The issue of divorce has two aspects that need to be considered:
1) how divorce should function according to our understanding of egalitarian halachah
2) how to deal with the practical reality that not all of the Jewish world shares our understanding of egalitarian halachah.
Now #2 is an important issue that should not be minimized (at least for opposite-sex couples). If a divorce is not recognized by all parts of the Jewish world, and then the woman subsequently has children with another man, those children (and their descendants) can face serious future consequences. And unlike having a marriage or conversion that is not recognized by some, there is no remedy.
However, as important as #2 is, there is an unfortunate tendency to consider only #2, to the point that it overshadows #1. For example, as mentioned in previous posts in this series, the model of marriage that Rachel Adler proposes in Engendering Judaism goes to great lengths to emphasize that it is not kiddushin. For example, the order of the service begins with a cup of wine, but Adler writes: "In the kiddushin ceremony, this blessing would be followed by the erusin blessing, and only the couple would drink from the cup. ... To distinguish this cup from the erusin cup, it may be passed to all those around the huppah." Why is it so important to do various things with a shinui [change] to make clear that kiddushin is not happening, even egalitarian kiddushin? "[B]ecause any kiddushin requires a get, a halakhic divorce, for its termination, it is important to establish that b'rit ahuvim is not equivalent to kiddushin. For if b'rit ahuvim is not a halakhic marriage, then it can be dissolved without a get." And why is requiring a get a bad thing? "Greedy husbands blackmail and extort in exchange for granting divorces. Vindictive husbands withhold divorce for years, leaving their wives in limbo. ... In our own time, religious courts research and record marriages and divorces with unprecedented assiduousness in a central computer bank." In other words, because the process of gittin in other parts of the Jewish world is undeniably disgusting (both because of the potential for abuse and because, even when the system is "working", it creates a huge power imbalance), we should rule out anything in our part of the Jewish world that resembles gittin, and therefore anything that resembles kiddushin. (The possibility of egalitarian gittin isn't even considered.)
I find this approach problematic for two reasons: First of all, it results in a definition of marriage that is too negative ("not kiddushin") and cedes too much to other Jewish subcultures. In the name of avoiding the abuses of vindictive husbands and the Orthodox rabbinic establishment (a laudable goal), the unfortunate result is defining our own marriages as "not a halakhic marriage". When we make decisions about the wedding ceremony because of how it may appear to someone else outside our community, we lose the opportunity to decide what we would do if we were the only Jews on earth. That freedom might result in fixing kiddushin and gittin to eliminate these abuses and power imbalances. But yes, even under those circumstances, many would still oppose kiddushin, even egalitarian kiddushin, for other reasons. Adler herself rejects it as "equal opportunity commodification". But there's still a difference between the choice not to use the framework of kiddushin (a choice that could be made for many reasons) and the choice to essentially wave red flags saying "This is not kiddushin" (which seems motivated only as a reaction to external perception).
Second of all, despite all the efforts to say "this is not kiddushin" in order to avoid the need for a get, I'm not convinced it would actually work. That is, no matter how many ways you try to mark your wedding ceremony as "not kiddushin", there's always the possibility that some corrupt functionary in the Israeli rabbanut will deem that kiddushin has taken place anyway, and therefore that you need a get. And then you're back where you started, when it comes to practical concern for descendants. So since there's no way to avoid this possibility completely, it seems to me that it's not worth trying too hard.
And I don't mean to pick on Adler specifically. She does outline how the berit ahuvim is to be dissolved. Others, in contrast, don't consider the question of divorce at all, or if they do, they frame the issue as creating "rituals" rather than as determining the legal procedure by which the marriage is terminated; the legal procedures are either ceded to someone else or deemed unimportant.
Instead, I would advocate fighting the war as if there is no White Paper, and so forth: figure out how divorce works in the context of marriage as we understand it, then figure out whether any additional provisions need to be included to address the rest of the world. We did both of these, but kept them separate: the former was addressed in the structure of the sh'tar kiddushin, while the latter was addressed by a provision in the ketubah. The remainder of this post will look at each piece of this.
The first question is how egalitarian kiddushin should be terminated. As discussed in Part 3, the procedure for divorce (one partner writing and giving the other a document) appears explicitly in the Torah, while the procedure for marriage does not. Some of the laws of kiddushin (especially kiddushin bishtar) are derived from the laws of gittin. So it would appear that kiddushin and gittin are inverses of each other, and a get is required to terminate a kiddushin.
How does this apply to egalitarian kiddushin, in which there are two separate acts of kiddushin? The simplest answer might be that each kiddushin requires a separate get to terminate it. But this interpretation leads to problems in practice: Many liberal Jewish weddings include two acts of kiddushin, and very few divorces include two gittin. If we take egalitarian kiddushin seriously (which, fortunately in this case, the Israeli rabbanut doesn't), then this interpretation forces us to conclude that many individuals (usually men) who were married via egalitarian kiddushin and gave a get to their ex-spouses, but did not receive a get, are unknowingly agunim! (Perhaps one could address this bedi'avad through a less condescending (and more tautological) version of Moshe Feinstein's infamous ruling that marriages officiated by non-Orthodox rabbis don't really count and therefore don't require a get: one could say that if they were of the belief that their kiddushin didn't require a get, then ipso facto they didn't really have the intent to do kiddushin, and thus kiddushin never occurred. Still, lechatchilah we shouldn't be thinking that our kiddushin doesn't count.) It also leads to problems in theory: If both partners must give a get before the marriage is completely dissolved, this doubles the potential for abuse rather than eliminating it, since either partner can now hold the other hostage.
But maybe this interpretation (that a get is required for each kiddushin) is not the correct one. After all, a get does more than terminate kiddushin: it also terminates nisuin, a mutual relationship that has no other formal termination. So perhaps a get terminates not just a single kiddushin, but ALL marital relationships between the two parties, including the kiddushin in the reverse direction. If this is the case, then if one partner gives the other a get, then it would automatically terminate both kiddushins.
The second way certainly seems like how things should work: either party can initiate divorce, neither party can take hostages, and no new agunim are created. But how do we know how it does work? This would require much research to figure out, since the question wasn't directly asked in the classical sources.
In the meantime, to ensure that it would work this way, that one get from either party could terminate both kiddushins, we stipulated it explicitly. As discussed in Part 3, our sh'tarot kiddushin included a condition that they were dependent on each other, so that if one kiddushin is terminated, the other is too. It's possible that this condition was unnecessary (except in determining when the kiddushin first went into effect), if that's the way it would work anyway, and it's also possible that the condition was necessary; by including the condition, we covered all our bases. As discussed in Part 3, this condition could also be used by couples doing other modes of kiddushin; it would just have to be in a separate document or spoken.
So that should cover us in regard to our own understanding of egalitarian halachah (#1 above). That is, if (chas v'shalom) either of us gave the other a get, we would both consider ourselves divorced. But, as discussed above, that's not good enough. What about the practical realities in the rest of the Jewish world?
Ideally, the non-egalitarian portions of the Jewish world shouldn't accept our kiddushin in the first place, and therefore there should be no issue: one kiddushin is from a woman to a man (and therefore not recognized on those grounds), and the other kiddushin (from a man to a woman) is conditioned on the validity of the first kiddushin. However, it's dangerous to count on that working out in practice. There's always the possibility that someone at the rabbanut or wherever would consider the condition itself to be invalid, and therefore consider one of the kiddushins to be valid. So it's still important to be prepared for that possibility.
Also, the rabbanut or whoever would only accept a get from particular batei din, likely a proper subset of the set of batei din that we would consider valid. And we wanted to deal with the consequences of that reality, without suggesting that we consider only batei din in that subset to be valid.
To address this, we adapted a solution developed by our friends Debra Cash and David Fillingham (which has been included in newer editions of The New Jewish Wedding). They signed a document committing to a particular process in the case of divorce. We didn't use their whole process, but did take one piece of it: "[W]e pledge that neither will create an impediment to the other's Jewish remarriage. Specifically, we pledge that ... if one of us requires a formal get from the other, in accordance with any denomination of the Jewish world that the requester identifies, the other will not be recalcitrant and will participate in person or by a proxy acceptable under the norms of that denomination's practice." In other words, this requirement that a get be granted from any denomination requested is not the act that dissolves the marriage, but is part of the divorce settlement.
We adapted this provision and translated it into Aramaic: ודאי בעא חד מינייהו או תרוויהו למפרש דא מן דא ח"ו יהון פורשין בספר תירוכין באפי בי דינא דיתרעי כל חד מניהו ולא מעכבן דא לדא. ("And if a time comes that one or both of them desire to dissolve their marriage, they will separate with a get, granted by either to the other in a bet-din requested, without delay.") This clause appears in our ketubah, in the same paragraph quoted in Part 4 that talks about dividing possessions, etc. It was an intentional choice to put this in the ketubah and not in the sh'tar kiddushin, so that it is far away from the conditions dealing with what it takes to actually dissolve the marriage, and is instead categorized with the responsibilities that we have towards each other during and after the marriage. This is in the same spirit as Cash and Fillingham's making this provision part of the divorce settlement.
Ok, now let's get real: this requirement that a get be granted, above and beyond the get (or other act) that actually terminates the marriage, is very unlikely to be invoked by "either party" in "any denomination". In practice, the likely scenario is very specific: a woman would ask a man to give her a get in an Orthodox beit din, and this clause would require him to grant it. However, by keeping the language general, we can deal with the situation without hard-coding either the gender inequality or the denominational politics into our ketubot.
Of course, this sort of clause would be unnecessary for same-sex couples. However, they might still choose to include it, in solidarity with straight allies for whom it is necessary (since some straight allies do many things in solidarity with same-sex couples).
We didn't give this clause teeth, e.g. state that if one party fails to abide by it, s/he will be subject to a financial penalty, or that the kiddushin will be retroactively annulled. While we know that we are honest people and would take our commitments seriously with or without teeth, I realize that's not the point: the point of teeth is to establish precedent for the whole Jewish world, so that even less honest people would be subject to that precedent. So why didn't we do it? I guess it's because we had already done our Kantian duty in establishing precedent by doing egalitarian kiddushin that can be terminated by either party: if everyone did what we did, there would be no agunah problem, and this additional clause is just a concession to the fact that not everyone does what we do.
The case that we haven't dealt with is the original agunah situation: what to do if one partner is missing and presumed (but not proven) dead, or analogously, the Terri Schiavo scenario. The solution to this (for which there is historical precedent) is probably to write conditional gittin, which will go into effect only if one of those horrific scenarios takes place. But this doesn't need to happen at the time of the marriage; it can be written any time.
So that's all for now. I obviously hope nothing in this post has to be implemented, whether in my marriage or in yours, but I hope kicking off this discussion is still helpful. This is the end of the series, but if you have questions about anything, please post away in the comments.