Richard says; Jameson, I support your move to ignore someone who doesn't listen.
Debate is supposed to involve actually wanting to understand other points of view. But some people aren't really interested in understanding. [endquote] This vagueness--implications about unspecified instances of not listening or disinterest in understanding the other person's point--is hardly surprising, coming from someone who has shown himself to be not so good at saying what he means in physics. I'm going to comment on the above-quoted statements, because they refer to, and exemplify, some of the most typical and common deficiencies in EM discussion. Richard says that I don't listen. ...to what? I'm not just kidding when I say that saying what you mean, and knowing what you mean, are really all-important in discussions at discussion forums, including EM. It's tempting to want to make angry noises, to imply errors or wrongdoings. But specifying what one actually means or is referring to is a bit more demanding, and so many people, including Richard, are tempted to delude themselves that they've adequately made their point when they've been entirely vague about what they mean. I've sometimes been criticized, at various forums, on various topics, for introducing new definitions...sometimes of new terms, sometimes supplying a clear and precise definition for a term already in use. Sorry, but clear, complete and precise definitions are needed for discussions. For example, Richard complained that I'd defined new criteria, and introduced new terms (to refer to new distinctions). Some of the criteria that I've defined are in wide use. The winning-votes Condorcet class of methods, introduced and initially advocated by me, has widespread popularity (...unfortunately. I no longer recommend it). But it isn't important what definitions we use, as long as we're consistent, and as long as we're quite clear about what definition we're using. That's the problem, of course--we often aren't. Anyway, when you want to say that someone doesn't listen, that statement is entirely without meaning unless you can be a bit more specific. What, in particular, hasn't that person listened to? Richard continued: Debate is supposed to involve actually wanting to understand other points of view. But some people aren't really interested in understanding. [endquote] Another instance of flinging charges around, without any support, justification, or specification of an instance of what the speaker is referring to. Because Richard made that comment in regards to a discussion with Jameson, then I'll point out that I've made as much effort as can be reasonably expected, to try to find out what Jameson was trying to say, to try to find out what his point was, and how he justified it. How many times did I ask him for that justification? Several times I clarified, and re-clarified, that it would be great if a method as easily-counted as Majority-Judgment (MJ) could significantly avoid, alleviate, ameliorate or solve the chicken dilemma. So, interestedly, I asked Jameson to tell how it would do that (He'd previously claimed, a few times, that MJ accomplishes that). Evasion doesn't accomplish anything worthwhile. The only worthwhile way to answer an argument is to answer in terms of what the person meant, in terms of what that person believes &/or means. I've always answered in that way. And, unlike Richard and Jameson, I haven't evaded answering. I've answered everything. And whenever someone claimed that I didn't answer what he meant, I asked him to better specify what he meant. Richard says I don't try to understand other people's point of view? Before starting my "Properties" article-series at Democracy Chronicles, and for the purpose of that article, I asked some questions at EM. Wanting to give a fair opportunity to state their position, to advocates of traditional unimproved Condorcet (TUC), I asked them, at EM, to answer some questions about what advantages TUC has, to counter some specified disadvantages that TUC has, in comparison with Approval, and in comparison with Symmetrical ICT. I asked. But only you can answer. I did my part by asking. Receiving no answers for a fair amount of time, I began the article-series without the TUC defenses that I'd invited. Afterwards, one person did cite two advantages of (some but not all) TUC methods: Clone Independence and Reversal Symmetry. I immediately answered that. I told why I didn't consider those criterion-failures to be important for Approval, Score, ICT and Symmetrical ICT. Citing criteria isn't enough. Someone should always be prepared to tell why the criteria that he cites are important...specifically what problems must result from failing those criteria. Telling why I didn't consider the criterion-failures to be important is an essential part of the use of criteria to evaluate methods. I briefly mention here that Kemeny fails Clone-Independence. I took the time, and devoted the article-space, to answering that person's statements about criterion-failures. Have I neglected to make the effort to understand Richard's point of view? Regarding what? If Richard would be more specific, then I'd answer whatever point of view he's referring to. I can't very well answer it without knowing what he's referring to. My main point at EM has been that drastically preference-distorting strategy-needs shouldn't be disregarded. Favorite-burial need makes nonsense of election-results--now, with Plurality, but also with other FBC-failing methods. Some of us discussed the matter of how often TUC will fail FBC. I listened to the claim, the point of view, that TUC's FBC-failure won't be a problem, because the failure doesn't happen often. Instead of ignoring that point of view, I respected it enough to reply to it. Always reply to someone where they are, speaking to the beliefs that they have. So I pointed out that our mass media consistently hammer home the belief that the winner will always be a Democrat or a Republican, and the assumption, constantly supported by the media, that, though there's corruption, it's universal and inevitable, and therefore acceptable, and that, in particular, the Democrats are acceptable, even if (to a Democrat-preferrer) a Republican might be unacceptable. I pointed out that , given those beliefs, promoted by media, and continually expressed by voters, the optimal strategy with TUC is to vote the Democrat alone in 1st place. We could argue endlessly (as some TUCists have) about how people will vote with TUC. But what we _can_ say with certainty is what the optimal strategy is (given the beliefs that voters are known to have). So I pointed out that what we can say isn't something favorable to TUC. TUCists continue to try to wriggle, make optimistic self-serving assumptions, etc. but the fact is that a look at optimal strategy doesn't look good for TUC. One must wonder why some want to devote so much effort to evasion about that, when there are methods that don't fail FBC. Simple and particularly easily-counted ones, like Approval and Score. Approval is the uniquely simple and elegant voting system, and the most easily counted, after Plurality. Additionally, there are methods that additionally meet CD, and automatically avoid the chicken dilemma. Symmetrical ICT is a Condorcet method that, while meeting FBC, also meets legitimately-defined Condorcet Criterion, and automatically avoids the chicken dilemma, and effectively share's Approval's Later-No-Help compliance (by complying with 0-info Later-No-Help). Additionally, for the general case, TUC's optimal u/a strategy is unknown. Comparison of strategy and strategy-incentive doesn't look good for TUC. I mention this topic because it's the main thing that EMers and I disagree on, and so that makes it a good example, to bring out discussion-ethics. I've made every effort to elicit and reply to the arguments and points of view of others. In particular, I've asked Richard how he expects anyone to accept a voting system for which his definition requires several pages. I've asked him what about it is importantly good enough to justify Kemeny's NP Hard count computation. I've asked him how long it would take to handcount it, with 100 count-workers in each precinct, and 1000 at central count headquarters, in an election between 20 or 25 candidates, with 100 million voters. In fact, how long would it take to count a 25 candidate, 100 million voter, election, using a typical home computer? I don;t ignore people's point of view. I often ask for it. But some people would rather just make angry noises and unsupported assertions. Mike Ossipoff ---- Election-Methods mailing list - see http://electorama.com/em for list info
