facts and the law, the lawyer should advise the accused with complete candor concerning all aspects of the case, including his candid estimate of the probable outcome.”
This standard is explained in Commentary a:
“Advice on the plea.
“The duty of the lawyer to investigate fully the facts of the case, regardless of the anticipated plea, is discussed in section 4.1, supra. The lawyer’s duty to inform himself on the law is equally and often more important; although the client may sometimes be capable of assisting in the fact investigation, he is not educated in or familiar with controlling law. The responsibility to know the law which the lawyer bears is a heavy one, given the rapid pace of change in many areas of criminal law and procedure.” (Emphasis supplied.)
Although merely holding oneself out to be a “lawyer” by implication imposes the requirement that a “lawyer” be fully informed on the law pertinent to a case, we expressly adopt sec. 5.1(a), supra, and the portion of the commentary which we have quoted. Certainly, a prudent lawyer, skilled in the criminal law, would be certain to be informed of the elements of heat-of-passion manslaughter prior to undertaking a first-degree murder trial, especially when, as in this case, there was evidence of provocation.
*507The court of appeals correctly concluded that the failure to consider the heat-of-passion manslaughter defense cannot be said to be a matter of trial strategy. As we pointed out above, “strategy” or “tactics” carry with them the connotation of familiarity with available choices as a prerequisite for a rational determination of a course of action based on pertinent law and facts. While this court is loath to interfere with a lawyer’s exercise of professional judgment by a hindsight evaluation, we are satisfied that requiring lawyers to inform themselves of the relevant law prior to formulating a defense or determining a strategy or tactic will promote the exercise of rational, informed, and considered judgment.
There are, of course, a multitude of cases in which a lawyer’s failure to inform himself of a particular defense could in no way be prejudicial, because that defense would not be pertinent to the facts. In the context of the failure to inform himself of the law, which prevents the exercise of professional judgment, and which prevents the opportunity to make a rational choice of trial strategy, prejudice does exist if the facts presented at trial or in the postconviction hearing would justify the submission of a defense or of a lesser-included offense to the jury. The facts revealed in this case must be such that the lawyer would, had he known the law, been able to introduce sufficient evidence to raise the heat-of-passion issue. State v. Lee, 108 Wis. 2d 1, 11, 321 N.W.2d 108 (1982).
The burden upon the defendant where a heat-of-passion defense is projected is merely the burden of production as opposed to the burden of persuasion. It is for the accused to come forward with some evidence in rebuttal of the state’s case — evidence sufficient to raise the issue of the provocation defense. The burden of persuasion, of course, always remains upon the state. *508The provocation issue is sometimes implicitly, or explicitly, sufficiently raised by the prosecution. Hence, if it appears from the evidence, either that produced by the state or by the defense, that a jury, under a reasonable view of the evidence, could conclude that the state has not sustained its burden of disproving provocation beyond a reasonable doubt, the heat-of-passion manslaughter instruction must be given to the jury. See, State v. Lee, supra, at 11. We have pointed out that the heat-of-passion manslaughter defense is dual-faceted. We said in State v. Lee, supra, at 12:
“To be legally adequate, the heat of passion resulting from . . . provocation must satisfy both an objective and a subjective test. In State v. Williford, supra, 103 Wis. 2d at 113, we wrote: ‘The heat of passion element of this offense must be the result of adequate provocation and the defendant’s state of mind at the time of the commission of the homicide; thus, heat of passion has both an objective (provocation) and a subjective (state of mind) facet.’ In other words, the provocation must be such that would cause an ordinary, reasonable person to be overcome with emotion to the degree discussed in the Johnson line of cases. Furthermore, this provocation must have actually caused such a reaction in the particular defendant.”
Although the court of appeals correctly concluded that the failure to raise the heat-of-passion manslaughter defense could not be overlooked as being trial strategy, it concluded, erroneously, that the failure in this respect was not prejudicial because neither the objective nor subjective tests set forth above could be satisfied. Because the facts would, had the law been known to counsel, have been sufficient to warrant an instruction on heat-of-passion manslaughter, the omission of counsel was prejudicial. The facts properly presented were such as would satisfy the accused’s burden of creating an issue in respect to the defense of heat of passion.
*509In State v. Williford, 103 Wis. 2d 98, 113, 307 N.W.2d 277 (1981), we said the provocation that would reduce murder to manslaughter must be sufficient to:
“ . so overcome and dominate or suspend the exercise of the judgment of an ordinary man as to render his mind for the time being deaf to the voice of reason: make him incapable of forming and executing that distinct intent to take human life essential to murder in the first degree, and to cause him, uncontrollably, to act from the impelling force of the disturbing cause, rather than from any real wickedness of heart or cruelty or recklessness of disposition. . . ”
The court of appeals erred when it held that the objective test could not be met by the evidence produced at trial or the evidence which the postconviction hearing demonstrated could have been produced. It reasoned that, although Robert Felton’s actions were violent and despicable, the ordinary person would not have responded to them in the way Rita Felton did. The court of appeals appeared to think that it would have been erroneous to consider how others of a similar background as Rita’s would have reacted, but only to consider whether a person ordinarily constituted would have acted in the heat of passion in response to the immediate provocation. It declined to consider how an ordinarily constituted person who was a battered spouse would have reacted to the provocation. In this we consider the court of appeals erred. While it is true that a defendant’s background is not in general relevant to the objective test for heat of passion, the question is how an ordinary person faced with a similar provocation would react. The provocation can consist, as it did here, of a long history of abuse. It is proper in applying the objective test, therefore, to consider how other persons similarly situated with re*510spect to that type, or that history, of provocation would react.
This question is not one of first impression in this court. Twenty years ago this court found it appropriate to instruct on heat-of-passion manslaughter in the case of battered spouses. In State v. Hoyt, 21 Wis. 2d 284, 128 N.W.2d 645 (1964) (rehearing), the defendant wife had been beaten, choked, and humiliated by the husband during the marriage. This court said:
“If we look solely at the action of Mr. Hoyt in the last few minutes before the shooting, it seems clear that such actions would not be sufficient to produce the required degree of disturbance in an ordinarily constituted person not previously subjected to the treatment visited upon Mrs. Hoyt by her husband and disclosed by the record. On the other hand, it seems reasonable that the treatment to which Mrs. Hoyt had been subjected for a long period of time, and the public humiliation of her within the previous hour would have a cumulative effect upon any ordinary person so that the provocation just before the shooting would be greatly magnified.” 21 Wis. 2d at 291.
Thus, this court has held that the objective test may be satisfied by considering the situation of an ordinary person who is a battered spouse. Williford, at page 122, summarized the situation in Hoyt, stating the defendant:
“. . . over a long period of time, had attempted to sublimate or repress her pent up frustrations with her husband’s frequent insults and abuse and became something of a human time bomb . . . .”
The history of the relationship which has been recited earlier in the opinion provides evidence of provocation sufficient to support a conclusion that the objective portion of the test was sufficiently raised. Had the heat-of-passion manslaughter defense been asserted, a trial judge could reasonably give that instruction and a jury *511could return a verdict based on facts satisfying the objective portion of the test for heat-of-passion manslaughter.
It should also be noted that, at the postconviction hearing, some evidence was introduced that Robert Felton had sexually abused two of his daughters, one of them just shortly before the shooting. The judge limited testimony on this issue, but it appears that, had Rita Felton’s attorney been aware of the heat-of-passion manslaughter defense and had that defense been submitted to the judge, his ruling might have been markedly different.
It seems clear that the history of abuse, plus the provocation which occurred on the day of the shooting, was clearly sufficient under the rationale of Hoyt to raise a jury issue as to the objective facet of heat of passion. Moreover, Dr. Kenworthy, the psychologist, testified at the postconviction evidentiary hearing that Rita Felton reacted in a manner that she would expect any ordinary person eventually to react as a result of a course of abuse.
While the court of appeals, in concluding that the standards for heat-of-passion manslaughter could not be met in this case, attempted to analogize this case to Williford, we do not think the cases similar. In Williford, the physical and verbal abuse was mutual, and none of the provocative events occurred within two weeks prior to the shooting. We held that, under those facts, there was a lengthy cooling off period, and hence the objective test was not satisfied. Here, at the most, the last assault or provocation occurred within two or three hours. We cannot conclude, as did the court of appeals, that, under these circumstances, as a matter of law, the heat of passion in a reasonable or ordinary person would have dissipated in less than three hours. The objective facet of the heat-of-passion test was satisfied.
A jury could also have found that the subjective test was satisfied. The court of appeals reasoned that the *512evidence showed that Rita Felton sat down and debated with herself whether to kill her husband, that she considered the consequences of her action, and that she finally determined that her action was necessary in order to save her own life and to improve the lives of her children. Under some circumstances, certainly this evidence would tend to show a deliberateness consistent with malice of forethought or the specific intent to kill another person, which is essential to murder in the first degree and which would negate a subjective provocation. However, there was evidence in this case that Rita Felton, just before the shooting, was overcome with extreme anger and terror. She testified that she was shaking and in a daze. As the testimony of Dr. Kenworthy pointed out, a person in the situation of Rita Felton could not in fact explain her motives and intentions, and that Rita was in a daze and was acting uncontrollably from the force of the disturbing provocation.
Although Rita Felton testified that she intended to shoot her husband, she testified that she had to shoot him. She was unable to recount much of what happened shortly before and after the time of the shooting. We said in Lee, supra, p 11:
“The heat of passion upon adequate provocation does negate the ‘distinct intent to take human life essential to murder in the first degree.’ ”
But we pointed out that intent or the volitional determination to take the life of another does not in itself abrogate heat-of-passion manslaughter. We stated in Lee, p. 9:
“The purpose of sec. 940.05(1), Stats., is clearly to punish homicides less severely if they result from the heat of passion. . . . This very intent is typically the result of the heat of passion. . . . This intent results from 'the impelling force of the disturbing cause rather than from *513any real wickedness of heart or cruelty or recklessness of disposition.’ ”
It seems clear, therefore, that there was evidence to show that, in respect to Rita Felton, the provocation could have been found sufficient to have caused a mental disturbance that overcame, dominated, or suspended her ability to exercise judgment, and rendered her mind temporarily deaf to the voice of reason and which caused her to act uncontrollably rather than from wickedness of heart. Because the evidence shows that trial counsel failed to inform himself about the law of heat of passion and because the evidence adduced, or which could have been adduced, clearly raised the jury issue as to whether the defendant was guilty of sec. 940.05(1), Stats., rather than first or second degree murder, the defendant was denied effective assistance of counsel. Clearly, the inability of Rita Felton to assert that defense was prejudicial.
We also note that the special materials volume of Wis. JI — Criminal SM 6, points to the problems that confront a trial judge in a situation such as the one that arose during the trial of this case. Although a trial judge wishes to avoid impinging upon defense decisions, nevertheless there are circumstances where a court may sw sponte offer instructions. We said in Price v. State, 37 Wis. 2d 117, 130, 154 N.W.2d 222 (1967) :
“ [T] he battle might be so unequal due to the disparity of the skill of counsel that justice would require, in the unusual case, that such instructions be offered for counsel’s consideration.”
It is interesting to note from the record that, at the very beginning of the trial, the judge asked the lead attorney whether he would introduce expert testimony on the defendant’s capacity to form a specific intent. The lawyer responded that he would. The judge then asked, *514“Now, you are sure it’s intent, not some other area such as 'provocation?” (Emphasis supplied.) Counsel’s response was that it was only on intent and self-defense. Had either trial counsel or the judge followed up on the “provocation” question when it became apparent that provocation was an issue in the case which might implicate the heat-of-passion manslaughter defense, a new trial might have been avoided.
The other issue'posed on the motion for new trial to demonstrate ineffective counsel involved the failure of trial counsel to consider asserting the defense of not guilty by reason of mental disease or defect. Trial counsel entered that plea and then withdrew it without consultation with his client. The A.B.A. Defense Function Standard 5.2(a) provides:
“5.2 Control and direction of the case.
“(a) Certain decisions relating to the conduct of the case are ultimately for the accused and others are ultimately for defense counsel. The decisions which are to be made by the accused after full consultation with counsel are: (i) what plea to enter; (ii) whether to waive jury trial; (iii) whether to testify in his own behalf.”
This course of conduct was in itself violative of the standard expected of a prudent lawyer. It deprived the accused of the opportunity of decision-making where the decision is uniquely one that should be the client’s, not the lawyer’s.
In the instant case, there is no intimation that, at the time of trial or at the time of consultations with the lawyer, Rita Felton was incompetent or unable to make a rational decision. She was fully able to discuss the case with her attorney. The withdrawal of the plea was undertaken in open court, and she gave her acquiescence to that withdrawal. However, the record proves that there was no consultation with her whatsoever on the meaning or *515consequences of the original plea or of its subsequent withdrawal. There is some indication in the record that the original entry of the plea was not taken seriously by counsel and that it was made only for delay. It would appear, however, that in the setting of this case, mental disease or defect could have been an important defense and could have posea a choice to counsel and to the defendant which should have been considered seriously by both of them.
As stated above, we have heretofore adopted in Harper sec. 4.1 of the Standards for the Defense Function in respect to the duty of the lawyer to conduct an investigation of the case and to explore all avenues leading to the facts relevant to guilt. We explicated this duty to investigate in Roe v. State, 95 Wis. 2d 226, 239, 290 N.W.2d 291 (1980):
“The duty to investigate extends to the exploration of whether the defendant possessed the requisite intent or capacity to commit the offense charged when it is not disputed that the defendant performed the act in question.”
As the court of appeals pointed out, counsel’s investigation was perfunctory at best. Prior to trial, trial counsel asked Dr. Fosdal, the state’s psychiatrist, whether he thought the defendant had any mental disease or defect at the time of the offense. Fosdal told him that the defendant was not at that time suffering from any particular mental disease or defect. However, in an affidavit submitted by Dr. Fosdal at the postconviction hearing, the state’s psychiatrist said that, had he been asked, he would have testified that, at the time of the offense, Rita Felton evinced a physical and mental disturbance, which was manifested by her despair, frustration, resentment, and fear, which impaired her judgment and which a jury could have considered to be a mental disease or defect. He stated that he would have recommended a plea of not *516guilty by reason of insanity, because the determination of nonresponsibility was for a jury and that a jury under the facts could have come to the conclusion that Rita Felton was not responsible for her conduct.
Also, Dr. Kenworthy stated that there had been no consultation with her by counsel with respect to the possibility of the defense of mental disease or defect. She acknowledged that she was not a forensic psychologist, and without that prior consultation and instruction by counsel, she would not have given an opinion with respect to criminal responsibility. She stated at the post-trial hearing, however, that with her understanding then of the meaning of mental disease or defect, she could state, on the basis of the evidence in the case, that in her opinion the defendant suffered from a severe personality disorder that constituted a mental disease or defect and that Rita Felton, accordingly, did not have the substantial capacity to appreciate the wrongfulness of her acts or that she was unable to conform her conduct to the requirements of law.
It is apparent that the court of appeals correctly concluded that, because counsel’s investigation was inadequate and violative of a prudent lawyer’s duty to investigate, the determination to abandon the defense of not guilty by reason of mental disease or defect was not a considered judgment based on the facts and, therefore, its abandonment could not be considered a matter of trial strategy. It was not a reasoned and considered choice. By abandoning this defense without proper investigation, Rita Felton was deprived of a defense which under the facts could have been submitted to the jury. She was prejudiced by the deprivation, and counsel was therefore not effective.
We affirm the court of appeals’ holding in this respect, and accordingly the defendant is entitled to a new trial *517at which she may, if she wishes, assert the defense of not guilty by reason of mental disease or defect. We point out, however, that, having determined that counsel was ineffective for failure to assert the heat-of-passion manslaughter defense, the defendant is entitled to a new trial on all issues. In this respect, the decision of this court differs from that of the court of appeals, which declined to order a new trial on the issue of her guilt.
The defendant also contends that the trial judge unduly limited the testimony of other witnesses in respect to the violent acts committed by Robert. We agree with the court of appeals’ conclusion that, if this was error, it was harmless, because the reputation testimony which was admitted and the testimony of the children supported Rita Felton’s testimony about Robert’s persistent abuse and repeated acts of violence. Hence, additional testimony on that issue was not crucial to Rita Felton’s defense.
A statement was taken from Rita Felton following her arrest and in violation of the strictures of Miranda. The essential part of her statement was that she had been contemplating killing her husband for “a long time.” This was elicited by interrogation following the Miranda warning and after she indicated that she did not wish to answer. The question was improper under the circumstances ; and, as the court of appeals determined, the response should have been suppressed. The statement is not to be admitted at retrial. In the context of our present decision, which reverses the entire verdict on other grounds, the statement is, of course, irrelevant.
There was also an objection based upon what the parties refer to as a “bridging” instruction. The jury was instructed to try to agree that the defendant was not guilty of first or second degree murder before considering manslaughter self-defense. Wis. JI — Crim. 1120. Ordinarily, a lesser-included offense does not require *518proof of any fact in addition to those which must be proved for the crime charged. See, sec. 939.66(1), Stats. However, the situation is different with lesser-included types of homicide. Wis. JI — Crim. 1140 recites that the elements of first or second degree murder must exist as a prerequisite to finding a defendant guilty of violation of sec. 940.05(2), Stats., manslaughter imperfect self-defense. Thus, it is argued by counsel for defendant on this appeal that it is only after finding a defendant’s conduct conforms with the elements of first or second degree murder may the jury go on to mitigating conduct in the nature of manslaughter. It is asserted that these instructions are likely to prevent a jury from ever considering manslaughter and are likely to have the effect of coercing the jury to return a verdict of first or second degree murder when the appropriate conviction would be on a lesser degree of homicide.
Because of our disposition of other issues in this case, the propriety of the bridging instruction is not necessary to the disposition of this review, and we do not address that issue as a determinant of this case. Furthermore, we note that revised jury instructions on murder and heat-of-passion manslaughter were adopted by the Criminal Jury Instructions Committee on October 15, 1982, and have as of the date of the writing of this opinion been distributed to the bar and bench. The new instructions are directed to the very problem to which defendant’s counsel points in the present review. Revised Jury Instruction 1130, after reciting the elements of first degree murder, carries with it the admonition:
“In this case, first degree murder also requires that; the defendant was not acting in the heat of passion caused by reasonable and adequate provocation.”
The instruction for second degree murder also carries that same admonition that there cannot be a conviction for second degree murder if the defendant was acting in *519the heat of passion caused by reasonable and adequate provocation.
While a problem may remain in respect to possible jury coercion, it appears that the revised jury instructions eliminate the particular coercive effect allegedly caused by the instructions used in the instant case. Hereafter, a jury will be instructed to reject first or second degree murder if the elements of heat-of-passion manslaughter are present. We do not address ourselves to whether the instructions used in the present case were coercive or tended to impel the jury to overlook the mitigating defenses; nor are we in a position to give our advisory approval to the proposed instructions. We merely note that different instructions will be available and may be used on the retrial of this case.
Because we find counsel’s conduct did not rise to the standard expected of a prudent lawyer reasonably skilled and versed in the criminal law and that the conduct of counsel prejudiced the defendant by depriving her of important defenses, we hold that counsel in this case was ineffective.
We reverse that portion of the court of appeals’ decision which affirmed the trial court’s judgment finding Rita Felton guilty, and affirm that portion of the court of appeals’ decision ordering a new trial on the question of criminal responsibility.
By the Court. — Decision reversed in part and affirmed in part; cause remanded to the trial court for a new trial on all issues.
Beilfuss, C.J., took no part.“4.1 Duty to investigate.
“It is the duty of the lawyer to conduct a prompt investigation of the circumstances of the case and explore all avenues leading to facts relevant to guilt and degree of guilt or penalty. The investigation should always include efforts to secure information in the possession of the prosecution and law enforcement authorities. The duty to investigate exists regardless of the accused’s admissions or statements to the lawyer of facts constituting guilt or his stated desire to plead guilty.”