On this blog I will give detailed criticism of political texts (e.g. opinion pieces), debates and speeches from an argumentation theoretic perspective. Examples of the kinds of things I will point out include:
- Failure to present evidence for factual claims
- Failure to present arguments for controversial ideas
- False statements
- Misleading statements
- Invalid arguments
- Failure to address obvious counter-arguments
- Vague or ambiguous language, especially when that makes it more difficult for the reader to evaluate the author’s claims
In short, I will criticize authors who breach what the philosopher Paul Grice termed the co-operative principle (a number of more specific maxims, which broadly correspond with the opposite of the above bullet points, fall under this general principle).
Make your contribution such as it is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.
Though Grice claimed that people usually do follow this principle, it is easy to see that it is sometimes breached. This is especially true of political debates, which can be seen as a sort of prisoner’s dilemma. More often than not, defecting from the co-operative norm brings you an advantage (for instance, false or misleading claims can be used to defuse counter-arguments to your thesis). At the same time, as a collective we would, presumably, be better off if everybody followed the co-operative principle. This is precisely the kind of situation that the prisoner’s dilemma describes: while the collectively optimal situation is that where everybody follows the co-operative principle, each individual is better off if they defect (regardless of whether the opponent defects or not).
One way to escape the prisoner’s dilemma is by increasing the costs of defection or cheating. Breaching the co-operative principle is only effective in case this is relatively unnoticed. If every false statement came with a little tag saying “this statement is false”, the number of false statements in political debates would presumably drop radically (given that there is, as Grice rightly points out, a strong norm to the effect that we should not breach the co-operative principle, either through lying or in any other way). Hence pointing out argumentative errors can be a very effective means of forcing people to adhere to the co-operative principle.
Already today there are costs attached to breaching the co-operative principles. Salient breaches of the co-operative principles tend to get pointed out not only by political opponents but also by impartial media. Here factchecking sites such as factcheck.org and politifact.com and Washington Post’s Fact Checker blog deserve special mention. They meticulously go through politicians’ statements of fact, which gives politicians incentives not to utter any falsehoods.
TV and radio journalists who interview politicians also try to make breaches of the co-operative principle as costly as possible by pressuring politicians who are misleading or evasive. A good example of this strategy is that used by the BBC program HARDtalk. In HARDtalk, the interviewer takes a more “activist” approach compared to many similar programs. Interviewees who fail to answer the interviewer’s question or who make dubious statements are frequently interrupted and pressured (hence the name).
These efforts can only be applauded. My hunch is that if it were not for these and other journalistic criticisms, the public debate would have been far worse than it currently is. (It’s not easy to prove this hunch right, though one possible method could be to look at correlations between standards of political debates and standards of critical media cross-nationally and cross-temporally, while controlling for other factors.) If this is right, it makes sense to try to raise the price of breaching the co-operative principle even further, since that should improve the standards of political argumentation (which surely aren’t where we want them to be at the moment).
It seems to me that the standards of journalistic criticism gradually are improving (e.g., HARDtalk and the various factchecking sites are comparatively new). Things are moving in the right direction. My aim is to contribute to these endeavours by giving detailed criticisms of all breaches of the co-operative principles that I find in various political texts, speeches or debates. Hence I will not only criticize factual errors, but also the other kinds of argumentative errors mentioned above.
Now in order for an argument-checking website such as mine (or a fact-checking site, for that matter) to raise the standards of political debates, it needs to have three qualities.
- It needs to have a wide readership.
- It needs to have a comprehensive coverage (i.e. you need to cover a large part of the political debate)
- It needs to be perceived as fair and accurate
My hunch is that if you reached all of these goals – i.e. if an argument-checking website perceived as fair and accurate offered a comprehensive coverage and got a wide readership – then you could improve the standards of political debate greatly. However, getting to that point promises to be very hard (perhaps it’s even impossible). My goal here is to test the waters, as it were – to check whether a grand-scale argument-checking project is feasible, and what methods it should apply. I’m therefore grateful for any input concerning my analyses of political texts, debates and speeches, as well as on the overall project.
For more discussion concerning this project, see this blog post of mine at the LessWrong website. I got a number of comments there, and especially liked this one:
There have been many proposals like this before. My favorite idea (which I cannot recall the name of right now) was a browser plugin that would overlay annotations onto arbitrary webpages. People could make it highlight certain questionable bits of text, link to opposing viewpoints or data, and discuss with each other whether the thing was accurate. Imagine a wiki talk page, but for every conceivable site.
I don’t know whether that’s technically and legally possible, but it’s a very nice idea. In effect you would indeed put a tag saying “this statement is false” on all false statements under this plan (and similarly for other argumentative errors). That is bound to have quite an effect.