SPEECHES
Print This Page | Email This Page


Joint Session AEI-Brookings Institution
Washington, DC
January 25, 2005

We need better mechanisms to determine science policy. Michael outlined several issues before a joint meeting of liberal and conservative think tanks.

You can also watch a video of the speech.



As some of you may know, I have spent the last several years exploring various environmental issues, particularly global warming. I have been deeply disturbed by what I have found, largely because the so-called evidence for so many environmental issues is often shockingly flawed and unsubstantiated.

But more troubling, to me, is the degree to which the political process seems to have captured and often corrupted the integrity of the scientific research that is used to formulate policy, and inform policy decisions.

I am also troubled by the insensible and distracting contentiousness that seems to inform so much of current political debate - especially when environmental issues are involved. As a result of this political friction - which is all heat and no light - policy is often established by way of litigation, rather than negotiation and legislation.

From these observations, I conclude that as a society we lack the tools and methodologies we need to resolve thorny science-policy issues promptly, equitably and constructively.

We're having this trouble because we have not developed mechanisms for decision-making that we all agree are fact-based and judicious; so that the results of such processes will be generally perceived to be fair and equitable.

As a result, as I mentioned, we often resolve environmental disputes through litigation, which is neither good public policy nor a sound basis for administrative rule-making.

So we ought to establish new mechanisms for determining social policy. And I believe that in the not too distant future, we will.

Today I am going to focus on six major problems that will confront science policy in the 21st century, and then consider briefly how we might resolve each of them.

SLIDE: Six Questions for the 21st Century
How do we obtain good and unbiased information?
How do we set policy in uncertainty?
When do we choose to prevent, and when adapt?
How do we promote desirable technologies?
How do we regulate a knowledge society?
Can we manage complex natural systems?

Let's begin with the first:

HOW DO WE OBTAIN HIGH-QUALITY UNBIASED INFORMATION?


Traditionally policymakers have trouble getting good information. This problem is especially acute with scientific decisions, because the issues are complex and policymakers are not usually trained in science. In addition, the staffs feeding policymakers often give them deliberately-biased information in an effort to make a partisan case. In the process these staffs are doing us a double disservice. They are both preempting the policy maker's traffic cop role; and they are violating the integrity of the firewall that should always stand between those gathering information, and those setting policy based on it.

The issue here is not simply the avoidance of bias. The issue is how to avoid bad information. In areas of contention, critical and profoundly influential information is often stunningly flawed.

By way of illustration I'm going to discuss a recent example from climate science, and also show you a graph - the so-called "hockey-stick" graph. Many of you will be familiar with this.

Here's how the hockey-stick graph came to the public's attention.

In 1998, an American climate researcher named Michael Mann, along with his co-workers, published an estimation of global temperatures from 1000 to 1980. They arrived at this estimate by combining the results of 112 previous proxy studies. By "proxy studies" I mean tree-ring and isotope and ice core studies that are intended to provide an indirect measurement of temperature in the time before thermometers existed. Mann's results appeared to show a spike in recent temperatures that was unprecedented in the last one thousand years. As a result, his report achieved immediate and world-wide fame. It also formed the centerpiece of the U.N.'s Third Assessment Report, in 2001, of the Intergovernmental Panel on Climate Change.

Mann's assessment of the data was criticized on several fronts. The first was historical fact: his chart didn't appear to show the well-known medieval warm period, or the so-called little ice age that began around 1400. This his advocates explained away by saying that those were European but not global phenomena. That started a hunt through historical records in China and elsewhere. And now, I think, many people are inclined to believe that the sharp rise and equally sharp fall in medieval temperatures were, indeed, global phenomena.

The next chapter in the story began when two Canadian researchers, McIntyre and McKitrick, obtained Mann's data and repeated his study. They found numerous grave and astonishing errors in Mann's work, which they detailed in 2003. For example, two statistical series in Mann's study shared the same data. The data had apparently been inadvertently copied from one series to another. In addition, nineteen other series had had gaps in the data, which Mann's team had then filled in - a fact that had not been disclosed. In addition, all 28 tree ring studies had calculation errors - and so on and so forth. Such that in the end, the Canadians' corrected graph looked quite different:

The corrected graph suggests that the global temperature today is very far from the warmest it has been in the last thousand years.

But there were more problems to come. Mann's statistical approach to the data was somewhat unusual, and raised questions about the validity of the formula he had used to do his metastudy. When researchers tested Mann's formula, they discovered that a table of trendless numbers (generated by computer) would invariably create the hockey-stick shape.

Here you see a series of hockey-sticks. One is Mann's original graph; the others are created by sequences of trendless numbers. So it appears that Mann's formula will turn any data into a hockey-stick - and had apparently never been tested by Mann prior to its use in his study.

Mann's work has since been attacked by a number of laboratories around the world:

Slide
"An artifact of poor data handling, obsolete data and incorrect calculation of principal components."
Stephen McIntyre and Ross McKitrick, Energy and Environment, v 14, 6:2003.

"The graph contains assumptions that are not permissible...Methodologically it is...rubbish."
Hans von Storch, quoted in Der Spiegel

"A real shocker...the hockey stick, the poster-child of the global warming community, turns out to be an artifact of poor mathematics. How could it happen?"
Richard Muller, MIT Technology Review, 15 October 2004

Mann's work has been called "phony" and "a shocker" and "rubbish" by climate scientists who believe in global warming, and who are concerned that such sloppy work might undermine the legitimacy of the claim that global warming is a dangerous and alarming fact; as indeed it has undermined it-although I would say, very little.

But to my mind, the real point of the story is that the UN's Intergovernmental Panel on Climate Change, the IPCC, accepted Mann's study without question and without independent review. And therein lies the real warning to policymakers. Because even the most widely-touted and allegedly reputable studies may be significantly less reliable and substantive than they initially appear to be.

And if I may digress for a moment: one explanation for the public's credulity may be the assumption - widely held by non-scientists - that scientific reports are repeated and therefore verified by other labs. But the reality is that most studies are not. A very, very small number of them are verified. So one can't just assume that because a study has been published, it's accurate.

So bad data is out there, and severely biased studies are out there, and policymakers have the unenviable task of separating the wheat from the chaff. (You remember what Adlai Stevenson said, that newspapers have the ability to separate the wheat from the chaff - and then print the chaff.)

In that vein, I know of only three strategies that are useful in the effort to improve the rigorousness of data, and verify its integrity. Two of these strategies are already well-known and established. And the third will, in my view, soon become a reality.

I call the first strategy the "FDR Tactic" - let the participants air their views and slug it out.

Franklin Roosevelt was famous for inciting conflict and confrontation between his advisors. And he made sure that when his people fought the issues out, they did it in front of him.

In addition, he always maintained two different sets of advisors. The first was made up of the members of his Cabinet. The second was comprised of assorted friends, mentors and cronies - FDR's kitchen cabinet. The variety of views, prejudices and motives thus exposed made for an incisive and effective information management technique.

But the closest we now come to that level of inquiry, and its exposure of partisan bias and equivocal advocacy, is the dog and pony show of Congressional testimony. And those hearings tend to be more concerned with the scruples of deference than the investigation and determination of fact. They reflect the chair's obsession with process, not product. Questions are neither penetrating or challenging. Nor are the answers that members accept either instructive or informative. Congress lobs softballs at witnesses - questions designed to elicit a specific response that will prove the member's point, support the member's stance, and placate the member's constituents.

But this is a cruel farce. This is show-biz, not the people's business; this is vaudeville, and not democracy.

Far better for policymakers to create a forum in which opponents can engage in direct debate - the much-touted free marketplace of ideas. Insisting the debates be public is also a good idea, as sunlight always has a sanitizing effect. And a prolonged series of debates, in which opponents knew they would face each other again, would be extremely helpful.

Why? Because to cite just one example: there is at present no good public forum in which to debate and evaluate climate data, in an atmosphere of aggressive and penetrating inquiry, full of challenge and true debate.

A second procedure of the FDR variety would be to give grants for research to multiple laboratories at the same time. I really don't know why this isn't done. In areas where policy is very important, you don't give the research to just one lab. You give it to three, and you make sure that two of them are strongly opposed to each other.

All three should know they will have the right to inspect each other's data and procedures. All three should know that their results will be published in concert, simultaneously. Such simple procedures would make everyone clean up his act real fast.

Of course such a procedure is more expensive. But let's face it: bad information can also be very expensive, especially when it leads to bad policies.

But more to the point: the notion that a single study by a single research team can be used to set policy is really outdated. We just can't do it any more.

And I want to make the argument-I think it's already true in many areas-that government-funded data ought to be publicly available, except in circumstances where privacy issues take precedence. Otherwise, I don't see why data isn't on the net at the time of publication. I don't see why anybody has to sue a laboratory to release its data. The public has paid for it, the public owns it, and the public has an absolute right to access it.

It strikes me as odd that in terms of availability and accuracy of data, we are ready to hold the heads of corporations to a higher standard of conduct than we do the heads of laboratories. We're sending corporate heads to jail, with the clear understanding that it will have a bracing effect on other CEOs. Frankly, if at some point we sent the head of a laboratory to jail, it would probably have a similarly bracing effect on the management of information in other laboratories. And I suspect that sometime in the 21st century, that will happen.

Let's move on.

A second method of securing reliable data is one we might call the "FDA Strategy" - a methodology aimed at systematically removing all bias from the process that gives us data we wish to use. I know the FDA is having some troubles at the moment. We can speak of them in a sort of idealized way. The core FDA procedure is the requirement for double-blind studies of drug efficacy.

Let's review what a double-blind drug test is. The drug and the placebo are bottled by one group. A second group-that does not know the first-administers the drug to patients. A third group evaluates the patients. A fourth group tabulates the results. None of the groups ever meet. They're in different cities and preferably different countries.

We know from experience that this is what we have to do to get bias out of the system. But many areas of research are not held to such rigorous standards. And I can tell you that if there were a double-blind assessment of climate models, the global warming debate would have been over yesterday. I can tell you further that if a blue-ribbon panel of disinterested non-scientists were convened to review the global temperature record, we would also witness a swift end to the current debate. Why? Because at the moment climate science is an insider's game, and serious outside scrutiny has never taken place.

I find this inexplicable. We're talking about spending trillions of dollars to control carbon emissions on a global scale because computer models of climate predict a dangerous future. And yet nobody is willing to subject these climate models to the kind of rigorous testing that we require to license a drug.

But so it goes. Let's move on.

There is third method for vetting data that is on the horizon. It's not here yet, but I am convinced it is coming. I'm talking about product liability for information.

We live in an information society. Nearly one American in three is a knowledge worker. More people are knowledge workers in this society than are engaged in manufacturing. By and large, what these knowledge workers do is generate information. And our society is totally dependent on the integrity of information. Yet we still do not define information as a product. And as a result it has evaded the Quality Revolution that has transformed other industries.

But product liability is already enforced for maps and charts, and will soon be applied to other information products as well. It's absolutely essential for the future.

Now let's turn to the next question:

HOW DO WE SET POLICY IN UNCERTAINTY?

Samuel Butler said that "Life is the art of drawing sufficient conclusions from insufficient premises." And that's even more true today than it was when he said it, 180 years ago. Because many science policy decisions cannot be made with confidence-not because the data is biased, but because the data doesn't exist at all.

An example that I will speak of only briefly-because a lot of people in this room know about it, or have worked in it-is arsenic regulation. Arsenic is a naturally-occurring carcinogen found in three parts of the US--the Northeast, Michigan, and the Southwest. It thus affects only a fraction of the population.

Since the 1920s the arsenic level has been set at 50 parts per billion. But new data from Taiwan, where arsenic levels are extremely high, suggested to many people that the US level should be lower. The Clinton Administration set it at 10 parts per billion shortly before leaving office. The Bush administration thought it should be 20; it went to the National Research Council, which suggested it ought to be only 3 parts per billion. I believe at the moment it is 10 ppb, and there has been a good deal of litigation.

My point here is that based on the current data on arsenic, a good case could be made for 50, 20, 10 or 5 parts per billion. We don't have decisive data telling us what to do. We know the cost differential between setting a level at 20 ppb and 3 ppb is about three quarters of a billion dollars.

The argument that I would make is that simply because we don't have the information is not a reason for us to think we can't get it. We might reduce contentiousness if we set a policy and simultaneously initiated an epidemiological study to tell us if the policy was correct.

Now, in the case of arsenic, this is going to be a long-term study. Arsenic cancers develop late in life; many are not fatal, so we're talking about a hundred year study. But so what? And if it costs you a million dollars, it's still a bargain. Because it's a lot better to spend a million and set your level at a certain higher point and review it again in 20 or 30 years, than it is to commit to 750 million dollars now. Because, as we know, there is a cognitive illusion that all human beings demonstrate and that is our willingness to spend vastly more for what we regard as 100 percent safety than we will for 95 percent or 98 percent safety. It's a very irrational and ineffective tendency, but it is in our brains and very well demonstrated in the psychological literature. As policy people, we have to fight this tendency.

Linking a regulatory level to a research program implies more flexibility in review of standards, but that already happens, in a quiet way. Sixteen years ago, when I bought an old house, I had men in spacesuits taking the asbestos out. It required two weeks. Last year, when I bought an old house, the asbestos was removed in a couple of hours without any spacesuits at all.

Let's move on to the next question:

WHEN DO WE PREVENT, AND WHEN ADAPT?

This is a very contentious issue, especially in environmental circles where the precautionary principle is popular.

There are some instances in which prevention is obviously the best strategy: oil slicks, radiation leaks, exposure to lead and pathogens. Thus the conversion to double-hull tankers and vaccination. It all makes sense. But inevitably things still go wrong. And when they do, we adapt. Indeed, we're so adaptable that we often demand a crisis in order to address the root cause of a problem.

Whatever its putative virtues, the precautionary principle is, I notice, erratically applied. Many of my friends who want to label or ban genetically modified foods because they have not been adequately tested, communicate with fellow advocates by cell phone, even though cell phones haven't been adequately tested. Certainly they've never been proven safe.

So, over time, I have actually developed an affinity for adaptation, as opposed to prevention, both as a coping mechanism and as a policy predicate. I believe it's the better strategy. As Mark Twain said, "I've seen a heap of trouble in my life, and most of it never came to pass."

In addition, I remind you of the work of the late Aaron Wildavsky, who argued, in a very complicated analysis that you can find in "Searching for Safety," that the strategy of prevention favors the elite; adaptation favors the average person. Certainly if you look at who is advocating which strategy, it seems clear that Wildavsky was right.

HOW DO WE PROMOTE DESIRABLE TECHNOLOGIES?

My answer is, we don't. This is a tough and expensive lesson. I don't know why we haven't learned it. We've had two major crash programs in technology that were hugely successful. The first was the Manhattan Project leading to the atomic bomb, the second was Kennedy's push to land men on the moon within ten years. Neither involved broad scale changes in the society or the adoption of some kind of new consumer product.

As far as I know, every other program to promote technologies has failed. Whether it is Lyndon Johnson's war on cancer, or California's attempt to legislate a threshold number of electric cars on the road by 1998 - or 2004 - government cannot and should not presume to mandate the creation of knowledge, the creation of consumer preference, or the creation of mass markets.

Indeed, since governments are notoriously inefficient, it's odd that they should ask others to do what they cannot. My sense is that this kind of legislation is usually a cover up for a failure of government to act.

In the case of California, the issues are quite stark. California has a number of options to reduce emissions and improve air quality. It can raise gasoline prices; it can impose luxury taxes on gas guzzlers; it can get older cars off the roads; and it can formulate land use policies that would discourage long commutes. But as Wall Street auto analyst Maryanne Keller pointed out, California had the guts to do none of these things. Instead, the legislature demanded electric cars; and when that mandate didn't transpire, they cried conspiracy. But the truth is that consumers simply didn't want electric cars, at least not the ones that manufacturers could build at that time. And today, the waiting list for hybrids-a car consumers do want-is six months.

This is, as I said, a lesson that we can't seem to learn. I think probably everybody understands that legislators ought to specify outcomes and not procedures. Yet time and again, they decide they want to be in the car business. Or some other business.

Let's move on.

HOW DO WE REGULATE A KNOWLEDGE SOCIETY?

This to me is the most critical of all the questions discussed today, in the sense that it seems the least examined. When we look back at concepts of the future in the early 20th Century, you find they almost always involve the assumption that society will be centrally planned - the economy centrally-controlled - and basic goods and services such as housing and transportation centrally designed and distributed.

Here is a still from "Things to Come," a 1930s movie, and it shows, along with a Roman vision of elite future fashion, a clear presumption of central planning, inherent in every aspect of the image.

But by "Blade Runner" in the 1980s, a different image of the future had emerged - a hodge-podge city that had grown organically, and was full of chaotic disconnects. It envisioned an Asian model of urban growth, and indeed many urban landscapes today look as if they are right out of Blade Runner.

The model of central control assumes a variety of conditions that no longer hold in the real world. Arguably you can attempt to control quality in a fixed manufacturing plant. But we have learned sadly that we can't control the technologies themselves, even one as complex and expensive as nuclear power. Control is even less likely in the case of bioengineering, which can be done in a garage with equipment costing a few hundred dollars. What, really, are we going to do about genetic engineering? What can we do? Once the rhetoric stops, this is a difficult problem that we have yet to engage. It is very possible that there is nothing we can do to control it.

What, then, is the role of central planning in a technological society? I would argue that, increasingly, there will be no role; and that the very concept of central planning and control is an anachronism of the industrial age of the 19th century. It will vanish completely in the 21st century.

And so we turn to my final question:

CAN WE MANAGE COMPLEX NATURAL SYSTEMS?

This is the issue of greatest concern to me, personally. The intellectual basis of my problem with most environmental thought is that it's enormously out of date in terms of science. If you date the start of environmental awareness with the first Earth Day in 1970-just to put a number on it-it is clear that since then, science has changed phenomenally, as a result of our new understanding of nonlinear dynamics and complex systems.

We used to believe there was something called the balance of nature. And this balance of nature-a Greek notion-informed almost all our thinking. It lay behind the notion that the way to preserve nature was to get people out of it. If you could remove people from natural places and leave those places alone, everything would be fine.

It is now quite clear that that's not true at all. And that there is no balance of nature. There never was a balance of nature and if you leave a forest alone, what will probably happen is it will grow, decay, become filled with pests and burn down with such heat that it will sterilize the ground and something completely different will grow up in its place.

We also know that the landscape white men saw when they first came to the New World was something they didn't understand at all. They thought they were seeing wilderness, but in fact it was a landscape entirely altered by native peoples who were there at the time. They were setting fire to the plains. They were burning down old-growth forest. There is more old growth forest in California today than there was in 1850. The Indians didn't like old growth because it didn't support enough game. So they burned it down.

What visitors saw as a natural world had been very much managed by a people who were true students of nature. And we've come in and imposed our old ideas. We don't live in nature any more. We live in cities and we don't really know anything about nature and those of us who even go camping are in for a huge shock. So we're imposing intellectual notions on a landscape that, by the way, doesn't care what we think of it. It doesn't respond to our thoughts.

And the result has been disastrous. If you read Alston Chase's history of Yellowstone Park, "Playing God in Yellowstone," it's a horror story and it remains a horror story for a hundred years. We now have raw sewage bubbling out of the ground in Yellowstone. We must be doing something wrong.

We need to understand that, left alone, wilderness changes, often catastrophically. And if that's the case, the whole relationship of man-to-nature is revised. This becomes a very difficult problem for present environmental groups to address because it turns so many prior conceptions upside down.

It implies that if we are going to have a picture-book wilderness-if we're going to have a landscape that's good for hiking or fishing or for viewing wild animals-then we are going to have to shape it. Yellowstone Park was so beautiful that it was made our first national park. But it was beautiful precisely because the Indians hunted the elk and moose to the edge of extinction. That's why it looked so good.

And as soon as the white men came in and stopped the hunting, they changed the ecology, and triggered a long cascade of errors continuing decade after decade, and ending in raw sewage. Although the end is not in sight.

Because of our ineptitude, some people have asked whether human beings are capable of managing complex systems. There's been a lot of philosophical discussion of that. The German psychologist Dietrich Dorner actually did research. And he found that yes, we can manage such systems. This is what he did.

He built computer models of a number of environments-a Saharan cattle environment, a town in Maine, and so on. And he brought in various academics to run these environments for a period of ten computer years.

What happened is that almost everybody ran the environment into the ground and made it vastly worse than it had been when they started. Only a few people were successful. So, Dorner looked at what strategies distinguished the people who succeeded and how they differed from the strategies of the people who failed.

And what he found was that the people who succeeded waited a little before they started. They gathered information. Not too much, but some. Once they began, they looked for unexpected consequences. They looked for something to pop up at the edge of their field of view, something they didn't think was going to happen. As time went on, they made more and more decisions. They were increasingly interactive.

The people who failed came in with a philosophical point of view. They established their procedures based on what they believed and then they left the environment alone. As their plan started to fail, they blamed whatever unexpected conditions arose. They also made fewer and fewer decisions because, by now, they were in a failure mode. It doesn't work to manage complex systems according to a philosophical view.

Now, I would say that we all know this. Because the one complex system that almost all of us have knowledge of is children. And anybody who just applies Democratic or Republican principles to their kids and then walks away will soon be visiting those kids in jail. Complex systems require continuous interaction. You have to be watching constantly; you have to be adjusting and revising; as the system does this, you do that. And it's a never-ending process-these days, really never ending. (A friend of mine once said that it's only after your kids get out of college that they really start to cost you money.)

It's Dorner's conclusion that there are strategies to manage complexity: that they can be learned; they can be taught; but they're not necessarily natural to us. And I think the final consequence of this new way of looking at the environment that I'm talking about is that it's going to be stupefyingly expensive. It's going to cost a lot of money to manage 2.2 million acres in Yellowstone and make it look good. You can get a pass to all the national parks in the United States for a year for $50, but you'll spend more than that for one day of fishing in Scotland, where they have this kind of management I am talking about. So, there's a huge expense in the future management of complex natural systems.

Mark Twain once said, "It's a terrible death to be talked to death." But I want to mention one final topic before I close. I've been talking about my book for about a month, in a number of countries and I'd like to report to you something I found increasingly odd, which is that we now seem to live in a world in which almost no one can think about the environment in any terms except political terms. The most common thing that I've been asked is, aren't you just supporting George Bush's agenda?

And I say, well, no, actually not, though I happen to think he's right on this point. But it's an accident. I'm just following the data, going wherever it leads. And people look at me as if to say, What kind of an idea is that? It seems very strange to them.

So you start to talk about the data, the evidence, and show them graphs and charts, and they look at you as if you're doing something odd. You know, like, why do you want me to look at a graph? Or you're tricking me. But the fact is, this is data. Data's not Democratic or Republican, it's data.

So it's an odd world we live in and I would really like to get the political psychodrama out of decision making, at least in the environmental area. I'm sure that's a really naive hope, but it is my hope nevertheless. And, thank you very much for listening to me, and thanks for having me here.

DR. CRICHTON: We'll take questions.

[Questions and answers have been slightly edited for clarity. MC]

PARTICIPANT: Would you comment on the impact Hollywood has on research laboratories? How do people react in a research environment, based on what they see in the movies?

DR. CRICHTON: Well, it's funny because I spoke to the triple AS [American Association for the Advancement of Science] about this subject a few years ago. They were wound up about the image of science in movies. And I told them to relax. You know, in the famous words of Hitchcock, "It's just a movie." (Laughter) But the deeper reality is that the scientific process can't be well represented in a movie. Science is a search. It's a process of discovery that goes on for a very long period of time, in laboratories where increasingly there is little to see except a bunch of white boxes on countertops. None of this works for movies. Movies don't like searches; searches are boring, and static environments are boring. What movies really want is a chase, preferably really fast, involving people with very little clothing. (Laughter)

So, science is ill-suited and movies tend to short-circuit it. You've probably noticed that in a scene where someone is lecturing, you always enter the scene at the end of the lecture and the professor is saying, "See you next Tuesday and bring your assignments..." And that's how the movie scene begins. (Laughter)

PARTICIPANT: I am just curious, having worked in the Hollywood community myself, whether you experienced any shunning by your Hollywood colleagues because of your latest book and the stance that you've taken, and do you think it'll risk any future nominations for Academy Awards for yourself? (Laughter)

DR. CRICHTON: Yeah, it's interesting, the temperature's dropped, at least around me. But Diogenes always said, what good is a philosopher who doesn't annoy anybody? (Laughter) I view most of what Hollywood does as conventional. Hollywood is a very conventional place. And I'm contrarian by nature. I don't believe most of what I'm told, so I'm already considered odd. But my experience is that, in general, Hollywood is like the rest of the country in terms of the environment, which is that they have an attitude, but they don't have any information.

PARTICIPANT: Have you put any thought into the ways there might be financial and social incentives to get activists and other people who make a living from dealing with these questions to move in your direction?

DR. CRICHTON: I can't do everything, I'm actually just a novelist. (Laughter) But your question is a good one. One of the things I didn't talk about was how government policy can be formulated in an era of single-issue advocacy. Because when you have all kinds of groups who just want one thing and don't care about the broad spectrum of social needs, it's a very odd and difficult circumstance. Every household has to budget for many conflicting needs. But I look at some of the decisions that we've made in the regulatory arena and I think: we live in a country where 40 percent of high school graduates are functionally illiterate. We live in a country where kids go through metal detectors on their way into school. And I think that we're wasting a lot of money in a lot of areas and unfortunately, people who have just one issue that they care about, don't care about all the rest. I don't know the solution to that.

PARTICIPANT: I am curious and perhaps it might be fruitful in thinking of how to persuade other people-was there a moment for you when you furrowed your brow and asked, why was all that I was taught in the past seem to be in error? Was there a pivotal moment or influence on your and your thinking that seemed to draw you in the direction you have taken on these issues?

DR. CRICHTON: I suspect my attitude probably has to do with just getting old. (Laughter) And living through a lot of revisions of the conventional wisdom. When I was a kid, if you looked at a map of the world you might notice that Africa seemed to fit rather nicely in the coast of South America, such that they might have once been a single piece of land and had then moved apart. If you asked, you were told by your teacher that, yeah, the continents looked as though they might have once been joined, but that was simply an appearance. It had no meaning. Because the continents were fixed and did not move.

I had trouble with that, it just looked too close to be coincidental. And by the time I was in college, we knew that there was something called continental drift, the continents were on plates, everything was floating and moving, and had been for most of the planet's history. And that's an enormous conceptual change.

When I graduated from medical school, one teacher told us that one-quarter of what we'd learned was already wrong. (Laughter) So I went out into the world with the expectation that a lot of what I read might not be right.

PARTICIPANT: I was struck by the quotation from Mark Twain in your book to the effect that in science you can get a lot of conjecture from a trifling investment of fact. I'm wondering about your first question, about how you get unbiased data. As a philosopher of science, it seems to me we live in an age in which we think that it's not just that theories are not independently verifiable, we're told they're not verifiable at all. And the facts, we are told, are value- and theory-laden, and somehow this wholesale conjecture is really based on trifling investment of facts. I am wondering whether the politicization of science may be a consequence of something we have learned about science. We came into science in the twentieth century with the idea that physics will be over in a few years and we have certainty about everything, and we come out of the century with the idea that it's just a matter of what paradigm you adopt, and one paradigm is as good as another. So I am wondering whether the idea that we just follow the data, is really a sufficient policy to follow. Isn't it rather the case that there is so much contradictory evidence, coming out of so many apparently credible sources?

DR. CRICHTON: No, I don't think so. I think what you say is true, up to a point. Part of the problem that you're describing derives from the fact that most of academia has read too much French philosophy. (Laughter)

But I come from a very much harder tradition that says science is the business of generating testable hypotheses. I know that science has demonstrably changed from that position. Starting in the 1960s, with our interest in SETI, which is, as many people have said, the study of a subject without an object. (Laughter) And then you have string theory, which I'm told is completely untestable. But we're now on our second generation of physicists who are working on string theory.

Even so, I would remind you that the verifiable, testable hypothesis still exists. It's of interest to me, for example, that if I decide I'm going to drop a piano off the Empire State Building, I can ask, how fast will it be going and how long will it take to hit the ground? And if I send that problem to 20 labs around the world, they will give me the same answer within a very small percentage of difference. They'll be in complete agreement. And, of course they'll be using S equals ½ at squared, it's Newtonian mechanics, I know all of that. But the fact is if I give that problem to labs around the world, they'll give me a single answer.

If I ask what will the temperature will be in the year 2100, I'll get an enormous range of answers, which suggests to me that the state of climate science is a very different proposition. I can rely on the answer of Newtonian mechanics, I know to get out of the way of that falling piano. But I'm not sure whether I ought to get out of the way of the temperature in the year 2100.

PARTICIPANT: You talk about how central planning is really not possible in today's knowledge society and at the same time you talked about forcing responsibility for knowledge and around here, forcing means a lot of times federal laws, which is more or less central planning. So, how do you expect a marketplace, as opposed to central planning, to force responsibility for knowledge which you said has apparently avoided the quality revolution that manufacturing has been doing.

DR. CRICHTON: Your question implies that the legal system is centrally managed, which I'm not sure I would agree with. I don't know enough to answer this, but it seems to me that the growth of legislation and the ideas of what's litigable and what's not is, in broad overview, more like "Blade Runner," than any central plan. As an example of that organic growth I would point to the changes in the definition of free speech, which you know is vastly different from what it was 100 years ago. Not that I'm complaining about it.

But I don't think it would be appropriate for me to deny that we're still going to have central governments and we're still going to have some degree of central management of things. I'm just saying that I don't think people are really looking at how vastly changed our society really is. And the extent to which our assumptions that we carried with us for so long, may not really hold anymore, that's all.

PARTICIPANT: How do you enforce responsibility?

DR. CRICHTON: The same way that you enforce any kind of product liability, it's a law. Maybe it's a good law, maybe it's a bad law. My fantasy is that some day I'm going to put on a baseball hat and it's going to have a huge warning label on the inside: "Caution, if you pull down too far, may block vision; may be lost in wind...and then the final comment, do not use as a hat." (Laughter)

PARTICIPANT: I'm from the Center for American Progress, but I was previously a staffer on Capitol Hill. And one of your earlier points was that sometimes staffers bring bias-and my experience from Capitol Hill was that almost more than I could talk to my boss about an idea, he came in with something that he heard on the news or read in the newspaper or even read in a book, and so, I wondered if you could touch on the responsibility of mass media to communicate unbiased information and maybe how well you think they're doing?

DR. CRICHTON: You probably know how well I think they're doing. (Laughter) I gave a talk to the National Press Club in '93 in which I told them that they were out of the quality revolution, that they were in desperate trouble. They didn't care then and they probably don't care now. I operate on the assumption that the mass media will never be accurate. I don't think they ever have been. When did Yellow Journalism start? Almost at the beginning of the American newspaper. And I don't see any reason for media to change. I mean, the great dictum of journalism is "Simplify and exaggerate"-which is exactly what Walt Disney told his cartoonists. (Laughter)

I do believe that there will come a time and it may come quite soon when, because of the Internet, people will be willing to spend a lot of money for verified information.

PARTICIPANT: The New York Times this week referred to the fact that the ice shelves are melting, and I guess I'm willing to believe that that's not true, but I find it hard to believe that the reporter, the editor, and the science editor are colluding to advance a global warming agenda.

DR. CRICHTON: Work on it. (Laughter.) But "collusion" is a strong word. Let me reframe. I think that there are certain kinds of stories that certain journalists find simply irresistible, whether they're accurate or not. I'll give you a recent one. A historian of science named Naomi Oreskes, was invited to write up an essay in "Science" magazine which has been widely reported since. She claimed that she'd inspected the abstracts of 928 articles on climate science from 1993 to 2003 and she had not found a single one which disagreed with the notion that climate change was human caused.

Now, the first thing to recognize is that if you are a reporter following climate science, when you hear 928 articles in 10 years, you immediately know something's wrong. Because the number is far too low. And, in fact, the number of climate articles in the last ten years is closer to 12,000. So, somehow her keyword search was inadequate.

The second thing is that because the exact number of 928 was reported, it is possible to work out which keywords she used to get that number and, therefore, to go back and obtain the actual abstracts that she said she had read. People are doing this. And I'm told, as a preliminary finding, that the claim that none of them contain any negative comment about global warming is far from the truth.

PARTICIPANT: Given that, if we did have product liability on information such as that, wouldn't that also stymie the creative process and the marketplace of ideas?

DR. CRICHTON: Well, I'm not a lawyer, although I pretend to be one. (Laughter) I think you would have to make distinctions between the reporting of information and the generation of the information. (And you would have to define opinion as not information, which in itself might be salutary.) But it's clearly a complex issue. All I wanted to do was flag for people in this room that I think product liability for information is coming. If you want to know the exact details of how it might be implemented, you probably ought to talk to somebody else, because cases are being litigated at this moment, I believe. Last question.

PARTICIPANT: Last fall scientists accused the Bush Administration of politicizing science and the Bush Administration accused the scientists of being political themselves, and being Democrats, and I wondered what you made of that debate and what you think it shows about the interface of science and politics.

DR. CRICHTON: I didn't, I-frankly, I didn't pay any attention to it at all. The kind of debate I'm talking about is on a very different level. Let me give an example. When I was at the Salk Institute, a conference was chaired by Salvador Luria, who won the Nobel Prize in chemistry and whose mother had cancer all her life and had become a heroine addict because of her pain. And Luria had difficulty obtaining heroine for his mom for all these years. As a result, he had a great interest in drug policy and he put together a conference on amphetamines and marijuana. And it was attended by all the important scientists in that area and a number of policy people. And sparks flew like nothing you ever saw. But it happened because he was bringing the actual researchers together. They all know each other; they all know each other's positions, and you got a fierce debate.

Having debates by the people who actually do the work is very different from having a letter written to the white house by a group of politically active scientists. That does not seem to me to involve debate of any sort. It's pure politics.

As for the question, is the Bush Administration politicizing science? Well, it's a long tradition. They are not the first, they didn't invent it. And I think that scientists, which is my area-You know, I'm interested in science. I'm actually not very interested in politics. I hate to say it in this room, but the tradition that I was raised in held that politics was a sort of inferior occupation. (Laughter) If you're not smart enough to do science, you can do politics. (Laughter) I understand, nevertheless, that politics is very important. But I'm concerned that science, having ascended to a phenomenal position of power within our society, has provided a temptation for some highly intelligent individuals to join in the political fray, where they really don't belong, where they do it really badly, and where they don't acknowledge they are damaging science as an enterprise. Because science needs to be kept separate from politics.

It's astounding to me that this is not widely recognized. I opened up "The Washington Post" today, and the lead editorial is about global warming. And it says in effect that even if the Bush Administration doesn't believe the science, they ought to get with the program because of political considerations.

The front page of that same newspaper marks the 60th anniversary of Auschwitz. And Auschwitz exists because of politicized science and it exists because of politicized science that started at the turn of the 20th century in the United States-policies that were supported by Theodore Roosevelt, the Justices of the Supreme Court, Oliver Wendell Holmes, Louis Brandeis, Woodrow Wilson, Winston Churchill, George Bernard Shaw, dozens and dozers of eminent people.

And it can be phenomenally dangerous when you start to take as policy something you want to happen and begin to claim it's science-based. Science has to stay independent, it has to stay focused on the data and it cannot be involved in where this is going to lead. In those days it was immigration policy and the "gene pool." Now it's something else. But it's a dangerous, dangerous gangplank to walk down and I hope we don't go further. We need science. Keep the politics out of it.

Thank you very much.

NOTE: Speeches contained on this site are the property of Michael Crichton and may not be reproduced, copied, edited, published, transmitted or uploaded in any way without express permission. For information about reprinting this speech please email speechrequest@crichton-official.com and be sure to put "Attention: Permissions Dept. / Michael Crichton" in the subject box.

© 1997-2010 Constant C Productions. All rights reserved. Home | Contact Us