Human Error Part 5

"Forget them! Take care of the Senators and the Project and you'll be doing enough."

Harper arrived shortly, nervous in spite of his attempt to appear composed. But he was put at ease when they took him to the laboratory of complex testing equipment a.s.sembled by Nat Holt.

Paul indicated a seat in the middle of the ma.s.s of equipment. "As near as we've been able to make it," he said, "this simulates the landing procedure of a rocket craft. There are a hundred and thirty-five distinct actions, observations and judgements involved. A taped voice will lead you through the sequence, asking you to press b.u.t.tons and make adjustments to indicate your observations and responses. When you can do all this to your satisfaction, you will turn off the tape and continue for as many cycles as you can."

"How long? A man could do that for a month, provided he didn't have to sleep."

"I think you'll be a little surprised. You will continue until your acc.u.mulation of errors becomes so great that the entire procedure collapses."

"It still looks like a kid's game to me," Harper said confidently.

"Let's get started."

Carefully, they fitted the multiple electrodes of the electro-encephalograph recorder to his skull. The tape instructor was turned on, and Harper began the first cycle.

Behind the one-way gla.s.s of the observation room, Paul sat with Nat Holt and Professor Barker and two a.s.sistants, watching. The rocket engineer began jauntily, contemptuous of the simple actions required of him, impatient to have it over with and get back to his duties at the take-off stand.

The instructions coming over the speaker had some variations from the normal handling of a ship, including the items necessary to record observations and responses. Harper listened to these for a half dozen cycles. Then, confident that he could breeze through the procedure for the rest of the day if he had to, he switched off the tape and settled back to take it easy.

One by one, he watched the meters, noted their information, made the proper adjustments, added compensations, waited for results, checked and re-checked--

"He'll go a long time," said Nat Holt confidently. "He's had top training. If it breaks down, we may find out a few things."

"c.u.mmins had top-drawer training, too," Paul said. "His break point seemed to have no adequate antecedents. I don't think we're going to find Harper holding out very long."

After an hour, the att.i.tude of contempt had left Harper's face, and he was proceeding with obvious boredom. He had made no error yet, but there was evident a faint trace of anxiety as he concentrated on the instruments and levers.

At two hours and a half Harper reached for a b.u.t.ton and withdrew his hand in abrupt hesitation. Then it darted out again and pressed decisively. At three hours he was making two such hesitations every cycle.

"Not so good," Barker commented. "Not for a man who battles himself the way Harper does."

Nat Holt remained silent, watching critically the wavering dials and graphs showing the engineer's physical condition and reaction.

At four and a half hours, Harper's hand reached for a lever in the center of the board. But it didn't get more than a third of the way. In mid-air it froze, as if paralysis had suddenly struck it. Harper regarded it in seeming dumb astonishment. His face grew red, and sweat broke out upon his forehead as if from the physical exertion of trying to put his hand to the lever.

Paul grabbed a microphone and switched it on. "Touch the lever," he commanded. "Draw it toward you."

Harper looked around as if in panic, but he completed the motion. He sat staring at the panels for a full two minutes while alarm eyes went from green to yellow to red.

"Alarm red!" Paul exclaimed into the microphone. "Correct course!"

Harper turned and glared about with hate in his eyes as if to find the source of the sound. He began tearing at the wires and contacts fastened to his head and body. "To h.e.l.l with the course!" he cried. "I'm getting out of here!"

He hurled the wiring harness at the panels. Then, he stood in a moment's further paralysis and slumped finally into the chair. He put his arms and head down on the instrument desk and began sobbing deeply.

Paul put away the microphone and moved to the door. "That's the end of that," he said. "I hope our record is good. Harper might not like to go through that again."

Nat Holt was still staring through the window at the sobbing engineer.

"I don't understand," he murmured. "What made him break down like that for no reason at all?"

One by one, the top engineers of the Base went through the breakdown test. Some broke down with an emotional storm as Harper had, others simply ended in a swirl of confusion that put lights flashing all over the panels. But all of them had a breaking point of some kind that could be measured in a small number of hours.

The test was a stab in the dark. It was based on an old and well-known principle that repeated tactile contact under command will break down the motor responses of the body in a matter of hours. Paul did not know whether it would actually provide a fertile lead to the problem of error or not, but it seemed the closest possible approach at present.

Nat Holt, however, was astonished at the reaction of the men. He insisted on trying it himself, determined that he would not break down no matter what happened. He lasted six hours before the panel lit up like a Christmas tree.

He subjected the resulting curves to an a.n.a.lyzer, and to his own he gave the most detailed attention. At the end of a full week of study on it, he called Paul with an excitement he could not suppress in his voice.

"It looks like you owe that dinner," he said. "We've got what we were looking for!"

"What are you talking about?" Paul demanded.

"We've got proof that a human being is nothing more nor less than a simple cybernetic gadget. It's a laugh--people trying to build a mechanical man all these years. That's the only kind there is!"

"You still aren't making sense."

"Come on over and see for yourself."

Puzzled and irritated, Paul left his office and went down to the a.n.a.lyzer laboratory. There he found Holt and his staff in a buzz of excitement.

The multiple recorder sheets were laid out on long tables, being studied intensely. Paul followed Holt to one series that was separated from the rest.

"We didn't know we had anything at first," said Holt. "The pulse was so low in amplitude that it was hard to pick out of the noise, but the a.n.a.lyzer showed it was consistently present under certain conditions of the subject."

"What conditions?" said Paul.

"At the exact moment of committing an error! I should say it occurs between the moment of making the decision to carry out an erroneous act and the triggering of the motor impulse that executes it."

Paul frowned. "How can you be sure it doesn't occur at any other time as well?"

"Because we've run every set of charts through the a.n.a.lyzer and this particular impulse comes out no other place."

"It looks very interesting," Paul said. "But why did you say you've got proof that a human being is nothing but a cybernetic gadget? I don't see what this has got to do with it."

"I didn't give you quite all the story," Holt said smugly. "I should have said that the pulse occurred every time there was an _intent_ to perform an error. Sometimes that intent was not carried out."

"I don't understand."

"That pulse is nothing more nor less than a feedback pulse indicating that an action matrix has been set up which is in non-conformity with the previously chosen pattern of learning or intent. It's a feedback alarm carrying the information that an error will result if the proposed action is carried out. When the feedback is successfully returned to the action matrix a change is made until there is no feedback and a correct action is taken. When the feedback is blocked or ignored, an error results. It's as simple as that! Your complex human being is nothing but a fairly elaborate cybernetic machine operating wholly on feedback principles. The only time he fails and breaks down is when he ceases to act like the cybernetic machine that he is!"

Holt's eyes shone triumphantly as he patted the long strips of paper on the table. Paul followed the motion of his hand and remained staring at the graphs in a kind of stunned recognition. There must be some mistake, there _had_ to be. Holt's interpretation was wrong, even if the data were correct. Man, a feedback response mechanism--! If that were true a vacuum tube structure could eventually be devised to do _anything_ a man could do.

"I think we'll hold off on that dinner a while yet," Paul said. "The data are interesting and, I'm sure, important--but I can hardly agree with your conclusions." Inwardly, he cursed the stiltedness he felt creeping into his voice, and his irrational resentment of Holt's continued smug grin.

"Take all the time you want," Holt said, "but when you're through you'll come up with the same answers I've got. Man is a machine and nothing else. Our only job now is to discover why the feedback sometimes fails, and to set it back on the job."