46 Encounters with mountaineers
In junior-high and high-school, I devoured the literature of mountaineering, from John Muir to Lionel Terray. But that interest lagged in college, and I stopped following mountaineers and mountaineering.
In the late 1980s I wanted to switch from studying lizards to studying fruit flies (Drosophila) so I could study the experimental evolution of thermal sensitivity. Al Bennett and I use crude phylogenetic approaches to study evolutionary patterns of thermal sensitivity of Australian scincid lizards, but Al and I both saw a need to switch from descriptive (historical) to manipulative approaches.
Drosophila appealed to me for such studies, as I’d been impressed by Michael Rose’s studies on senescence. My first “thought” experiment was to set up stocks of flies and rear them in the lab for several years at different temperatures. Then compare their thermal sensitivity (e.g., heat tolerance). If thermal sensitivity is evolutionarily labile, then the files that had been evolving at high temperature (that is, evolving by ‘natural selection in the laboratory’) should do better at high temperature than would flies evolving at lower temperature. But at this time, no one know whether flies had the genetic capacity to evolve quickly.
A problem for me is that I had no experience with Drosophila. However, I’d previously met Linda Partridge (now Dame Linda Partridge) at a behavior conference in London, and I knew she had switched birds to flies. She seemed like the perfect person to judge whether my proposed fly experiment would be both feasible and of interest. I saw that she would be attending a meeting at the University of British Columbia, so I drove up to seek her advice.
Linda said the experiment was feasible and of interest. Moreover, she had flies that have been evolving at 18 °C or at 25 °C for about three years, with three replicate populations at each temperature. She asked whether I like to study her stocks?
I was stunned as Linda had invested a lot of time setting up and maintaining those stocks. I said that I’d be delighted to work on those flies, but I had to admit that I had never even sexed a fly. So I asked whether she could teach me the basics and then whether we could collaborate on a study. She said sure, and I got a small grant from Burroughs Welcome, enabling me to make two trips to visit her lab in Edinburgh.
When in Edinburgh, I stayed with Linda and a friend. One night we were all invited to a party, but I wasn’t interested. So I stayed in her flat and looked for a book to read. I spotted The Challenge, by Reinhold Messner. I knew that Messner was a great climber, and that he and Peter Habeler had been the first to summit Mt. Everest without using supplemental oxygen. But that was all I knew about him.
I started reading this book, which is an account of a revolutionary climb that Messner and Habeler made in 1975 of the northwest face of Hidden Peak, an 8000-m peak in Pakistan. They climbed light and fast – “alpine style.”
That book described a pioneering adventure, but what attracted me most was Messner’s philosophy and his advocacy of “alpine style” climbing. This light and fast style was a radical shift from the huge, military-style (“siege-style”) campaigns typical of most prior Himalayan assaults.
Then about 1995, Messner came to Seattle for a lecture. He talked for perhaps two hours and gave the best lecture I’d ever heard. It blended adventure, tragedy, triumph, and philosophy. He held the audience – and especially me – from start to finish.
The lecture hall was packed, and many attendees were young. I remember thinking, Messner’s daring exploits will inspire some of these kids to go out and do really dangerous things. However, as Messner proceed through his presentation, I realized that Messner cleverly and effectively chose his words to blunt that possibility.
Messner obviously knew that his exploits could inspire young (and even old!) climbers to attempt climbs well over their abilities. He could have warned them by saying that adventures like this should be done only by trained professionals. Instead, when describing how he prepared for some new challenge, he repeatedly used a Tibetan expression that translates as “little by little.” In effect, Messner was giving his audience a template for how to prepare for a new challenge. Take little steps, and don’t attempt leaps. I was impressed. Messner was not only a great climber, but he was sensitive to the impacts of his words.
But what interested me most was Messner’s philosophy of Himalayan climbing. He argued that using supplemental oxygen was unethical. For him, climbing Everest or any big peak while using supplemental oxygen was physiologically like climbing a peak thousands of meters lower. I had to agree! Supplemental oxygen would raise the partial pressure of oxygen in the lung, mimicking the partial pressures lower on the mountain.
Also he argued against traditional siege-style expeditions, which were expensive, slow, and put lots of people at risk. Instead, he advocated climbing in a small team, eschewing fixed camps and ropes, and just going fast and light for the summit. He argued that “alpine style” climbing should be safer and more successful than the traditional siege-style climbs in the Himalayas.
I remember thinking, OK, but alpine style must be very dangerous. If a storm hits, and if you don’t have fixed ropes and established camps (with sufficient supplies), you are in deep trouble. Moreover, if one or more climbers become sick or injured, having team members to help seemed like a really good idea. I remember wondering, had Messner spent too much time at high elevation?
I don’t recall whether Messner gave a formal justification for his position. But I was intrigued by his arguments and gradually deduced his rationale as he talked. Most importantly, a big team (with fixed ropes and camps) inevitably has inertia. Moreover, a big team will have lots of people spending lots of time on the mountain, thereby exposing themselves to avalanche, rockfall, or medical issues. Thus, the probability of an accident (or ailment) should be high because of the number of people exposed as well as their long exposure on a mountain. In contrast, a small (if skilled) team can wait for a break in the weather and then scamper up and down the mountain, thereby minimizing overall risk. This was the essence of Messner’s approach, and it struck me as beautifully logical (several linked steps), but high risk.
Messner’s view did conflict with my “intuition.” [Note: recall I was not a mountaineer!] I distinctly remember sitting in the audience and debating his views as he talked. I codified our opposing views by “drawing” two graphs in my mind. In both graphs, the X axis was size of the climbing team, from ‘alpine’ style to ‘siege-style.’ The Y-axis was either the probability of success or the probability of death.
For Messner, the probability of success should be inversely related to the size of the team, whereas the probability of death should be positively related to the size of the time. For me, the relationships should be just the opposite!
I went home that night wondering which view was correct and whether data were even available to evaluate these views. This was a pin-ball moment. I was excited and couldn’t shake the idea of mountaineering analyses.
I soon discovered an amazing resource for data on Himalayan climbs. Miss Elizabeth Hawley was an ex-patriot American, but had lived in Kathmandu since about 1960. She was a Reuters correspondent and had been compiling data on mountaineers in Nepal, starting with the American Expedition of 1963. Then I also learned about Xavier Eguskitza (Basque), who had been compiling mountaineering records for K2. I corresponded with Xavier, and asked whether he’d want to collaborate on some mountaineering analyses.
He did, and our first project analyzed death rate of climbers descending from the summit of Everest (Hawley’s published data) or from K2 (Eguskitza’s unpublished data). Xavier was untrained in science and statistics, but had a natural feel for analytical challenges and numbers.
Messner had argued that the use of supplemental oxygen was cheating, but did it nonetheless enhance safety? No one had ever checked. So Xavier and I evaluated two competing hypotheses. If the low oxygen at altitude was physiologically limiting, then we’d expect that climbers who didn’t use supplemental oxygen would have relatively high death rates on descent from the summit. On the other hand, the only climbers who can reach the summit of Everest should be the best and most experienced climbers in the world. Therefore, all else equal, we’d expect that climbers who can summit without supplemental oxygen should have the lower death rate on descent.
Xavier and I found that climbers who never used supplemental O2 had markedly higher death rates descending from the summit of either peak than did climbers who had used supplemental O2, especially on K2. The input data weren’t randomized (climbers of course choose whether on not to use supplemental O2), but the results strongly suggest that supplemental O2 enhances safety. I was surprised by the clarity of the results, given that conditions even on a single route can vary dramatically from day to day and year to year.
We submitted our paper to the prestigious Journal of the American Medical Association (“JAMA”). This was a long shot, but we knew one of the editors was an ex-mountaineer. He liked it and accepted our paper with minor revisions.
That paper launched several other papers with Xavier, using mainly Miss Hawley’s published data on Everest.
During this period I’d written several emails to Richard Salisbury, a Himalayan mountaineer and database expert (University of Michigan). Richard was converting Hawley’s interview notes to a computerized data base, The Himalayan Database. I would ask him for data, which he kindly provide. But because he was not a statistician, he wanted to design his database to be maximally useful for potential statistical analyses. So he suggested that we worked together. This was win:win. I got access to a rich and carefully curated database, and Richard was able to get feedback on database design and development, which would make it easier for us and others to explore this remarkable database.
This led to a series of papers, looking for example at whether age and sex influence the probability of summiting and also the probability of dying. We also compared at how success and death differed from the Nepal vs. Tibetan sides or in different seasons. By so doing, we provided prospective climbers with quantitative data on the odds of success and death.
Before I started, there was only one significant paper analyzing Himalayan deaths – John Town (1986, in Mountain). And many of the quantitative statements in the literature were bogus. This is still not a cottage industry, but The Himalayan Database continues to grow and serve as the primary resource for journalists, mountaineers, and statisticians. I have personally enjoyed exploring mountaineering epidemiology and of meeting many mountaineers. They are a special group, even if their understanding of the laws of probability seems lacking.
I once showed the supplemental-oxygen data to Messner, who had eschewed using supplemental oxygen. Messner thought for a few seconds and then said that climbers need to decide for themselves.