How much does informed consent actually waive?

One of the things I've been looking at recently is Neil Manson and Onora O'Neill's "Rethinking Informed Consent in Bioethics", which puts forward the idea of dropping autonomy as the justification for informed consent (with a bunch of compelling arguments, the most compelling being that we really just don't know exactly what autonomy means), and instead using informed consent procedures as a waiver of normative expectations. For example, taking a knife and stabbing it into a person's body is not usually something you would expect to be acceptable, but surgeons will commonly do so given that the patient has gone through a particular set of informed consent procedures. So informed consent here is being used as a waiver of the expectation that one would normally not get stabbed in the chest by someone else.

This sort of idea is obviously already in use (most medical informed consent procedures for example require a certain amount of disclosure of information, some measure perhaps of how much the consenter understood the information, perhaps also some sort of competency test, and then some sort of signing of a waiver that absolves the practitioner of responsibility), but what Manson and O'Neill do is cut to the chase and rub out any lofty goals of achieving autonomy and full understanding and rather difficult-to-obtain idealistic goals like that, and get straight into what it is that informed consent wants to achieve for all practical purposes.

So it is with this in mind that I find this article today (excuse the terrible pun in the headline), about a (rather stupid) man and his then (also rather stupid) girlfriend who wanted to scam a local council out of compensation by claiming that a wall fell on the girlfriend's leg, breaking it. Of course, what does the scheming couple do? The girl puts her leg up on some bricks and the guy jumps on it, breaking it in several places. Oh yeah, and their friend films it on a mobile phone camera.

He claims they filmed it so he had proof she consented to it, and she says she gave her consent to him jumping on her leg (and even suggested they do it in the first place). At the trial of the man, the judge seemed to think this consent was not a mitigating factor.

When jailing Thomson, Judge Francis Gilbert said: "This is an extraordinary case. You broke Miss Hingston's leg deliberately at her request and with her consent in order to make a false claim against the city council. Whether or not she consented or suggested the scheme, it is no defence. You inflicted really serious harm to another person deliberately and with pre-meditation for a wholly financial motive."

-- http://www.smh.com.au/news/national/chef-burnt-as-he-overcooks-compo-plot/2008/04/11/1207856774622.html

I find this quite an interesting case, because although obviously they're both quite stupid people (I mean come on...) she did give her consent. However there could be an argument to say that the man just took advantage of her stupidity, but how are we to know that for sure (unless he confesses to it)?

So to put that back into Manson and O'Neill's informed consent procedure, she acknowledged and agreed to a waiver of the jumping on her leg for a particular purpose. Now it seems that the actual jumping on the leg is not theoretically the issue at stake, but the reasoning behind it that the judge takes issue with. So how should we take that into account? Should we perhaps have waivers of waivers? Or restrict legitimacy of waivers to waiving "reasonable normative expectations"? What would constitute a reasonable normative expectation then? Obviously to this couple, it was a perfectly reasonable thing to do.

Some things to think about!

By the way, the couple are apparently no longer together!