Souter, J.

Under Rule 34 of this court, the United States District Court for the District of New Hampshire has sought our opinion on the following question of law:

“Under the law of New Hampshire, may a life insurance policy provision which purports to exclude payment of benefits when the death of the insured is the result of suicide be avoided by proof of the insured’s insanity at the time of death?”

*396The parties have since agreed that the exclusion applies only when the insured was sane when he committed the acts that caused his death, and that proof of the insanity of the insured will defeat the exclusion and result in coverage. Therefore, in their briefs and arguments they have addressed only the issue of defining insanity in the context of the contract of insurance.

The parties do not dispute the facts. The defendant issued a group life insurance policy, under which the plaintiff’s late husband was an insured and the plaintiff was a beneficiary. The policy excluded coverage “[i]n the event of the suicide of an Insured within two years from the effective date of the Insured’s individual coverage.” Within that period the insured killed himself by shooting. In the action to recover under the policy, the defendant relies on the quoted exclusion in denying coverage. The plaintiff claims that the exclusion is inapplicable because the decedent was insane at the time he caused his own death.

The definitional issues before us arise not from any ambiguity of the contract, but from its silence. Our obligation, therefore, is to supply reasonable definitions. Spaulding v. Concord Gen. Mut. Ins. Co., 122 N.H. 515, 516, 446 A.2d 1172, 1173 (1982). At the outset we should note that we accept as reasonable the parties’ limited agreement that insanity as judged by some standard will defeat the exclusion. That has been the result in a body of cases from other jurisdictions, and it is consistent with this court’s recent statements about the nature of suicide, which we take as our starting point here.

In McLaughlin v. Sullivan, 123 N.H. 335, 461 A.2d 123 (1983), we spoke of the common understanding of suicide as a deliberate and intentional act. When contracting parties have not otherwise defined a term, common usage controls. Baker v. McCarthy, 122 N.H. 171, 175, 443 A.2d 138, 140 (1982). We conclude, therefore, that the policy before us should be read to define suicide as deliberate and intentional. This definition implies that one who commits suicide within the meaning of the terms of the policy must understand the natural physical consequences of his act to produce death and must have the capacity to choose effectively to do or not to do the act. Since insanity in the present context is thought of as a mental condition inconsistent with the capacity for suicide, the characteristics of insanity should at the least negate the possibility of one or the other of the elements of understanding and choice that suicide presupposes.

We take this position as our standard for accepting or rejecting the tests the parties have urged us to adopt. The cases they have *397cited present varieties and combinations of three basic alternative descriptions of the insane mind as lacking the capacity to appreciate physical consequences of the act, lacking the capacity to appreciate the moral character of the act, or lacking the capacity for choice to act or not.

Defining insanity as an incapacity to appreciate the physical nature and consequences of the act that produced death has had a long history in insurance litigation. E.g., Dean v. American Mutual Life Ins. Co., 86 Mass. (4 Allen) 96 (1862); Borradaile v. Hunter, 134 Eng. Rep. 715 (1843). The desirability of its adoption does not require extended argument. If the mind cannot understand that the finger is pulling the trigger or that the bullet will pierce the skull, there can be no intent to kill by these physical means.

The second alternative definition of insanity, as an incapacity to appreciate the moral character of the act, has had a history nearly as long. The cases commonly trace its source in American law to the statement in Life Insurance Company v. Terry, 82 U.S. (15 Wall.) 580, 591 (1872), that if the insured causes his death “when his reasoning faculties are so far impaired that he is not able to understand the moral character ... of the act he is about to commit” the death is not by the decedent’s own hand within the meaning of a policy exclusion. The plaintiff urges us to adopt this definition and cites authority from a minority of jurisdictions that have done so. E.g., Garmon v. General American Life Ins. Co., 624 S.W.2d 42 (Mo. App. 1981) (holding proof of such moral incapacity sufficient to defeat the exclusion); Hathaway’s Adm’r v. National Life Ins. Co., 48 Vt. 335 (1875) (apparently holding that proof of such moral incapacity is necessary though not sufficient to defeat the exclusion).

To apply this definition, a finder of fact must first conclude that the act had a “moral character” that the decedent was unable to appreciate. If the definition is to be applied consistently from case to case, there must be a consensus among finders of fact about this moral character. When the Supreme Court of the United States announced the Terry standard as federal common law, there was such a consensus reflected in statutes or common-law rules making suicide a criminal act, “to which is necessarily attached the moral responsibility of taking one’s life voluntarily.” Dean v. American Mutual Life Ins. Co., supra, at 98. That consensus has been lost to the law today, however. When the legislature enacted the present Criminal Code in 1971, it chose not to make suicide a criminal act, and we have been given no indication of agreement on the subject.

Without such a consensus that suicide is morally wrong, the application of any legal standard that refers to the moral quality of the act will vary from case to case with the varying moral positions of *398decedents and fact-finders. One judge or jury could decide suicide was morally acceptable in the circumstances and conclude that a decedent suffered from no incapacity if he believed it was right to kill himself. Another judge or jury could hold that suicide was wrong and conclude that a decedent who believed otherwise probably lacked capacity to appreciate this moral position.

A standard that is subject to such inconsistent application is unable to provide the predictability that a contract of insurance requires. Accordingly, we decline to follow the urging of the plaintiff’s counsel that we adopt the Terry standard of moral capacity, and instead join those jurisdictions that have not adopted that test. E.g., Dean v. American Mutual Life Ins. Co. supra; see Franklin v. John Hancock Mut. Life Ins. Co., 298 N.Y. 81, 80 N.E.2d 746 (1948).

The third possible criterion of insanity in the context of insurance law is probably the most widely adopted, the incapacity of the decedent to choose not to commit the act. Among cases already cited, the court in Hathaway’s Adm’r v. National Life Ins. Co., 48 Vt. 335 (1875), spoke of the insane idea as so controlling the reasoning faculties that the decedent had “no power to resist the insane impulse to take his life.” Id. at 355. The Court in Terry similarly spoke of “an insane impulse which he has not the power to resist.” Life Ins. Co. v. Terry, supra at 591.

We take issue with these formulations only to the extent they may restrict insanity to instances of impulsive killing. It is not clear why we should recognize incapacity to choose only in cases of impulsive behavior, as distinguished from actions long considered. We therefore hold that insanity for present purposes may take the form of mental incapacity to resist the influence upon the decedent to kill himself.

Before giving a final summary statement of the two alternative definitions of insanity that we adopt in this context, we should add a word about terminology and a word of caution. It is common to speak of insanity in terms of a disease or defect. See, e.g., State v. Jones, 50 N.H. 369 (1871). Here we have chosen not to speak of insanity as a disease or defect characterized by one or another incapacity, because the reference to disease or defect would add nothing but words to the definition. We are not speaking in the terms of clinical psychology, and there would be nothing useful for a jury in positing the existence of a disease or defect underlying the forms of incapacity that we have spoken of directly.

We urge caution to understand the limits of our holding. Insanity in the civil contractual context before us may consist of one or another of two functionally abnormal mental states. Proof of either one will defeat the application of the suicide exclusion. We do not *399hold, however, that no other abnormal mental state may qualify as insanity in such a civil context. Rather, we have confined our decision to the possibilities raised by the parties and actually in issue in this case.

In summary, we give the following answer to the district court’s question. An exclusion of coverage in a life insurance policy for death by suicide may be defeated by proof that the death resulted from the decedent’s insanity. Such insanity may be either a mental incapacity to understand that the physical consequences of one’s act will be or will cause death, or a mental incapacity to resist the influence to cause one’s own death.

Remanded to the United States District Court for the District of New Hampshire.

All concurred.