One thing I like in John Stuart Mill’s On Liberty is his consistent distinction between what deserves criticism from what deserves legal prohibition. Divorcing without sufficient cause someone who has been encouraged by fair promises to build their life on the assumption that a marriage will be lasting is a good example. Here is what he says in paragraph 11 of On Liberty “Chapter V: Applications”:

Baron Wilhelm von Humboldt, in the excellent essay from which I have already quoted, states it as his conviction, that engagements which involve personal relations or services, should never be legally binding beyond a limited duration of time; and that the most important of these engagements, marriage, having the peculiarity that its objects are frustrated unless the feelings of both the parties are in harmony with it, should require nothing more than the declared will of either party to dissolve it. This subject is too important, and too complicated, to be discussed in a parenthesis, and I touch on it only so far as is necessary for purposes of illustration. If the conciseness and generality of Baron Humboldt’s dissertation had not obliged him in this instance to content himself with enunciating his conclusion without discussing the premises, he would doubtless have recognised that the question cannot be decided on grounds so simple as those to which he confines himself. When a person, either by express promise or by conduct, has encouraged another to rely upon his continuing to act in a certain way—to build expectations and calculations, and stake any part of his plan of life upon that supposition—a new series of moral obligations arises on his part towards that person, which may possibly be overruled, but cannot be ignored. And again, if the relation between two contracting parties has been followed by consequences to others; if it has placed third parties in any peculiar position, or, as in the case of marriage, has even called third parties into existence, obligations arise on the part of both the contracting parties towards those third persons, the fulfilment of which, or at all events the mode of fulfilment, must be greatly affected by the continuance or disruption of the relation between the original parties to the contract. It does not follow, nor can I admit, that these obligations extend to requiring the fulfilment of the contract at all costs to the happiness of the reluctant party; but they are a necessary element in the question; and even if, as Von Humboldt maintains, they ought to make no difference in the legal freedom of the parties to release themselves from the engagement (and I also hold that they ought not to make much difference), they necessarily make a great difference in the moral freedom. A person is bound to take all these circumstances into account, before resolving on a step which may affect such important interests of others; and if he does not allow proper weight to those interests, he is morally responsible for the wrong. I have made these obvious remarks for the better illustration of the general principle of liberty, and not because they are at all needed on the particular question, which, on the contrary, is usually discussed as if the interest of children was everything, and that of grown persons nothing.

Nowadays, people are often advised that they must plan as if a marriage won’t last. That is not costless. For example, there are many important forms of relationship-specific human capital–knowledge and skills that are valuable if one stays with one’s current spouse but would not carry over to the next spouse. Acting as if a relationship won’t last inhibits the accumulation of such relationship-specific human capital, which in turn can make the relationship more likely to break up, in at least a partially self-fulfilling prophecy.

It is quite awkward to talk about divorce, since speaking of divorce as a failure seems like speaking against one’s divorced friends. But a divorce is a failure–just as a bankruptcy is a failure. People often pick themselves up and achieve success after a business failure, but that doesn’t make a business failure a good thing in itself. Similarly, people often pick themselves up after a divorce and make a good life for themselves thereafter (sometimes involving a second marriage and sometimes not), but the divorce itself was not a good thing unless something else worse than a divorce happened or was revealed before the divorce.

Of people I know who have divorced, I definitely think many of them have made the right decision. Yet I wonder how often everyone’s knowledge that divorces are relatively common on the part of at least one of the two spouses helped make things go bad enough that they got to that point–either because someone made an incautious choice of mate because they knew they could get out of the marriage, because the acquisition of relationship-specific human capital was inadequate, or because feasible efforts to work things out didn’t seem worth the heroic efforts required.

I wrote one post of marriage advice on Valentine’s Day 2014: Marriage 101. I consider research on how to make marriages better and more stable extremely important.

It is said that two heads are better than one. Two hearts are also better than one. Marriage has the potential to bring human beings to a higher plane. But its ability to do so is severely compromised if people bail out too quickly. Among many other things, a spouse is a mirror, who helps us see ourselves more clearly–the ugly parts of our souls as well as the beautiful parts. Besides its many other practical consequences, bailing out too quickly from a marriage makes it much less likely that one will deal with the ugly parts of one’s soul.