Drift into failure, p.5
Drift Into Failure, page 5
It is precisely because a system is a complex web of relationships that it can be creative and adapt to environmental challenges. This is how it survives – which is of course the story of evolution. Such survival should not be mistaken for progress (or regress). Complexity doesn't allow us to think in linear, unidirectional terms along which progress or regress could be plotted. Drift into failure surely invites the idea of regress (less margin, less room to recover, a decline in norms, a regression from higher forms of organizational stature to lower ones) as its central doctrine, and encourages the interpretation of history in sequences (as in the increase in lubrication intervals in Alaska 261, see Chapter 2). But complexity suggests that it is about constant adaptation to what happens now (or shortly ago), and any larger historical story that we put on it is ours, not the system's.
Why We Must not Turn Drift Into the Next Folk Model
Drift into failure may be an evocative title for a book – it calls to mind something inevitable, something slow and unstoppable. It may suggest that failure is always an option. And in complex systems, failure is always an option. But it is a better title than a metaphor. After all, adaptive behavior in complex systems probably moves in all kinds of ways, not just in one direction. Failure is one option; other outcomes are always possible too. A particular direction becomes visible only from the position of retrospective outsider, when we are able to lift ourselves out and look back on what has happened. Thus it may only be possible in hindsight to see how decisions and trade-offs accumulated and reverberated to create a particular trajectory toward an accident. The drift is thus a construct in retrospect, after the failure has already happened. Or, at the most, it is a construct that can be applied to a complex system from the outside. From the inside, drift is invisible. That actually also offers a lever for thinking about preventing drift and preventing it from becoming failure (see Chapter 7). Outside perspectives and new perspectives are important, as is the diversity of perspectives and taking a long view of things.
There is a deeper reason to be suspicious of drift into failure as a metaphor for what happens with and inside organizations. This has to do with its sense of inexorable decline or regress into something worse than what was before. This would be the inverse of the narrative that the West has always liked to tell about itself: the story of unstoppable progress. Newton and Descartes lie at the roots of this narrative (and, as Chapter 3 will show, society had good reason to put them there). Progress as a central concept to Organize Western history around reached its peak in the nineteenth century with the triumphs of the Industrial Revolution and Victorian imperialism and expansionism. It was made literal in a linear image in the U.S.A. where white settlers pushed the frontier ever further west. The further, the more progress. The fuller these settlements became, and the more supported by Industrial Revolution technology (for example, steam trains), the more progress people saw.
More recent decades show, at least in places, a reversal of this central Western narrative. Progress as the prevailing doctrine in the interpretation of historical sequence is being replaced, at least in some places, by a doctrine of regress. This has been infused by pessimism about human nature after having come out of the most murderous of centuries (the twentieth). People see regression in steady ecological destruction, in the decline of America and its market-driven capitalist model, in secularization and in the depreciation of the West as a whole. Specifically, we find seductive the idea of moral fragmentation, of growing societal anomie, a lack of social or ethical standards in individuals and groups.22 This, in many societies, is seen as drift into failure on a grand scale: the decline of norms, the irrelevance or unenforceability of Compliance, the Collapse of traditional systems of normative observance. In other words, a regressive metaphor for thinking about safety (like drift into failure) could tumble into well-prepared psychological soil.
But such recognition should not be confused with the truth value of the metaphor. Evolution has no direction, at least no visible direction when you're in the middle of it. Evolution, progress and regress in complex systems are perhaps nothing more than an illusion, a cultural, psychological imposition that comes from us, not from the system or its history. Stephen Jay Gould called progress "a noxious, culturally embedded, untestable, nonoperational idea that must be replaced if we wish to understand the patterns of history."23 If history is seen as progressive (or, for that matter, regressive, as in drift), then it becomes all too easy to see it as directed, as following a vector – going at a particular speed in a particular direction. One consequence is that it becomes tempting, once again, to see a hand behind that vector. A "cause" for its speed and its direction. Which would be a Newtonian idea, an idea that clashes with complexity.
The risk of a folk model exists when we grasp a label only superficially and subsequently start to recognize it everywhere in our world. This book wants to avoid introducing a folk model. So it is not an attempt at exegesis. It is not claiming that we can read drift and normal decisions and their resulting drift (or, on the other hand, fraudulence and amorality and their bad decisions) out of the story we encounter. Exegesis makes the assumption that the essence of a story is already in the story. It is there, waiting and ready-formed for us to discover. All we need to do is read the story well, apply the right method, use the correct analytic tools, set things in context, and the essence will be revealed to us. Such structuralism sees the truth as being "behind" or "within" a text. Some stories will show a trajectory of drift, others will not. Simply because the drift is already there or it is not –all we need to do is look whether we can find it.
Rather, in encouraging us to think about drift into failure, this book makes an appeal to eisegesis, of us reading something into the story. Any directions or vectors that we discern in such adaptation are of our making –Gould's "culturally embedded idea" of progress (or regress). Foucault too was always skeptical of linear trajectories (either up or down) as descriptive of the social order, including the process of, and perspectives on, knowledge or science. His position, and in many ways the position of complexity, is that of post-structuralism, the broad intellectual movement that swept social and other sciences synchronous with the growth of complexity theory and systems thinking in the latter half of the twentieth century. Post-structuralism stresses the relationship between the reader and the text as the engine of "truth." Reading in post-structuralism is not seen as the passive consumption of what is already there, provided by somebody who possessed the truth and is only passing it on. Rather, reading is a creative act, a constitutive act, in which readers generate meanings out of their own experience and history with the text and with what it points to.
As post-structuralism sees it, author and reader aren't very different at all. Even authors write within a context of other texts (data from a cockpit voice recorder, for example, or documents about an organization's quality assurance system) in which choices are made about what to look at, what not – choices that are governed by the author's own background and institutional arrangements and expectations. The author is no longer a voice of truth. Text, or any other language available about events or accidents, has thereby lost its stability. Nietzsche, for example, would have been deeply mistrustful of the suggestion that while everybody has different interpretations, it is still the same text. He, and poststructuralism in general, does not believe that it is a single world that we are all interpreting differently, and that we could in principle reach agreement when we put all the different pictures together. More perspectives don't mean a greater-representation of some underlying truth.
In a complex system the contrary seems to be at work instead. More perspectives typically mean more contradictions. Of course, there might be some partial overlap, but different perspectives on an event will create different stories that are going to be contradictory – guaranteed. The reason, says complexity science, is that complex systems can never be wholly understood or exhaustively described. If they could, they would either not be complex, or the entity understanding them would have to be as complex as the whole system. The complexity of the whole system would have to present in that part. Which the local rationality principle says is impossible. We (people, computer models, regulatory agencies, risk assessments) can only understand parts of a complex system, and pretty local parts at that. For what happens further away we have to rely on the perspectives and interpretations of Others. And we can only learn about those interpretations from those people – or from yet other people. By the time we learn about them, such interpretations will likely have gotten modulated, adapted, suppressed, enhanced, changed.
Implicit in this stance is that there are no final conclusions. The events we want to study (just like any complex system) can never be fixed, tied down, circumscribed conclusively with a strong perimeter that says what is part of it and what is not. Even if the system has pretty much died, and all we have is archeological data (for example, after a plane crash which caused the airline to go defunct), it will never be possible to say that we have all the data, that we have every perspective, that we have considered all angles. We simply can't know.
We may never be able to successfully argue that some accidents are the result of drift into failure, and that others are not. Of course it would be convenient, and quite "scientific," perhaps, to carve up the world into categories and say, "this is a drift into failure accident, and this is not." But post-structuralism doesn't really allow us to say whether something is or is not in the text (in the accident, in the event) itself. It is up to the reader, the observer, to interact with the text – the event, the accident – to interrogate it, to read a particular background and questions into it, and then bring out what is relevant, what is "true" to that reader. By reading drift into a particular failure, we will probably learn some interesting things. But we will surely miss or misconstrue other things.
If we don't believe that a particular accident is the result of a drift into failure, but rather the result of a momentary, sudden stochastic contraction of elements in time and space that had no antecedent, no history, no diachronic trajectory, no path dependency – fine. We can certainly go ahead and build a persuasive story around that conviction. But by not reading drift into it, we might also miss things and misconstrue others. In the end though, whether there is drift into failure depends not on what is in the story, but on what is in us. It depends on what we bring to the story, what we read into it. It depends on how far we read, how deeply, how far back, what else we read. And it depends on the assumptions and expectations we bring about knowledge, cause, and ultimately morality.
References
1 Rasmussen, J., and Svedung, I. (2000). Proactive risk management in a dynamic society. Karlstad, Sweden: Swedish Rescue Services Agency, p. 14.
2 Woolfson, C. and Beck, M. (eds.). (2004). Corporate responsibility failures in the oil industry. Amityville, NY; Baywood Publishing Company, Inc.
3 Epistemology is the branch of philosophy that asks how we know what we know. By epistemological space, I mean a space in which certain ideas are seen as logical or possible, while other ideas are seen as ridiculous or even fall completely out of consideration because the epistemological space contains no language, no knowledge to put them in.
4 This is of course unfair to both men, not only because their own ideas are more complex and conflicted and paradoxical than what I make them out to be, but also because subsequent people (e.g. Pierre Simon Laplace, 1749—1827) played a greater role in popularizing some of the ideas more then they themselves did. And, of course, there were people before them (like Plato, 429—347 bce and Socrates, 469—399 bce) who may have been even more instrumental in laying a philosophical
foundation for these ideas and their opposite. There is, however, a case to be made for the existence of a set of ideas that can roughly be called Newtonian—Cartesian, which is what I do in this book.
5 Page, S.E. (2008). Uncertainty, difficulty and complexity. Journal of Theoretical Politics, 20(2), 115—49, p. 115.
6 Broder, J.M., and Calmes, J. (2010, June 18). Chief of BP, contrite, gets a scolding by Congress. International Herald Tribune, p. 1.
7 Mufson, S., and Kumblut, A.E. (2010, June 15). Amid claims of BP's 'shortcuts,' Obama speech to stress action, prevention. Washington Post.
8 Mufson, S., and Kumblut, A.E. (2010, June 15). Ibid.
9 Mufson, S., and Kumblut, A.E. (2010, June 15). Ibid.
10 Green, J. (2003). The ultimate challenge for risk technologies: Controlling the accidental. In J. Summerton and B. Berner (eds.), Constructing risk and safety in technological practice. London: Routledge.
11 This is Paul Cilliers' pithy insight. See: Cilliers, P. (1998). Complexity and postmodernism: Understanding complex systems. London: Routledge.
12 Associated Press (2010, July 18). Number of homeless families grew in 2009, report says. International Herald Tribune, p. 4.
13 Broder, J.M., and Calmes, J. (2010, June 18). Ibid., p. 3.
14 McLean, B. and Elkind, P. (2004). The smartest guys in the room: The amazing rise and scandalous fall of Enron. New York: Penguin, p. 133.
15 As Snook explains in his 2000 book, sudden contractions or stochastic fits in system coupling (indeed the crises where all correlations go to 1) occur in systems that have undergone practical drift, where locally grown logics of action have made individual units slide apart. Once crisis, or a sudden contraction in coupling, puts them back together with demands for quick and safe cooperation, the fissures and gaps that have developed become acutely active and sometimes even visible, leading to misunderstandings, an increase in risk and to failure. See: Snook, S.A (2000), Friendly fire: The accidental shootdown of US Mack Hawks over Northern Iraq. Princeton, NJ: Princeton University Press.
16 Heylighen, F., Cilliers, P., and Gershenson, C. (1995). Complexity and philosophy. Brussels: Vrije Universiteit.
17 Heylighen, E, Cilliers, P., and Gershenson, C. (1995). Ibid., p. 8.
18 Vaughan, D. (1996). The Challenger launch decision: Risky technology, culture and deviance at NASA. Chicago: University of Chicago Press, p. 37.
19 Vaughan, D. (1999). The dark side of organizations: Mistake, misconduct, and disaster. Annual Review of Sociology, 25, 271—305.
20 Mufson, S. and Kumblut, A.E. (2010, June 15). Ibid.
21 Urbina, I. (2010, June 25). Some call a new BP project risky: 'On-shore' rig is located on an artificial island, and oversight is criticized. International Herald Tribune, p. 2.
22 Giddens, A. (1991). Modernity and self-identity: Self and society in the Tate Modem age, London: Polity Press.
23 Gould, S.J. (1987). Opening remarks of the conference on evolutionary progress at Chicago's Field Museum. Quoted in Lewin, R. (1999). Complexity: Life at the edge of chaos, Second Edition. Chicago, IL: University of Chicago Press, p. 139.
2
Features of Drift
In the early afternoon of January 31, 2000, Alaska Airlines flight 261, a McDonnell Douglas MD-80 took off from Puerto Vallarta in Mexico, bound for Seattle. The pilots had just taken over the airplane from the incoming crew, who had nothing special to report about the status of the airplane.1
A bit into the flight, the pilots contacted the airline's dispatch and maintenance control facilities in Seattle on the radio. This was a shared company radio frequency between Alaska Airlines' dispatch and maintenance facilities at Seattle and its operations and maintenance facilities at Los Angeles International Airport (LAX), They had run into a pretty serious problem: the horizontal stabilizer, which helps control the aircraft's nose attitude while in flight, appeared to be jammed.
"Understand you're requesting diversion to LA?" Seattle maintenance asked the pilots at 3.50 p.m. "Is there a specific reason you prefer LA over San Francisco?"
"Well a lot of times it's windy and rainy and wet in San Francisco and uh, it seemed to me that a dry runway, where the wind is usually right down the runway seemed a little more reasonable."
A few minutes later, a dispatcher from Seattle provided the flight crew with the current San Francisco weather. The wind was light, out of the south (180 degrees), and the visibility was good (9 miles). But, the dispatcher added, "If you want to land at LA of course for safety reasons we will do that, we'll tell you though that if we land in LA, we'll be looking at probably an hour to an hour and a half we have a major flow program going right now," referring to air traffic control restrictions that would make it hard to get the aircraft out of Los Angeles again.
"I really didn't want to hear about the flow being the reason you're calling us, because I'm concerned about over flying suitable airports," the captain replied. He did, however, discuss with his first officer potential landing runways at SFO, and finding a discrepancy; "One eight zero at six ... so that's runway one six what we need is runway one nine, and they're not landing runway one nine."
"I don't think so," the first officer replied.
The captain then asked Seattle dispatch if they could "get some support" or "any ideas" from an instructor to troubleshoot the problem.
He received no response.
"It just blows me away," the captain then said to his first officer, "they think we're gonna land, they're gonna fix it, now they're worried about the flow. I'm sorry, this airplane's not gonna go anywhere for a while. So you know."
"So they're trying to put the pressure on you," a flight attendant replied.
'Well, no. Yeah."
The Seattle dispatcher had not given up on San Francisco. He informed the flight crew a few minutes later that the landing runways in use at SFO were 28R and 28L and that "it hasn't rained there in hours so I'm looking at ... probably a dry runway."
The captain replied that he was waiting for a requested center of gravity (CG) update (for landing), and then he requested information on wind conditions at Los Angeles. The dispatcher replied that the wind at LAX was out of the west (260°) at 9 knots.
