(dissenting). I respectfully dissent. The majority opinion clearly and accurately lays out the relevant facts and case law; I have little to add to either. Nonetheless, I cannot join the majority’s conclusion that this mother was not denied “impartial justice.”
The trial judge’s frustration with the Department of Children and Families is understandable. Faced with a request to remove an eleven month old child from the mother while leaving a four month old child in her care — and no explanation for the different positions — the judge spelled out his concern to the attorneys. Receiving no adequate response, he then, pursuant to G. L. c. 119, § 51 A, ordered an investigation of the circumstances surrounding the mother’s care of the four month old.1 I have no quarrel with that, and I agree that there was no error in denying the recusal motion.2 However, given that history, it was particularly important for the judge to conduct the hearing on the termination of the mother’s parental rights with some care, paying scmpulous attention to providing her with a fair trial.
This, he failed to do. As the majority points out, it is often *549necessary for a trial judge to ask questions in order to elicit crucial information. This is particularly true in a case where the safety of children is at stake. In this case, however, the judge asked more than 1,000 questions, compared to 725 for the four attorneys combined; more importantly, his questions, particularly those posed to the mother, do not seem designed to seek information. On the contrary, they look very much like an effective cross-examination by a skillful adversary. The following exchange, from early in the trial, captures the flavor of the entire proceeding:
The court: “[Ajnswer the question. Did you not visit with [your son] for those eight months you were in Florida?”
The witness: “No.”
Counsel for mother: “When you say ‘no,’ what do you mean?”
The witness: “I was in Florida and, as I had said before, my doctors told me, due to medical reasons, they advised me it wouldn’t be a good idea for me to travel in my condition.”
The court: “What was your medical —”
Counsel for children: “Objection. Move to strike.”
The court: “Overruled. What was your medical condition?”
The witness: “My pregnancy to [my daughter] at the time. And, of course, towards the end, they didn’t really want me to be leaving, you know, the state because they already had something set up to do the C-section for [my daughter] and it was advised at the time —”
The court: “So you missed eight — what were you getting for visits? Once a month?”
The witness: “I think so.”
The court: “So you missed eight visits with the child? . . . Do you think that had any impact on the child?”
The witness: “I’m sure it did at that point.”
The court: “What kind of impact — do you think that it had a lasting impact?”
*550The witness: “Yes.”
The court: “What kind of impact do you think it had on the child?”
The witness: “Not a very good one. I felt bad for him every day, you know.”
The court: “So do you acknowledge, then, that it was harmful for the child not to see you for those eight months?”
The witness: “Yeah, but can I say, um, that I tried to work on the interstate compact plan, which I felt was the best thing.”
The court: “Well, that would have been getting the child down in Florida?”
The witness: “Yeah.”
The court: “Did that interstate compact plan get approved?”
The witness: “I guess [the social worker] can reflect on that.”
The court: “Well, I’ll ask you. Did it get approved?”
The witness: “I don’t remember how that went.”
The court: “In any event —”
The witness: “It took a long time to get it to go through.”
The court: “[T]he child never went to Florida?”
The witness: “No, he never had.”
The court: “Proceed, counsel.”
After the mother’s attorney asked two questions, the judge took over the questioning again; he proceeded to question the witness, virtually uninterrupted,3 for the next twelve pages of transcript. Some of his questions during this part of the hearing seem, at least in hindsight, particularly unfair. For example, the judge asked the mother if her husband had a criminal record and, when she said, “Not that I even knew of,” he asked her if she had checked.4
*551This case is very different from Commonwealth v. Campbell, 371 Mass. 40, 44-45 (1976), on which the majority relies. In that case, “it appeared] that in questioning [the witness] some 1,725 questions were asked by counsel, with an additional eighty-five being asked by the judge. Reference [was] also made to another witness of whom the judge asked some twenty-five questions, with counsel asking a total of 901.” Ibid. Here, as the majority notes, the judge asked more than one-third again as many questions as all four of the attorneys combined. In addition, unlike the majority, I discern, even in a bare transcript, a tone in the judge’s questioning that can be described fairly as “aggressive.” The majority points out that there was no objection by mother’s counsel. However, failure to object to the judge’s questions, or to the manner in which he conducted the trial, cannot carry the same implication of acquiescence as a failure to object to the questions of opposing counsel. See Commonwealth v. Ragonesi, 22 Mass. App. Ct. 320, 322 n.4 (1986) (“It should be apparent to anyone who reads the transcript that the judge would not have tolerated any further objection”).
These cases are among the most challenging a trial judge can hear; they often present with heartbreaking facts, prompting well-founded concerns for children’s safety. “A judge has the right and in some circumstances, the duty to participate in the examination of a witness. Commonwealth v. Festa, 369 Mass. 419, 422 (1976)” (emphasis supplied). Adoption of Seth, 29 Mass. App. Ct. 343, 351 (1990). It is also true that sometimes counsel could be more effective. However, in this case, the attorneys were given little opportunity to participate. This judge repeatedly interrupted each examination, and frequently the questions and answers themselves, often to make, or to reinforce, a point of his own. This is unacceptable. It contributes neither to the truth-seeking process nor to the perception of fundamental fair*552ness that is essential if litigants and the public are to have confidence in our system of justice.5
Despite her considerable shortcomings, this mother was entitled to a fair trial, with representation by counsel and an opportunity to present her side of the story. On this record, I cannot say that she received one. And, under all of the circumstances, I cannot agree to uphold the termination of her parental rights simply on the ground that a fair trial might produce the same result.
The care and protection petition that was filed was dismissed, and the child returned to the mother. Eventually, proceedings involving both children were consolidated for trial in the present case.
I agree with the majority, however, that the judge would have been wise to express his frustration more temperately.
The children's attorney interrupted once, saying that she could not hear the answer.
In addition, the judge questioned the mother extensively — for pages of transcript — about whether she had seen signs of violence or “an anger management problem” in her husband before he assaulted her after her second child was bom. The clear implication in the tone of the questioning was that she was either lying about the history of violence in her relationship or *551neglectful when she got married, because she had not noticed warning signs. Either, or both, of those things may be true. However, the literature on domestic violence is replete with examples of episodes of violence that first appeared only after the relationship had been established. See, e.g., Browne, When Battered Women Kill (McMillan Free Press 1987) (“Typically — in 72 percent to 77 percent of the cases — violence occurs only after a couple has become seriously involved, is engaged, or is living together; rather than in the early, more casual states of dating. Victims have difficulty interpreting assaultive behavior from someone they thought they knew so well”).
This court has addressed this issue previously with this judge in an unpublished opinion issued pursuant to rule 1:28. See, e.g. Adoption of Nurit, 71 Mass. App. Ct. 1104 (2008) (“In this case, the judge asked approximately 2,925 questions as compared to the 2,623 questions asked by . . . all four trial counsels for the department, mother, father, and children. The number of questions asked by the judge was substantial, and we think that the better practice would have been to leave questioning to the parties’ attorneys”).