Tuesday, 21 May 2013

Lord Tebbitt commits some Logical Fallacies

In a bid to make myself feel like a good samaritan I bought a copy of The Big Issue today. In it, Lord Tebbitt, who is, shall we say, slightly opposed to the idea of gay marriage, makes several charming statements:
“When we have a queen who is a lesbian and she marries another lady and then decides she would like to have a child and someone donates sperm and she gives birth to a child, is that child heir to the throne?’
“It’s like one of my colleagues said: we've got to make these same sex marriages available to all.
“It would lift my worries about inheritance tax because maybe I’d be allowed to marry my son. Why not? Why shouldn't a mother marry her daughter? Why shouldn't two elderly sisters living together marry each other?”

What a nice man. I write about this today though, not to make any comment on the issue itself (I know people who have strong opinions in either direction) but because it provides an excellent platform for discussing something I find interesting - logical fallacies.

First off, the obvious - The 1949 Marriage act has a completely different set of rules for marrying relatives; they're listed out in the Table of Kindred and Affinity, which quite sensibly states that not only can you not marry your son, you also can't marry his wife if he is still alive. Also there are seperate incest laws, not to mention the possibility of getting accused of tax evasion, although ultimately I discount the idea that Lord Tebbitt genuinely wants to marry his son and is instead not being serious. This is our first fallacy, and it's called the Appeal to Ridicule - presenting an opponents arguments in a way that makes them seem absurd. 

If I was being cheeky, I could also call the middle sentence an Appeal to Anonymous Authority - if this colleague said it, why must it be true?

But the first paragraph is truly a gold mine. Here, he posits a series of occurrences, which leaving our fallacies for a minute deserves a closer look. The question posed at the end relies on a number of things happening, but let's for a minute assume that he could be talking about the Duke and Duchess of Cambridge's baby. Below, I list the assumptions his question has to make, followed by my best statistical estimation of them actually occurring.

  • The baby is a girl
    • 50%, give or take. The couple may be genetically predisposed one way or the other (but we don't know), and there is a slight difference in the rate of boys to girls at the moment, but not enough to shift the percentage.
  • The girl is gay or bisexual
    • Bit trickier, this one, but the UK census suggests a rate of 1.1 % for the former and 0.4 % for the latter. This obviously only includes those who self-identify as gay, but I realized this isn't a problem because Lord Tebbitt assumes/implies that this Queen must have self-identified at some point.
  • The girl is married
    • This is where it gets tricky. Assuming that all those in a civil partnership would get married if the opportunity arose (a statistical leap in itself) and that the rate stays the same (again, bit of a stretch) there would be around 53,000 same sex marriages in the UK at this mythical point in the future. Assuming it takes two to tango, that puts the number of people (in total) in same-sex marriages at around 0.16 % and (assuming straight people aren't in same-sex marriages) the number of gay people at around 14.5 %.
  •  The girl wishes to have children via artificial insemination
    • This is apparently about 41%, according to one survey. I am not going to cite all of my references, as these are all approximate.
  • It is successful
    • This is tricky to estimate, but assuming she's under 35, the success rate is about 32.5 %, according to the NHS, and the percentage declines after that point.

How likely is Lord Tebbitt's hypothesis? By multiplying all of these together, the odds come out somewhere around 0.1 %. I think the primogeniture question of the throne is pretty safe.

So which fallacy is this he's using? This is called using Misleading Vividness. He has described a  scenario in such detail that you see the problem before considering how likely it all is. Interesting stuff.

And if you think, even after all this, that this particular argument is one that still needs to be thought about, I direct you to your own personal fallacy: the Appeal to Probability. Just because something can happen, doesn't mean it will, even if it is likely. This is very, very unlikely.

The conclusion we can all draw from this is that this doesn't exactly add anything to the debate about gay marriage on either side, just that we can safely ignore Lord Tebbitt and instead look at this cool diagram of logical fallacies.

1 comment:

  1. By request

    Boy/girls stats: http://www.telegraph.co.uk/news/uknews/1511041/More-boys-than-girls-born-each-year.html

    Rates of Homosexuality:

    Marriage/Civil Partnerships:

    Children desire:

    IVF success rates: