The purpose of marriage

I listened to the second Reith lecture this year
http://www.bbc.co.uk/iplayer/episode/b00l0y01/The_Reith_Lectures_The_Reith_Lectures_2009_Morality_in_Politics/
which is entitled 'Morality in Politics'. Professor Michael Sandel discusses how politics should not attempt neutrality on vexed moral questions, but needs to engage in debate and resolve them.
He also mentions how the attempt to be distanced from moral debate has led us to a flawed tolerance of disagreements, in which we pretend to accept others' opinions whilst finding them fundamentally disagreeable. This chimes nicely with a quotation which I have cherished for some time (and which I took from a games designer called Mark Rosewater, but which is not lessened by its unusual pedigree) and which ends like this:
'When did everyone get so afraid of conflict existing? Conflict is a wonderful thing. It causes resolution.'
We need proper debate about these problems, and that is the one thing that we are not getting. Another news source today mentioned that a few decades ago a politician got about 43 seconds to make a point. A few years ago it was 9 seconds and it is now 6 seconds. The article was attempting to blame computer games and video for shorter attention spans, but whatever the cause, it is clear that we are getting less substantive debate, not more.

Professor Sandel went on to talk about one thing especially: how the state cannot avoid being involved in moral questions. His first example was one in which a disability law made the question of the essential nature of golf a legal question, to be decided by judges, and his second was the subject of homosexual marriages.
States currently all support heterosexual marriages, and Professor Sandel clearly distinguishes between the options of equally supporting all types of marriage, supporting homosexual relationships in a lesser way or under a different name, and not supporting any marriage. As he says, marriage is patently not about procreation, despite many people's protestations.
In his opinion, marriage is about honouring the mutual and exclusive dedication of two people to each other. This honour is conferred not only by name but also through various legal benefits.

His analysis is clear and concise, if a little slowly spoken. It is about other points that we differ. Whereas he believes that the state must support homosexual marriage as an equal with heterosexual marriage, since the individuals have made identical commitments, I think that his examination is enough to show us that the whole institution of marriage is unsound.

Firstly, I see no reason for the state to honour mutual and exclusive dedication at all. Why this is a special form of relationship entirely eludes me; I have no arguments against it, but fail to see justification also. It simply appears ludicrous.
Secondly, I think that legal benefits are better given as support for procreation, if given at all. The mutual dedication is good only because it might (I doubt it) be always better for a child. In this case marriage is flawed because we recognise marriage intrinsically, not parents with children. Thus in some way I actually agree with those who are categorised as right-wing loonies and Christian nutters.
Thirdly, the 'mutual and exclusive' tag can be applied to any number of people. The addition of 'between two people only' is as ludicrous as 'between a man and a woman' if it is only the feelings that we are honouring. This is similar to Professor Sandel's dismissal of the 'simplistic' argument that people should be free to choose their partner. He says that it is not that simple, because of the option of the state not supporting a partner system at all. Rather than choose to remain in a clearly untenable system of thought, as he does, I take this problem as a demonstration that marriage is indeed an arbitrary and unjustifiable legal imposition.

We already know that our states are predicated on the dispensation of as much freedom as possible; it is commonly held that it is good for the state to impose no more on people's choice than is necessary to prevent others within or without society imposing on people's free choice. To support the institution of marriage is therefore beyond the legitimate scope of a government, since it is an arbitrarily chosen good, dependent solely on belief.

The 'right-wingnuts' or whatever they choose to call themselves are quite right. Once marriage is dissociated from child-rearing its value is lost. Where they are wrong is to suppose that it currently is linked to child-rearing and that supporting homosexual marriage will break this link. Until marriage necessarily involves a child then that link does not exist. We have plenty of married couples without children, and plenty who no longer have children.

I'm not even sure that giving any support for child-rearing is a good thing, given the problems of global over-population and people having children just for the benefits. Gregory Clark makes a compelling argument (implicitly in his work) that because child subsidies are proportionally greater for the poor, the poor will have more children, who will have a reduced quality of upbringing the more children the poor have. The rich, who can afford to care for a child better, will feel less incentive.
He also argues that the long-term trickle down of greater fertility in the rich generated the work ethic and culture of personal virtue that started the industrial revolution. While the direct evidence is weak, the possibility is absolutely clear. Although any idea of attempting to control fertility of any group is anathema to modern society, it being claimed that everyone has the right to as many babies as he chooses, this seems silly to me. Soon we will have to indulge in population control anyway.
Furthermore, child subsidies already disproportionately affect the poor. It can be argued just as cogently that we are already therefore favouring fertility in this group, and that changing this system will remove our immoral influence. Of course, this is delving into subjects that could merit a whole discussion of their own.

The last point on this subject (that I have in mind at the moment) is something that Professor Sandel mentioned in his first Reith lecture: that financial incentives replace moral incentives, rather than adding to them. If putting a price on pregnancy (as he mentions in the second lecture) is wrong, then surely a price on child-rearing is wrong. Children should be raised to be moral citizens because it is necessary for society, not because society pays for children and has punishments in place for their immorality.
I would support a centralised child-rearing scheme. If anything is essential to society then future members are. If working members of the public supported both young and old then there might be less of a pensions problem: if there are too few workers to support the aged at some point in the future then it is because those aged did not support enough children when they were working (in a stable society: society with changing life expectancies but not retirement ages, or a society with a shrinking income would suffer anyway).