The issue in this criminal case is whether the evidence supports the trial court’s finding that the defendant, who *70pleaded not guilty by reason of insanity, is still suffering “from mental disease and that it would be dangerous for him to go at large.” See Laws 1977, 180:1 (current version at RSA 651:11-a (Supp. 1982)).
On December 12, 1979, the defendant entered a plea of not guilty by reason of insanity to two charges of recklessly causing serious bodily injury to another in violation of RSA 631:2. After a hearing pursuant to RSA 651:9-a (Supp. 1981), the Superior Court (Mullavey, J.) found the defendant dangerous to go at large and committed him to the New Hampshire Hospital. On December 11, 1981, a hearing was held pursuant to section 180:1 of the 1977 session laws (current version at RSA 651:11-a (Supp. 1982)) to determine whether the defendant then suffered from a “mental disease” and whether it would still be “dangerous for him to go at large” so as to warrant a renewal of the committal order for another two years. But cf. RSA 651:11-a (Supp. 1982) (orders of committal now valid for five years).
The Trial Court (Nadeau, J.) made express findings “beyond a reasonable doubt” and renewed the committal order. Cf. State v. Gregoire, 118 N.H. 140, 144, 384 A.2d 132, 134 (1978) (the court found that a statutory amendment reducing the burden of proof for recommittal to the preponderance of the evidence was unconstitutional). The court, however, requested that a plan be submitted to increase gradually the defendant’s privileges under reasonable supervision in order to test his ability to function outside the institution. The defendant appealed.
The defendant claims that the criminal commitment statute, section 180:1 of the 1977 session laws (current version at RSA 651:11-a (Supp. 1982)), with its language “that it would be dangerous for him to go at large,” sets a more difficult standard than the civil involuntary commitment statute, RSA 135-B:26, with its language “as to create a potentially serious likelihood of danger to himself or to others.” We disagree. The language of RSA 135-B:26, “a potentially serious likelihood of danger,” means nothing more than that it would be dangerous to allow the person to go at large. It is a different way of saying the same thing, and only an overly technical and strained construction could make it mean otherwise.
Moreover, it is not likely that the legislature intended to make it easier to confine those who have not committed criminal acts than to confine those who either have been found to have committed such acts or have admitted them by their plea of not guilty by reason of insanity, as the defendant did in this case. See Gibbs v. Helgemoe, 116 N.H. 825, 827, 367 A.2d 1041, 1042 (1977); Jackson v. *71 Indiana, 406 U.S. 715, 723-24 (1972); Baxstrom v. Herold, 383 U.S. 107, 111-12 (1966).
We further hold that section 180:1 of the 1977 session laws (current version at RSA 651:11-a (Supp. 1982)) is not confined only to danger to persons but extends also to property. That statute contains no words of possible limitation, and it is unlikely that the legislature would have limited its application so as not to include a mentally ill arsonist or anyone else whose mental illness creates a significant risk of substantial damage to valuable property. The people of this State are as entitled to protection of their property as to protection of their persons. N.H. Const. pt. I, arts. 2 and 12.
On the question whether the defendant suffered from a mental disease, a psychologist testified that the defendant had a “borderline personality disorder” and that this was a “recognizable mental disorder.” This was supported also by the testimony of a physician. The evidence furnished ample support for the court’s finding that the defendant suffered from a mental disease, a finding which need not be based on medical testimony.
The defendant contends that the evidence does not support the finding that it would be dangerous for the defendant to go at large. It has been made clear that the determination of mental disease and dangerousness are matters for the trial court to decide and that the opinions of experts are not necessarily dispositive. State v. Rullo, 120 N.H. 149, 152, 412 A.2d 1009, 1012 (1980); cf. Roy v. Perrin, 122 N.H. 88, 95, 441 A.2d 1151, 1155-56 (1982) (standard for assessing competency). The trial court’s findings on these issues will be upheld unless it is determined that no rational finder of fact could make the findings. Id.
We need not dwell long on the meaning of the phrase “that it would be dangerous for him to go at large.” As we have said before, it is not necessary that the court find that the defendant will commit crimes if he is not confined. State v. Hudson, 119 N.H. 963, 967, 409 A.2d 1349, 1351 (1979). It is enough if the risk of harm to society is great enough to outweigh the subject’s right to freedom. Dangerousness is, as we have said before, a legal concept and not a medical one, id. at 966, 409 A.2d at 1351, and no precise measuring device can be applied in making the determination. It is for the trial court to weigh all the evidence and to decide whether in its judgment the release of the defendant into society would create such a risk that it would be a dangerous thing to do.
The defendant next contends that the evidence does not *72support the court’s finding beyond a reasonable doubt that it would still be dangerous to allow him to go at large. The State’s burden is satisfied “if the court is convinced beyond a reasonable doubt that there is a ‘potentially serious likelihood’ that he will be a danger to himself or others.” Opinion of the Justices, 122 N.H. 199, 203, 442 A.2d 594, 596 (1982) (citations omitted). “Recommittal orders ‘need not be based on specific acts, because commitment may be based on a pattern of prior action and testimony relating to the question whether or not any cure for the defendant’s condition has been effected.’” Id. at 204, 442 A.2d at 596 (quoting State v. Hudson, 119 N.H. at 967, 409 A.2d at 1351) (emphasis in original).
Although there was expert testimony that the defendant’s condition had changed from schizophrenia to a borderline personality disorder, there also was evidence that the latter condition involves unpredictability, impulsiveness, limited ability to delay gratification, and limited tolerance when confronted with obstacles, and includes assaultive criminal behavior. There was further testimony that it was difficult to predict the defendant’s responses to various situations, that there was still a need for the structured environment of an institution with people who could help him, and that he still exhibited aggressiveness, impulsiveness, and anger. None of the people who testified could predict what his behavior would be if he were released. Moreover, there was testimony from lay personnel at the hospital of specific instances of lack of self-control on the part of the defendant.
Considering the fact that it was dangerous to allow the defendant to go at large two years earlier and that no one could presently predict his behavior outside the institution, coupled with the evidence that he still needed the structured atmosphere of the hospital, we cannot say that no reasonable person could find as the trial court did.
Affirmed.