The 1895 Montparnasse Train Derailment and Understanding the Human in the System
The 1895 Montparnasse Train Derailment may be the most visually striking examples of its kind. Immortalised forever in the sepia-toned image that marks the event, it remains a classic example of how crucial it is to understand the human in context with the complex technological systems that surround them.
On October 22, 1895, the Granville-to-Paris express, carrying 131 passengers spread across 10 carriages, approached the Gare Montparnasse in Paris. Though it departed Granville on time, just over seven hours later, it was running late.
Driver, Guillaume-Marie Pellerin, with 19 years of experience, was a railway veteran, but this didn’t stop him from approaching the station at cruising speed, somewhere between 40 and 60 kilometres per hour.
Too late, Pellerin activated the Westinghouse air brake, meant to slow the train. However, the brake either wasn’t adequate to slow the train’s speed, or it malfunctioned.
Conductor Albert Mariette, who was also responsible for the train’s safety, was distracted with paperwork, which meant he failed to pull the handbrake until the train had reached the end-of-line buffer.
The Granville Express broke through the buffer, slid 30 meters across the platform and then cantilevered 10 meters to the street below. There it lay, a smoking, angled heap when the now iconic photograph was taken.
Falling masonry struck and killed magazine seller Marie‑Augustine Aguillard, who worked below the station, but thankfully there were no other fatalities.
The easy solution, in any complex sociotechnical disaster, is to blame the humans for the failure of the system. In this case, that’s what the law chose to do. The driver Pellerin was fined 50 francs and sentenced to two months in prison. The conductor was fined 25 francs. Later, both fines and sentences were suspended.
The more enriching approach to making things and making things better, especially when dealing with complex technology, is to understand how humans fit into the system, and where training, process, and technology may have let them down.
This is based on a fundamental assumption; technologies are always more malleable than humans. If people need more support with a complex task, it’s more ethical and results in better performance, to change the technology, the training, or the process.
After all, humans are humans. We are a bundle of complex perceptions, thoughts, emotions, and action capabilities. When we’re part of a larger complex system, mistakes emerge from the gaps between people’s capabilities and the system that enwraps them.
Though the events at Gare Montparnasse were more than a hundred ago, we can ask questions about how the driver and conductor behaved, in the context of the wider system around them, including:
How serious and/or punitive were the train company incentives for late arrivals and how might this have affected pressure on the driver to increase the speed?
Why was the conductor occupied with paperwork? If his role was safety critical, then shouldn’t this have been removed from his job?
Were the driver and/or conductor trained in emergency scenarios, and did they have time to practice them?
How did the driver and conductor communicate or work together?
How often was the emergency equipment inspected and to what standard?
Why was the end-of-line buffer so short? G. Richou an engineer writing about the accident at the time, noted it only had 0.8 meters of travel.
G. Richou also wondered, even with a short end-of-line travel, why weren’t there absorbent buffers, like those with artillery cannon on naval ships, that can absorb large forces over short distances?
Was the placement of the train tracks above a busy thoroughfare really a good design choice?
This is only a selection of prompts that spiral out into a wider human-centred exploration of the disaster. And they evoke an interesting conclusion.
It’s easy to blame people, and much harder to understand how people, when working in the context of a complex system, make mistakes because of how technology is designed.
It’s a good reminder that we have to be both Human-Centred and System-Wide. Starting with people and then spiralling outward, picking up all the challenges and opportunities with an existing or new system.
That’s how we help avoid future disasters.
