|
APOCALYPSE WHEN?:Determining and Comparing Catastrophic Risks Steven D. Dolley, University of Vermont 1986
- Fertile Ground : The Agriculture Debate |
|
"We teeter, teeter, teeter, teeter on the
brink of . . . of something I can't even say. The fuse is lit and it could blow any day! " - Bobby Weir In the 1985 Debater's Research Guide, Danny Povinelli did some admirable on the too long neglected subject of risk analysis in academic debate. 1/ While recognizing the vital importance of considering and analyzing apocalyptic risks, Povinelli lamented that the strategic utility of propagating as many scenarios for world disaster as possible often crowds out cogent analysis of the likelihood and relevance of such risks. Few with any amount of exposure to recent academic debate in either college or high school can reasonably dispute this thesis. Techniques of risk analysis, though extensively discussed and developed in the field of policy making, find only the most skeletal expression in our activity. I hope to extend on Danny's fine theoretical treatment by suggesting some practical applications. The world we live in is indeed a troubled one. Countless crises arise in every area of our daily existence. Given modern communications technology and the interdependence of every person on the Earth, even apparently limited problems can have profound and dangerous planet-wide implications. More than ever, it is imperative that the truly concerned "think globally and act locally." Povinelli stressed, and I strongly agree, that the ostrich-like response of arbitrarily choosing to ignore apocalyptic risks is no solution. Besides being unworkably subjective - when does a risk become a generic meatball rather than a relevant policy concern? - this approach risks teaching the leaders and decision makers of tomorrow who are debating today that they can safely neglect threatening second - and third-hand effects of an action, merely because such effects do not reveal themselves as intuitive or apparent at the first glance. Therefore suggestions such as that made by Thomas A. Hollihan 2/ regarding turnarounds - that judges should unilaterally discount counter intuitive positions as contrary to common sense - are, to put it politely, misdirected. What we need is not judges wearing blinders to keep them focused on somebody's idea of the core issues of the topic, but debaters whose analytic ability allows them to determine which risks are salient and which are not, and persuade their auditors that they are correct. As is true of most issues, the determination and comparison of risks ideally should be done by the debaters in their speeches. Not only does this allow the participants to learn more about the issues being discussed, it also minimizes the type of intervention likely to spark grumbling that the judge is "debating the round over again" for her or his self. The following collection of hints is hardly exhaustive, but I hope that it will enable rebuttalists to do more than sputter "We outweigh!" or - my least favorite plea - "We have five nuclear wars; they only have four!" WHAT IS RISK? HOW IS IT DETERMINED? Many flaws in debate risk analysis stem from basic misunderstandings about the nature of risk itself. Simply put, risk analysis is a method, formal or informal, of determining the seriousness of a given hazard, and the relative importance of avoiding it. One often-ignored fact is that risk is a function of both the harm magnitude and harm likelihood. Similarly, a hazard with no likelihood of occurrence, regardless of the magnitude of its harm, carries no risk. Many speakers attempt to win debates by conjuring up images of global catastrophe. Frequently they choose to ignore or misrepresent the true nature of risk, relying instead on the type of subjective psychological process documented by Paul Slovic and his colleagues. Slovic et al. contend that "the very discussion of any low-probability hazard may increase the judged probability of that hazard regardless of what evidence indicates." 3/ This is so because "from a statistical standpoint, convincing people that the catastrophe they fear is extremely unlikely is difficult even under the best conditions. Any mishap could be seen as proof of high risk, whereas demonstrating safety would require a massive amount of evidence." 4/ Realizing that the catastrophes they caution aga7inst have essentially no likelihood of occurrence, they will frequently substitute some of the following assertions and labels for coherent risk analysis. 1. The impact is absolute. Any risk is enough. Many disasters such as a nuclear war or total destruction of the Earth's ecosystem are likely infinite-magnitude harms, in that they would engender the extinction of humanity and most other forms of life. Any likelihood of such an infinite-magnitude harm creates an infinite risk, since infinity times anything greater than zero is infinity. Disarmament activist Johnathan Schell gave an eloquent summary of the "any risk" position, which sounds hauntingly familiar to those who have heard dozens of second negative rebuttals.
Mathematically this argument is valid, but it becomes difficult or impossible to apply in practice. First, it is a rare debate indeed where only one team is claiming to avoid an infinite-magnitude harm. This will be especially true on this year's agricultural policy topic. When both teams claim to avoid total catastrophe, the likelihood of occurrence must be sketched sharply enough that likelihood can be compared. After all, what can a judge do if either choice s/he is presented with carries infinite risk? Under such circumstances, why not just get drunk and wait for the mushroom clouds? Second, the risk analyst must discern if any likelihood at all adheres to the infinite-magnitude harm. The logic of the "any risk" position itself requires this, since even infinity times zero equals zero. Quite frequently, scenarios for global destruction are so poorly documented, extended, or linked to the issue at hand that essentially no likelihood of occurrence exists. Certainly it would be cavalier and foolhardy to ignore apocalyptic threats, but it seems equally silly and futile to live our lives cowering in fear of a million imaginary Rube Goldberg-like disasters that have no chance of occurring. If I get up now, walk over to my cassette blaster and pop in a tape, there is an ephemeral chance of my being electrocuted even before Jerry and the boys begin to play "Sugaree." After all, my blaster is crackling with 120 volts of alternating current. Merely touching it could, if there were a short circuit, end my career as a forensic wise guy in one big blue flash. Nevertheless, I routinely discount that minimal-likelihood hazard. Its risk is so low as to be ignored. So, excuse me for a moment; the music has stopped. 2. The impact is linear. This popular slogan contends that the danger being discussed is incremental. Each amount of a given action causes a proportional amount of the harm to occur. Second negative rebuttal assertions to the contrary, infinite-magnitude harms cannot be linear. Either we are all dead or we are not all dead. Of course many actions would cause some of us to die while others lived, but this is a harm of finite magnitude. The questions then become: How many of us would die, and how likely is this to occur? Unemployment harms are an example of a finite, linear impact. Some studies have contended that each one percent increase in the U.S. unemployment rate kills thousands of people via stress-related diseases such as hypertension, mental illness, alcoholism, etc.. 6/ The mere fact that an impact is linear does not absolve a debater of the duty to attach a level of probability to it. We must deduce what amount of harm will be triggered, and what likelihood exists that the triggering incident will occur. 3. The risk is linear. This claim is quite different from claim two. Debaters confused about the definition of risk will often use these arguments interchangeably, however, creating yet more confusion. When one says that a risk is linear, s/he means that each occurrence of a given action compounds the likelihood of the harm's occurrence. For example, a small slit on your finger may draw blood, but it is unlikely to cause you to bleed to death. Each time you add a slit while the others remain unhealed, you are compounding the chance that you will suffer a fatal hemorrhage. Since the damage is cumulative, you would do well to avoid bleeding slits on your body as a general rule; the risk of your death, which for you Is an infinite-magnitude harm, is linear. A global policy example might be the practice of sending spy planes, there is a possibility that they will be mistaken for attacking bombers, triggering an accidental nuclear conflict. If that is always true, then each and every flight increases the risk of nuclear war. In a way, the claim that "the risk is linear" begs the question. One must still figure out by what amount each action increases the likelihood of catastrophe, in order to treat it with the seriousness it warrants - and no more. 4. We are on the brink now. The claim here posits that the current situation is so treacherous, so close to the precipice of the catastrophe being discussed, that any additional occurrence of the causal action will set off the infinite-magnitude harm, in all its destructiveness. The economy provides some good brink examples. A disadvantage could prove that interest rates for loans now average 13.9 percent. Coupled with evidence providing that interest rates of 14 percent would destroy the U.S. farm sector by denying it access to needed loans, this would document a very low threshold for a loan disaster. (Don't quote these figures by the way - I make them up as an example. They're probably way off base.) Proving we are on the brink of a given catastrophe is often extremely difficult. Quantified thresholds of the type exemplified above are the most clear and make for the most accurate risk analysis, but such exact evidence is hard to locate. However, the appropriate remedy is meticulous and extensive research, not unproven thresholds or second-rate risk comparison. 5. This is a priori issue. It must be considered first. This argument is completely different from the previous four. The Latin phrase a priori literally means "from the one before." As is often the case the term "solvency" jumps to mind - debate blithely ignores literal meaning. (Topicality sticklers take note!) When a debater claims an issue is a priori, s/he is positing it as a sort of decision rule (another popular and abused catch-phrase). S/he is claiming this issue must be considered before any others, and not traded off against competing considerations. 7/ The underlying assumption in a claim that an issue Is a priori is that the value in question is a paramount one, always more important than other considerations. This assumption itself is what the astute debater will subject to risk analysis. Suppose a second negative accused a plan which mandates the use of motorcycle helmets of inhibiting the cyclists' freedom of choice, and reads evidence saying, "Personal freedom is paramount; nothing can be more important; no practical considerations justify an infringement on freedom of choice." If this decision calculus holds, the affirmative's claim that saving hundreds of lives each year is far more important than Joe Honda's preference is irrelevant. What must be challenged is the assumption that personal freedom is always most important. What is the harm of neglecting this decision rule, and what is the likelihood that such harm will occur? In conclusion, decision rules always contain value assumptions, and these assumptions need to be unearthed and analyzed when disputing a so-called a priori issue. EVIDENCE ANALYSIS Often debaters become frustrated with opponents who attempt to avoid intelligent risk analysis by reading an armload of quotations which purportedly prove that they will avoid some horrible global holocaust. Indeed, it is not too difficult to gather an imposing batch of powerfully-worded warnings from many sensationalistic scholars, and some cautious ones as well. Those debates who can overcome their inbred faith that a name and date make a statement holy writ will want to consider the following problems in evidence discussion catastrophic risks. 1. Qualifiers. These are words and phrases such as may, possibly, under some circumstances, could conceivably, etc.. Qualifiers limit the scope and strength of the statements that contain them, and card-worshipping debaters often observe an unspoken agreement to ignore them. Hence, many speakers will not challenge an opponent's label "X guarantees nuclear war," even when the card below that label reads "X could conceivably, in certain circumstances of rising tensions, increase the admittedly slight chance of superpower conflict." Most likely they are afraid that such challenges would lead to uncomfortably close scrutiny of their own overclaimed evidence. Authors do not insert qualifiers in their works for decoration. They are used to show that the conclusion is not a rash overgeneralization, but only probably true, or possible, or true under some circumstances. Ignoring qualifiers misrepresents the source's analysis by attaching a greater likelihood of occurrence to the event discussed than the author believes in proper. 2. Scenario assumptions. Since one cannot read an entire book or article into a speech, short quotations are trimmed from the whole to document a specific point. While this can be done without distorting the authors conclusion if caution is exercised, the entirety can never be completely conveyed by two or three isolated sentences. if it can be, the author has certainly wasted a lot of effort writing the rest of the world. Frequently quotations will be the conclusion of a lengthy section of analysis, several pages or even an entire chapter, which postulates numerous assumptions and disputable prior conclusions en route to the final contention. Suppose your opponent reads a quotation from Professor Unit of the Food Study Center which says, "U.S. food aid to the starving masses overseas will kill ten times more people than it saves."8/ Certainly a devastating indictment if true - but why would this occur? Perhaps Professor Unit's position is that food aid increases population growth by keeping more people alive, putting pressure on limited resources and leading to a Malthusian catastrophe somewhere down the road which would starve many more people than are hungry now. Perhaps he feels that food aid lowers food prices overseas, destroying the incentive for local farmers to grow crops and precipitating a disastrous food shortage. Perhaps he is worried about accumulated grain dust in storage elevators, which could spark a series of fatal explosions. Who knows? Only Professor Unit, and those who have read his book. If you do not wish to wind up boxing shadows, it is important to probe for hidden scenario assumptions. 3. Rhetorical flourishes. Always eager for that killer card, debaters are willing to ignore Aristotle's distinction between logos, the appeal to logic, and pathos, the appeal to emotion. 9/ Many speakers and writers make extravagant pleas to the gallery which, while not necessarily lies, may overstate their cause in an attempt to persuade. When a senator says that a vote against aid to the Nicaraguan rebels is a vote for Communist tyranny, or a representative argues that voting to approve air raids on Libya is tantamount to pushing the nuclear button yourself, we are witnessing this phenomenon. To persuade his readers of the evils of discrimination, sociologist Philip Carey once wrote that racial and sexist discrimination pose as great a danger as nuclear war, because all those phenomena threaten the breakdown of social cohesion. His comparison was powerful, but I don't believe he meant that racism and sexism will leave billions of people dead under radioactive rubble. Keep an eye open for quotations that likely popped out in the heat of oratory and bear little or no relation to reality. Your opponents will win rounds with such cards if you let them. 4. Unclear Metaphor. Somewhat to rhetorical flourishes, this problem arises when impact cards describe a harm metaphorically. 10/ When evidence says certain actions will cause a catastrophe, holocaust, apocalypse, Armageddon, or even Ragnarok, what exactly does it mean? What happens, and how many people die? Strongly encourage your opponents to clarify vague metaphorical referents in their impact evidence. If you feel that their explanations overclaims the evidence, press for additional evidence from the same source defining the disaster image used. 5. Ballpark figures. Sometimes authors will round off numbers so that they look nice and even. Sometimes they will attempt to attach numbers to phenomena that cannot be quantified. Either action can produce the notorious ballpark figure. A certain biology professor who writes on a wide variety of issues is fond of warning that certain types of ecological damage will lead to the deaths of one billion people. It seems surprising that such a precise figure can be attached to the consequences of ozone layer depletion. As Alice would say, it's curiouser and curiouser when the very same round figure turns out to be the fatality toll predicted for Southeast Asian monsoon diversion, population growth-induced famine, and extinction of endangered species as well. 11/ Another writer, not to be outdone, decided that the trend toward a colder global climate would kill two billion people. 12/ This is not to say that these-authors, or other-writers postulating huge numbers, are necessarily wrong. Nonetheless, you would do well to ask opponents reading such evidence to explain how the author reached her or his conclusions, and exactly how the number quoted was calculated. By the way, don't be scared off just because the number was pulled out of a computer rather than magician's top hat. As any hacker will assure you, the "GIGO" (Garbage in, garbage out) principle holds true: these machines' conclusions are no better than the quality of their input data and programming. In the words of the immortal Mr. Rogers, "People are much more important than machines. Much more!" 6. Unquantified catastrophes. This is the opposite of the "ballpark" problem. When evidence talks about famine, war, economic depression, plague, pestilence and so forth, we know what it means, but we do not get a clear sense of the scope of the disaster. A nuclear war could well kill everyone. On the other hand, no one has died since 1945 in the legally unended "war" between Japan and the Soviet Union. A "famine" means people are going hungry, but does not tell us how many, or how hungry they are. Seek a more precise delineation of the extent of damage when such inexact words are employed. 7. The ever-popular composite catastrophe. Say your opponent reads a card from source A saying your plan will cause an economic depression, a second card from source B saying depression causes famine, and a third card from source C saying famine causes nuclear war. Have they thereby proved that your plan causes nuclear war, or merely come up with an original but low-credibility fairy tale that neither A nor B nor C would find believable? Povinelli captures this problem well with his image of all the authors of the evidence from one disadvantage meeting at a coffee shop, startled by the bizarre conclusion drawn by some second negative from their highly diverse and often incompatible scholarship. 13/ The mere fact that the evidence is from different sources does not prove the argument invalid, but it does raise the possibility of inconsistency. A sharp debater will explore this possibility, either in cross-examination or by incisive rebuttal presses. THE BIG PICTURE Unfortunately, almost all of us succumb occasionally to what former Dartmouth debate great Tom Isaacson has termed "lines-and-arrows debating." This is the tendency to view each line of argument in isolation, never stepping back to note how the issues relate and interact. Lines-and-arrows debaters can be quite proficient technically, answering each of their opponents- responses and saving that crucial five seconds at the end of their last rebuttal to gasp, "Overall, my partner smashes case. We win D.A.s one and two cold; they outweigh. Vote Neg." While such terse eloquence may allow a lines-and-arrows speaker to muddle through and achieve moderate success, a more consistently effective strategy accounts for the synergistic relationship of various positions and scenarios. Though this rare skill must be acquired rather than taught, a few tips seem in order. 1. Seek a holistic perspective. At the most fundamental level, this requires that you make sure your positions are consistent. Do not claim that the plan increases the federal budget deficit and destroys the economy, if you are also claiming that economic growth is bad. Be sure the first negative positions are consistent with your various disaster scenarios as well. For instance, you should not argue that a portion of the plan will not work, if one of your disadvantages depends on the success of that provision to gain its link. 2. Account for systemic influences. Policy systems such as affirmative plans and counterplans would create a new situation if implemented. Various effects discussed in the round may not account for the influences, positive or negative, that such new systems create. Perhaps an agricultural policy of bureaucratic streamlining within the Department of Agriculture would enhance administrative planning and foresight, allowing the agency to avoid disastrous effects that such overhauls might otherwise entail. This would be an example of a positive systemic effect. A possible negative systemic effect might be the plan's usurpation of resources - money, personnel, or what have you - that otherwise would be available to ameliorate the effects of a disadvantage. Bear in mind that a plan is not merely grafted onto the present system which otherwise remains static. Once the plan is added, feedback and ripple effects are bound to transform the whole in both good and bad ways. As the ecologist tell us, we can never do just one thing. Examine the assumptions and causal linkages in arguments opponents offer against your system; these assumptions and linkages may not hold true or operate the same way under a new and different system. 3. Be award of time frame questions. To barrow common second negative lingo, different effects have different "fuses." A spending disadvantage may occur almost immediately, or at least as soon as the money is allocated. A population growth disaster, on the other hand, may take many decades to reach the point of planetary catastrophe. Other things being equal, we should be more concerned about those crises that threaten us most immediately. Of course, such a rule should be tempered by the considerations previously discussed, such as magnitude of harm and likelihood. But it makes sense to lose more sleep over a threat that could explode the world this week than about one that will not arise for another twenty years. This is true not only because the long-time frame gives us more opportunity to avert the disaster, but also because the time before occurrence is intrinsically valuable, given we are alive during it. Faced with the choice of being shot in the head tomorrow at dawn or being shot in the head thirty years from dawn tomorrow, most of us will take the latter option. 4. There are several independent levels of probability in every chain of events. Most complex disaster scenarios assume many effects all occur, culminating in destruction. For instance, the population growth or Malthus position assumes that food aid decreases mortality, that these additional people put intolerable strain on world resources, that this strain triggers some sort of devastating tragedy, and that the tragedy will be greater in magnitude than the current tragedy of starvation. Each assumption has its own likelihood of occurrence which must be assessed. The likelihood of Malthusian disaster is a cumulative likelihood, that is, the likelihood that all these effects will independently occur. The cumulative likelihood will almost always be much lower than any of the separate levels, and an accurate risk analysis will need to account for this. CONCLUSION Apocalyptic debate has based its success on Yeats' poetic prediction that "the center cannot hold." Indeed, "mere anarchy is loosed upon the world" in most every constructive speech - sometimes literally, if the speakers are Kropotkin fans. Be that as it may, it cannot be true that every action is malignant, that all paths lead off a cliff. Michael Mankins recently wrote of a metaphorical crossroads, one fork leading to doom, the other to Nirvana. 14/ Though it is not as simple as that - as Mankins would no doubt agree - all but the most hardened nihilists can conceive of benign courses of action that we should pursue. Presumption of unknown horrors must not paralyze our lives, as Richard Berk emphasized: In all this discussion about the dangers and risks of policy making, the logical conclusion seems to be that we should recommend no changes at all, because there will always be unforeseen dangers, circumstances and problems. In being progressive there is always the risk that we might mess things up. But that is part of the price a society pays when it tries to make things better. If we worry about all of the terrible things that might happen, we will never do anything. 15/ Debate is one of the most promising forums for exploring such positive courses and possibilities. Exacting risk analysis, biased neither toward rosy optimism nor toward apathetic defeatism, can help us avoid the many catastrophic risks we all fear and discuss so often. ---------------------------------
1/ Danny Povinelli,
"Catastrophe Without Cause: Escaping the Paradigmatic Disadvantage,"
Debater Research Guide 1985: Clarifying Water Policy, pp. 4-9.
One noteworthy exception is this neglect of risk analysis in the context
of academic debate in Vincent Follert's well-documented article in the
Journal of the American Forensic Association, Fall 1981. |