The purpose of this paper is to analyze claims made about the plutonium on the Cassini-Huygens spacecraft by mission supporters and opponents on the new media terrain of the Internet. Analysis focusses on claims made about the degree of risk involved, citizen control over risk exposure, fairness in the distribution of risks and benefits from the mission, and trust in governmental institutions responsible for risk assessment and management. Variations in framing by gender and by discernible self-interest are also addressed. Of particular interest is the alteration of the relationships among risk assessors, risk managers, and the public by the debut of a new form of media, the Internet.

Risk assessment science and risk management policy presumably inform one another in natural and technological hazard situations. Policy toward any given hazard is forged in sometimes contentious dialogues between risk assessment scientists and risk management decision-makers (many of whom are elected politicians with risks to their own careers possibly riding on these debates). Impinging on these two sets of players are influences from companies and agencies with varying interests in the outcome and citizen pressure. Citizen pressure is generated by public interest activists, many of whom are quite sophisticated at educating the public about their takes on issues and adroit in stimulating political activism among the newly-informed. ¹

Many contentious debates play out in the often distorting presence of print and broadcast media. Media play a crucial role in the social construction of a given hazard. That is, the media portrayal of a given hazard or disaster affects individual perceptions and agency reactions to a given situation or event and the meaning people place on it. Unfortunately, media have goals that may not dovetail with the information needs of society or the communication needs of risk assessment scientists.

Media have been roundly criticized for the sensationalism ² and biases ³ they display in covering disasters and hazardous situations. Risk assessment scientists, risk management policy-makers, and activists cannot control the media, and producing and distributing their own information in forms they like are costly beyond their resources. Dependent on traditional print and broadcast media, their own needs to communicate and the public's need to know do not necessarily harmonize with the media's need to attract audiences that will in their turn attract advertisers. Activists are in a slightly better situation, since they can more appropriately generate events that might appeal to the media's need for sensation and human drama, such as demonstrations. Of growing importance, however, is the increasing use of Internet media in these discussions to generate awareness and political activism. These interactive media allow technical experts and activists to bypass media they do not control to get their messages out in forms they can control to relatively large audiences.

The various facets of the Internet may alter the nature of information acquisition in hazards communication and thereby affect the balance of power among various stakeholders in a risky situation. E-mail, UseNet, listservers, chats, and Web pages have very modest costs of entry, so it is difficult for a few powerful media businesses to govern content, whether consciously or inadvertantly. More importantly, the Internet also potentially enables the exponential expansion of communication through the chain-mail dynamics of the forward button. For the first time, ordinary citizens and risk assessment scientists might have the ability to communicate their messages with nearly the efficiency of any media company. Will depolarization of power in communication amplify or attenuate perception of risk? How will it alter the social meaning of a hazard and behavior towards it, including political behavior? How might the Internet, then, affect the interaction between risk assessment science and risk management policy in a democracy?

This paper traces the impact of the Internet on the dialogue between risk assessment and risk policy in the case of the plutonium on board the Cassini-Huygens mission to Saturn and Titan. ⁴ This technological risk controversy is particularly noteworthy for the degree to which it was carried out on the Internet, spanning as it does the period during which the Internet exploded into a pervasive medium competitive with television, radio, and print media in the United States. The next section provides a background on the mission itself, the bases of NASA's decision to use plutonium and an Earth gravity-assist, the objections raised by anti-nuclear activists, and the impact of activism in the controversy.

The Cassini-Huygens mission to the Saturn planetary system is physically the largest, scientifically the most ambitious, and organizationally the most international project ever undertaken by NASA or its partners, the European Space Agency (ESA) and the Agenzia Spaziale Italiana (ASI). At over 5,600 kg, the 7 m high and 4 m wide orbiter and probe house eighteen instruments supporting 27 different investigations of the planet, its rings, its icy satellites, its magnetosphere, and Titan.

NASA dismissed solar power for mission instrumentation and temperature maintenance needs for several reasons. First, the 5,655 kg mass of the Cassini (Saturn orbiter) and Huygens (Titan probe) combination and its navigational fuel, even without massive solar arrays, was already very close to the 5,760 kg launch limits of the largest American expendable launch vehicle (Titan IV/Centaur combination). Second, the duration of the cruise to Saturn (seven years) and tour at Saturn (four years) and extremity of the conditions out at Saturn require absolutely dependable and durable power sources, conditions NASA concluded solar panels could not satisfy. Third, the orbiter will make repeated movements through the ring system, and solar cells are highly susceptible to impact damage (and provide large targets for such damage). Fourth, NASA put a premium on minimizing moving parts that can fail, after the disastrous failure of the high-gain antenna to deploy on the Galileo spacecraft in April 1991 during its cruise to Jupiter, which markedly reduced data return from that mission. Fifth, Saturn is located about 9.5 astronomical units (AU) or 1.4 billion kilometers from the sun. Applying the inverse square law, solar energy receipt is about 1 percent that at Earth. Solar panels in optimum condition would, therefore, have to be about 500 square meters in size. They would thus be enormously heavy, entail many moving parts to deploy and struts to support, and require heavy batteries. ⁵

For these reasons, NASA decided on the compact radioisotope thermoelectric generator (RTG) and radioisotope thermal unit (RHU) design. These generate heat and, in the case of the RTGs, electrical power through the alpha radiation emitted by ceramicized plutonium-238 dioxide. Pu-238 is non-fissile and cannot sustain chain reactions the way Pu-239 and Pu-241 can. ⁶ The collisional energy of the large alpha particles striking the ceramic in which the Pu-238 is embedded generates heat, which maintains the instruments at operating temperature in the deep cold around Saturn and is converted into electricity to power their operation. The RTGs and RHUs became a point of controversy by 1995, due to their incorporation of 33 kg (72 lb.) of plutonium dioxide. Plutonium was characterized by John Gofman "the most fiendishly toxic substance ever known" in 1974 ⁷ and subsequently popularized by Ralph Nader "the most toxic substance known" in 1975. ⁸

Adding further to the controversy, NASA further opted for a Venus-Venus-Earth- Jupiter Gravity Assist (VVEJGA) trajectory. ⁹ Gravity assists entail exchanging angular momentum with a planet by swinging a spacecraft past it in the direction of the planet's revolution. The gravity assists would boost Cassini-Huygens to the velocity it needed to reach the Saturn system during the careers of its science teams. The Earth gravity assist element led to accusations that Cassini could suffer a mishap and strike Earth or otherwise break up and explode in the earth's atmosphere, distributing its plutonium throughout the earth's atmosphere. ¹⁰ If the plutonium were thereby vaporized into particles small enough to be inhaled but large enough not to be exhaled, the radiation from an embedded particle could indeed induce cancer through the concentration of its energy within the very small penetration radius of alpha radiation. ¹¹

The questions are how likely is such an accident scenario, what would happen to the RTGs and RHUs in such an accident, what would be the number and size distribution of plutonium dioxide particles released by an explosion, where would fine particles (if any) fall, and what is the probability of a person being at that location to inhale the appropriately sized particles in that distribution.

Risk assessment performed for NASA as part of the Environmental Impact Statement for the mission characterized the risk of plutonium release on launch or a swingby accident as negligible and acceptable, at levels that would not be statistically observable over a five decade timeframe. ¹²

By 1995, opposition to the mission had begun to organize itself. ¹³ Anti-Cassini activists were skeptical of any risk assessment performed for NASA, questioning the independence of the outside agencies and individuals consulted by NASA. The activists came up with their own figures, ranging from over 200,000 deaths ¹⁴ through 1 million deaths (attributed to John Gofman by Grossman) ¹⁵ and 4.6-9.2 million deaths (Ernest J. Sternglass quoted on the NoFlyBy web site) ¹⁶ to as many as 40,000,000 deaths (attributed to Sternglass by Grossman). ¹⁷ The timeframe of these predicted surplus deaths is not specified. The opponents further claimed that NASA was imposing an unnecessary risk, in light of improvements in solar technology: They argued that solar power would have been an option, even out at Saturn, where incoming solar radiation is one percent that at Earth. ¹⁸ They openly wondered if Cassini were part of some military conspiracy to acclimate Americans to "nukes in space" so that space could be "weaponized." ¹⁹

From 1995 through October of 1997, the opposition movement concentrated its efforts on creating pressure to abort the October 1997 launch of Cassini. ²⁰ The launch went forward, so the movement then focussed on aborting the flyby, scheduled for August of 1999. ²¹ The movement was unsuccessful in stopping either of these events, but it did generate an enormous amount of controversy and a lot of pressure on Congress, the White House, and courts. ²² Several senators and representatives signed a public petition against the mission, and California Senator Barbara Boxer commissioned a study by the U.S. Government Accounting Office entitled, "Space exploration -- power sources for deep space problems." ²³ State and local government representatives also received pressure to declare their jurisdictions in opposition to the launch or flyby. Several responded, including the Massachusetts House of Representatives, which passed a resolution to abort the launch, as did the Newton, Massachusetts, City Council, the Santa Cruz, California, City Council, and the Marin County Board of Supervisors. ²⁴ The movement may not have achieved its original goals, but it was highly effective in making RTG and RHU use controversial, which may affect the design, authorization, and funding of future missions. ²⁵

Besides a number of print media and television pieces on the controversy, ²⁶ most of the day-to-day activism took place on e-mail and listservers, the Web, and on UseNet. ²⁷ Indeed, much of the traditional media pieces germane to the subject wound up as e-mails and Web pages, and Internet activism resulted in demonstrations that were then covered by conventional media.

The dense and rapidly increasing network of connections among people in cyberspace clearly offered a way for the politically active to bypass the constraints of conventional print and broadcast media to get their messages directly into the hands of readers. Readers could easily propagate the messages themselves, so that, ultimately, a handful of activists accessed hundreds of thousands of people indirectly through the math of the forward button. The Cassini controversy exemplifies the power of the Internet in information propagation and activist recruitment.

My purpose in examining this controversy included testing several hypotheses from the hazards perception literature. One of the most common statements in the literature is that there is a marked difference in hazard perception on the part of risk assessment experts and the lay public. These often consist of exasperated descriptions of the public as ill-informed and, therefore, offering input of little value in risk management policy decisions. ²⁸ A variant on this vision of the public as poorly informed is also expressed in statements that, lacking the time and training to evaluate certain risks, people will form their opinions about them through deference to reference groups they trust and, once these opinions gell one way or the other, they become very resistant to contrary evidence. ²⁹

Douglas and Wildavsky elaborated a model based on cultural predispositions toward risk assessment and tolerance. ³⁰ They differentiated a "center" approach, associated with people at or near the social core of political power, who are characterized by respect for rational risk assessment, acceptance of the hierarchical structure of modern society that privileges them, and willingness to tolerate a certain degree of risk if it is offset by greater benefits enabled by assumption of that risk. Countering that, they defined a "border" approach, associated with people peripheral to social power, who are suspicious of the center, fearful of pollution of the natural environment by the self-serving power elite, and unimpressed by a rationality seen as serving central interests. They are generally described as egalitarian in social values and oppositional in politics. Basically, in this model, expert opinion is centrist and lay opposition to technology is borderer in character. This extension of the ill- informed public hypothesis has generated an enduring debate between those who would dismiss the public and those who are troubled by the incompatibility of this attitude of dismissal of the public and the premise of a democratic society.

Many of the Douglas and Wildavsky critics argue that lay hazard perception may, in fact, reflect a more nuanced and multifaceted consideration of risk than does the narrowly focussed mortality and morbidity calculations of risk assessment experts. ³¹ The gist of this literature is that the public will overestimate risks and perhaps even be aroused into political activism against them if one or more of the following dimensions are present: lack of control, lack of familiarity, unfairness, dread, or mistrust.

People will accept quite high levels of risk if they perceive themselves as controlling their own risk exposure (e.g., choosing to drive fast or choosing to live along a scenic stretch of river) but will become very upset at even the most trivial risk if they feel it is being imposed on them. ³² Too, people are more tolerant of risks that have become familiar to them, e.g., cars versus planes. ³³ The public will accept higher risks in situations they see as benefiting them, but they will become incensed if they perceive that they are assuming the risks and someone else is getting the benefit. ³⁴ Dread attaches to any hazard, no matter how tiny the probability, if the imaginable consequences include sheer horror, particularly frightening diseases (e.g., cancer or AIDS), and transmission of harm to future generations. The horror of Hiroshima and the slowly growing realization that even minute exposures to radioactive substances can result in mutation and cancer ensures that anything nuclear evokes dread. ³⁵ Social concern over certain hazards may become amplified far beyond the risk assessed by experts if the institutions responsible for assessing risk and protecting the public are themselves under suspicion. Public trust of governmental institutions in general seems to have hit a decades-long slide in the wake of revelations of gross government misconduct, conspiracy theorizing from both left and right, and a pervasive "X Files" mentality. ³⁶ Lastly, many studies have found differences in risk perception or tolerance along ethnic, age, and gender lines, perhaps reflecting the relative situatedness of these aspects of identity along the fairness, control, familiarity, and trust axes of this and many other social issues. ³⁷

As stated earlier, the purpose of this paper is to analyze the claims made about the plutonium on board the Cassini-Huygens spacecraft, as expressed on the Internet. Specifically, ten questions derived from this literature are evaluated:

1. Do Internet debates on Cassini bring up the issue of perceived control over plutonium?
2. Are issues of fairness in the social distribution of mission risks and benefits raised by the mission's opponents and proponents in their debates online?
3. Do Cassini opponents focus on expressing or eliciting dread?
4. If so, since dread is a more compelling emotion than the scientific curiosity and the romance of space-exploration motivating mission proponents, are opponents more numerous and more vocal than proponents?
5. Is suspicion of governmental institutions generally or of NASA specifically used by opponents to discredit the risk assessment performed by or under contract to NASA?
6. Can demographic variations be discerned among the people contributing to this debate online, and, if so, are these demographic differences associated with different percentages of authors in opponent and proponent stances?
7. Does self-interest affect the propensity to contribute to this debate, i.e., does the removal of identifiably self-interested individuals significantly change the balance of authors and messages between support and opposition to the Cassini-Huygens mission?
8. Is there evidence for the claim that people rely on trusted reference groups to form political opinions on technical issues beyond their training, and does this vary between mission opponents and proponents?
9. Can Douglas' and Wildavsky's center/border dichotomy be discerned in the values and concerns raised in the content of the messages themselves, with mission proponents expressing more confidence in science and mission opponents expressing more counter-cultural perspectives?
10. Does the use of the Internet for political communication create an effective alternative channel for risk management agenda-setting to the traditional print and broadcast media and lobbying?

The Internet varies in the degree to which it can be interrogated for research purposes. For example, the Web contains a huge population of sites devoted to one aspect or another of this controversy, but this population is quite unstable through time, as people change their websites and discard materials no longer of immediate interest. Listservers number in the dozens, if not hundreds, of thousands, ³⁸ but, while the lists themselves can be located through several search services, postings on them are archived (or not) by the individuals running each list. Chats are by their nature ephemeral. UseNet news groups became the focus of this analysis, because they were archived by Déja.com in a fully searchable form. ³⁹ Messages can be searched by keyword, author, dates, and newsgroup.

Using Déja.com's search engine during April through August 1999, I searched for the keyword, "Cassini." This resulted in a population of 19,853 messages that had been posted from April of 1995 through March of 1999, which contained the word "Cassini" anywhere (posted message, subject line, author name). Since the debate unfolded over four years, I ensured longitudinal representativeness of messages by using "Cassini" as the keyword and specifying dates by month, working backwards from April 1995 through March 1999. The number of postings varied from a low of three comments in a month (April 1995) to a high of 4,385 (October 1997, the month of the spacecraft's launch).

I then sampled the discussion by viewing up to the top 250 messages in each month, working backwards. Déja.com sorted messages by "relevance," basically the number of times the keyword turns up in a message. From this maximum of 250 messages per month, I extracted only those written in English on the subject of the spacecraft, saving their authors' names and e-mail addresses and an abstract of their messages in a word processor. Excluded were postings on "Cassini" topics unrelated to the mission. ⁴⁰ Furthermore, I saved only the most recent posting by a given author and then categorized the author's stance on the basis of this comment as proponent, opponent, or neutral. If author stances were not decipherable from these comments, I would then search on these authors' names and "Cassini" to read their other messages on the subject, until I could classify their stances. Additionally, I did searches on all author names and "Cassini" to identify numbers of postings on the subject per author as a crude indicator of interest level. If people are highly interested in a debate, chances are greater that they will contribute to it repeatedly.

This sampling process yielded comments by 937 authors who had, among them, posted 8,020 messages on the subject or forty percent of the "Cassini" messages dating from the four year study period. The authors were classified by stance, central concerns raised, gender (the only detectable demographic variable), affiliation with NASA and related institutions or with opposition organizations, and whether their messages were largely original compositions or largely forwards from someone else. Classification was done solely by myself, as funding did not allow the training and hiring of teams of coding assistants. Single coder content analysis may suffer from reduced reliability of coding of each message, because it lacks a mechanism for soliciting possibly divergent interpretations of that message and then democratically deciding among such interpretations. On the other hand, single coder content analysis does guarantee consistency in coding across messages as one person applies a single approach to all messages rather than training others to apply that approach with varying levels of motivation and understanding.

Later, this text database became the basis of a spreadsheet database including fields for author name, login nickname, e-mail, gender (from given names), basic stance, a single phrase or sentence summarizing the central concerns raised in their messages, and number of postings an author had made on the subject. I also noted whether the message had been originally composed by the author or was basically a forward of someone else's writing. Most of these fields are straightforward but, in light of the one-coder methodology, it is appropriate to show a few of the messages, randomly chosen, and how their stances and central concerns were classified.

To that end, a handful of representative posts were extracted from this database and listed in Figure 1, which resides at the following Web address: https://home.csulb.edu/~rodrigue/cassini/risk01figure.html. To extract them, I used a random number generator and looked up the message corresponding to a given random record number in the spreadsheet. I then found the corresponding message in the text database of UseNet messages and pasted them onto the web page. With some of these, more than one concern is found in the message: I classified them by the concern on which the author expended the most writing. The database developed in this manner was sorted on various of its fields to yield the results summarized in the following tables. Significance levels are generated from difference of proportions Z tests, with prob-values < 0.05 considered significant.

The great majority of UseNet authors were supportive of the mission: 60 percent were supporters; 21 percent were opponents; and 19 percent were neutral (Table 1). Given the emotional power of nuclear dread, this result was surprising. The opponents were, however, considerably more vocal than the proponents or neutrals: The 21 percent of authors that were opponents posted 31 percent of the messages (prob<0.001). The 19 percent of neutral authors posted 13 percent of messages, significantly less than their numbers (prob<0.001). The 60 percent of pro-Cassini authors posted 56 percent of messages, which is signficantly less than expected (prob=0.005).
 

                         Table 1 
                        Stance by gender
_________________________________________________________

Stance       Gender           Individuals           Posts
                                  #     %        #      %
_________________________________________________________

Neutral      female               7   3.9       10    0.9
             male               139  78.1      930   87.5
             organization         4   2.2       14    1.3
             unknown             28  15.7      109   10.3
19.0% of authors                178 100.0     1063  100.0
13.3% of posts                                           
---------------------------------------------------------
Opponent     female              16   8.2      103    4.1
             male               132  68.0     2067   82.4
             organization         6   3.1      121    4.8
             unknown             40  20.6      217    8.7
20.7% of  authors               194 100.0     2508  100.0
31.3% of  posts                                          
---------------------------------------------------------
Proponent    female              19   3.4      154    3.5
             male               468  82.8     3946   88.7
             organization         3   0.5       24    0.5
             unknown             75  13.3      325    7.3
60.3% of authors                565 100.0     4449  100.0
55.5% of posts                                           
_________________________________________________________

 937 = n (authors)                                       
8020 = n (posts made by these authors)                   
_________________________________________________________

The only demographic difference discernible among the authors was gender (Table 2). This debate was overwhelmingly a male preserve: Fewer than 5 percent of authors were female, and they contributed only about 3 percent of the posts. The Internet has been perceived as a largely male space, but the disparities between the percentages of women contributing to this discussion are drastically lower than those of women using the Internet itself, according to the statistically representative telephone survey series by CommerceNet/Nielsen Media Research Internet Demographic Study. ⁴¹ In 1995, at the beginning of the study period, women made up 33 percent of Internet users in the United States and Canada (versus 7 percent of identifiable participants in the Cassini debate from April 1995 through March 1996). In 1999, at the end of the study period, women constituted 46 percent (versus 5 percent of Cassini discussion participants from April 1998 through March 1999). While gender disparities in user numbers are disappearing online, there remain differences in the interests women and men pursue in their online shopping and purchasing activities, with men likelier than women to pursue technological interests. This may reflect the same factor as likelihood to participate in an online technological risk debate. ⁴²

Both genders in this debate were likelier to support Cassini than to oppose it. Even so, there was a gender-gap. Only 45 percent of the women were mission-supporters, compared to 63 percent of the men (prob=0.019); 38 percent of the women were opponents, while only 18 percent of the men were (prob=0.002). Had the genders been equally represented among the authors, the proponents would still have been in the majority (54 percent of the 81 gender- identified authors, had they been equally male and female), but the disparity would not have been so extreme (opponents would have been 28 percent and neutrals, 18 percent).
 

                        Table 2
                       Gender by stance
_________________________________________________________

Gender       Stance           Individuals           Posts
                                  #     %        #      %
_________________________________________________________

Female       neutral              7  16.7       10    3.7
             opponent            16  38.1      103   38.6
             proponent           19  45.2      154   57.7
4.5% of authors                  42 100.0      267  100.0
3.3% of posts
---------------------------------------------------------
Male         neutral            139  18.8      930   13.4
             opponent           132  17.9     2067   29.8
             proponent          468  63.3     3946   56.8
78.0% of authors                739 100.0     6943  100.0
86.6% of posts
---------------------------------------------------------
Organization neutral              4  30.8       14    8.8
             opponent             6  46.2      121   76.1
             proponent            3  23.1       24   15.1
1.5% of authors                  13 100.0      159  100.0
2.0% of posts
---------------------------------------------------------
Unknown      neutral             28  19.6      109   16.7
             opponent            40  28.0      217   33.3
             proponent           75  52.4      325   49.9
16.0% of authors                143 100.0      651  100.0
8.1% of posts
_________________________________________________________

 937 = n (authors)
8020 = n (posts made by these authors)                   
_________________________________________________________

The authors' stated concerns and stances suggest what activated authors to contribute to the social debate over Cassini (Table 3). Opponents were dominated by three subtypes: 24 percent simply passed on messages originating from about half a dozen people or organizations, often without comment; another 24 percent wrote messages consisting of independent expressions of concern about the risks of plutonium in general or during the launch and flyby phases of this mission in particular; and 21 percent were people interested in Nostradamus and astrology, who expressed great fear that Cassini was the "King of Terror" that Nostradamus had predicted would come from the skies and destroy Earth in summer of 1999 (the Earth flyby took place on August 17-18,1999).

Proponents, given their much larger numbers, discussed a wider range of issues and concerns, with no one issue commanding as many as a fifth of the authors. The most common statement (17 percent) was that the opposition was very small if very vocal and unqualified to comment. Sixteen percent opined that the risk of the mission or of RTGs was being grossly overstated. Thirteen percent simply enthused about the mission and its goals. Another 10 percent engaged in rather nasty "flaming" of the opponents (slang for ad hominem attacks). Only 6 percent forwarded on other people's or organizations' messages, usually something from a NASA publicity office. The content of these postings, then, emanate from people enthusiastic about science and technology and comfortable with conventional risk assessment analyses and results. That is, they seem more centrist in point of view, in Douglas' and Wildavsky's usage.

Neutral messages fell into two main types. The most common of these were questions or answers about very technical issues related to the spacecraft's design, instrumentation, and objectives (40 percent). A distant second were very elementary questions and answers, such as which planet was Cassini's destination or why does it take so long for a spacecraft to get there (11 percent). Another eight percent were forwarded messages, and another seven percent each were basic questions from Nostradamus fans or queries about the risk involved in the mission. The balance of the concerns expressed, then, emerges as more centrist than borderer.

Opponents were significantly more likely simply to pass on others' messages (24 percent) than were the other two groups (7 percent of the pooled 743), with a prob-value of <0.001. Forwarding on others' opinions reflects the tendency for time-pressed citizens to rely on trusted reference groups' opinions in forming their own in complex issues unfamiliar to them. That opponents were likelier to resort to this form of political communication than proponents may speak to the center/border dichotomy in access to and trust in conventional technological risk assessment information. Proponents were far likelier to engage in individually-targeted "flames" (ten percent) than were either opponents (two percent) or neutrals (four percent). The difference between the proponents and the other two stances is significant at a prob- value well below 0.001 and carries a whiff of center snobbery towards the border.

Within the Nostradamus debate, opponents and neutrals were fans of the sixteenth century astrologer-poet, while proponents were his critics. Of those who commented on Nostradamus, 21 % of the opponents discussed him in connection with Cassini, as did 7 % of the neutrals and 4 % of the Cassini proponents. Opponents, therefore, are significantly more likely to come to this technological risk debate from an ascientific and countercultural (border) perspective than the neutrals and the proponents (prob<0.001).
 

                        Table 3
               Central concerns raised by stance
_______________________________________________________________

Neutral Issues                                        #       %
_______________________________________________________________

   Technical questions/answers                       72    40.4
   Asking/providing basic information                20    11.2
   Passing on others' messages                       14     7.9
   Nostradamus fan asking basic question             13     7.3
   Risk question                                     12     6.7
   Flames                                             7     3.9
   Costs, taxes                                       6     3.4
   Politics/bureaucratization                         5     2.8
   Privatization of space                             4     2.2
   Vulnerabilty of big mission                        2     1.1
   Other                                             23    12.9
sum                                                 178   100.0
_______________________________________________________________

Opponent Issues                                       #       %
_______________________________________________________________

   Passing on others' msgs                           46    23.7
   Risk                                              46    23.7
   Nostradamus/astrology/666 fears                   41    21.1
   Calls to action                                   11     5.7
   Costs, scale, opportunity costs                    9     4.6
   Censorship by media                                7     3.6
   Conspiracy/militarization of space                 6     3.1
   Flames                                             4     2.1
   Privatization of space better than NASA            3     1.5
   Other                                             21    10.8
sum                                                 194   100.0
_______________________________________________________________

Proponent Issues                                      #       %
_______________________________________________________________

   Opponents a small # unqualified Luddites          95    16.8
   Risk overstated, disproproportionate              91    16.1
   Enthusiasm for the mission and space              73    12.9
   Flames                                            59    10.4
   Orbit/trajectory aimed to be safe                 36     6.4
   Passing on others's messages                      36     6.4
   Past nuke/RTG failures didn't kill life on Earth  27     4.8
   Solar not feasible                                22     3.9
   Big missions=big results                          20     3.5
   Nostradamus critiques                             23     4.1
   Cass budget doesn't allow for cruise science      16     2.8
   Opportunity costs of opponent activism            11     1.9
   Media censorship/bias against science              9     1.6
   Calls to action                                    8     1.4
   Privatization critique for large-scale missions    4     0.7
   Other                                             35     6.2
sum                                                 565   100.0
_______________________________________________________________

937 = n (authors)

This sample may not be a representative sample of all those on the Internet with an opinion on Cassini: It is more than likely that people who bestir themselves to contribute to the debate are in some way self-interested in its outcome. These may be employees of NASA, the ESA, the ASI, or employees of their subcontractors or, conversely, committed and activism-prone members of opposition organizations.

To examine self-selection bias on the basis of self-interest, I removed all 165 people with e-mails originating from the space agencies, companies doing contract work for them, and academic institutions having sizable grants with them (according to the campus web sites), as well as those 7 people who posted from activist organization addresses. It remains possible that individuals working or volunteering in these organizations also maintain private e-mail accounts not associated with their organizational affiliations and, so, would not be culled in this manner. This is especially likely to be true of opponents, who are less likely to have e-mail accounts on their organization's domains (many Internet service providers offering domain-hosting services for small businesses and organizations usually limit the associated e-mail accounts to about twenty). With these caveats, the easily-identifiable affiliates made up eighteen percent of the authors. Suggestively, they contributed 26 percent of the messages, a disproportion (prob=0.022) suggestive of their passion on the subject (Table 4).

By removing them, the database dropped to 765 individuals and 5,912 messages originating with people having no discernible ties with Cassini and the organizations that oppose it. Of these remaining authors, 20 percent are neutral, trivially more than was the case with the full database (prob=0.469). They posted 16 percent of messages, however, a somewhat greater percentage than did the neutrals in the original database (prob<0.001). Twenty-three percent of the authors in the reduced database are opponents, a trivially greater percentage than in the original (prob=0.319), but they posted fully 38 percent of the messages, which is quite a bit higher than was seen in the full database (prob<0.001). The public left in the database who oppose the mission, then, emerge as more likely to communicate their feelings. The percentage of proponents in the revised database dropped insignificantly, from 60 percent to 57 percent (prob=0.156), but these are less communicative about their sentiments than was the case when identifiable employees of NASA and related institutions were left in. That is, the percentage of posts from non- self-interested proponents dropped to 46 percent from the 56 percent seen in the original database (prob<0.001).

In all, the public left in the database were basically indistinguishable from the full database in terms of the proportions of individuals adhering to the three positions. Those individuals left in the database who oppose the mission, however, are more communicative about their views, which offers support to the expectation that the emotional basis of opposition, dread of nuclear contamination, is more compelling than that of support for the mission. This is particularly evident when people whose livelihoods may depend on the mission or on the space program (a potent emotional driver!) are removed. Indeed, though supporters left in the database dominated as individuals, their support was considerably more tepid emotionally than when identifiably self-interested persons remained in the database, at least as judged from the number of posts they offered on the subject.

The impression created by separate examination of the unaffiliated public is heightened by examination of those persons working for NASA and its affiliated institutions and those with e-mail addresses in opponent organization domains. Of the 165 people employed at NASA or affiliated institutions, fully 79 percent are proponents, and 86 percent of their 1,965 messages were supportive of the mission. All 7 people with e-mail accounts on the opponent organization domains were, not too surprisingly, mission opponents, and they posted 162 messages. Where, on average, the 765 unaffiliated authors posted 7.7 messages on the topic of Cassini, the 165 NASA-affiliated authors averaged 11.9 messages, while the 7 people identifiably in opponent organizations averaged 23.1 messages. Again, the disparity in numbers of messages in this debate suggests the difference in intensity of motivation: Self-interest and, especially, nuclear dread produce greater frequency of communication.

Particularly striking with respect to the nature of UseNet communication was the great influence of a very small number of individuals in this debate. Some 24 percent of opponents were passing along messages from others. I traced these back and found that they mainly originated with about half a dozen individuals!

The whole controversy began in print media articles by Karl Grossman starting in 1995 that highlighted the amount of plutonium on board Cassini-Huygens, questioned NASA's safety record, implied a military connection to get Americans used to "nukes in space," and drew attention to the large corporations (including NASA contractors) with stakes in nuclear power. ⁴³ In this, he was continuing a line of articles he had done on the Galileo mission to Jupiter. His articles were posted widely online, which he encouraged. ⁴⁴ His writings would be credible to many environmental and peace organizations, which themselves serve as reference groups, because his work has often been cited in the progressive Project Censored's ⁴⁵ annual list of news stories that received insufficient coverage.

In March of 1997, a very energetic individual who owns a software company, Russell Hoffman, became concerned by Grossman's arguments and began to push them online very vigorously, in a long series of Web-based newsletters (Stop Cassini Newsletter). ⁴⁶ Many UseNet postings were forwards of this newsletter. Jane Wardlow Prettyman, editor of The Real News Page (now called American Review), began a web page devoted to Cassini, called Disingenuous Digest: Analysis of News Media Handling of Cassini. ⁴⁷

Three organizations began publicizing this cause: the Florida Peace and Justice Center (Bruce Gagnon), ⁴⁸ and the Lovearth environmentalist network, ⁴⁹ and the NoFlyBy group (Jonathan Mark), ⁵⁰ These early activists then recruited individuals with credentials in physics, such as Michio Kaku, ⁵¹ John Gofman, ⁵² Ross McCluney; ⁵³ medicine or health physics, such as Helen Caldicott, ⁵⁴ Earl Budin, ⁵⁵ Ernest Sternglass, ⁵⁶ and Horst A. Poehler; ⁵⁷ and a former employee at Cape Canaveral, Alan Kohn, ⁵⁸ who were willing to make public statements or speeches at demonstrations encouraging opposition to Cassini, which then were sent around UseNet and other Internet venues. Kaku, McCluney, Budin, and Poehler were willing to write and post materials for the cause, while Gofman, Caldicott, Sternglass, and Kohn permitted themselves to be quoted on activist web sites. Many of these individuals are well-known and widely respected in progressive political and environmental circles, making their statements on this issue credible among people who admire and trust them in other situations. So, this whole controversy basically started with from 1 to 11 people, depending on how far back one goes.

On the proponent side, nearly 6 % of UseNet authors simply forwarded on others' messages. These were generally posts from NASA publicity office people, notably Ron Baalke ⁵⁹ and Mary Beth Murrill. ⁶⁰ Eight percent of neutrals also forwarded on others' messages. The influence of key individuals, thus, was proportionately and absolutely less in the case of neutrals and proponents. Only the Cassini opponents were able to make effective use of the exponential math involved in chain-mail forwarding. Chain-mail depends for its success on somehow catching the attention of its readers and then motivating them to forward the message to others. The dread elicited by opponent postings in a manner of speaking substituted for the "bad things will happen to you if you do not immediately send this to ten of your friends" curse that typically accompanies ordinary chain-letters (and their contemporary e-mail versions).
 

                  Table 4
         Stance by self-interest
____________________________________________

Those with no traceable affiliation

Stance           Individuals           Posts
                  #        %        #      %
____________________________________________

Neutral         156     20.4      968   16.4
Opponent        174     22.7     2233   37.8
Proponent       435     56.9     2711   45.9
____________________________________________
 765 = n (authors)                         
5912 = n (posts made by these authors)     
____________________________________________

Those affiliated with NASA

Stance           Individuals           Posts
                  #        %        #      %
____________________________________________

Neutral          22     13.3      138    7.0
Opponent         13      7.9      136    6.9
Proponent       130     78.8     1695   86.1
____________________________________________
 165 = n (authors)                         
1969 = n (posts made by these authors)     
____________________________________________

Those affiliated with opponent organizations

Stance           Individuals           Posts
                  #        %        #      %
____________________________________________

Neutral           0      0.0        0    0.0
Opponent          7    100.0      162  100.0
Proponent         0      0.0        0    0.0
____________________________________________
  7 = n (authors)                          
162 = n (posts made by these authors)      
____________________________________________

In the following paragraphs, these findings are related back to the hypotheses generated from hazards perception literature. Table 5 organizes the research questions and their outcomes.

First, contrary to the expectations of hazards literature, this technological risk debate does not seem driven by the issue of control over the plutonium exposure, not even among the opponents. Fairness questions are often raised as an explanation for public activism over technological risk, but only two percent of authors in this debate on UseNet brought up the issue of fairness and that in a manner tangential to the risk of plutonium exposure.

Second, there was a significant gender gap, which has emerged in other hazards perception studies, with women more doubtful about the assumption of this technological risk. It should be noted that, even so, support for Cassini commanded the highest percentage of the few women in this debate among the three stances. The most striking manifestation of gender in this debate, however, was the near invisibility of women among the 937 authors (only 42): Online pace-related technological risk discussion remains a male preserve.

Third, and perfectly in accordance with prior literature, dread is the central axis in this hazards debate. Two thirds of opponents expressed dread of nuclear contamination, and the Nostradamus discussants seemed quite terrified that Cassini would bring about the predicted end of the world. Over a quarter of the proponents addressed the dread factor, too, mainly by trivializing the probability of an accident and the consequences of an accident should one occur.

Another factor mentioned in hazards literature is mistrust of public institutions, and it is a secondary theme in this debate. Six opponents say that there is a NASA conspiracy to militarize space and the plutonium on Cassini is merely the camel's nose in the tent, while another 7 stated that the media were censoring the plutonium risks of Cassini. Both of these arguments are, however, often cited in the 46 messages forwarded by opponents. Even a few proponents (9) said they thought the media were biased towards the opponents and were not letting NASA have a chance to defend the mission and its goals. So, mistrust of national government and of media is present among opponents of the mission and, in the case of the media, this mistrust is shared by a few proponents, too.

Fourth, surprisingly, given the level of dread attached to plutonium, more UseNet contributors supported the mission than opposed it, even when self- interested people were removed. Self-interest did affect individual authors' propensity to communicate in this controversy, however. When discernibly self- interested persons were removed from the database, there was no significant change in the balance of authors among the opponent, proponent, and neutral stances towards Cassini-Huygens, but there was a very significant shift in the volume of communications coming from each stance. The more loquacious proponents in the original database seem to have been those in Douglas' and Wildavsky's center, people involved in the mission or in the space program, closer to the center of political power in this issue, and obviously much more comfortable with conventional risk-assessment and trade-offs between risk and gain. With the removal of more "central" people, a more communicative and passionate periphery is seen, skeptical of risk assessment science, mistrustful of government, and oppositional in communication style.

Fifth, suggestions of the center/border dichotomy also show up in this particular debate in the form of Nostradamus. Nostradamus' predictions appeal to some non-scientists, especially those inclined to New Age and other classically counter-cultural movements. These movements explicitly question scientific logic and method, denigrating science as merely a privileged but not an epistemologically superior way of knowing the world and making decisions. ⁶¹ The New Age and Nostradamus, then, fit the counter-cultural quality of Douglas' and Wildavsky's border and, so, it is not surprising that fans of Nostradamus were prominent among the opponents to Cassini-Huygens ⁶² and that all but 2 of the 13 proponents who addressed Nostradamus' "King of Terror" were debunkers of astrology and the New Age.

Te tendency to forward others' messages is indicative of both the center/border split and the tendency for time-pressed people to rely on the judgments of reference groups they trust. The amount of research necessary to understand the details of the plutonium controversy is beyond most people's time and energy, and yet in a democratic society it is an important social issue. Few outside the borders of the space program and its contributing institutions are familiar with the technology, while those more central to the program may be both familiar with it and self-interested in the outcome of the debate. Those on the outside must, therefore, necessarily depend on other people's opinions, judgments, and analyses; hence, the opposition more strikingly used the forwarding of other people's messages to communicate its concerns and suspicions.

Finally, the vanguard of this movement utilized border themes and complaints to recruit others to spread the message. The forwarded messages appeal to suspicion of NASA, the military-industrial complex, bodily pollution being imposed by greedy powers, science as the corrupt servant of insane elites, astrological concerns about the "grand cross" of 18 August 1999, and claims that the "Ancients" had detailed knowledge of plutonium. The messages, thus, were embedded in a matrix of pre-existing border beliefs that made them credible to many of their recipients. Incorporated in that matrix were elements shown specifically to bear on perception of technological risk and arouse concern: dread, above all, and mistrust of public institutions responsible for informing the public about and protecting the public from hazardous technologies.
 

                                Table 5
                          Summary of Findings
________________________________________________________________________


Research questions                            Outcomes                 
------------------------------------------------------------------------
1. Perceived control over exposure .........  Not a focus
2. Fairness of risks and benefits ..........  Not a focus
3. Dread ...................................  Central axis of debate
4. Opponents more numerous and vocal .......  Less numerous, more vocal
5. Suspicion of government .................  Some evidence both sides
6. Demographic variations (gender) .........  Few women, plurality does
                                              support mission, gender 
                                              gap: fewer women support
                                              mission than men do
7. Self-interest ...........................  Public supports mission,
                                              self-interest is evident
8. Reliance on reference groups ............  Opponents forward others'
                                              messages more than 
                                              proponents and neutrals do 
9. Center/border dichotomy .................  Opponents more counter-
                                              cultural (Nostradamus),
                                              more reliant on reference
                                              groups
10. Internet as effective political medium .. ~Half dozen people started
                                              impressive movement that
                                              made Cassini very contro-
                                              versial at national and
                                              local levels
________________________________________________________________________

Risk assessment is a probabilistic, statistical science, not a deterministic, experimental one. Its conclusions inescapably carry the hazards of Type I and Type II errors, and the minimization of one generally raises the probability of the other. We assume a hazard exists or a technology is dangerous unless shown otherwise by tests with very low prob-values. High confidence in the name of human safety may exact opportunity costs in knowledge or economic benefits foregone, however, while minimizing opportunity costs may increase danger.

Standards in hazard assessment science are, thus, inherently political choices. It is a policy decision to manage a hazard to promote human safety and accept opportunity costs or to manage it so as to minimize opportunity costs of regulation and accept lower levels of human safety. Assessment science and management policy must inform one another, but the relationship is unavoidably controversial. Media can create, enhance, or obscure controversy on any given hazard assessment and decision-making process through their coverage of the interaction between experts and activists, which may result in activist recruitment among the public to pressure risk management policy- makers.

Traditional print and broadcast media wring out the sensation and drama in a disastrous event or hazardous situation and then move on to other, more "newsworthy" stories, sometimes leaving information needs unmet. Responding to their own goals and constraints, such media are out of the control of risk assessment scientists and risk-management policy-makers. Activists are only marginally more capable of reliably hooking coverage.

The Internet is an emerging way to get information past the control of traditional media decision-makers. The Internet requires a vanishingly small price of entry compared with that required in the highly oligopolistic conventional media. It is also growing explosively, if unevenly, into a densely interacting global community. The Cassini controversy demonstrates the power the Internet offers to political activists to affect the agendas of policy decision-makers, particularly if the channels chosen include those not dependent on an audience actively looking for information on a given topic (as the Web does). Using e-mail, listservers, newsgroups, or chats, a handful of people can alert others to an issue of concern and enlist them to spread the news. The population passively receiving these notices expands exponentially and, even if just a small percentage of those exposed to the idea responds politically, the result can be tremendous political pressure.

This kind of audience-passive Internet communication offers a counterweight to the political influence of great corporations and wealthy individuals, which normally dominate the traditional print and broadcast media because of the cost of entry. Democratization is the great strength of the Internet in the sense that it allows a wider cross-section of society to generate effective political pressure.

The demagogic use of the Internet, however, remains the shadow of democratic empowerment. Sensationalism, conspiracies, ad hominem attacks, exaggeration, and other emotionally-manipulative devices are abundant in the Cassini debate, particularly among the opponents but also among flame-prone proponents.

The technological hazard case of Cassini raises issues of expert qualification, acceptability of risk, and the tension between democracy and demagoguery in cyberspace. Independent risk assessment of the plutonium on board Cassini deemed the hazard vanishingly tiny in probability and relatively trivial in consequence. This assessment did not comfort those who deeply dread nuclear technology in any form whatsoever or who profoundly mistrust government, and they found experts to claim high probabilities of disaster and extremely serious consequences, some of them speaking out in areas far from their expertise and without benefit of peer-review.

As this battle raged on in the listservers, chat rooms, and UseNet groups, it thus had all the appearance of dueling risk assessment experts. Expertise was, thus, delegitimated. People encountering the messages over the issue were left to find their own resources to decide if Cassini were a mortal danger or not, with an array of experts among whom to cherry-pick in support of their decision after the fact.

The complex nature of Cassini and of many other both technological and natural hazard controversies makes them baffling to the average citizen. The citizen yet must decide whether to act politically about this or similar situations or, worse, for a democratic society, remain uninformed and apathetic.

This is a dilemma we all face as citizens, scientists or not: We have to make political judgments, and we simply do not have the time to look into issues far from our training and everyday concerns. So, there are shortcuts to opinions -- we tend to defer to the opinions of people and organizations we trust. New media make it possible for a handful of people to tap this mechanism of trust and, through the ease and exponential expansion of activism-by-the-forward button, mobilize a lot of us into a politically potent movement over a socially amplified hazard, perhaps deflecting us from taking effective political action on more significant hazards. Hazard misperception exacts an opportunity cost in civic organizing time and energy.

What does this controversy teach about risk assessment and risk management decision-making when this new medium has emerged to shape hazard perceptions on the part of the public, from which all democratic legitimacy proceeds? For risk assessment scientists, the advent of the Internet means more effective opposition to the technology being proposed. This puts the onus on scientists clearly to state and justify all assumptions, procedures, and logic used to assess a risk. Among those assumptions must be a statement of the policy informing the assessment: How has the proper balance between human safety costs and scientific or economic opportunity costs been chosen?

There is now an urgent need to make all this transparent in documents written specifically and clearly for the lay public and placed online, and this public communication in many ways is harder than the original risk assessment. It would help if risk communicators became very familiar with the axes most likely to trigger public credibility anticipate how they will play out in a given situation. This may entail hiring of outside consultants, as most risk assessment scientists do not have the communications background or inclination for such a project.

The alternative is to have that process explained to the public by a hostile party. Jet Propulsion Laboratory (JPL) did put its various environmental impact statements online, ⁶³ and effort was made to justify the risks involved at a very elementary level. ⁶⁴ The intermediate level of explanation was dominated by the critics who were able to situate their explanations and criticisms within the larger suspicions of an alienated public.

Risk management decision-makers, particularly politicians will be hearing more frequently from a larger sample as the Internet becomes more and more ubiquitous. In this sense, Internet communications may better represent the feelings of the general electorate by dint of reducing small sample effects. As the effort involved in constituent communication becomes ever smaller, however, it is becoming harder and harder to discern just how much political commitment constituents have on an issue. Traditionally, elected officials assumed that one paper letter, because it took so much bother, might represent the feelings of a much larger number of constituents without the time to write. A swarm of e-mail, however, might come from a sample of people to whom the issue means just barely enough to hit a forward button.

In the case of a given technological or natural hazard debate, this sample, whatever it represents, may be responding to demagoguery, self-interest, or the well-informed consideration of risks and benefits: The source of political pressure may not be too apparent when decision-makers consider policy to manage a hazard. While one would hope they rely on risk assessment science in framing their responses, they must do so in an atmosphere of political risk and uncertainty, with its own Type I and Type II hazards to their own careers.

For activists, the Cassini controversy illuminates the possibility for tremendous empowerment of individuals. In this case, a handful of people became an effective vanguard for a mass movement that might just have been able to shut this mission down (and it might well have been successful had Internet organizing been available when the mission was most vulnerable to cancellation, during the 1992 economic crisis and the ensuing Congressional cost-cutting frenzy). On the other hand, the electronic frontier does not guarantee success: Cassini-Huygens was launched and the Earth gravity-assist went forward, both safely. Despite the exponential transmission of a handful of people's oppositional messages, the majority of people online remained unconvinced, despite appeals to the end of the world. There are, apparently, limits to success in online organizing. It is in the interest of activists to find out just what those limits are and respect them.

The Internet clearly can confer a measure of political agenda-setting power on a wide cross-section of society. Individuals highly placed in corporate management or at the head of major organizations and major political contributors may find their own power more often and more effectively contested (the WTO "Battle in Seattle" comes to mind) and diffused among "Netizens." In this sense of diffusing agenda-setting power, Internet organizing promotes democratization. Demagoguery, however, remains the shadow of empowerment, and its hallmarks are in this debate. How the dance of empowerment and demagoguery play out in the hazard debates of the future, it is well to remember Jefferson (letter to William C. Jarvis, Monticello, September 28, 1821): ⁶⁵

I know no safe depository of the ultimate powers of society but the people themselves; and if we think them not enlightened enough to exercise their control with a wholesome discretion, the remedy is not to take it from them, but to inform their discretion.