At the conclusion of plaintiff’s evidence the trial court directed the jury to find for defendants, appellees here. This appeal followed.
In January 1964, appellant, plaintiff in the circuit court, ran into a large excavation on the shoulder of Deepwood Drive in Radcliff, Kentucky. Appellees were at that time the contractors with the city for the construction of a sewer system.
Appellant was traveling north on Deep-wood Drive, a street 14 to 16 feet wide without curbing, when he met another vehicle traveling south. Appellant testified that the unidentified and unknown operator of the southbound vehicle crowded him, the appellant, off the hard surface of the road; that he was familiar with the shoulder of the road at that point and knew it to be flat and without any obstruction; that he confidently drove off the hard surface onto the shoulder about four or five feet from the edge of the hard surface when suddenly his jeep landed in the middle of a large excavation; that appellees negligently failed to place any road signs in that area warning of construction and failed to put up a barricade or sign as notice to the public that there was such a cavity in the earth so close to the traveled part of the highway; and that the dormant, muddy ground in the area of the wreck gave no evidence of having been disturbed to put him on notice that the shoulder of the road might be dangerous.
Appellant claimed he sustained serious personal injuries in the wreck, for which he claimed $69,000 damages.
Appellees take the position on this appeal that (1) they were under no duty to place barricades, signs, or other warnings around the hole; (2) the lack of such warnings was not the proximate cause of the wreck and resulting damages; and (3) plaintiff was guilty of contributory negligence.
Appellees admit that their liability, if any, is the same as the liability of the city of Radcliff, apparently because of a con*882tract they made with the city which provided that it was appellees’ duty to “employ all necessary day and night guardsmen” and to “erect all necessary barricades for the protection of the work and the safety of the public.” This eliminates the need for any discussion of the distinction between the liability of a city and the liability of an individual who contracts with the city. In City of Lancaster v. Broaddus, 186 Ky. 226, 216 S.W. 373 (1919), it was held that cities are under a duty to keep their streets and the areas immediately adjacent thereto in the margin of the traveled part in a reasonably safe condition. This duty includes erecting “such guards or barriers along the margin of the traveled portion of the highway, or even along the external portion of the highway as laid out, so that if a traveler is injured in consequence of a want of such barriers, he may have an action for damages against the municipality, although the injury in fact took place outside the limits of the traveled path.” City of Lancaster, supra, was cited with approval in Central City v. Snodgrass, 234 Ky. 396, 28 S.W.2d 467 (1930).
It is appellees’ contention that the holes could easily be seen and that such an open and apparent condition required no warning. Dillingham v. Department of Highways, Ky., 253 S.W.2d 256 (1952), is cited for the proposition that the state is not liable for its failure to maintain the shoulders of the highway in a reasonably safe condition for travel except as to defects which are obscured from the view of the ordinary traveler and are so inherently dangerous as to constitute traps. The City of Lancaster and the Central City cases were distinguished on the basis of the fact that in those cases the holes next to the road were hidden by weeds and tall grass and were not obvious to the traveler.
Under the Dillingham rule, the question becomes one of fact whether the hole next to the road was hidden from view. The pictures in evidence show that the area around the holes was muddy in the case at bar. The dirt taken from the holes was piled next to them on the side away from the street. The pictures show that the holes were difficult to see because of the way they blended in with the surrounding muddy area.
The testimony of the witnesses also indicates that the holes were difficult to see. Appellant says he never saw the holes before the accident even though he lived on Deepwood and had driven by them several minutes earlier. Appellant’s wife and son testified to the same effect. The driver of the wrecker which removed the jeep from the hole said that he never noticed the second hole of like size and shape even though he worked there for an hour, and it was only 20 feet away. A neighbor who lived on Deepwood said that she had driven by the holes and never noticed them until they were pointed out to her. One of the appellees, Alden Knapp, said that after he was told of the accident he and an associate drove to Deepwood to look for the hole. His testimony indicates that he was looking for a hole in the pavement, but he said that he did not see any holes anywhere, on or off the pavement. It is difficult to understand how appellant could be expected to see the holes while he was driving and concentrating on an oncoming car which was in the wrong lane when ap-pellee did not see them when he drove along looking for holes, even if he was looking primarily on the pavement.
The only indication that there might have been a hole beside the road was the dirt pile, but even that does not show the exact location of the hole because in construction areas dirt is often dumped at a distance from the excavation.
Also, as a matter of policy, it should be mentioned that appellees could have erected some type of warning device easily and inexpensively.
We conclude the facts in the present case classify it along with City of Lancaster and Central City, supra, and presented a factual jury question. In this respect, *883the basis of the action of the trial court in directing a verdict for appellees was error.
Appellees next contend that conceding arguendo that there was a duty to warn and a failure to do so, same was not the proximate cause of the wreck and that the failure to warn was merely a remote cause or condition for which they cannot be held liable.
Coinage of the legal term “proximate cause” is charged to Lord Chancellor Bacon. It is said in Maxims of Law, Reg. I, Bacon: “In jure non remota causa, sed próxima, spectatur.” (In law the near cause is looked to, not the remote one.) Modern dictionaries define “proximate” as very near, immediately adjoining.
Prosser on Torts, 3rd Ed., Ch. 9, in writing on “proximate cause” says: “There is perhaps no other one issue in the law of torts over which so much controversy has raged, and concerning which there has been so great a deluge of legal writing.” At page 289.
Over the ages, numerous formulae have been suggested for determining “proximate cause,” such as Bacon’s “near cause,” long since found unsuitable; “last human wrongdoer,” advocated by some courts; “the substantial factor test,” proposed by Jeremiah Smith in 1911 in Harvard Law Review, 103, 223, 229; then came Professor, later Judge, Edgerton, to maintain still another formula, to wit: “Justly attachable,” was workable. This formula simply said that the essential question is whether the harm which has been suffered is “justly attachable” to the defendant’s conduct. This formula, like all others, will not work in all situations.
Prosser would abandon the “fruitless quest for a universal formula” and propose an approach to proximate cause as a “series of distinct problems, more or less unrelated, to be determined upon different considerations.” He would list a few of the problems thus: “(1) What part has the defendant’s conduct played in bringing about this result; (2) the problem of apportionment of damages among causes; (3) the problem of liability for unforeseeable consequences; (4) the problem of intervening causes; and (5) the problem of shifting responsibility.”
Some of the problems listed by Prosser have multiple variations.
For the present case, the part defendants’ conduct played in bringing about the result and the element of foreseeability of plaintiff’s conduct may be the dominating consideration. On the question of what part the defendants’ conduct has played in bringing about the result, it is reasonable to believe that had warning been given of the hole, the accident would not have happened. Insofar as the element of foreseeability is concerned, it may be stated that if appellees could foresee or anticipate that vehicles might pull over onto the shoulder of the road “four or five feet” from the vehicles’ right edge of the hard surface of the road, it would be reasonable to impose liability upon them for their failure to place some barricade or warning of the condition.
This court has clearly recognized the relationship of foreseeability to proximate cause in Miller v. Mills, Ky., 257 S.W.2d 520, 522 (1953), in which it was said:
“We think it is clear that so far as foreseeability enters into the question of liability for negligence, it is not required that the particular, precise form of injury be foreseeable — it is sufficient if the probability of injury of some kind to persons within the natural range of effect of the alleged negligent act could be foreseen.”
See also 100 A.L.R.2d 944.
It seems logical that appellees should have foreseen that vehicles would have occasion to pull over on the shoulder of the road for various reasons.
Of course, if appellant would have driven into the hole even though there had been a barricade or warning, appellees would not be liable. The evidence does not *884lead to this conclusion. Appellant saw the oncoming car when it was three hundred feet or more away and then in the middle of the road. He chose to slow down and pull off, which appears reasonable. Appellant, by necessity, kept his eyes on the oncoming car, but it was also possible for him to see a barricade or sign out of the corner of his eyes, had there been such warning, and either stop before getting to it or drive around it.
These are questions for the jury.
Appellees raise one final argument with regard to proximate cause. They contend that the negligence of the driver of the oncoming car was an intervening cause which cuts off their liability. Prosser, in his discussion of intervening causes divides them into two groups: 1. Foreseeable intervening cause — which is “one in which in ordinary human experience is reasonably to be anticipated.” 2. Normal intervening cause — which is one “which could scarcely have been contemplated by any reasonable man in the place of the defendant at the time of his conduct, but which are nevertheless to be regarded as normal incidents of the risks he has created.” The difference between these two boils down to the degree of foreseeability.
It is not necessary to decide which of these categories covers the case at bar because defendant could be found liable under either category. They are pointed out only to show that the intervening cause need not be reasonably anticipated in ordinary human experience but can be only a normal incident of the risk-created by the defendant. Certainly appellees created a risk when they did not put barricades around the holes. In other words, while it could be argued that they could not reasonably anticipate that one car would force another off the road into the hole, yet the accident was a normal incident to leaving an unbarricaded hole next to the road.
The final question raised by the briefs is whether appellant was contributorily negligent because he did not see the hole. The standard of conduct used in questions of contributory negligence is that of the reasonable man under like circumstances. Restatement of Torts, § 464. Appellant found himself on a street 14 to 16 feet wide with another car which was partially over in his lane. Under such circumstances, the reasonable man is more likely to be concerned with the danger presented by the oncoming car than with the condition of the shoulder of the road. He may assume that the shoulder of the road is reasonably safe for travel by vehicles in case of emergency. Moreover, appellant would probably have been negligent if he had not watched the oncoming car.
We conclude that the question of proximate cause was one for the jury and that the trial court erred in directing a verdict for appellees.
The judgment is reversed for a new trial consistent herewith.
STEINFELD, MILLIKEN, and PAL-MORE, JJ., concur.