CRM: THE MISSING LINK
On September 8, 1970, a Trans International DC-8-63F (Ferry Flight #863)
crashed on takeoff at JFK. The NTSB found the cause to be:
"...loss of pitch control caused by the entrapment of a pointed,
asphalt-covered object between the leading edge of the right elevator and the
right horizontal spar web access door in the aft part of the stabilizer...an
apparent lack of crew responsiveness to a highly unusual emergency situation,
coupled with the captain's failure to monitor adequately the takeoff,
contributed to the failure to reject the takeoff."
The elevator was jammed to almost the full-up position. The crew could not
have known that, as they started the takeoff roll, because the plane was
design-certified (incredibly) without a control indicator in the cockpit.
However, they had adequate time to realize pitch control was lost as the plane
began to rotate at a speed of 80 kts. The tail skid began dragging the ground at
91 kts. and could be heard on the CVR (cockpit voice recorder). The F/O (first
officer) was flying and the planned Vr (rotation) speed was 124 kts. The tail
skid left its marks on the runway, beginning at 1,550 ft. from the takeoff end
and continued for an additional 1,250 ft. The plane left the runway (dry, with
over 14,000 ft. available) at a point 2,800 ft. from the takeoff end. Twelve
seconds after the unwanted pre-mature rotation and 11 seconds after the sound of
the dragging tail skid, the captain words of "let's take it off" were
recorded on the CVR. Two and one-half seconds later the F/O replied, "Can't
control this thing, Ron." ("Ron" was the captain's name) One
second later the sound of the stick shaker began.
They took off in daylight with a headwind and clear weather. Clearly, the
captain made the wrong decision by deciding to continue the takeoff instead of
aborting. It is also clear that if the F/O had overruled the captain and
initiated an abort himself, when it became obvious that he had no pitch control,
the plane would not have crashed and those 11 crewmembers would not have died.
On June 7, 1971, an Allegheny Airlines prop jet Convair 340/440 (flight #485)
crashed into "the upper portions of three beach cottages at a height of
approximately 29 feet m.s.l. (mean sea level),..." The plane was making a
VOR (non-precision) approach to the Tweed-New Haven Airport in heavy fog. The
MDA was 380 ft. The NTSB determined the probable cause
"...was the captain's intentional descent below the
prescribed minimum descent altitude under adverse weather conditions, without
adequate forward visibility or the crew's sighting of runway environment. The captain
disregarded advisories from his first officer that minimum descent
altitude had been reached and that the airplane was continuing to descend at a
normal descent rate and airspeed. The Board was unable to determine what
motivated the captain to disregard prescribed operating procedures..."
There was no doubt about the descent below MDA being intentional. The flight
recorder revealed AL #485 had made 3 VOR approaches to the Groton Trumbull
airport (the enroute stop prior to the Tweed-New Haven stop), when reported
weather was well below that required to initiate the approach. The MDA for the
Groton approach was 610 feet, yet the flight recorder showed a descent to 175
feet, for the first two approaches, before a missed approach was initiated. On
the third approach, the plane descended to 125 feet, yet still failed to land.
The captain finally landed at Trumbull by making the 4th try as a "contact
In its "human factors" discussion of that accident, the NTSB noted
both the Allegheny Airlines procedures relating to the responsibilities of a F/O
and also the testimony of the surviving F/O:
"These procedures further state that:
'All crewmembers must realize that the captain is in complete command of the
airplane and his orders are to be obeyed, even though they may be at variance
with written instruction...'
"First Officer Walker, in response to a direct question, that at any
time during the approach after passing minimum descent altitude did he
consider taking over control of the airplane, the first officer replied
that '...There was a thought in my mind..., It's better one man flying the
airplane in perfect control, than than [sic] two men fighting over it...Had he
been incapacitated in any manner, I mean, I would have, because that is the only
time that I can take an airplane away from a captain.'"
"Considerable testimony was developed during the public hearing and
during the interviews with the first officer. Much of this testimony was
oriented toward the fundamental question: 'Why did the first officer not
take more positive action or possibly take over control of the airplane when an
extremely low and dangerous altitude was reached?'
"...The captain could also be classified as an authoritarian who enjoyed
absolute command. By contrast, the first officer appeared to be the quiet,
submissive type, not one who would question a superior or his authority."
"These personality profiles, combined with the apparent friendly
relationship that existed between the captain and the first officer, would be
conducive to a situation wherein the first officer would not challenge the
judgment of the captain under virtually any operational circumstances."
"...The Safety Board fully appreciates the most difficult dilemma of the
first officer in this case and recognizes the possibility of grave consequences
in questioning the captain's command authority under a situation as developed in
the case of AL 485."
"The Safety Board is concerned with the apparent delegation of authority
for operational control to the Pilot-in-Command without a concomitant
system to assess the effectiveness of how that authority is exercised in view of
the air carrier operator's duty to perform the operation with the highest of
"Inherent with delegation is a responsibility to assure that the
delegation is effectively fulfilled. In this instance the Captain's deviation
from the regulations governing the operation and the air carrier's operating
certificate was one that the operator could not control at that moment."
"The concept of command authority and its inviolate nature, except in
the case of incapacitation, has become a a [sic] tenet without exception. This
has resulted in second-in-command pilots reacting differently in circumstances
where they should perhaps be more affirmative. Rather than submitting
passively to this concept, second-in-command pilots should be encouraged under
certain circumstances to assume a duty and responsibility to affirmatively
advise the pilot-in-command that the flight is being conducted in a careless or
"...The second-in-command is an integral part of the operational control
system in-flight, a fail-safe factor, and as such has a share of the duty and
responsibility to assure that the flight is operated safely. Therefore, the
second-in-command should not passively condone an operation of the aircraft
which in his opinion is dangerous, or which might compromise safety. He
should affirmatively advise the captain whenever in his judgment safety of the
flight is a [sic] jeopardy."
"...The Board recognized that there is a dearth of guidelines
regarding the circumstances and manner in which a flight crewmember should take
affirmative action, which in turns leads to uncertainly [sic] in his mind
when an actual dangerous situation [sic]. For this very reason, and in
light of the circumstances of this accident, the Board believes that management
and pilots' organizations should reexamine the relationship between the
captain and flight crewmembers with a view toward enunciating the
responsibilities in circumstances where the aircraft is being operated unsafely."
"The Board believes that it is incumbent upon the air carrier's
management to devise and carry out a system that would enable it to continually
assess the pilot-in-command's performance in executing the carrier's operational
control responsibility which it must rely, to a great extent, upon the
pilot-in-command to fulfill."
While it is not certain the plane would not have crashed, if
the F/O had taken the controls and forced a climb back to a safe altitude (since
the reaction of the captain, in that scenario, can only be speculated upon), it
is certain the plane did crash because the captain was allowed to
have his way.
On January 11, 1983, a United Airlines DC-8 freighter, crashed on takeoff at
the Detroit airport, primarily because the stabilizer was mistrimmed. The
takeoff warning horn did not activate because (incredibly) the plane was
design-certified without the stabilizer position being tied into the warning
horn system. The plane was destroyed and all 3 crewmembers died.
The NTSB also noted the captain allowed a non-qualified pilot (the second
officer) to occupy the F/O's seat and make the takeoff (pilots refer to it as
musical chairs). That second officer had failed to qualify as a DC-8 F/O and had
lost his qualification as a 737 F/O and was permanently removed from all pilot
duties, by mutual agreement with the company. The NTSB concluded the crash could
have been prevented if the flying pilot had immediately applied nosedown trim
and forward elevator when the plane began to over rotate -- a likely scenario,
if the flying pilot was qualified and familiar with the normal control forces of
the DC-8, which was not the case with the incompetent second officer.
It is likely, therefore, that the captain and F/O would have noted and
corrected the stabilizer mistrim, during the running of the pre-takeoff
checklist, had they not been distracted from their normal flow of duties by the
seat swapping that occurred just one minute prior to takeoff. Or, if they had
still failed to catch that error, after running the checklist, it is likely the
highly qualified F/O would have immediately taken corrective action when he felt
the plane rotating prematurely, if he had refused to relinquish his seat
to the incompetent second officer.
On January 13, 1977, a Japan Airlines DC-8-62 freighter, carrying live beef
cattle to Japan, crashed shortly after takeoff from the Anchorage, Alaska
International Airport. The plane was destroyed and the three crewmembers and two
cargohandlers aboard were killed. The NTSB determined:
"...the probable cause of the accident was a stall that resulted from
the pilot's control inputs aggravated by airframe icing while the pilot was
under the influence of alcohol. Contributing to the cause of this accident was
the failure of the other flightcrew members to prevent the captain from
attempting the flight."
The NTSB, once again, became very concerned about why the other crewmembers
stood by and did not prevent the accident even though they must have realized
they were getting into a very precarious position:
"In view of the overwhelming evidence of the captain's condition, the
Safety Board must consider the lack of action by the other crewmembers."
"...It is extremely difficult for crewmembers to challenge a captain
even when the captain offers a threat to the safety of the flight. The concept
of command authority and its inviolate nature, except in the case of
incapacitation, has become a practice without exception. As a result,
second-in-command pilots react indifferently in circumstances where they
should be more assertive. Rather than submitting passively to this concept,
second-in-command pilots should be encouraged to affirmatively advise the
pilot-in-command that a dangerous situation exists. Such affirmative
advice could result in the pilot-in-command's reassessing his actions. The
Safety Board has previously stated, and continues to believe, that the
second-in-command is an integral part of the operational control of a flight, is
a fail-safe factor, and has a share of the duty and responsibility to
assure that the flight is operated safely. Therefore, the second-in-command
should not passively condone any operation of the aircraft which might
compromise safety. He should affirmatively advise the captain whenever,
in his judgment, safety of flight is in jeopardy, particularly when the
safety problem is detected before the flight is airborne. The Safety
Board could not determine what transpired between the crewmembers before they
boarded the aircraft, but there is little or no evidence that the
second-in-command or the flight engineer expressed any concern about the safety
of the flight. In addition, there is no evidence that they took any action
to prevent the flight from proceeding as planned."
On March 27, 1977, the worst carnage in airline history occurred
when a KLM
captain insisted on commencing a takeoff at Tenerife, in heavy fog, without a
takeoff clearance and with the knowledge that another 747 (Pan American) was
taxiing down that runway and had not yet reported clear of the runway. Both the
F/O and second officer knew they had not received clearance to takeoff and that
they had not confirmed Pan Am had taxied off the runway. They tried to convey
those concerns to the KLM captain, but he insisted on commencing the takeoff
because he was concerned about running out of legal duty time for the flight
Had the F/O jammed on the brakes and yanked the throttles back to idle,
in direct defiance of the KLM captain, 583 people would not have suffered a
terrible death by fire.
In 1990, a 747 relief F/O advised his captain they were departing the gate
(starting the first leg of a long international trip) illegally as certain
"no go" maintenance items had not been resolved. The captain refused
to listen, even though one of the items was related to the takeoff configuration
As the conversation continued, it became clear the captain understood the
takeoff warning configuration horn might sound when the throttles were advanced,
because one of the unresolved maintenance items affected that system. It also
became clear the captain intended to continue the takeoff, even if the
configuration horn sounded. Such a decision, as well as the attitude behind it,
was clearly wrong, dangerous, irrational and illegal. Few actions by a captain
could be more threatening to the safety of an airliner and its passengers. The
takeoff warning horn is an absolute "no go" item. No passenger
aircraft can be operated without its functioning properly and no pilot can
lawfully continue the takeoff if the warning begins below a safe abort speed.
The critical importance of the takeoff warning configuration horn is
highlighted by past accidents:
March 21, 1968, at ORD, a United 727-QC freighter, crashed on takeoff because
the flaps were not properly set. The takeoff warning horn sounded early in the
takeoff roll but an abort was not initiated until after lift-off. The aircraft
was destroyed. Two of the 3 crewmembers on board escaped without serious injury,
while the captain did require hospitalization. "The Safety Board
determines that the probable cause of this accident was the failure of the crew
to abort the takeoff after being warned of an unsafe takeoff condition."
December 26, 1968, Anchorage, Alaska, a Pan Am 707-321C freighter, crashed on
takeoff because they failed to extend the flaps. The takeoff warning horn was
not activated because the temperature was so cold the throttles were not pushed
forward enough to engage the switch (the colder the air, the more thrust for a
given throttle position). The aircraft was destroyed and all three crewmembers
August 16, 1987, at Detroit, a Northwest MD-82, crashed on takeoff, killing
156, because the flaps were not set to a takeoff position. The warning horn
failed to activate, because the pilots or maintenance personnel had deliberately
opened the circuit breaker, or because the circuit was automatically tripped by
a brief overload, or because the circuit breaker itself was faulty.
August 31, 1988, at DFW, a Delta 727-232, crashed on takeoff because the
flaps/slats were not properly configured. The takeoff warning system failed to
activate, probably because of a faulty switch. Eleven passengers and two flight
attendants died and the aircraft was a total loss. The NTSB said:
"Contributing to the accident was Delta's slow implementation of
necessary modifications to its operating procedures, manuals,
checklists, training and crew checking programs, which
was necessitated by significant changes in the airline...
"Contributing to the accident was the lack of sufficiently
aggressive FAA action to have known deficiencies corrected by Delta and the lack
of sufficient accountability within the FAA's air carrier inspection process."
That relief F/O subsequently reported the incident to the head of the 747
training department for that airline. That department head responded with anger;
not at the offending captain, but at the F/O making the report. He simply did
not want to hear or believe that one of his captains could have been so
irresponsible. He refused to accept the report or to investigate and take
appropriate action, despite the fact that the airline's flight operations manual
(an FAA required and approved document) stated:
"In order to maintain the highest level of safety and to permit the
Company to initiate follow-up action, it is the obligation of all
crewmembers to report any areas which could be detrimental to safe operations.
This includes, but is not limited to, items such as mechanical problems,
weather, crew proficiency, and airport or ATC problems."
Early in 1991, that relief F/O sent a 15-page letter to the head of that
airline's flight operations department, documenting the specific details of the
incident, his verbal report to the 747 department head and the requirements in
the 747 maintenance manual and flight operations manual. He concluded the letter
with analysis and recommendations. Excerpts:
"I believe that the accidents I have cited (as well as the one where a
JAL captain deliberately crashed a DC-8 and killed passengers -- because he was
mentally ill), demonstrates that, as a simple matter of logic, there are times
-- extremely rare though they may be -- when it is right, proper and necessary
for another cockpit crewmember to overrule a captain's decision and even take
control of the airplane away from him if he fails to respond in a rational
manner. My [cockpit resource management] training teaches me that; the
admonitions of the NTSB teaches me that; my own rational mind teaches me
"That management devise a program or system to implement what the NTSB
has been advocating for almost 20 years:" [He then quoted the NTSB's
commentary -- see above -- from the 1971 Allegheny accident report]
"Such a system should encourage cockpit crewmembers to make written
reports to management when they observe a pilot-in-command making decisions that
clearly endanger the safety of the operation. And that system should provide for
positive follow-up so that [the airline's management] duty to provide the
highest degree of safety is not short-circuited by a captain who continues to
make dangerous and irrational decisions."
"That [cockpit resource management] training be expanded to include a
thorough discussion of the responsibilities and proper actions of cockpit
crewmembers who observe a captain making dangerous decisions. The "human
factor" accidents, described above, should be required knowledge of all
pilots and guidelines should be established as to how and when control should be
taken away from a captain so as to prevent an accident."
"Without such guidelines, the danger continues to exist that we will
have more accidents that could have been prevented, like the ones enumerated
above; or, that improper actions might be taken by a crewmember, who is
legitimately concerned about the safety of the operation, because of the stress
of the moment and the lack of clear training."
Copies of that letter were sent (all via certified mail with return receipt
requested) to the following:
1. Mr. Samuel Skinner, Dept. of Transportation, Washington, D.C.
2. Mr. James Busey, F.A.A., Washington, D.C.
3. Mr. James Kolstad, NTSB, Washington, D,C.
That relief F/O never received a written reply from any of the above. The
only feedback he got was via the verbal grapevine: the management of the airline
he worked for wanted to fire him. They didn't, of course, because all the rules
required him to make that report and it was quite proper for copies to be sent
to government agencies that are charged with responsibility for airline safety.
Cockpit resource management training (referred to as CRM or CLR) is one of
the most valuable safety tools we have today. It has contributed significantly
towards the prevention of "pilot error" accidents; it has saved
airplanes and lives. But there seems to be great reluctance to even discuss the
idea that there are times when the captain's authority must be countermanded. It
is clear, from the accidents noted above, crashes could have been prevented if
the second-in-command had overruled the captain.
CRM training should, therefore, be expanded to highlight that issue. The
cited accidents should be known and thoroughly discussed by all pilots, and
specific guidelines set out so that they will know when and how to act, if a
captain makes decisions which are clearly irrational and dangerous. Such training
will, in itself, probably prevent most future incidents because captains will
know they will be countermanded if they attempt to take a course of action,
endangers the lives entrusted to them.
See also, the guest editorial on CRM.
[All bold emphasis
in this editorial is that of the author]
Robert J. Boser
Return to Editorials Page
The Editor of this Web Page, now retired, was an airline pilot for
33 years and holds 6 specific Captain's type-ratings on Boeing Jet Airliners.