This was a moment of decision, immeasurable in time, for the master of the Andrea Doria. There were so many miles of ocean around and now so little room. His lifetime knowledge of the sea and ships had to be used for the correct instinctive maneuver, if he were to have any hope of avoiding a collision. Should he go right? Left? Straight ahead? Stop? The decision made, Capt. Calamai called out, “Tutto sinistra. . . all left.” – Alvin Moscow in “Collision Course”
Professional mariners all realize that the in extremis situation is to be avoided. If it cannot be avoided, all aboard the imperiled vessels must hope that those on watch and the captains they have summoned to the bridge have a thorough grasp of the Collision Regulations (ColRegs) and emergency shiphandling fundamentals and the judgment to apply them correctly. Regrettably, many mariners either do not understand the rules and fundamentals or fail to apply them correctly. Some blindly maneuver according to dubious “rules” of emergency shiphandling they memorized years earlier. Analyses of vessel collision data reveal that in many cases the vessels involved — like Andrea Doria and Stockholm — actually maneuvered their way into, rather than away from, a collision, perhaps after performing what Capt. John Trimmer has labeled the “dance of death” in his book How to Avoid Huge Ships.
The in extremis doctrine is commonly thought to include three concepts: (1) a rough definition of the “in extremis situation,” by which is meant a situation where two vessels have approached so closely that collision can no longer be avoided by one ship acting alone; (2) a release from strict compliance with the Collision Regulations when necessary to avoid immediate danger, and (3) the more lenient standard of care applied in the subsequent legal proceedings to an innocent vessel that was confronted with a sudden emergency, which did not allow time for a carefully considered decision before taking avoiding action. It should be noted that the second and third consequences do not follow in all extremis situations. Collision might be avoided while still complying with the rules, obviating a departure (particularly where the rule itself is qualified by language such as “if the circumstances of the case admit”). Second, it might be the case that neither of the vessels were “innocent” or lacked the time necessary for a reasoned decision, thus disqualifying both from the more lenient standard of care.
Over the years, two shiphandling approaches to what the British label the “agony of collision” have been variously championed or vilified. One is the more formulaic rules-based approach, most fully developed by Cmdr. Davis Lott of the U.S. Navy. The second is a flexible approach espoused by Capt. Robert Slack and several other master mariners, who reject the notion that there is a single correct maneuver for any given extremis situation. The increasing capability of and expectations for pre-programmed or automated solutions to a variety of tasks require us to once again question the extent to which fixed rules or algorithms will replace human judgment in choosing, executing, evaluating and adjusting collision prevention maneuvers.
Collision Regulations, rules of thumb and uniform rules
As vessels approach each other so as to create a risk of collision, the Collision Regulations generally require one or both vessels to maneuver. Which vessel maneuvers, and the choice of whether to change course or speed, depends on the vessels’ relative positions and approach angles, and may be influenced by the nature of the vessels’ employment, degree of maneuverability and whether the vessels are in sight of one another. In general, the requirements can be grouped in two categories: the single action maneuvering rules, where the give-way vessel maneuvers while the stand-on vessel holds her course and speed, and dual action maneuvering rules, where both vessels are required to maneuver to avoid collision. It is important to note that in both cases, no ship is any longer forbidden to take avoiding action until the situation reaches extremis. Even in the single action approaches, the stand-on vessel is permitted to maneuver—before the situation has reached extremis—when the requirements of Rule 17(a)(ii) are met, subject to the restriction in Rule 17(c) against power-driven vessels turning to port toward another power-driven vessel approaching on her port side. Indeed, some courts have found the stand-on vessel at fault for failing to take permissive action under Rule 17(a)(ii) where the evidence demonstrates that a reasonably prudent mariner would have done so.
The Mariners’ Museum
Andrea Doria lies on its side before sinking following its collision with another passenger ship, Stockholm, off Nantucket, Mass., on July 25, 1956. Analysis of the accident indicates the ships may have maneuvered their way into rather than away from each other. As a result, 52 people died.
Under the 1972 ColRegs (as well as the U.S. Inland Rules), two rules are commonly associated with the in extremis doctrine: Rule 17(b), which requires a stand-on vessel to take such action as will best avoid collision if it finds itself so close to the other vessel that collision cannot be avoided by the action of the give-way vessel alone, and Rule 2(b), which permits (and may require) a vessel that finds itself in “special circumstances” to depart from the rules if necessary to avoid immediate danger. Courts have found that some in extremis situations constitute a special circumstance under that rule (and some do not, because a departure from the rules is unnecessary). Some commentators also cite Rule 8 on actions to avoid collision, including Rule 8(e), which requires vessels to slacken speed or take all way off, if necessary to avoid collision or allow more time to assess the situation. It should be noted, however, that nothing in Rules 17, 19 or 2(b) specifies what maneuver the respective vessels must take when they find themselves in extremis.
Several writers have over the years suggested general and pragmatic rules of thumb for collision-avoidance maneuvers. The canonical U.S. Navy text for many years was Capt. Ray Crenshaw’s Naval Shiphandling, which includes a list of collision prevention tips (largely reproduced in Command at Sea by Admirals James Stavridis and William Mack), including “turn early and turn plenty.” Few would quarrel with most of that advice (Rule 8 supports the admonition to turn “early” and “enough to be readily apparent,” though “plenty” might be too much in some cases), but it is not particularly helpful in the extremis situation. Crenshaw also advises that when collision is unavoidable, and the goal is no longer to avoid collision but rather to minimize the damage, turning toward the danger and backing emergency is the most effective maneuver (at least for your own ship). What about the situations between these “bookend” tips: when you didn’t or couldn’t turn early and you are not yet at the “collision-inevitable” point? Are there “uniform rules” to avoid collision in those intermediate situations?
Cmdr. Lott, founder of the Navy’s Emergency Shiphandling School at Pearl Harbor, thought so. In his 1947 book, Collision Prevention, Lott boldly declared that “uniform emergency shiphandling procedures have been evolved and they have been so standardized that skippers of any two ships in imminence of collision can know in advance what maneuver the other will most likely take.”
To be sure, Lott deserves praise for his important contributions to collision prevention theory. Before he embarked on his extensive analysis of collision reports, the most often-cited recommendation for the extremis situation was to always turn toward the other vessel’s stern. Lott’s studies exposed the weakness of that approach. In his book, he went on to provide the reader with guidance for determining any given vessel’s minimum safe maneuvering distances (the range at which the situation should be deemed to be in extremis) and set forth recommended extremis maneuvers appropriate to the various approach situations. For most approach situations, Lott’s recommended extremis maneuver was, not surprisingly today, a course change to the right. Indeed, at about the same time, the U.S. Second Circuit Court of Appeals opined that “it is indeed the instinctive response of a master in an emergency to put his rudder hard right.” While it might be an instinctive response, it is not always a prudent choice.
For many years, Lott’s method for determining the distance at which a watchstander could declare the situation in extremis and his maneuvering rules were widely taught in rules of the road and shiphandling courses. A summary of his rules has been added to David Sears’ chapter on collision avoidance in the current edition of William Hayler’s Merchant Marine Officers’ Handbook. New deck officers (the author included), who were taught to respond to common ship emergencies by executing bridge checklists, were often encouraged to memorize Lott’s rules for the various approach situations and execute them if he found himself in extremis.
We must bear in mind that Lott was writing before the 1972 ColRegs Convention and its flexible Rule 17(a)(ii) was adopted, permitting a stand-on vessel to maneuver when it becomes apparent that the give-way vessel is not taking timely avoiding action. In those pre-ColRegs days, many mariners seized on Lott’s book as a rational and legally defensible basis for determining when their ship was in extremis, and thus no longer compelled to maintain course and speed. Anyone who doubts Lott’s contribution to mariners before 1972 should visualize himself as a stand-on vessel in a crossing situation (or meeting a less restricted vessel under the Rule 18 priorities) with an inattentive or stubborn give-way vessel and without the flexibility of Rule 17(a)(ii).
What should today’s mariners make of Lott’s “rules” for extremis maneuvering? If followed, are they likely to prevent the collision, as his book title suggests? If not, will the Lott maneuver at least be deemed a reasonable choice in the inevitable inquiries that will follow? Has someone come up with a better set of emergency shiphandling rules? Or is what Lott called the “universal” rules approach now a relic of that pre-ColRegs era, before the advent of ARPA, AIS and decision support systems and aids like “trial maneuvers” capable of executing collision-avoidance algorithms in nanoseconds? Now that Rule 17(a)(ii) has been added to the rules, do we care anymore?
Assessing the universal rules approach
Several flaws in Lott’s universal rules approach to shiphandling in extremis suggest themselves. First, the possible extremis situations are too numerous and multifaceted and are altered too often by the other ship for fixed approaches to ensure the safety of either. As MacElrevey writes in Shiphandling for the Mariner, there is no substitute for intimate knowledge of your ship’s turning and stopping characteristics. Indeed, the STCW Convention and Code and the U.S. Navigation Safety Regulations require as much. MacElrevey lists a number of such factors that might cast doubt on a rules-based approach to emergency shiphandling, such as the effects of wind and current; the tendency of a ship to turn more quickly away from the wind and swell rather than into it; differences in some ships’ turning diameter for a turn to port versus a turn to starboard; the effect of acceleration and deceleration on turning diameter, particularly with single-screw vessels; shallow water effect on turning diameter; and the effect, sometimes dramatic, of differences in trim or loading on the ship’s turning characteristics. Lott’s recommended rules on evasive turns arguably also fail to account for situations where there is insufficient sea room or the presence of a third or fourth vessel makes the formula maneuver untenable. Finally, there is the sense that, given the real danger that any evasive maneuver might be negated by conflicting action on the part of the other ship, experienced mariners directly observing the developing situation will be in a far better position to assess those possibilities in real time and select the most appropriate course of action than shore-side algorithm programmers. Similar arguments are raised against suggestions that VTS operators should be given the authority to direct ship maneuvers.
Putting aside the putative “universal” rules on turns, what about the often-repeated rule that, when in extremis, one should always slow or even back emergency? While it is true that slowing will lessen the resulting damage if a collision occurs (the force of impact varies with the square of the speed), a study by the Massachusetts Institute of Technology (MIT) demonstrated that slowing or backing is not always a prudent maneuver to avoid collisions. In fact, citing studies since 1875, the MIT study concluded that, in general, “the crash astern appears to be an extremely ineffective means of avoiding collision.” Even Rule 8(e) only requires slowing or stopping if necessary to avoid collision or permit greater time to assess the situation. A mariner would therefore be justified in not slowing, and perhaps even in accelerating, if necessary to avoid the collision.
Most mariners accept as their prime directive the duty to take all reasonable measures to avoid collision, or to minimize the damage of a collision which is no longer avoidable. In choosing the course of action the starting point is, of course, the applicable Collision Regulations. The arguments for additional, unofficial maneuvering rules — for “universal” rules to be memorized and then executed to extricate your ship from an extremis situation when the Collision Regulations are largely silent — are, at present, unpersuasive. The better view seems to be that such putative rules should be considered mere guidelines. Like ARPA and other decision aiding technologies, they provide information to be used judiciously, but not blindly. If there are others out there who believe otherwise, I invite you to make the case.
Craig H. Allen Sr. is the Judson Falknor Professor of Law at the University of Washington. He is a retired Coast Guard officer and the author of Farwell’s Rules of the Nautical Road.