(concurring in part, dissenting in part)—I concur in the majority's resolution of the Christen case; the record lacks sufficient evidence to raise an issue of fact regarding obvious intoxication. In the Long case, however, I find the majority's approach to tort law untenable. It unnecessarily narrows liability for serving an obviously *511intoxicated person which is not required by the facts or the law.
Neither the majority nor McDougall's claims there is insufficient evidence that the latter breached its duty to not overserve Mr. Coates. Majority, at 491. The only issue is whether the "foreseeability" principle limits the scope of McDougall's liability. As the majority correctly points out, foreseeability is an issue for the jury unless reasonable minds cannot differ. Majority, at 492; see Bernethy v. Walt Failor's, Inc., 97 Wn.2d 929, 933, 653 P.2d 280 (1982); Rikstad v. Holmberg, 76 Wn.2d 265, 270, 456 P.2d 355 (1969). Instead of relegating the current issue to the jury, the majority holds, as a matter of law, that it is not foreseeable that an overserved, obviously intoxicated minor patron might use a switchblade knife he had previously shown to the bartender.81 Following the Rikstad case, decided by this court, I would send that issue to the jury.
In Rikstad, an intoxicated driver had driven his pickup truck over an intoxicated man who had passed out in the grass beside a campsite. The trial court had held, as a matter of law, that a reasonable man could not have foreseen that a person would be lying in the grass. Rikstad, at 265. In reversing, this court explained that it is not the unusual sequence of events in any given case but the general field of danger that affects liability. The court found that "reasonable minds can well differ" on whether the injury was within the foreseeable scope of risks arising from the driver's duty to operate his truck in a reasonable manner. Quoting precedent, this court emphasized that "it is for the jury to decide whether the general field of danger should *512have been anticipated . . . [by defendant](Italics mine.) Rikstad, at 270 (quoting McLeod v. Grant Cy. Sch. Dist. 128, 42 Wn.2d 316, 324, 255 P.2d 360 (1953)).
The majority's attempt to limit the jury's function is unpersuasive. The main thrust of the argument focuses on the court's recent trend to foreclose the development of tort law in this area. Because the court has never ruled on this issue before, the majority argues, it should decline to find potential liability. See majority, at 494-95.
Other courts have properly recognized two principles. First, repeal of a strict liability dramshop act leaves fundamental principles of negligence unimpaired. Rappaport v. Nichols, 31 N.J. 188, 156 A.2d 1, 8 (1959). To hold otherwise provides an unwarranted measure of immunity for tavern owners and other furnishers of alcohol. See McClellan v. Tottenhoff, 666 P.2d 408, 412 (Wyo. 1983); see also Ontiveros v. Borak, 136 Ariz. 500, 667 P.2d 200, 205 (1983). Second, it is the court's duty to develop the common law consistent with the needs of a changing society:
The genius of the common law is that it is constantly expanding to meet new and unique conditions. The spirit of the common law is not dead. . . .
Colligan v. Cousar, 38 Ill. App. 2d 392, 414, 187 N.E.2d 292, 302 (1963). "When the ghosts of the past stand in the path of justice, clanking their medieval chains, the proper course for the judge is to pass through them undeterred." United Austl., Ltd. v. Barclays Bank, Ltd., [1940] 4 All E.R. 20, 37 (H.L.), quoted in El Chico Corp. v. Poole, 732 S.W.2d 306, 315 (Tex. 1987).
In support of narrowing the duty to not overserve an obviously intoxicated person, the majority cites Burkhart v. Harrod, 110 Wn.2d 381, 755 P.2d 759 (1988), Wilson v. Steinbach, 98 Wn.2d 434, 656 P.2d 1030 (1982), and Shelby v. Keck, 85 Wn.2d 911, 541 P.2d 365 (1975).
Both Burkhart and Wilson involved the question of social host liability and are not, therefore, controlling. In Burkhart, in addition, all opinions agreed the facts did not *513indicate that Burkhart was obviously intoxicated. Therefore, the defendants did not breach a duty not to serve obviously intoxicated persons. The majority's remarks concerning social host liability were therefore mere dicta. Dicta cannot serve as precedent for the resolution of the instant case. Moreover, the majority's position received scathing criticism:
[T]he Burkhart court not only failed to articulate a convincing rationale for failing to recognize such a duty but also declined to fulfill its proper role in defining the norms of responsible social behavior.
Recent Cases, Negligence—Social Host Liability—Social Hosts Not Liable for Accidents Caused by Intoxicated Guests.—Burkhart v. Harrod, 110 Wash. 2d 381, 755 P.2d 759 (1988), 102 Harv. L. Rev. 549 (1988) (hereafter Recent Cases).
As stated in my concurring opinion in Burkhart, if the facts showed obvious intoxication, I would find, applying ordinary principles of negligence, that a social host has a duty to take reasonable measures to prevent an intoxicated guest from driving on public roads when the host has furnished an obviously intoxicated guest with alcohol. Burk-hart, at 394 (Utter, J., concurring).
Many of the principles discussed in my Burkhart concurrence apply equally here. Courts have a responsibility to fine tune the common law to be more just. The "deferential posture adopted by the Burkhart court. . . amounts to an abdication of judicial responsibility." Recent Cases, at 555. As a public policy issue, courts consider imposing liability to provide incentive to avoid the loss altogether. W. Keeton, D. Dobbs, R. Keeton & D. Owen, Prosser and Keeton on Torts § 4, at 25 (5th ed. 1984) (hereafter Prosser and Keeton). The drinking establishment, not the assault victim, is in a better position to avoid the loss by refusing service.
The next inquiry, then, involves applying negligence principles, beginning with duty limited by foreseeability. In Burkhart, I stated that "beyond question" a traffic accident *514is a foreseeable result of "mixing gasoline and liquor". Burkhart, at 395 (Utter, J., concurring). In contrast, here I do not find that a criminal assault is "beyond question" a foreseeable result of mixing alcohol and weapons. Rather, I find it an issue about which reasonable minds can differ. Therefore, it is a question for the jury. Shelby v. Keck, supra, relied on by the majority, does not support a contrary position. Shelby merely requires that principles of negligence, not strict liability, apply.
In Shelby, the plaintiff sought to recover for the wrongful death of her husband who was accidentally shot by a tavern patron. She alleged the tavern owner was negligent in two respects: first, for serving liquor to a man when the tavern owner knew or should have known that the man was carrying a loaded weapon; and second, for permitting Keck to become intoxicated. Shelby, at 915. In resolving the first issue, this court affirmed a directed verdict in favor of the defendant. The plaintiff failed to present sufficient evidence that "would alert a reasonable man in the defendant's position that Keck was likely to be armed and thereby posed a threat to the safety of his other patrons ..." Shelby, at 915. None of the tavern employees knew the patron was armed. By inference, the carrying of a weapon might be sufficient to provide notice of an enhanced risk to others. The court found no liability on the second charge because there was no evidence that Keck was obviously intoxicated.
I do not dispute the majority's contention that Shelby requires notice of risk. The majority, however, holds that prior aggressive acts of an armed patron are the exclusive means of providing notice of increased risk to others. Majority, at 491. By so holding, the majority unnecessarily narrows the concept of notice. The Supreme Court of Florida, resolving a premises duty case such as those relied on by the majority, specifically rejected the majority's position: "specific knowledge of an individual's dangerous propensities is not the exclusive method of proving foreseeability." (Italics mine.) Allen v. Babrab, Inc., 438 So. 2d 356, 357 *515(Fla. 1983) (foreseeability of assault may be shown by proving that the proprietor, based on the establishment's history, "should have recognized the likelihood of disorderly conduct by third persons in general"). In a more recent case, that court elaborated by stating that the indicia listed for providing notice of general or specific harm were not exclusive. Hall v. Billy Jack's, Inc., 458 So. 2d 760, 762 (Fla. 1984); see also Kerby v. Flamingo Club, Inc., 35 Colo. App. 127, 532 P.2d 975, 978 (1975).82
Moreover, in the premises duty cases relied on by the majority, at 497-98, this court did not hold as a matter of law that a tavern owner is not liable absent notice provided by acts of the patron. In Moore v. Mayfair Tavern, Inc., 75 Wn.2d 401, 407, 451 P.2d 669 (1969), this court addressed whether the jury was justified in finding that the assault was not foreseeable. Moore, at 405-06. Similarly, in Potter v. Madison Tavern, 74 Wn.2d 704, 446 P.2d 320 (1968), we considered whether a case fully tried to a court without a jury should be overturned on appeal. Potter, at 704. In neither case did this court say, as a matter of law, that criminal assaults are not foreseeable absent notice provided by previous conduct. In both cases the court merely considered whether there was sufficient evidence to uphold either the jury's or the trial court's findings. As in Moore, the present question should be an issue for the jury.
Furthermore, cases involving a premises duty do not and should not resolve the question of foreseeability for the duty not to overserve. In those cases, not only is there no requirement that the assailant be obviously intoxicated, there is no requirement that the drinking establishment ever served him at all. Where the drinking establishment has not created the risk, as it would in breaching a duty not to overserve, this court has required and should require *516more in the way of notice of future problems. See Moore, at 407 (patron consumed two schooners and part of a third, slapped another patron, was ejected from the tavern, returned with a gun and shot the other patron); Jones v. Leon, 3 Wn. App. 916, 478 P.2d 778 (1970) (assailant entered the bar, ordered no drinks, and shot his girl friend's dancing partner), review denied, 78 Wn.2d 997 (1971); Potter, at 705 (trial court found as an unchallenged fact that the patron was not intoxicated); see also Getson v. Edifice Lounge, Inc., 117 Ill. App. 3d 707, 453 N.E.2d 131 (no discussion of intoxication, in fact evidence conflicts whether the lounge served the assailant at all), review denied, 96 Ill. 2d 559 (1983). It is the combination of factors that leads to foreseeability, not any one factor in isolation. Because Welch v. Railroad Crossing, Inc., 488 N.E.2d 383 (Ind. Ct. App. 1986), ignores this consideration, I find that analysis unpersuasive.
Examining underlying principles of supplier liability clarifies both the question and its resolution. In Bernethy v. Walt Failor's, Inc., 97 Wn.2d 929, 653 P.2d 280 (1982), this court ruled that owners of a gun shop could be found negligent, under the theory of negligent entrustment, for selling a gun to an intoxicated man who immediately returned to a tavern to kill a person. The majority rejects using Bernethy by analogy, finding no implication of the doctrine of negligent entrustment. Majority, at 499. Prosser and Keeton, however, have grouped such cases together in discussing the liability of suppliers:
[A] seller who delivers a dangerous thing to one whom he should know to be unfit to handle it has been held liable for the harm which results to others—as where, for example, a gun is sold to a child, or an automobile to a known incompetent driver. There are by now a good many cases of so-called "common law dramshop" liability, in which a seller has been held liable to third parties for the sale of intoxicating liquor to a minor, or to an intoxicated person. It is the negligent entrusting which creates the unreasonable risk; and this is none the less when the goods are conveyed. . . .
*517(Footnotes omitted.) Prosser and Keeton § 104, at 718. Other jurisdictions have also recognized the similarity of principle. See, e.g., McClellan v. Tottenhoff, 666 P.2d 408, 415 (Wyo. 1983); El Chico Corp. v. Poole, 732 S.W.2d 306, 312 (Tex. 1987); Ontiveros v. Borak, 136 Ariz. 500, 667 P.2d 200, 207 (1983); Alegria v. Payonk, 101 Idaho 617, 619 P.2d 135, 138 (1980) ("The 'negligent entrustment' tort ... is a recognition of the risk of injury which exists when two ingredients are combined . . ."); Rappaport, 153 A.2d at 7.
Although there is a difference between providing a weapon to an obviously intoxicated person and providing additional liquor to an obviously intoxicated person who already has a weapon, similar principles undergird a policy of finding potential liability for both. While hard documentation of the relationship of alcohol and violence is incomplete, statistics and logic suggest a causal connection.
According to statistics available in 1981, alcohol was involved in about 10,000 murders annually. Panel on Alternative Policies Affecting the Prevention of Alcohol Abuse and Alcoholism, National Research Council, Alcohol and Public Policy: Beyond the Shadow of Prohibition 82 (1981). The panel noted the absence of pharmacological studies substantiating a physiological connection between alcohol and violence but, nonetheless, concluded a causal relationship exists:
Nevertheless, even without a pharmacological connection, there is a strong causal connection between alcohol and violence: the culture tells people this is so, and in acting on this belief they make it come true. Research designed to sort out the specific quantitative contribution of cultural belief to alcohol-involved violence is in its infancy, but this connection requires consideration even in the absence of good epidemiological estimates.
Alcohol and Public Policy, at 83.
A study guide prepared for the Department of Justice discussed the relationship between drinking and crime:
Alcohol abuse is one of the Nation's gravest health and social problems. It is also one of its most serious crime problems. . . .
*518There is a close relationship between alcohol and crime, and there is good reason to believe that consumption of alcohol causes crime. The most obvious crimes in which alcohol plays a role are drunk driving and public drunkenness, but alcohol is also a factor in large percentages of violent crime and intrafamily crime.
. . . Alcohol also plays a major role in violent crime. In Marvin Wolfgang's classic study of homicide in Philadelphia, he found that the killer, the victim, or both were drunk in more than half the cases. That finding seems also to apply to rapes. Needless to say, many aggravated assaults involve alcohol; indeed, a large number occur in taverns or among drinking companions. Alcohol also plays a prominent role in crimes within the family. A high percentage of wife beatings, child abuse, and child sexual abuse are committed by men who are drunk. Recent research sponsored by the National Institute of Justice found that a majority of new prison inmates had severe alcohol problems.
That so large a percentage of perpetrators of serious crimes drink heavily beforehand does not prove that alcohol causes crime. People may first decide to commit their crimes and then get drunk to muster courage or allay fears. For some violent offenders heavy drinking and [emphasis in original] savage behavior are both symptoms of deep psychological problems. Still the relationship between drinking and violence is not subtle or difficult to understand. It is common knowledge and scientifically demonstrable that alcohol releases inhibitions and distorts judgment. The "fighting drunk" is a regular member of tavern culture . . .
(Italics mine.) U.S. Dep't of Justice, Drinking and Crime (1983).
Even a cursory review of the literature reveals numerous publications that either provide statistics demonstrating an alcohol-crime relationship or present attempts to explain the relationship. See, e.g., R. O'Brien & M. Chafetz, Encyclopedia of Alcoholism (1982) (evidence of alcohol use has been reported in 24 percent to 72 percent of assailants in assault cases); Weston, Alcohol's Impact on Man's Activities: Its Role in Unnatural Death, 74 Am. J. Clinical Pathology 755, 757 (1980) (Table 5 indicates that 72 percent of all stabbings in the United States are alcohol related); Temple & Ladouceur, The Alcohol-Crime Relationship as an Age-Specific Phenomenon: A Longitudinal *519Study, 13 Contemp. Drug Probs. 89 (Spring 1986) (suggesting an indirect relationship between crime and alcohol during adolescence that diminishes with age); Collins, Alcohol Use and Criminal Behavior: An Empirical, Theoretical, and Methodological Overview, in Drinking and Crime 288, 289 (1981) (finding a significantly strong alcohol-assaultive crime empirical association to justify the inference that sometimes alcohol and serious crime are causally related).
We must not lose sight of the central issue—the issue is foreseeability. There is no need for a detailed socio-psychological-pharmacological understanding of the alcohol-crime relationship. All that is necessary is to acknowledge that there is some relationship, a relationship recognized by laymen. Historically, laymen have recognized a relationship. The pre-prohibition era, an era not focused on alcohol-vehicle accidents, provides voluminous material. In denouncing intemperance as "the great source of crime," one criminal lawyer of the 1880's quoted numerous literati:
Man, with raging drink inflamed, Is far more savage and untamed; Supplies his loss of wit and sense With barbarousness and insolence.
A. Richmond, Leaves from the Diary of an Old Lawyer: Intemperance, the Great Source of Crime 28 (1880) (quoting Hudibras).
Courts also have recognized a connection between alcohol and aggression. "Alcohol plays a major role in the incidence of death and injuries on our highways, the battered women's syndrome, and assaults and batteries in taverns, on the streets and in the home." Clendening v. Shipton, 149 Cal. App. 3d 191, 198, 196 Cal. Rptr. 654, 658 (1983) (finding foreseeability of assault a jury question in a case filed before statutory revision); see also Carey v. New Yorker of Worcester, Inc., 355 Mass. 450, 245 N.E.2d 420, 422 (1969); Terry v. Markoff, 26 Ohio App. 3d 20, 497 N.E.2d 1133, 1135 (1986) (finding a commercial provider of *520alcoholic beverages can be liable for assault, under a negligence theory, when it overserves an obviously intoxicated patron).
Nonetheless, the majority affirms the Court of Appeals decision, holding that reasonable minds cannot find that a stabbing is a foreseeable result of overserving an obviously intoxicated minor possessing a switchblade. Prosser's guidance bears repeating: "As the gravity of the possible harm increases, the apparent likelihood of its occurrence need be correspondingly less to generate a duty of precaution." Prosser and Keeton § 31, at 171. In Long, I would reverse and remand for trial.
In the lower courts, the parties disputed whether there was sufficient evidence that Coates actually showed the knife to the bartender. In a summary judgment action, this court, like the trial court, independently reviews all evidence presented. Yakima Fruit & Cold Storage Co. v. Central Heating & Plumbing Co., 81 Wn.2d 528, 530, 503 P.2d 108 (1972). Nonetheless, the majority and both lower courts considered the issue immaterial to the legal analysis. It becomes material only if it makes a difference whether the bartender saw the knife. For the purpose of refuting the majority's analysis, I discuss only the legal issue.
Colorado's Dramshop Act applied only to serving habitual drunkards when the server had received notice that the patron was a habitual drunkard. See Note, Crespin v. Largo Corporation and the Legislative Response: The Turbulent State of Dram Shop Liability in Colorado, 57 U. Colo. L. Rev. 419, 420 n.10 (1986). The complaint in Kerby alleged three theories of recovery based on negligence.