From akriman@darwin.helios.nd.edu Sun Feb 9 03:51:25 2003 Received: from mxu1.u.washington.edu (mxu1.u.washington.edu [140.142.32.132]) by lists.u.washington.edu (8.12.1+UW01.12/8.12.1+UW02.12) with ESMTP id h19BpOCK039664 for ; Sun, 9 Feb 2003 03:51:24 -0800 Received: from mailspool.helios.nd.edu (mailspool.helios.nd.edu [129.74.250.7]) by mxu1.u.washington.edu (8.12.1+UW01.12/8.12.1+UW02.12) with ESMTP id h19BpKfV020617 for ; Sun, 9 Feb 2003 03:51:20 -0800 Received: from darwin.helios.nd.edu (darwin.helios.nd.edu [129.74.250.114]) by mailspool.helios.nd.edu (8.9.2/8.9.2) with ESMTP id GAA04444 for ; Sun, 9 Feb 2003 06:51:19 -0500 (EST) Received: (from akriman@localhost) by darwin.helios.nd.edu (8.12.6/8.12.6/ND-cluster) id h19BpJol011504 for classics@u.washington.edu; Sun, 9 Feb 2003 06:51:19 -0500 (EST) Date: Sun, 9 Feb 2003 06:51:19 -0500 (EST) From: Alfred M Kriman Message-Id: <200302091151.h19BpJol011504@darwin.helios.nd.edu> To: classics@u.washington.edu Subject: selector > Pay no attention to the man behind the screen! MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable X-Uwash-Spam: Gauge=IIIIII, Probability=6%, Report="SPAM_PHRASE_01_02, __EVITE_CTYPE" Since there was so much interest in the results of the moral philosophy selector (http://www.selectsmart.com/PHILOSOPHY/) on Friday, I investigated its innards. In the usual classroom situation, there are many test-takers and a single grader. You can rank the test-takers by their grades. This selector is like that, but when you visit you are the single test-taker and there are nineteen graders ("philosophies"). It is the different graders' scores that are "curved" (actually just scaled so the highest grade is 100%) and ranked. In a way this is much like the true school situation, where higher (i.e., easier) graders tend to be more highly ranked by students. A philosophy, so far as the selector is concerned, is a score sheet: a list of points to be assigned to possible answers. In a later posting I will explain how the overall scores are computed, but that information is not essential here. You can think of the score sheets as the results of a survey. In this survey, each philosophy is asked twelve questions, and for each question is offered two or more possible answers. The philosophy responds to each suggested answer on an eleven-point scale from 0 (Hell no!) to 10 (Amen!). However, I'm going to stick with the grading metaphor, because everyone is familiar with the idea of grading individual answers first and weighting and curving later. To review briefly what a visitor to the site observes: There are twelve numbered questions, each with three or more possible answers, labeled a-c or a-d or whatever. The answers are selected by radio button without a default selection. That is, when you first load and when you reset the page, no answer is selected for any answer. Once you select an answer, you can change it, but you can choose only one out of the choices given. The last answer for each question is "Doesn't matter/Dislike all answer choices," which I call the catch-all answer. For each question, you have the option of placing a priority on your answer: high, medium (the default), or low. I ignore that completely in this post. Upon submitting your answers you get grades for nineteen philosophies, mostly labeled by the name of some philosopher. (You can get the list in alphabetical order by pressing the submit button without answering any question.) What experiments reveal: Selecting the catch-all answer to a question is equivalent to not selecting an answer -- you get no points for this response. For the other answers, you get anywhere from zero to ten points. I will post the complete list. It is also temporarily at . By far the most common score is 0 -- most philosophies grade most of the answers as dead wrong. The most common nonzero score is 9. I guess if each philosophy had had a chance to write its own exam and choose its favorite wording, 10's would be more common. As it is, a 10 is like extra credit for being perfect. Question 7, for example, asks for your opinion of the categorical imperative (in so many words) and Kant will award you ten points for an unqualified "yes" (answer 7a). Prescriptivism will give you a 10 for answer 1c (agreeing that moral statements are "primarily... statements of command"). Rand is by far the most generous extra creditor, giving 10 for answers 4d ("society should ... not be an influence on a person when one is trying to find virtue"; this doesn't quite address the question it "answers"), 9e ("acting in one's own self-interest [and against others'] is morally essential"), as well as 11c and 12b. On Q11, about ascetic life, Rand gives a 10 for "physical comforts are fine, they may even be rewarding." Epicureanism only gives this a 9, and I guess I can see an argument for that or less. On the other hand, Cynicism gives a 10 for the answer that "it is essential to live this [ascetic] way." Perhaps someone can weigh in on this one. ( I never think of the cynics as principled ascetics -- I think of them as performance artists. The story is recorded by Diogenes Laertes, that Diogenes of Sinope only took to the tub when he couldn't get real property. Did Diogenes turn over a new leaf after being caught adulterating the currency? Yes, later he apparently showed real commitment to staying poor, offending Alexander, teaching his masters' sons to do for themselves.... But thinking of this now, I remember the ostentatious humility of the hermits Twain lampooned in _Connecticut Yankee etc._ And Dostoyevsky described fiercely competitive hermit piety in _Brothers Karamazov_. If that were all that ever became of asceticism, only then would the cynics qualify, ISTM. ) An interesting question to test the test is Q6: "Will using morality properly [sic] necessarily result in maximum happiness?" I'm happy and VERY surprised to report that most (14 of 19) philosophies gave 8 or 9 points for the answer 6a ("yes"). Most of the rest gave high scores to "no, not necessarily." Intriguingly, however, Nollings gives zero points to both answers. I don't see much middle ground between "yes [necessarily]" and "no, not necessarily." I'm pleased to report that Aristotle backs me up on this (apparently excluding the middle as he scores "yes" 9 and "no" 0). I suppose I can understand the Nollings answer as riding the "doesn't matter" side of the double-edged catch-all answer, although that strikes me as more of a stoic answer (yet Stoics align with Ari on this). The real problem here, as I see it, is not logical but quantitative. Nollings may be right to shun both yes and no, and go for some sort of jein, but you don't catch any flies with that vinegar. There are no points for the catch-all answer, so agreeing with her position doesn't score any points, whereas agreeing with either more conventional position does score points for or with other philosophies. Nollings doesn't give any points on answers to Q7 or Q11 either, but there's worse. If you want a tough, tough grader, if you want to see lots of red on your exam paper, you're asking for the slasher: Ockham. I can just see him proctoring an exam, playing absent-mindedly with his strop: "Let us not multiply answers unnecessarily, boys." On one third of the questions (Q4, Q7-9), he gives everyone a zero. He doesn't multiply grades either. So much for the tough graders. On the other end of the spectrum we have the philosophies that complaisantly accept more than one answer as "pretty much correct" (not a direct quote). On the same happiness question (Q7), Bentham gives 8 points for "yes" and 6 for "no," rather straddling the fence, if you ask me. Sartre gives 6 points for either answer. If hell is an eternity with friends, then I suppose this answer makes sense. But Sartre is master of and-you're-right-tooism. On Q12 (describing a virtuous person), he awards 7 points for any of six answers. And it may well be that a virtuous person is strong, powerful, restrained, humble, passionate, very rational, spiritual, caring, loving, and concerned with others. Jean-Paul Sartre, Boy Scout. In principle, I have no problem with partial credit, but I expect simultaneously correct answers to resemble each other. This can easily happen when there are a number of partially overlapping answers, as in Q2. ("Does each person have a moral purpose/morally ideal way to live?") Epicureanism accepts both 2b and 2c (9 pts. for either "yes, but..." answer) and Sartre accepts 2d and 2e (9 pts., either "no, yet"). Nietzsche awards 9 pts. for these: (2b) Yes, but the way to live in order to meet that purpose is unique for each individual (2d) No, yet there are ways to act that are inherently more conducive to the self-interest of the person who is acting I feel it is morally incumbent upon me to confess, honestly, that I know approximately squat about moral philosophy. Safe-cracking is more my line. Just going, then, by the ten that Nietzsche scores for 9e ("and acting in one's own self-interest is morally essential"), I guess answers 2b and 2d are not too inconsistent. When many answers score highly for a philosophy, it introduces a bias. Other philosophies might score as highly if their preferred answers were given in detail, rather than hidden in the zero-scoring catch-all ("dislike all answer choices" _offered_). A way to quantify this effect is to sum all the points for all the answers given by each philosophy. That's done in the bottom row of the appendix table. The numbers range from 92, 114 and 116 (Ockham, Prescriptivists, and Nodding, resp.) to 188 (Spinoza) and, with Christian grading charity abounding, 225 (Aquinas). Many are called, few are chosen? This might explain some unexpected Aquinas matches, as well the fact that no one reported matching Ockham, Prescriptivists, or Nodding. AMK .