PERRETTA, J.

There has never been any question as to the paternity of the child whose custody was the sole issue in dispute in this action brought under G.L.c. 209C. At the conclusion of the trial on the complaint, the Probate Court judge awarded the parties shared legal custody of their son, with primary physical custody to the father. It is the mother's principal complaint on appeal that the judgment is tainted by a lack of comprehensive findings on the issue of the father's physical abuse of her as it relates to his fitness to have shared legal and primary physical custody of their child and by an erroneous deference to the child's expressed preference to be with the father. Because of the absence of detailed findings regarding the father's significantly inappropriate behavior and its effect on the child, we reverse the judgment.

1. Relevant pretrial proceedings. Although the mother and the father never married, they had lived together for over ten years in a relationship that had all the characteristics of a marriage. Their cohabitation was, however, marred by frequent episodes of abusive behavior, both physical and emotional. Ultimately, on October 1, 1992, the mother appeared in District Court and obtained an order requiring the father to vacate and remain away from the home and the mother's place of business and to surrender custody of their eleven year old child to her. See G.L.c. 209A. The day after the mother obtained the c. 209A order, the father commenced this action in the Probate Court seeking to establish paternity and to obtain custody of the child.

Within two weeks of the filing of the complaint, the parties stipulated to the entry of a judgment of paternity as well as to temporary orders of custody pending trial on that issue. Pursuant to their stipulation, temporary orders entered providing for joint legal and physical custody of the child. The parties also agreed to the appointment of a named licensed clinical psychologist to serve as a guardian ad litem (guardian) to do an evaluation and report concerning custody of the child.

About four months later, the guardian filed his report in which he recommended that the parties continue to have joint legal custody of the child but that physical custody be modified to provide that the child's primary home during the week be with the father and that weekends be spent with the mother. The basis for this recommendation was the guardian's opinion that the child's close relationship with his father was marked by "problems in the important areas of separation and individuation" which would be heightened by a "forced separation" and likely to cause "major psychological problems for both . . . [the child] and his father" and endanger any rapport that the child might achieve with the mother.

Shortly after the filing of the report, the judge entered a new temporary order. Although by that order the parties retained shared legal custody of the child, the father was granted sole physical custody of him with visitation rights extended to the mother.

2. The trial. It is necessary to recite the evidence in some detail. The father and the mother met in Maine in 1977. Soon thereafter, the father moved in with the mother and her two children from prior marriages, a boy then five years old and a girl who was nine. At that time, the mother was a real estate salesperson during the day, and she worked nights as a cocktail hostess. The father worked odd jobs. In 1981, the family moved to Massachusetts, settling in what is regarded as a summer resort area. The parties' son was born the following year.

Until the child was about five years old, the mother was his primary caretaker. The father started his own business involving property maintenance work, such as carpentry, painting, and landscaping. As the child became older, the mother resumed her work as a realtor, putting in especially long hours during the summer months. Gradually, the father became primarily responsible for the child's care while the mother provided the economic support for the family.

In the beginning of their relationship, the parties would drink to excess. The mother joined Al-Anon in 1978, and, since that time, has limited her alcohol consumption to social occasions. The father became a member of Alcoholics Anonymous in 1985. It is undisputed that throughout their cohabitation, the parties used physical force against each other. The mother, who is 5'7" tall and weighs 150 pounds, described how the father, who is 6'5" in height and weighs 285 pounds, would fly into rages, threaten her, and physically abuse her. He frequently bruised her, and, on one occasion, he struck her with such force that she was rendered unconscious. The mother had to seek medical and police assistance on numerous occasions. She also related that the father verbally abused her; he would threaten her, call her obscene names, ridicule her physical appearance, and make derogatory comments about her French-Canadian heritage. The father related how the mother, without provocation, would kick, knee, elbow, scratch, and bite him. Each used foul language during their arguments.

Many of these incidents of abuse and profanity occurred in the child's presence. They were also witnessed by the mother's other children. The older son stated that he also was the object of verbal and physical abuse by the father who would poke and kick him, slap him with an open hand, and often lose his temper about trivial matters. The daughter testified that she saw the father use physical force against her mother and heard him make threatening statements to her. She described the father as a big man who, when playing with her and her brothers, "liked to play rough."

There was also evidence of highly questionable physical activity by the father in respect to the mother's daughter and the child. The daughter, who left the home when she was seventeen, testified that the father would kiss her with his tongue in her mouth, that he had attempted to fondle and "feel . . . [her] up," and that he had walked about the house nude in her presence. The father did not dispute that he and the child would shower together and give each other body massages with scented oil. The mother stated that when, in 1992, she complained to the father that the matter of oil massages was "getting out of hand" and that she objected to it, the father told her, "'This is between my boy and myself. You stay out of it.'"

Both the mother and the father are in counseling. The mother, who had been abused as a child and in her two prior marriages, is working to regain her self-esteem and confidence. The father had been diagnosed in the past as manic-depressive and had been prescribed lithium. At the time of trial, he was seeing someone once a week to work on issues which he described as having "to do with being in an abusive relationship and how I survived and what I do about going on now and surviving now."

It is the mother's view that the father is over indulgent with the child. There was consist testimony . . . to the effect that the mother, although supportive and loving, was the stricter parent. She would set limits upon the child's activities and time. As put by the trial judge, she was the "disciplinarian." While the mother acknowledges that the father truly loves the child, she is concerned about the child remaining with him because of his inconsistency in matters of discipline and his temper. She explained that she had originally agreed to shared legal and physical custody because she had feared that if she had refused, the father would harm her or run off with the child. She is no longer of the view that she and the father could share custody of the child because since being awarded sole temporary physical custody, the father has refused to discuss matters pertaining to the child with her, he has refused to seek counseling for the child as ordered, he has interfered with her visitation rights, and the child has become rude and physically aggressive toward her.

The father encourages the child's participation in little league, soccer, basketball, lacrosse, and tae-kwon-do. Over the mother's objections, the father has given the child a Honda 50 motorized bike, a Suzuki DS80, a pellet pistol, and a bee-bee gun. When with his father at a gun range, the child is allowed to use a "22 caliber high standard victor . . . target pistol." The father stated that he strives to keep his son intellectually stimulated as well as physically active. He has also provided the child with a personal computer which is "loaded with games of all sorts and levels, from microsoft windows to . . . some really high level games."

In his interviews with the guardian and the mother's expert, a child and family forensic psychologist with a subspecialty in battered women's syndrome, the child expressed a preference to live primarily with his father but not to the exclusion of living with his mother. As related by the guardian at trial, most of the people with whom he spoke in preparing his report, including the child, thought that the father had a bad temper and was intimidating. The child told the guardian that his father would get angry, red in the face, and yell at him. He also told the guardian that unlike in the past, he is no longer afraid of his father.

Although the guardian stated that he thought the mother was a battered woman and that the father had a problem controlling his temper and impulses, he was of the view that the child should reside with his father during the week and that he should be in the joint legal custody of his parents. As noted at the outset, the guardian reported that it was his opinion that a "separation and anxiety problem likely exists for . . . [the child] and his father . . . [T]he separation anxiety would likely be heightened by a forced separation and would likely cause major psychological problems for both . . . [the child] and his father as well as result in greater difficulty, not rapport, between . . . [the child] and his mother." The guardian testified that whether a child who witnesses family violence is himself a victim of violence depends upon the "interjections of the child," but he acknowledged that he had not analyzed the family relationship in respect to the characteristics to be found in a battered family.

3. The findings. Although the trial judge made numerous findings, those which give support to his ultimate decision are based solely upon the testimony of the parents and the guardian. The trial judge found that the father had been the child's primary care giver for the past five or six years, that "[b]oth the mother and father are batterers," that "[b]oth likewise have been battered," that "[b]ased upon the parents' volatile temperments, conduct and abusive language . . . [the child] could be said to be at risk with either one of them," and that both parents love the child who, in turn, loves both his parents. However, a substantial factor for the trial judge was the relationship between the father and the child. "While much of . . . [the father's] conduct and treatment of . . . [the child] may be questionable (i.e. the scented oil massages they give to each other and their showering together) as well as . . . [the father's] permissiveness with . . . [the child], the father and son have clearly developed a special relationship. The evidence does not reflect that . . . [the child] has been hurt by this relationship. To the contrary it establishes the very strong bonding between father and son."

Based upon that finding, the trial judge further found that to remove the child from the father's physical custody would be "counterproductive" to the child's best interests and "negative to [the father]." Visitation rights for the mother were set at alternate weekends with one additional overnight during the week, "as the parties may agree." On those weeks in which there is no weekend visitation, the mother has two consecutive overnights per week "as the parties shall agree." In addition to specified and apportioned holidays, the mother is entitled, upon at least sixty days advance notice, to have the child with her for two weeks during the summer.

4. Evidence of harm to the child. Although not the subject of findings by the trial judge, there was evidence of harm to the child by reason of his relationship with his father. The mother's expert witness, a child and family forensic psychologist having a subspecialty in battered women's syndrome and expertise in the field of family violence and its impact on families, testified that the child related to him during interviews that when he was not with his father, his father would be very sad and unhappy, that he could help his father with his temper by keeping him happy and calm. It was the psychologist's "clinical impression" that the child had taken on "an excessive sense of responsibility for his father's well-being. He was speaking about his father as if he was looking after his father, rather than his father looking after him."

Testing of and interviews with the child by this psychologist led him to the conclusion that the child suffers from some of the emotional and behavioral problems experienced by boys who have witnessed abuse of their mothers. More specifically, the child has a "sense of depression and sadness," "some anxiety," and "an excessive sense of personal responsibility for his father." The psychologist indicated that the child revealed that "he has opinions about his mother, which is [ sic] similar to boys who have witnessed violence and observed physical abuse or verbal or psychological abuse." Examples of the child's opinions are as follows: "He thought his mother exaggerates a lot. He thought that his mother was too fat. He thought that his mother had a number of difficulties." When the psychologist asked the child how he had arrived at these opinions, the child explained that "these are things that his father says about his mother."

When asked his opinion about the undisputed fact that the father and the child showered together and gave each other massages with scented oil, the psychologist stated that such activity was "inappropriate and a violation of appropriate boundaries in terms of contact between a father and a son." Based upon the father's behavior toward the mother's daughter as well as his showers and massages with the child, the psychologist was of the opinion that the father had problems respecting other people's bodies.

It was the opinion of the psychologist that the mother had been battered and abused by the father and that she suffered from battered women's syndrome. Although he acknowledged that the mother had used profanity and force against the father, he refused to accept the father's characterization of himself as the physically abused partner in the relationship. The psychologist regarded the majority of the mother's physical acts against the father as "defensive" and stated that her acts had to be evaluated in light of what had precipitated them. He stated that "in situations where there's family violence . . . women quite often will go to extreme measures to defend themselves," especially where, as here, the man is so much larger.

The psychologist did not agree with the guardian's diagnosis that the child would likely experience separation anxiety if removed from the father's custody. He was of the opinion that "separation anxiety" is not an "appropriate" diagnosis either for a child of this child's age or in a situation where a child is separated from a parent who has been assaultive. The psychologist's disagreement with the guardian's diagnosis was also based upon the testimony of the mother, the older son, and the daughter describing the child when with them as well as on the child's statements concerning his relationship with his mother. Those statements indicated to the psychologist that the child viewed his mother as giving him more independence than does the father. She allows him to make decisions. While the child expressed concern that the mother works hard and is away from him at times, he takes pride in her accomplishments. He also expressed a "sense of belonging to a family unit" when with his mother, brother, and sister.

It was also the opinion of the expert that the child should not be permitted to live with the father who had perpetrated violence against the mother because: "Children who live with batterers basically are learning that violence and threats of violence and intimidation is a way to get what you want. So, they're really having a number of inappropriate behaviors being modeled for them about the use or threats of violence both in general and, in particular, violence within relationships. Children of that age also learn that if the person who has been abusive — the parent who has been abusive during the relationship receives custody for them . . . it says that the violence has been condoned or reinforced in some way; that if somebody is really aggressive and intimidating, they can eventually get what they want." As related by the expert, research indicates that "80 percent of all men who batter their partners have witnessed violence in their family of origin" and that "boys tend to model their fathers . . . especially boys who identify strongly with their father try to model their father's behavior."

Notwithstanding his opinion that the child would be at risk with the father, the psychologist recognized the importance of fostering a relationship between the child and his father. The psychologist encouraged unsupervised visitation, during which time the father and child could pursue their many interests and activities, provided that the mother, the father, and the child were in counseling. If there were no counseling, "I think we're looking more at some sort of supervised visitation." The psychologist's insistence upon counseling for the father was based upon the opinion that it was important for the father to "start taking some responsibility for his role in the violence and try to start to explain that to his son."

5. Discussion. Neither G.L.c. 209C, § 10( a), as added by St. 1986, c. 310, § 16, nor G.L.c. 208, § 31, as amended by St. 1989, c. 689, specify or enumerate factors which must or should be considered in determining custody. Compare G.L.c. 208, § 34. Instead, custody orders and judgments are made on the basis of a determination of the best interests of the child, and the statutes leave it to the trial judge to identify and weigh any factors found pertinent to those interests in the circumstances of the specific dispute. "A Probate Court judge must settle custody in a manner that advances the best interests of the child." Bak v. Bak, 24 Mass. App. Ct. 608, 616 (1987), and cases therein cited.

Although the trial judge recited much of the testimony of the mother, her older son, the daughter, and the psychologist, he made no findings of fact on that testimony. The question squarely before us, then, is the consequence of the trial judge's failure to make detailed and comprehensive findings of fact on the issues of domestic violence and its effect upon the child as well as upon the father's parenting ability.

Notwithstanding the broad range of discretion afforded a trial judge in determining the best interests of the child, there is national consensus that significant consideration must be given domestic violence in making custody orders and judgments. See Statutes at Large: H. Con. Res. 172, 104 Stat. 5182-5183 (1990), where, because of the many effects of physical abuse upon children who witness the abuse of a parent and because "few States have recognized the interrelated nature of child custody, and battering and have enacted legislation that allows or requires courts to consider evidence of physical abuse of a spouse in child custody cases," Congress, by resolution, expressed its view that "credible evidence of physical abuse of a spouse should create a statutory presumption that it is detrimental to the child to be placed in the custody of the abusive spouse." See also Haralambie, Handling Child-Custody, Abuse and Adoption Cases § 7.13 (1993), where State statutes are collected and categorized in support of the statement that the "emerging trend is for legislative recognition that domestic violence, whether or not witnessed by the children, is a factor which should be considered or weighed against custody being awarded to the abuser." See generally Cahn, Civil Images of Battered Women: The Impact of Domestic Violence on Child Custody Decisions, 44 Vanderbilt Law Review 1041, 1062-1071 (1991).

Although our Legislature has left it to the discretion of the trial judge to determine custody on the facts presented, it remains open to the appellate courts to provide guidance on the issue of how domestic violence should affect any decision. "The legislature and/or appellate courts should make it clear that abuse of any family member affects other family members and must be considered in determining the best interests of the child in connection with any order concerning custody. If access to the child is allowed, judges should be directed to make arrangements to protect any family member from further abuse." Supreme Judicial Court, Gender Bias Study of the Court System in Massachusetts 4 (1989). See also, Cahn, Civil Images of Battered Women, 44 Vanderbilt Law Review at 1072.

We conclude that in cases where, as here, there is credible evidence of physical abuse to a household member by a person seeking custody of or visitation with a child, a trial judge must make detailed and precise findings of fact which demonstrate that the effects of the domestic violence on the child have been evaluated and, in the event physical or legal custody is awarded to the perpetrator of the abuse, how such an award advances the best interests of the child.

It appears from the trial judge's findings that because he found the mother and the father to have equally flawed parenting abilities, the relationship between the father and the child and the child's preference weighed the scales in the father's favor. However, we conclude that the most crucial of the trial judge's findings are clearly erroneous (see Capitol Bank Trust Co. v. Richman, 19 Mass. App. Ct. 515, 519 [1985]) for the following reasons.

In considering the many incidents of domestic violence in the present case, the trial judge found that both parents were batterers and that both parents had been battered. To the extent the finding is directed to the fact that both parents used force against the other, the finding is technically correct. Although we are mindful that credibility is "a preserve of the trial judge upon which an appellate court treads with great reluctance," Springgate v. School Comm. of Mattapoisett, 11 Mass. App. Ct. 304, 310 (1981), we nonetheless conclude that the finding is inconsistent with the evidence about which the trial judge made no findings, that is, that the mother suffers from battered women's syndrome and that most of her physical acts of force were defensive. See note 7, supra. In an area as critical as domestic violence, we are unable to conclude that the trial judge's simple finding of mutual battery implicitly rejects this evidence as incredible or irrelevant.

Moreover, this suspect finding, that both parents were batterers, appears to have formed the basis for the trial judge's imprecise finding that the child "could be said to be at risk with either" of the parents. A subject of such gravity calls for a more detailed exploration and analysis of the evidence and an explication of the degree and nature of the respective risks to the child.

Without the benefit of these unsupported findings, it becomes apparent that the custody determination rests primarily upon the report and testimony of the guardian, the father's role as primary care giver, and the preference of the child. However, those three factors, taken separately or in combination and considered in the light of all the evidence, do not, in these circumstances, support any custody determination. Because the guardian did not, as he acknowledged, evaluate the significance of the domestic violence or consider its effect on the child in recommending that the father be awarded primary physical custody, his opinion offers little basis for the trial judge's determination. Further, the father's role as primary care giver over the past five or six years should not be afforded substantial weight without a determination as to the quality of that care. "The fact that one parent has successfully been the primary care parent in the past obviously speaks to the issue of that party's continued ability to exercise care of the child in a custodial situation" (emphasis supplied). Kindregan Inker, Family Law and Practice § 1170 (1990). As for the child's preference, we need state only that the weight to which that preference is entitled is, on the evidence presented and on the law, see Bak v. Bak, 24 Mass. App. Ct. at 617, and cases therein cited, far from considerable.

Little discussion is required on the matter of the award of joint legal custody. Although the parties had originally stipulated to joint legal custody, that stipulation was in respect to temporary orders only. Moreover, as previously noted, the mother testified at trial that once the father was awarded sole temporary physical custody, the joint legal custody award became, in her view, unworkable. "[I]n order for joint custody or shared responsibility to work, both parents must be able mutually 'to agree on the basic issues in child rearing and want to cooperate in making decisions for [their] children.'" Rolde v. Rolde, 12 Mass. App. Ct. 398, 404 (1981). The award of joint legal custody is inconsistent with the overwhelming undisputed evidence of hostility between the parents and their disagreement on matters pertaining to the child. It is, therefore, also inconsistent with the express language of G.L.c. 209C, § 10( a). See K.J.M. v. M.C., 35 Mass. App. Ct. 456 (1993).

6. Conclusion. Because of the inadequacies of the findings on the evidence presented, the judgment is reversed, and the matter is remanded to the Probate Court for a new hearing on the issue of permanent custody of the child. Pending the new hearing, custody of the child shall be as provided for by the parties' previous stipulation (which effectively has been restored by the reversal of the judgment, see Felton v. Felton, 383 Mass. 232, 241 [1981], and the omissions in the guardian's report which, as discussed in this opinion, provided the basis for the trial judge's temporary custody order), subject to any appropriate and necessary orders of the Probate Court as dictated by circumstances which have developed since the trial in this matter.

So ordered.