Monday, April 28, 2014

Scientific Debate vs Political Debate, Victor Venema

A couple days ago I posted some thoughts about the difference in debate styles over at my WUWTW.  Reflections on my years long and oh so futile efforts to engage climate science "skeptic(s)" in an actual on-point constructive "debate" focusing on rationally comparing our respective 'scientific facts' and "observations" in order to explore which arguments carry the most weight. 

On the same day Victor Venema posted his own observations comparing a "scientific debate" with the public debate.  Since Victor is a real scientist he does a nice in-depth dissection.  It's definitely worth reading - mine is earthier given my own simpler "spectator's" perspective.  Still I like to think they make a nice complement and I'm happy to share his article over here at CC.   

Worth adding is Ezra Klein's "How Politics makes us stupid" looking at Dan Kahan's work on the politics of how we process scientific information.:
" The theory is particularly prevalent in Washington, where partisans devote enormous amounts of energy to persuading each other that there’s really a right answer to the difficult questions in American politics — and that they have it. 
But the More Information Hypothesis isn’t just wrong. It’s backwards. Cutting-edge research shows that the more information partisans get, the deeper their disagreements become. "   E.Klein
~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~

Sunday, 27 April 2014

Are debatable scientific questions debatable?

In a previous post I tried to make clear why debates are such a bad way to improve scientific understanding. In the comments Mark Ryan pointed to a great article by John Zyman: "Are debatable scientific questions debatable?" In the article, political debates are compared to scientific disputes. Zyman seem to agree with me: a short summary of his position could be that the only similarity between political debates and scientific disputes is that they are both public.

The main difference is that while debates are typically verbal, scientific disputes are resolved in the scientific literature. That makes a large difference. In the scientific literature the ideas have to be and can be presented in all their gory details. Detail that allows one so see its errors or improve upon the idea. A written dialogue is also slow, which helps to check details and to rethink ones position multiple times before answering.


In the article a distinction is made between political debates and scientific disputes.
I shall use this word ‘debate’ as a term of art for an argumentation ritual defined by the following features:

* A debate is a one-off, public event, initiated and completed within a limited period of time. Typically, it is conducted in a forum before an audience, most of whom take no direct part in the proceedings and may be quite ill-informed on the topic. ...
* A debate is conducted orally, in direct speech, by named participants. ...
* A debate has a specific topic, in the form of a ‘black or white’ proposition. This is typically understood to be a ‘question’ to be answered ‘yes’ or ‘no’ ...
* The argumentation is polarized, and balanced ritually between a protagonist and an antagonist, (‘proponent’ vs ‘opponent’, ‘plaintiff’ vs ‘defendant’ , etc.) ...
* The proceedings are adversarial, in that each ‘party’ endeavours not only to make their own case but also to negate the case of their opponents, even to the extent of attacking their credibility as expert advocates or cross-examining them severely on disputable points.
* A debate is normally conducted by a chair (or ‘speaker’), who simply ensures that the rules are observed. ...
* The debate almost always concludes with a decision — i.e. a ‘yes’ or ‘no’ answer to the question— determined typically by a simple majority of those who have (supposedly!) attended to the debate and are qualified to vote on such matters. Sometimes this includes the whole audience ...
A debate is about winning and convincing, it is not about improving our understanding of reality, of the problem at hand, not even of the position of the other party. As it is a one off event, presenting misinformation is an effective strategy, especially if you have no reputation to lose, like the climate pseudo-sketpics. Another effective debating strategy if you only want to win is a Gish Gallop, after the strategy of the creationist Duane Gish, to spew such a gallop of misinformation, that the other side would not have enough time to correct all the falsehoods.

Scientific dispute

The rules of the game are very different in science.
"Scientific disputation ... typically involves sequences of written arguments and counter-arguments in the official scientific 'literature', rather than oral exchanges in a public forum."
Scientists naturally do talk with each other a lot, but what counts is the literature. There are presentations at conferences and workshops, fitting to scientific presentations not being that important"a remarkably low level of verbal fluency is acceptable. ... The conventional presumption, emphasized by stock courtesies from the chair, is that the theme of the paper [i.e. presentation] is worth of serious attention, even if, at first hearing, it seems incomprehensible, wildly speculative, inconsistent, totally boring or otherwise unconvincing or uninteresting!" This statement is funny because it is so true. Not uncommonly you have to make quite an effort to understand the speaker and often a talk is mainly useful as an announcement of a new article.
Scientific papers ... [do not] avoid controversial statements. On the contrary, one of the norms of science is that a would-be contribution to knowledge must always be 'original'.
I am not sure if I fully agree with this. While scientific papers do not avoid controversy, they also do not aim for it. Papers should be original. Original and controversial are different. You can have a new way to compute something already known, you could analyze a new dataset about which no one had a preexisting opinion, etc.

I had to think some time to find topics under scientific dispute. That presupposes that two people or groups hold a strong opinion about something without sufficient evidence. It is more common that scientists keep all options open until the evidence is there. A scientist may even use two conflicting theories depending on circumstances, such as relativity and quantum mechanics. "Scientists learn by experience to hold simultaneously in mind a number of uncertain, perhaps inconsistent, ideas, without begin paralysed by logical gridlock or complete scepticism." Also if a paper changes our current understanding, if the evidence is strong, there is no reason for dispute, it is simply accepted. Or as Zyman writes:
Ideally, [innovative ideas] are presented as if they were already part of the (thus newly-achieved) consensus, in a form that is so unquestionable that it can be treated as camera-ready copy for a new chapter in the next generation of text-books!
The only scientific dispute in my surrounding that is not related to the political climate debate, is whether station measurements contain platform-like pairs of break inhomogeneities of short duration. More precisely, whether this is more common as one would expect from random breaks. That this became a dispute was because a decision with a deadline had to be made whether to include them in our HOME validation dataset for homogenization methods. Doing both was unfortunately no option because of man power. The need for a decision, like in politics, did not allow us to withhold opinion until enough evidence was gathered. Fortunately, it was a friendly dispute as no one expects these platforms to be too important. The other two disputes I have experienced, long memory in temperature series and naturally the so-called hiatus, I would view as intrusions of the political debate into science.

Another ritual is that "every paper starts with a very conventional 'survey of the field', including a lengthy bibliography that cites all the usual suspects. ... It can be interpreted functionally as a rhetorical device designed to show that the author is so completely at home in the field that their novel ideas are worthy of attention." It also helps the peer reviewers to judge the novelty of the work and avoids double work by forcing the researcher to be up to date. Its main role for scientific progress is that it aids think neutrally, because it forces the author to write about the work of others in an objective way. "Every author is well aware that one of the reviewers may support another side in such a controversy and is likely to turn down a paper that does not present fairly the case of that side." If that is the only peeve, the reviewer will more likely ask for a revision, rather than immediately request rejection.

The questions also have a different nature. A debate, in a formal sense, is normally about the acceptability of a specific proposition, but scientific controversies are seldom that simple. ... In reality, scientific progress does not follow Popper's recipe of successive theoretical conjectures being winnowed away by empirical refutations. As other philosophers have pointed out, there are always just too many other untested assumptions that might be the cause of the apparent disconfirmation.While I do not think that Karl Popper was that naive, it is clearly true that if a discrepancy is seen in science, it is not clear in advance where the problem lies. Finding that is the largest part of solving the problem.

Further differences are:
* The audience is assumed to be already very well-informed on the topic;
* The proceedings are not adversarial, as between officially identified protagonists and antagonists;
* Attacks on the personal motivation or credibility of the disputants are unacceptable;
* The discussion does not close with a formal decision or verdict.
Conventional 'debating' practises just do not fit into the evolution of scientific knowledge in its traditional academic mode. In effect, 'debatable' scientific issues are never actually 'debated'. One must therefor seriously ask whether much of value can be achieved by a formal procedure designed to bring scientists together for just such an activity.
All the more so for bringing scientists and anti-scientists together. It will certainly not help scientific progress. I fail to see how it would help in communicating our current understanding to the public. A documentary seems much better suited for that.

1 comment:

Victor Venema said...

Thanks for the repost. I should have thought of this post when Brian Cox was debating a right wing politician on climate change on Australia's television program Q&A. That would have been a good time to remind everyone of how impossible the task of Brian Cox was.