Opinion of the court by
JUDGE DuRELLEReversing.
Appellee company, in December, 1896, issued to appellant’s husband, Willis B. Campbell, who was then a policeman, a policy by which it insured him for the benefit of appellant in the sum of $1,000 '‘against bodily injuries sustained through external, violent, and accidental means.”
Clause 5 of the policy is as follows: “This insurance does not cover disappearances; war risks; nor voluntary exposure to unnecessary danger; nor injuries, fatal or otherwise; resulting from poison or anything accidentally or otherwise taken, administered, absorbed, or inhaled; nor' injuries, fatal or otherwise, received while or in con*664sequence of having been under the influence of, or affected by, or resulting, directly or indirectly, from intoxicants, anesthetics, narcotics, sunstroke, freezing, vertigo, sleepwalking, fits, hernia, or any disease or bodily infirmity.”
Campbell was shot and instantly killed in November, 1897, by a policeman named Duncan. Under the clause quoted the defense is made: (1) That the shot was not accidental, but intentional, and was flred by Duncan in his necessary self-defense, to save his own life, and to protect himself from death or'great bodily injury, and from an assault then being made upon him by Campbell. (2) That Campbell voluntarily exposed himself to unnecessary dan.ger, in that, late at night,' in the ‘streets of Shelbyville, when drunk, he did wrongfully, maliciously, willfully, and unnecessarily assault and attempt to kill Duncan, then a policeman on the Shelbyville police force, engaged- in the discharge of his duties, armed as and clothed with the authority of a policeman, all of which Campbell then well knew; and that Duncan then, in his necessary self-defense, shot and killed Campbell. (3) That when- Campbell was-killed he was under the influence of intoxicants. (4) That at the time and place at which Campbell received the injuries of which he died he received such injuries in consequence of, and they were the result, directly and indirectly, of his being under the influence of and affected by intoxicants. The affirmative averments o.f the answer were controverted of record, and a trial had, which resulted in a verdict for the company.
It appears that Campbell had been a member of the Shel-byville police force, serving as such for about two years with the man by whom he was killed, and who during that time was a special policeman. A. short time before his death he had been removed, and Duncan retained in his *665place. He had a grievance against the mayor and some of the board of town trustees, who were instrumental in his removal, but so far as the record shows, was on perfectly friendly terms with Duncan, and early in the evening smoked a cigar with him in an amicable manner. The evidence tends to show that Campbell had a drink of whisky before supper, and after supper, a litle after eight o’clock, left his wife at a neighbor’s, and went down town. The evidence also tends to show that between that time and his death he took a . drink of whisky at one saloon, a glass of beer at another, and four drinks of whisky at another, three of these being taken in rapid succession. Whether he took anything else does not appear. Just before the shooting — -which took place between half past ten and a quarter of eleven — he drank a glass of ginger ale. As to the extent, if any, to which he was under the influence of liquor, there is the usual conflict. There is evidence tending to show that he was somewhat noisy, one witness stating that he was more disorderly than drunk; tha-thewasonceor twice urged to go home; and that just before the killing, after leaving a saloon, he yelled on the street one or more times. For this he was reproved by Duncan. When Duncan for the second time urged him to stop his noise, and go home, referring to the fact that he had been a policeman, Campbell seems to have said it was a “damn lie,” that he was not hallooing, and to. have knocked Duncan’s billy from under his arm. As-Duncan stooped to recover it, Campbell rolled it out of reach with his foot. This was done three times. As Duncan stooped the third time, Campbell drew his pistol with his left hand — the pistol-finger of his right hand having been shot off some time previously — and presented it at Duncan’s head. The pistol according to some witnesses, *666“chawed” or made a clicking noise. When Duncan rose up, and saw the pistol at his head, he struck it up, and, at the same time, drew his own revolver and fired. When the doctor reached the spot a few minutes later, he found Campbell quite dead, with the pistol still in his hand, and his finger through the guard' and on the trigger.
We shall first consider the objections to the admission and refusal of testimony. The witness Skelton, a saloon keeper into whose place Campbell came on the night of the killing, was asked whether he was drunk or sober, replied: “I took him for sober. I didn’t take time to notice him very particular.” The witness was then asked: “Did you discover that he walked and talked like a sober man, or did he stagger, or did he stammer, or have any indications of a drunken man?” Before the objection to this question was passed upon by the court, the witness answered. “Did not. I didn’t have much time to pay any attention to him. I think if he had been very drunk I would have discovered it.” This answer, against objection, was excluded from the jury. That the question was proper — if it be not considered leading — we have very litle doubt.
In Brown v. Com., 77 Ky., 405, in an opinion which is an object lesson in clear statement, Judge Hines stated the doctrine on the admissibility of opinion evidence of nonprofessional persons as to various mental, moral, and physical conditions' — including intoxication — and applied: it to evidence upon the question of sanity. It was there held that by the expression used in some cases, “the witness must detail the facts upon which the opinion is based.” It is not intended that the admissibility of the evidence shall be made to depend upon the ability of the witness to state specifically facts from which the jury may, independent of the opinion of the witness, draw a conclusion of *667sanity or insanity, for it is the opinion of the witness that is the subject of the inquiry.” It has never been held that such specific facts might not themselves be stated to the jury.
On the contrary, Judge Hines, in the opinion quoted, continues: “The ability of the witness to detail certain facts which are in themselves substantive evidence of the condition of the mind-' may add very greatly to the weight of the opinion- given in evidence. . . .” But, while the question was proper, the exclusion of the1 answer was not, under the facts in this case, prejudicial. The witness 'had already given his opinion as to Campbell’s sobriety. The answer which was excluded itself shows he bad no sufficient knowledge or opportunity for observation to enable him to give any details upon which the opinion might be based, and he was subsequently permitted to go over practically the same ground covered by the question. The witness, Morgan, being asked whether Campbell was or not under the influence of liquor that night, said1: “I don’t know what you call under the influence. Q. Do you mean to tell the jury that, you don’t know w-hat is meant by under the influence of liquor? A. When you take one you are under it some, and when you take two, more.” After some other inquiries, the question was asked: “Now, tell this- jury whether that man was under the influence of liquor that night. If you don’t know, give us your judgment.” Against objection, the witness was allowed to answer: “He was under the influence some, but I would not call him drunk, judge.” Under the opinion in ithe Brown Case, supra,, it was proper to permit the witness to give his opinion, as he had shown that he had ample opportunity for observation of facts upon which to base his opinion.
*668The instruction's given are complained of. They are as follows-: “No. 4. If the jury believe from the evidence that Willis B. Campbell, not in his necessary or apparently necessary self-defense, attacked O. D. Duncan with a pistol, a deadly weapon, with the intention of killing him, the said Duncan, or doing Mm some bodily harm, and the said Duncan in his (the said Duncan’s) necessary or apparently necessary self-defense killed said Campbell, they should find for the defendant.” “No. 5. The jury should not exonerate the company on the ground that the decedent, while in consequence of having been under the influence of or affected by the use of intoxicants-, nor because such shooting resulted, directly or indirectly, from the use of intoxicants, unless they believe from the evidence that Campbell was so m-uch under the influence of intoxicants as to amount to drunkenness or intoxication.” “No. B. If the jury believe from the evidence that the dbath of Willis B. Campbell was not the result, directly or indirectly, of intoxication when killed (if he was intoxicated), or the result of his voluntarily exposing himself to unnecessary danger (if he did so expose himself), they should find for plaintiff the sum of one thousand dollars, with interest at 6 per cent, per an-num from December 28, 1898.”
Before considering the objections' to the- individual instructions, it seems' that we should consider the general principles of law applicable to the'facts disclosed by this record. The policy insures “against bodily injuries sustained through external, violent and accidental means'.” The bodily injuries insured against include those resulting in death, as is clearly shown by the remainder of the clause of the policy. That the injury which resulted in Campbell’s death was- sustained through external and violent means 'is certain. Was it, though the result of an *669intentional act on the part of Duncan, accidental, within the meaning of the policy? In an opinion by Judge Bennett, in Hutchcraft’s Ex’r v. Insurance Co., 87 Ky., 302, (8 S. W., 571), in which is given a classification of accidents, this question is answered in the affirmative. In discussing the class of cases in which one person intentionally injures another, this court said: “When the injury is not the result of the misconduct or the participation of the injured party, but is unforeseen, it is, as to him, accidental, although inflicted intentionally by the other party. ... In other words, we do not regard it as essential, in order to make out a case of injury by ‘accidental means,’ so far as the injured party is concerned, that the party injuring him should not have meant to do so; for, if the injured party had no agency in bringing the injury on himself, and to him it was unforeseen— a casualty — it seems clear that the fact that the deed was willfully directed against him would not militate against the proposition that, as to him, the injury was brought on by ‘accidental means'.’ ” Therefore, unless — to use the language of the opinion — “the injury is not the result of the misconduct or the participation of the injured party, but is unforeseen” (and such a state of fact seems to be covered by the other provisions under consideration), the killing of Campbell was accidental as to him. This view is supported by the argument of the opinion by Judge Pryor in Accident Co. v. Reigart, 94 Ky., 549, (23 S. W., 192), (21 L. R. A., 652), where the death of the insured had been caused by a piece of beefsteak accidentally passing into his windpipe, andl the defense was made that the means through which the injury occurred was not external or violent. The court held that the means was external, though the injury itself was not, and that the *670word “violent” in tlie policy meant “unnatural;” quoting with approval the rule laid down in May, Ins. (3d Ed.), section 175: “No rule, in the interpretation of a policy, is more fully established or more imperative and controlling than that which declares in all cases- it must be liberally construed in favor of the insured, so as not to defeat, without a plain necessity, his claim to indemnity which, in making the insurance, it wasi his object to secure. When the words, are, without violence, susceptible of two interpretations, that which will sustain and! cover the loss must, in preference-, be adopted.” So, in Accident Co. v. Carson, 99 Ky., 445, (36 S. W., 170), (34 L. R. A., 302), where the insured was shot and intentionally killed by another, it was said, in an opinion by Judge Hazelrigg: “While our preconceived notions of the term ‘accident’ would hardly lead us to speak of the intentional killing of a person as an ‘accidental’ killing, yet no doubt can now remain, in view of the precedents established by all the courts, that the word ‘intentional’ refers alone to the person inflicting the injury; and if, as to the person injured', the injury was unforeseen, unexpected, not brought ábout through his agency designedly, or was without his foresight, or was- a casualty or mishap not intended to befall him, then the occurrence was'-accidental, and the injury, one inflicted by accidental means', within the meaning of such policies.” The opinion cites Casualty Co. v. Johnson (Miss.), 17 South., 2, (30 L. R. A., 206), in which death by hanging at the hands of a mob was- held to be accidental within the meaning of a policy -similar to the one in the case at bar.
The provisions of the policy that it does not cover voluntary exposure to unnecessary danger is presumably the one under which the fourth instruction was given. This *671language has been construed! by tliis court in the very recent case of Insurance Co. v. Clark, (59 S. W., 7). In that case it was said, in the opinion by Judge Burnam: “To enable appellant to escape liability under the provision of the policy relied on, it is essential that it should show that the assured knew of and realized the danger to him in sleeping in such close proximity to the safety valve, and that with such knowledge he purposely and consciously exposed himself to such, risk;” citing Miller v. Insurance Co., (Tenn. Sup.), (21 S. W., 39), (20 L. R. A., 765); Insurance Co. v. Randolph, 24 C. C. A., 305; (78 Fed., 754); Burkhard v. Insurance Co., 102 Pa. St., 263. In the case of Insurance Co. v. Randolph, 24 C. C. A., 312, (78 Fed. 761), Mr. Justice Harlan reviewed a number of cases passing upon the question of the construction to be given to language of doubtful import in policies of insurance, and construed the language here under consideration as follows: “What do the words 'voluntary exposure to unnecessary danger’ in the contracts in suit import? In First Nat. Bank v. Hartford Fire Ins. Co., 95 U. S., 673, 679; (24 L. Ed., 563), it was said that, if a policy of fire insurance was so framed as to leave it doubtful whether the parties intended the exact truth of the applicant’s statements to be a condition precedent to a binding contract, the court should lean against a construction that imposes upon the assured the obligations of a warranty. 'Its attorneys, officers or agents,’ the court observed, 'prepart'd the policy for the purpose, we shall assume, both of protecting the company against fraud, and of securing the just rights of the assured under a valid contract of insurance. It is its language which the court is> invited to interpret, and it is both reasonable and just that its own words should be construed most strongly against itself.’ The *672same rule was recognized in Thompson v. Insurance Co., 136 U. S., 287, 297); (10 Sup. Ct., 1019); (34 L. Ed., 408), which was a case of fire insurance; and was upheld in Insurance Co. v. McConkey, 127 U. S., 661, 666; (8 Sup. Ct., 1360); (32 L. Ed., 308), as applicable in a case of life insurance. This court enforced the same rule in Indemnity Co. v. Dorgan, 16 U. S. App., 290, 309; (7 C. C. A., 581); (58 Fed., 945), where this court, speaking by Judge Taft, said that all language in life policies' limiting the liability of the company should be construed favorably for the insured; that all doubts or ambiguities should be resolved against the insurer. The words ‘voluntary exposure to unnecessary danger,’ literally interpreted, would embrace every exposure of the assured not actually required by the circumstances of his situation, or enforced by the superior will of others, as well as every danger attending such exposure that might have been avoided by the exercise of care and diligence upon his part. Rut the same words may be fairly interpreted as referring only to dangers of a real, substantial character, -which the insured recognized, but to which he nevertheless purposely and consciously exposed himself, intending at the time to assume all the risks of the situation. The latter interpretation is most favorable to the assured, does no violence to the words used, is consistent with the object of accident insurance contracts, and is, therefore, the interpretation which the court should adopt. One- of the accepted meanings of the word ‘voluntary’ is-, ‘done by design or intention; purposed; intended.’ Webst. Dígí. Judge Clark, who presided at the trial, instructed the jury that: ‘Mere negligence or inattention is not an exposure to danger within the meaning of the policy — mere thoughtlessness —but it requires a degree of appreciation of danger at *673the time to make it voluntarily assumed', andJ a voluntary exposure. ... If you find that standing on tbe platform, under all circumstances *of this case, taking into account his position on the train, the speed of the train, the track, and everything else that makes up the situar tion where the accident occurred, if you find that that was dangerous, and that, being conscious of the danger, he took a position that exposed him to it, and death resulted, your verdict should: be for the defendant, otherwise for the plaintiff, as to that issue.’ The company was not entitled to a more favorable interpretation pf the contract than this instruction indicated.” We think the law applicable to the case at bar upon this question is correctly laid down in the citations given, although the instructions to the jury should, of course, not be given in such argumentative form. Ini this view of the law, instruction No. 4 is obviously erroneous. It submits- to the jury the question of whether Duncan was justified in shooting Campbell, and that only. Duncan’s justification and appellant’s contract rights- may obviously be entirely independent. The instruction, moreover, does not present to the jury what we consider an essential element of the defense presented under this language uf the policy, viz.: the consciousness of the danger, and the voluntary exposure to it. That the assault upon Duncan was actually dangerous may properly be assumed from, the facts in this record. In this connection' it may be remarked that testimony as to the general character of Campbell for peace and quiet is entirely irrelevant and incompetent upon any issue presented by the case.
We shall next consider the provision of the policy with reference to intoxicants, and in doing this it is better to *674paraphrase the provision, so that the separate exceptions from the scope of the policy may be separately considered. There are excepted from the insurance contracted for: (1) “Injuries, fatal or otherwise, received while . . . under the influence of or affected by . . . intoxicants. . . .” (21 “Injuries, fatal or otherwise, received . . . in consequence of having been under the influence of or affected by . . . intoxicants. . . .” (8) “Injuries, fatal or otherwise, received ... . resulting directly or indirectly from intoxicants. . . .” The last cf these three exceptions included in the clause of the policy with reference to intoxicants may be eliminated entirely, as it does not apply to the facts in this case. We think it refers solely to the effect of intoxicants upon the system, and not at all to the acts of the insured done by reason of his being under the influence of, or his mind affected by, intoxicants, which is covered by the second one of the exceptions stated. The language, “while under the influence of or affected by intoxicants,” in the first of the separated exceptions, is clearly susceptible of two interpretations. Strictly and literally, it would include being under the influence' of liquor to any extent; and, as aptly said by one of the witnesses, “when you take one, you are under it some, and when you take two, more.” But in ordinary acceptation it does not mean that. It means being so far under the influence as to constitute a state of intoxication, and it ha» thus uniformly, as we think, been construed by the courts. In Joyce, Ins., section 2612, it is said: “Where the policy provides that the company shall not be liable for injuries which are received while the assured is intoxicated, or in consequence of his being binder the influence of intoxicating liquors,-' the phrase ‘under the influence of intoxicating liquors’ • *675is held to mean a condition amounting to intoxication. So an instruction to the jury that, if the insured was under the influence of liquor, whether intoxicated or not. the company is not liable, was held to be erroneous. Where a life policy was conditioned to be void in ease' death should occur ‘while the insured was, or in consequence of his having been, under the influence of intoxi-' eating drink/ and the insured, while intoxicated, was shot and killed, it was held that, if the insured was under the influence of intoxicating drinks, when he died, the .policy was avoided, and that it was immaterial whether or not drunkenness was the cause, proximate or remote, of the death.” In Shader v. Assurance Co., 23 Am. Rep., 65, (66 N. Y., 441), referred to in support of the doctrine laid down in the quotation, supra, it was said, in the opinion' by Judge Miller: “An .exact and accurate interpretation of the language employed manifestly conveys the idea that it was intended to comprehend all cases where injury' or death might happen while the assured was under the influence of or affected by intoxicants? This clearly does occur by reason of the use thereof. As to the first class of cases stated in the proviso, the words imply that it is not required that the use of intoxicating drinks should be the moving cause in producing the injury or death, and quite sufficient to avoid a liability that the person in whose favor the policy was issued was under the influence of such stimulants, without regard to the effect which might result from such a condition. The limitation in the policy related to the condition of the insured, not to the cause which might produce his death. And here lies the distinction which is to be drawn in its eoftstruction, for, by any other or different interpretation, the words used would not only be unnecessary, but *676meaningless, and without point. As the policy was rendered void if the assured was injured or killed while under the influence of intoxicating drinks, it was not essential, to work a forfeiture, that injury or death should occur in consequence of the use of the same*. . . . Accidental policies are issued principally to travelers or persons exposed to unusual peril and danger, and, the risk in such cases being extremely hazardous, it is 'by no means unreasonable that the insurer should require that the assured should be under no exciting influence which may affect his self-possession or judgment, or seriously interfere with the free, full and deliberate exercise of his faculties in protecting himself from accident or harm. It follows that the proposition laid down by the judge was erroneous, and he also erred in refusing to charge as requested.” This doctrine, which we think the correct one, is, in effect, that under a provision like the one now under consideration the policy holder is not insured while he is intoxicated. See, also, May, Ins., section 531.
The second exception presents a more difficult question, and one upon which there seems to be a dearth of direct authority. What injuries may properly be said to have been received in consequence of having been under the influence of or affected by intoxicants? This clearly does not include^ injuries such as disease resulting from the direct or indirect effect upon the system of the intoxicants, and such results are provided for by other language' used in the clause. We think this language applies to injuries received in consequence of the effect of intoxicants upon the nerves, the mind, or the disposition of the insured. For example, it applies where a man’s nerves are so unsteady from the use of intoxicants that he loses his balance upon the edge of a precipice, and falls; where his mind *677is so affected by tbeir use that he does not see a present danger, but deliberately walks into it; or where his passions are so excited, and his temper so warped, that he recklessly does acts in themselves dangerous, or which naturally tend to produce dangerous results from others. Other examples might be given, but these are sufficient to indicate the meaning of the court. If this view be the correct one, it follows necessarily that it is not necessary, in order to come within this exception of the policy. that the insured should be so far under the influence of intoxicants a>s to be in what is ordinarily termed a state of intoxication or drunkenness, but only so far under the influence of or affected by them that the incapacity, nervous or mental, or the excitement of passion, should be such that the injury results in consequence of it. Therefore the definition of “under the influence of intoxicants” which applies to the first of the exceptions does not apply to the second, because, fairly considered, such a definition could not have been contemplated by the parties to the contract. The language used is not fairly susceptible of two interpretations, as it is when used in connection with the condition of the insured at the time of receiving the injury; for by-no process of fair construction can the conclusion be reached that in excepting from the scope of the policy, injuries received in consequence of being under the influence of, or affected by, intoxicants, it was intended not to except any injuries received in con-' sequence of such a condition, when the condition was sufficient to cause the injuries. We can not say that this language, in this connection, means what is ordinarily considered being under the influence of intoxicants, for the words “in consequence of” show that the exception was not intended to be limited to a particular degree of in-*678fiuence, but was intended to cover all influence of which the injury might be the consequence. For this reason the fifth instruction was erroneous.
We are of opinion that the jury should have been instructed, in substance, as follows: (1) The shooting and killing of Campbell by Duncan was a bodily injury sustained through external, violent, and accidental means, and the jury should find' for plaintiff unless they believe from the evidence that defendant is excused from liability on the policy on the grounds stated in the second, third and fourth instructions, or any of them. (2) If the jury believe from the evidence that Campbell assaulted Duncan, and at the time of such assault realized that such assault was dangerous to him (Campbell), and with such knowledge and consciousness of such danger voluntarily made the assault which exposed him to the danger, they should find for the defendant. Í3) If the jury believe from the evidence that at the time Campbell received the injury which resulted in his death he was so far under the influence of intoxicants as to be in a state of intoxication, they should find for the defendant. (4) If the jury believe from the evidence that Campbell received the injury, which resulted in his death in consequence of having been under the influence of or affected by intoxicants, although not to such an extent as to amount to drunkenness, they should find for the defendant. For the reasons given, the judgment is reversed, and cause remanded, with directions to award appellant a new trial, and for further proceedings consistent herewith.
Guff-y, J., concurs in the reversal, but dissents from the reasoning.