Imagine that you are a specialist in climate science. Like 82 percent of your colleagues, you believe that “mean global temperatures [have] risen compared to pre-1800s levels, and … human activity [has] been a significant factor in changing mean global temperatures.” You worry about the consequences, which may range from acute suffering in the world’s poorest countries and loss of natural species to global catastrophe.

You also know what science is like–it is always uncertain and provisional. Every article has a “limitations” section, every data table has margins of errors and sources of bias, and rarely do two articles precisely agree. Nevertheless, you know that to change the course we’re on will require millions of people to alter their political and consumer preferences. But people are fairly selfish and short-sighted. Besides, we have lots of other things to worry about, from our day-to-day practical struggles to spiritual concerns, plus all the alarms we receive from the mass media: serial killers, terrorist attacks, corrupt politicians, swine flu.

Given all this clutter, you, the climate scientist, decide that you’d better become much more effective at communicating a sense of alarm. You are constrained by ethical limitations (no outright lying, for instance, even to save the planet), but simplification, evasion of complexity, exaggeration of certainty–all that seems necessary.

These are the habits that one can see in the leaked private emails of climate scientists. Their messages include mentions of “tricks” in the presentation of data, data withheld from direct public inspection, and references to skeptics as “idiots.” Reactions to the emails range from George F. Will (the documents “reveal paranoia on the part of scientists … [N]ever in peacetime history has the government-media-academic complex been in such sustained propagandistic lockstep about any subject”) to Paul Krugman (“all they show is that scientists are human, but never mind”).

In my view, the emails reveal a shift from one kind of communication to another. Borrowing a distinction from the contemporary German philosopher Jürgen Habermas, I’d distinguish strategic, instrumental, means/ends communication from deliberation or dialog. When communication is strategic, you know what your goals or ends are, and you use efficient means to convince others. When communication is dialogic or deliberative, you reason with the other party about what the goals and means should be.

The leaked climate emails show scientists becoming strategic rather than dialogic. The reason is clear: modern society is so structured that strategic communication generally beats dialog, at least in the short term. It simply works better.

Yet strategic communication is unethical, insofar as it tries to manipulate the other person’s reasoning capacity. It uses him or her as a means, not an end. It is also self-destructive in the long term. Our views of matters like climate change depend fundamentally on trust. I cannot directly sense changes in the climate, let alone their causes. Neither can scientists–despite their fancy equipment. An account of how and why the climate is changing requires aggregating the research of many scientists and collaborative teams. To use the aggregated information, you must trust all the contributors. Then, to make matters even harder, people like me don’t read any of the scientific literature on climate. We read what we regard as high-quality news coverage of the scientific literature, which means that we must trust some reporters, as well as the scientists they cover. And we must trust the reliability of the relationship between them.

All of this works if we assume that scientific discourse and high-quality journalism are not strategic forms of communication. They are not supposed to pre-judge the outcome and try to convince. Rather, they are supposed to explore the truth in the company of their readers. To the extent that they communicate strategically, they are just interest groups, basically like all the others. They have goals; they may be willing to negotiate; but they cannot persuade on the basis of trust.

This analysis suggests a real dilemma. Dialogic communication won’t change mass opinion, and counting on it may put the earth at risk. But strategic communication is unethical and ultimately self-defeating. It’s the nightmarish side of modernity.