dissenting.
Poor Robert! In order to “assure” that his relationship with his father “survive[s] and grow[s],” he has been placed in a state of custodial schizophrenia — six months with one parent and six months with the other. In my view, the trial court’s failure to bite the bullet and award custody to one parent or the other has placed a two-year-old child in a state of animated suspension, a custodial limbo. What we have is a probationary custody period of four years, to be reviewed when Robert is of school age. Obviously, that allows the court to defer making a tough decision, but it does not justify it and we ought not approve it. By condoning a practice that may provide instant gratification to a trial judge but real problems to a very young child for whom bonding and attachments to the primary caretaker are crucial, the majority avoids the horns of the dilemma that every trial *500court is faced with in a custody decision, and fails to provide any guidelines or benchmarks for trial courts to follow in resolving disputes between equally fit and loving parents over the custody of young children.
The trial court obviously disbelieved Chris’ evidence of abuse by Michael. That is within the trial court’s authority. However, as a consequence of the alleged abuse, a protective order had been issued restricting visitation between father and child. The trial court concluded that such restricted visitation was not in the child’s best interest. But instead of removing the restraints and ordering liberal and unsupervised visitation, the trial court divided custody. To me that is wrong as a matter of law and fact. I adhere to my dissent in Gravning v. Gravning, 389 N.W.2d 621, 625 (N.D.1986) (Levine, J., dissenting), that divided custody should not be a tool of punishment or coercion. But even the Gravning holding that a custody award may be used to punish a parent for bad behavior, is inapposite to the circumstances of this case. In Gravning there was a unilateral parental decision to restrict visitation; in the case before us, there was a court order restricting visitation. Therefore, even under Gravning, the trial court erred in dividing custody.
I am alarmed by the message sent by the majority. By affirming, this court parts company with the more circumspect jurisdictions that uphold joint custody only where there is parental agreement or cooperation. E.g. In re Marriage of Lampton, 704 P.2d 847, 850 (Colo.1985); Gatti and Gatti, 73 Or.App. 581, 699 P.2d 1151, 1154 (1985); Lemback v. Cox, 639 P.2d 197, 200 (Utah 1981). The current issue of “Trial” features an article on joint custody that discusses a recent study of families with joint physical custody and continuing disputes. In their study, Dr. Judith Waller-stein and Susan Steinman found developmental, behavioral and emotional problems in the children and verbal and physical aggression by the parents. The study’s authors caution “against encouraging or mandating joint custody or frequent access when parents are in ongoing disputes.” “Trial”, April 1989, Joint Custody by C. Rick Chamberlin [citing 4 Barclay’s California Family Law Monthly 480 (1988)].
Divided custody should be cautiously applied and approved because “shifting a child between homes forecloses a stable environment and the development of permanent associations and creates confusion regarding authority and discipline.” 2 Child Custody & Visitation Law and Practice § 13.04[2]. I share the Florida court’s concern and skepticism that “an infant three years old can develop normally and thrive if at the end of every six months he is removed from surroundings familiar to him and forced to become accustomed to new ones.” Hurst v. Hurst, 158 Fla. 43, 27 So.2d 749, 750 (1946). Like the Florida court of appeals, I too would reverse a custody award that places a child alternatively with each parent for six months at a time. Bienvenu v. Bienvenu, 380 So.2d 1164 (Fla.Ct.App.1980). The error is exacerbated when the parents who have been awarded such joint custody are “mutually antagonistic.” Id.
There may be circumstances in which divided custody is justified. See Lapp v. Lapp, 293 N.W.2d 121 (N.D.1980). However, as I stated in Gravning v. Gravning, 389 N.W.2d 621 (N.D.1986) (Levine, J., dissenting), when equally fit parents seek custody of young children, there should be a presumption in favor of the children’s primary caretaker, the parent who provides the child with daily nuture, care and support. Professor O’Kelly describes more persuasively and effectively than I can the rationale for the presumption favoring the primary caretaker of young children and the impact of the intimate interaction between young children and their primary caretaker:
“[Primary] caregiving creates strong psychological bonding and that protection of the psychological bond ... is more important than other relevant considerations in identifying the parents in whose custody children ‘will ... feel more loved or secure, or, in the long term, be more competent and effective as adults.’ ...” O’Kelly, Blessing the Tie *501That Binds: Preference for the Primary Caretaker as Custodian, 63 N.D. L.Rev. 482, 609 (1987).
Expert study indicates that this psychological bond is “the essential cornerstone for a child’s healthy emotional development” and is “critical to the child’s learning to place trust in others and to have confidence in her own capacities.” Id. at 513.
In recommending that North Dakota recognize a preference for primary caretakers, Professor O’Kelly concludes:
“... that it would serve the best interests of young children better than un-weighed considerations of all relevant factors. An explicit preference would more effectively protect children’s primary psychological relationships, reduce the risk of coercive misuse of custody issues in negotiation, and facilitate trial and appellate decisionmaking....” Id. at 533-34.
She suggests that only if the primary caretaker is not the primary psychological parent, a truly rare occurence, should the preference be disregarded. Id. at 534-37. O’Kelly thus proposes a “weak” presumption in favor of the primary caretaker. Id.
In this case, I believe divided custody is wholly inappropriate because it removes a two-year-old child from the care of his primary caretaker for an extended duration and it requires shared parenting from parents the trial court finds unable or unwilling to cooperate. It also deprives Robert of his significant attachments to and associations with his two half-sisters.
I am not alone in my opposition. In In re Marriage of Hickey, 386 N.W.2d 141 (Iowa Ct.App.1986), the court reversed divided custody stating that “[i]f children were like chattels, it would seem eminently fair to divide their time as equally as possible between their parents.” The appellate court awarded custody to the primary caretaker.
So too, the Minnesota court of appeals, in Bateman v. Bateman, 382 N.W.2d 240 (Minn.Ct.App.1986), reversed the trial court’s award of divided custody, finding that it was employed to coerce cooperation between the parents, which was contrary to the children’s best interest. Instead of granting joint custody because the parties could cooperate, the trial court granted joint custody because they could not. “Although ideally the parties should make major decisions concerning their children jointly, joint legal custody should not be used as a ‘legal baseball bat’ to coerce cooperation....” Chapman v. Chapman, 352 N.W.2d 437, 440-41 (Minn.Ct.App.1984).
The adverse effects suffered by children from divided custody as well as experts’ continued opposition to divided custody because of the instability and discontinuity, make the concept extremely problematic. Dodd v. Dodd, 93 Misc.2d 641, 403 N.Y.S. 2d 401 (N.Y.Sup.Ct.1978); Goldstein, Freud, Solnit, Beyond the Best Interests of the Child (1973). Joint custody requires shuttling the children back and forth which undermines not only the stability of their environment but more importantly, the continuity of their care, nurture and support by the custodial parent. It also is a ready source for misuse and abuse by the parents. Dodd, supra.
Divided custody is a tool that should be used sparingly in exceptional circumstances. Even the majority in Gravning acknowledged as much. I fail to see what is so exceptional in this case to warrant divided custody. No expert recommended it and the evidence does not support it.
The cases reversing divided custody abound. E.g. Courie v. Courie, 288 S.C. 163, 341 S.E.2d 646 (1986) [divided custody causes confusion and interferes with proper training and discipline of the child]; Sharpe v. Sharpe, 256 S.C. 517, 183 S.E.2d 325 (1971) [divided custody should be avoided and is justified only in exceptional circumstances]; Dodd v. Dodd, supra [divided custody has adverse effect on children and parents’ unwillingness to cooperate makes it inappropriate]. See also Annot., Comment Note — “Split,” “divided” or “alternate” Custody of Children, 92 A.L. R.2d 695 (1963) and Later Case Service.
Notwithstanding what was said in Gravning v. Gravning, supra, I view divided custody as the antithesis of “Solo-*502monic.” I note that Solomon did not offer the contestants who appeared before him divided custody of the sought-after child and I assume he refrained from doing so for obvious reasons. It may have provided an easy out but it would have resolved neither the underlying contest of parenthood nor the source of continuing upheaval in the continuity of the child’s life. Not that making a sandwich of the child is good for the child, but Solomon’s acuity was in his recognition that not even combative parents would demand so much. Thus, Solomon avoided letting the parents off the hook by resorting to divided custody. Instead, he forced them into resolving the issue based upon the health and welfare (best interest?) of the child. I would require as much from the trial court.
I would reverse the award of joint custody and because the trial court found “both parents are fit custodians,” young Robert’s custody should be awarded to his primary caretaker, Chris Kaloupek. I therefore respectfully dissent.