The abstracts for the planned talks at the Collective Intentionality VIII conference are now up on the website, including mine. There are some very interesting talks, its a shame that each speaker (in the parallel sessions) can only get 20mins to present them. The full (early draft) text of the paper I plan to present is below, I’ll have to think hard about which part to focus on in the short time. Any comments appreciated.
Early bug filled draft – Please Don’t quote
Love, Plural Subjects & Normative Constraint
“However and whenever infatuation begins, if given the opportunity it transforms itself into a continuing romantic love or else it disappears. With this continuing romantic love, it feels to the two people that they have united to form a new entity in the world, what might be called a we.”
Nozick, Loves Bond, p.232
1 – CRRs as wes
In his paper, Loves Bond1, Robert Nozick makes the case that ongoing loving relationships involve participants seeing themselves as united together such that, combined, they create a new entity; a ‘we‘. Now a quick snog behind the bike sheds might be characterised by nothing more than pure individual lust however on Nozick’s account moving beyond the stage of initial desire necessarily involves a feeling of union. The development of love is prone to fine combed distinctions especially in its early stages, such that it is possible to distinguish between ‘going out’, ‘dating’, ‘seeing each other’ etc. Marriage might be thought to be the clearest form of the move beyond these early stages, requiring as it does formal vows and declarations. This is true but the unions Nozick describes are, particularly in our modern age of long unmarried relationships, certainly possible long before and without even the prospect of marriage. For the sake of clarity I will speak of the stage Nozick seeks to understand, the stage that comes after ‘dating’ and can be made explicit by marriage but may exist without it, as ‘committed romantic relationship’, for brevity henceforth referred to as a CRR.2
Nozick’s account fits with how we commonly think of being in loving relationships but we might worry that while talk of being ‘united as one’ makes sense poetically or metaphorically, it can not be taken to be literally true. Even when we are most deeply in love it is indisputable that we remain distinct biological entities; we never literally become one creature with two heads. If we are to, as we probably should, refrain from making vague reference to unexplained social spirits or ontologically mysterious collective entities, can we successfully speak of love in terms of unity? For this task it seems obvious to look to the growing literature around what has been called ‘Collective Intentionality’. Of particular relevance to Nozick’s work is the ‘Plural Subject Theory'(PST) of Margaret Gilbert, for his talk of forming a we nicely mirrors her focus on “… the central sense of the of pronoun ‘we‘”.3 Unfortunately, though Gilbert gives some indication that she feels her account could play this role4 she has not as yet pursued this line of enquiry. Such a project has however, to a certain extent at least, been taken up by Andrea Westlund. In her paper The Reunion of Marriage5 she proposes an account of CRRs that is broadly along the lines of Gilbert’s PST account.6
Westlund suggests that CRRs (or as she calls them relationships of ‘companion love’7) involve the formation of a plural subject with a particular structure; a structure that respects the reciprocity that she sees as essential to love. While Westlund does engage with Gilbert’s PST she rejects Gilbert’s account of normative constraint as arising directly from being part of a plural subject. She believes that doing so is necessary to account for the necessary flexibility between parties that is at the heart of loving relationships. This is a key element of Gilbert’s theory and I will challenge Westlund’s rejection of it. However, while we should reject Westlund’s requirement for flexibility, I argue that CRRs do demonstrate a related phenomena that might be confused with Westlunds flexibility, this is, the possibility of variability of the pain of resisting normative constraint. In this way, while Westlund’s modified PST is too weak, I thus argue that Gilbert’s is too strong. My solution is to urge the replacing of Gilbert’s digital all-or-nothing voluntary pooling of the wills with a notion of wills becoming entangled through analogue levels of (necessary) identification with the plural subject. In this way, on my account, the strictness of lover’s constraint by the will of the we that they form is conditional on the extent to which their understandings of their own-selves has become entangled with their understanding of the agency of the we they together form.
2 – Westlunds Plural Subject Account:
Westlund is motivated to give a plural subject account by the following considerations. She notes that various accounts of love8 see being in a CRR as involving the participants forming a ‘union’ with each other and sees merit in this. However she worries that insomuch as these views, which she dubs ‘union views’, appear to suggest that in CRRs lovers become psychologically or ontologically melded,9 they seem contrary to another aspect that she believes is fundemental to the notion of CRR; reciprocity. A key part of healthy love Westlund claims is the idea that each agent value the other’s interests, the possibility of such reciprocity is, as she sees it, a central feature of non-pathological love. If the union view requires that the participants become psychologically or ontologically melded Westlund thinks that this implies that the interests of each party are dissolved into the interests of the whole. The worry then is that if each party is truly melded with the other this replaces the interests of each as a separate self with the interests of the singular melded entity, the interests of the we. The problem for reciprocity is that it does not seem possible to value the interests of the other if they have been dissolved in this way. Lets call this the ‘problem of reciprocity’.
Rather than give up on the possibility of a successful union view, Westlund thinks that the problem of reciprocity can be overcome by developing an account of the union in question that does not entail ontological or psychological melding. She takes this to be possible if the union is of the form (a la Gilbert) of the creation of a plural subject. So what is a plural subject if it is not something created from the melding together of multiple individuals? For Gilbert plural subjects are the willed creations of individual agents; they are the result of those agents jointly committing to, within a certain scope of activity, pool their agency. Gilbert gives the example of two people taking a trip as a plural subject, or as we might more naturally say, travelling as a pair. For Gilbert what makes travelling together in this sense stronger than merely travelling in proximity (both temporal and spacial) to one another, is that each agent can be said to have together manifested and accepted their “… willingness to constitute with the other a plural subject with the goal that they travel in each others company.”10 Plural subject-hood thus involves agents remaining as individuals, i.e. not becoming psychologically or ontologically melded. Rather than becoming melded they place some of their agency into a shared pool, a pool with a specific goal or stance. Westlund is not entirely clear on how introducing the notion of plural subjectivity helps but I think we can read her as saying something like the following: Plural subjectivity does not face the problem of reciprocity because the forming a plural subject need not destroy the self, i.e. the identity of each individual can remain intact.
Such plural agents can be transitory; the travellers in Gilbert’s example may have met for the first time at the train station and travel together solely for one trip, or they can have more open ended extended scopes; the travellers may be regular fellow commuters. A CRR would clearly be of the latter type, indeed in marriage the participants pledge to be as one until death parts them. So far so good but we might ask, given that plural subjects can have all kinds of implementations, what is the specific scope of a committed romantic relationship? Gilbert’s gestures towards an answer to this question in a minor remark in a footnote in her On Social Facts, saying that it might be a “… plural subject with a specific primary goal: something like the well-being of both parties equally”.11 As I have already noted, above, Westlund sees the reciprocity of each participant caring for the wellbeing of the other as a vital part of what it is to be in a CRR however for Westlund this reciprocity does not characterise CRRs by being their direct goal but rather it characterises the nature of how CRRs try to achieve their main goal. That main goal itself is far more general, it is simply living as one in a loving way. Thus at any one point those in a CRR can have all sorts of immediate shared goals; to go for a walk, to buy a sofa, to discipline their child etc. but “… the projects and plans undertaken by companion lovers are ways of realizing an overarching desire to be (and do) together.”12 Being a we is, then, not primarily some means to a further end but is rather, as Nozick puts it, “…part of what love intends”13
According to this picture then, being in a CRR is characterised by being a plural subject with an ongoing open ended scope. Of course we might say the same about those in certain sorts of business partnerships, or clubs. For Westlund the difference with lovers is that they must be engaged in forming this we in a loving way, that is, in a way that has reciprocity at its heart. What does this mean? Well Westlund claims that it has two elements; firstly the jointness must go down to the level of sharing reasons as well as sharing ends, and secondly that the participants must aim to do this in a way that engages with the individual interests of each party. Or to put it another way; a couple in a CRR form a plural subject that is engaged in an ongoing process of forging a shared practical perspective. That is a project of coming to have shared reasons and shared ends. Sharing in ends means together aiming at doing particular things as a couple. Sharing in reasons, means together taking certain considerations to count in favour for them, as a couple, doing those particular things. The difference between deciding what to do in cases of mere competition is that as “… lovers we are characteristically concerned in a different way with the mutual acceptability of the plans on which we settle”.14
Suppose that a couple in a CRR are planning to go on holiday; on Westlund’s account this different from, say, strangers who having won a game show must negotiate a single location to both be whisked away to. It is different in that for the couple the aim is not just to come to both independently converge on a single destination, rather the aim for the couple is to make the decision jointly such that they agree not just on the outcome but on the reasons that will jointly count as justifying their holiday.15 Rather than merely weighing up all their individual preferences and trying to just finding a holiday that maximises these are far as possible, the couple will try to find reasons that they can agree count in favour of a holiday for them jointly; “I love culture and my partner loves beer but we want to go to Majorca because we want to go somewhere hot”.16 So, Westlund says that the discussion ends not with the discovery of shared reasons assessed by merely weighing up individual preferences, but rather in creation of joint reasons through the joint acceptance of a consideration as a reason for them.17 Thus for Westlund the collective will is a “… joint product of collective deliberative agency”.18
3 – Gilbert’s PST & normative constraint:
Westlund’s attempt to apply plural subject theory to the nature of love has some compelling elements. However, as I will detail below, she rejects too hastily a central element of Gilbert’s account regarding the normative authority of plural reasons and ends (I shall speak of these taken together as the plural will).
For Gilbert the creation of plural reasons and ends involves the individuals jointly committing to pool their wills into a plural will. Doing so does not mean that they are then left with no individual will, for our wills are not finite substances, but it does mean that (within the scope of the area of concern of the we in which they are involved) they are rationally committed to whatever the plural will is committed to. In a sense we can say that by voluntarily taking part in the plural will they thus become joined to it. This is significant because the plural nature of this plural will means that it is under the control of the participants together. That is, no one participant is individually in a position to unilaterally reject its demands by a simple personal change of mind or declaration. So for example, Gilbert says of the couple walking together that one might well rebuke the other by saying “You can’t turn round, we said we would walk to the top”.
On Gilbert’s account the normative authority of plural reasons and ends thus arises directly from the nature of the joint commitment required for the construction of a plural subject – the plural will, formed collectively, can only be changed or rescinded collectively. This gives rise to what she calls the ‘obligation’ and ‘permission’ criteria. The obligation criteria says that each participant has a pro tanto obligation to promote fulfilment of that which is intended by the plural will. The Permission criteria says that participants understand that they are not (ordinarily) in a position to unilaterally ‘by a simple change of mind’ remove the constraints imposed on them by the obligation criteria.19 For Gilbert this normative constraint, is not some additional feature of morality, peer pressure or politeness, rather it arises directly from the collective will – it is akin to the normative constraint that as individuals we face with our own intentions; I.e. If I intend to give this paper today but I actually stay in bed, then I have done something normatively wrong and likewise if we intended to give a paper today by I stay in bed if have done a similar wrong. The key difference between the plural will and our individual wills is that we are in a position to unilaterally change our individual wills, however with collectives we are not.
Westlund does not to like this aspect of Gilbert’s PST in general – though she does not give a direct argument against it as a possibility for any collective.20 Her particular gripe is with it as an account of CRRs and in her alternative account there is no direct normative constraint as such. In contrast to a plural will that each participant becomes joined to, Westlund’s (not fully spelled out) view is that the creation and maintenance of a plural will involves an ongoing ‘dance of union, separation and reunion’21, that is, it involves individuals being engaged in an ‘evolving framework’ of plural stances to which they must continuously reaffirm. Westlund does not discuss Gilbert’s scenarios of dissent, such as the two walkers discussed above, but if she did she would not be able to recognise the picture Gilbert paints of their experience. This is because her view implies that with regard to collective plans each participant has “ …some discretion over her own follow-through”22. That is not to say that Westlund wants to claim that the plural will does not have any kind of sway over each individual. If this were the case then it would be hard to see how could would play the role that Westlund believes it does; i.e. how it would play the role of making the reasoning of couples in CRRs fundamentally different from that of the game show bargainers. However, Westlund explains this sway by claiming that it exists because there is a further commitment to a ‘robust form of mutual accountability’. This accounts for the normative authority of those plural reasons and ends. To put this in terms that Abraham Roth uses,23 Westlund believes that each participant has an ongoing bridging intention that ties them to the collective will – the bridging intention to see themselves as accountable for their collective actions together with the other.
4 – Westlund’s reasons for rejecting direct normative constraint (and a rejection of them)
Why does Westlund differ from Gilbert’s in this way? To understand this we have to recall the quandary that she evokes plural subjects in order to solve i.e. the problem of reciprocity. As I spelled out above, the concept of plural subjects in supposed to solve the problem of reciprocity because it allows the possibility of forming unions without losing each partner’s individual autonomy. Further, a key part of this for Westlund is that the joint deliberative agency by which the couple form their collective will necessarily involves an ongoing sensitivity of each to the practical perspective of the other. This, it might be thought, rules out each party being directly constrained by the nature of the collective will for their being so constrained might be thought to be a block on the other have sensitivity to their perspective. So for Westlund “… any balance that is struck between the parties must be regarded as defeasible in the face of further reflection and experience, such that a shared practical perspective is always a work in progress.”24 [my emphasis]. So, this flexibility of plural subject perspective is, Westlund believes, needed in addition to the non-destruction of the individual perspective in order for the possibility for reciprocity, and reciprocity is a central part of non-pathological CRRs. I will refer to this idea in what follows as the requirement of flexibility.
Should we accept Westlund’s rejection of Gilbert’s description of direct normative constraint? I think we should not for two reasons; firstly because of the importance of this constraint to Gilbert’s overall project, and secondly because it does not seem to me to fit with the phenomenology of being in a CRR. On the first point I cannot say that much here for a full defence of it would require an overall defence of Gilbert’s philosophical project that there is no space for here. What is worth noting is that Westlunds rejection of direct normative constraint loses an important element of Gilbert’s account, that is, its ability to explain the normative authority of plural subject attitudes as arising directly from the nature of joint commitment. This thus robs Gilbert’s account of that which makes it distinctly collectivist.
What about my second claim; here I want to challenge the idea that we experience CRRs as necessarily having the flexibility that Westlund describes. It is true that when love is progressing well, when a relationship is of a good kind, that each party tries to take the needs of the other into account when engaged in joint deliberation. In the holiday example we would say that if the couple, Mary and Clare, are in a well functioning CRR then each party should try as hard as they can to accommodate the desires of the other and to compromise when these do not fit with their own. Likewise, once they have come to their collective decision, i.e. once they have constructed their collective will, then if they truly care for each other each will be open to the possibility of further deliberation over the collective will. However, this does not mean that the collective will itself must be automatically reflective of any change in each individuals perspective. If Mary comes to realise that she hates hot nights, this does mean that (given Clare’s love for her) Claire ought to be open to re-engaging in joint deliberation and open to them jointly changing their minds. This requirement for to other partner to try to be flexible does not imply that the collective will must automatically ceases to hold sway over her as soon as she changes her own individual mind. In parallel with Gilbert’s walkers I suggest that if on the way to the airport Mary turned round and stated walking back home, Clare would be in her rights to say; “You can’t go home, we said we were going to Majorca!” What the existence of love demands is not that their collective perspective must automatically lose its force if Mary changes there mind, but rather that Clare should be willing to listen if Mary does air her change of mind, and further that she should be willing to engage in further joint deliberation to try to accommodate this change of mind.
Further to Westlund looking for flexibility in the wrong place in healthy CRRs, it seems to me that she idealises love by ignoring the fact that unhealthy CRRs still count as CRRs. Feelings of being a we do not necessarily disappear with failure of each party to treat the other with reciprocity. Once we allow the phenomenology of badly functioning CRRs into our picture then the fact that there is no necessary flexibility at the level of constraint by the collective will becomes all the more apparent. Given that Westlund seems to believe that being bound to the collective will requires an additional personal commitment she would have to say that the shared perspective merely dissolves in the face of one, or both, parties ceasing to continuously reaffirm the framework which takes the other into account but this just seems wrong as there is a sense in which even in bad relationships we feel trapped by the collective will. What Westlund calls ethically defective forms of love may well be ethically defective but she is wrong to bar them from consideration in forming a view about the essential nature of CRRs.
While Gilbert gets closer to the phenomenology here than Westlund, I will now suggest that she also misses the mark for there is something overly restrictive in her set up, it just that it is not the something that Westlund thinks it is. As we have seen for Gilbert once a will pooled it is forever pooled, unless collectively un-pooled. In this she is right but what she fails to acknowledge is that there are circumstances when the normative feeling of breaking with the obligation to the commitment to the collective feels worse than others. I will speak of this as the variability of normative pain. It is the existence of this phenomenon and the way in which it mimics a certain aspect of Westlund’s description of flexibility that accounts for the existence of a certain level of intuitive plausibility to Westlund’s rejection of Gilbert’s account of direct normative constraint.
5 – The variability of the ‘pain’ of breaking with the normative constraint of the collective will
Individuals in different CRRs experience different levels of what we might, as above, call ‘normative pain’ when they break with the obligations placed on them by the collective will. This is, I believe, evident in the cases described below. Unfortunately for showing this to be the case the direct normative commitment of the will that Gilbert describes is not only thing that provides reasons for the individual to be committed to the we in cases of love. There are of course also moral, romantic, practical and further reasons in play. My claim is not that these other reasons do not matter but rather just that Gilbert’s direct normative constraint of the will plays an important part in the mix.
I’ll proceed by giving three examples that I believe illustrate the variability of normative pain, they are a variation on the holiday example above:
Early days: Imagine first a newly formed embryonic couple, Bill and Ted. They have been together for a few months, seeing each around once or twice a week. They see themselves as very much in love, they are full of the strong feelings of lust and desire, but live in separate houses, have separate groups of friends and very different hobbies.
Long term and going strong: Now contrast this with our couple from the example above, Clare and Mary. Lets say that they have been together for 25 years, they have bought a house together, they have adopted and raised child together, they share friends and have the same hobbies. They feel still very much in love.
Coming apart: Patrick and Madeleine have been in a relationship longest of all, 30 years. They do live together, and pay bills jointly and have raised children. But they also do an increasing amount of things apart, they have separative friendship groups and they enjoy different social activities. They don’t really feel much romance toward each other and often find themselves attracted to other people.
Now imagine that each of these couples is on their way to the airport and one of each of them turns round and starts walking back home. Everyone of them will be open to a rebuke of Gilbert’s form: “You can’t go home, we said we where going to Majorca!” However, the act of rebelling against this, will feel different to each member. So when Ted turns round to Bill and announces that he is not going, the wrong that he is committing by violating the collective will is going to feel less strong than that which would be felt by Mary if she were to do the same. Similarly, if Madeline was to do so then her wrong would feel worse than Ted’s thought not as normatively painful as Mary’s deviation.
I am not claiming that any of the couples face no obligation towards their collective wills rather just that the feeling of wrong, what I am calling the normative pain, of breaking these obligations will be different in each of these cases. Of course this could all just be a function of variance in other factors; the different level of moral, romantic, practical obligations felt by each party. However, that the rebuke may feel more painful to Madeline (who is not really getting that much out of her relationship any more) than to Ted (who is very much infatuated) suggests to me that the difference lies in the nature of the we rather than the personal feelings of each party. Further, given that the rebuke makes direct reference to the will of the we, I want to suggest that we should look for variance towards this relation to explain this phenomenon.
So, how can we make sense of the variability between the couples? Well, my proposal is that we must replace Gilbert’s digital vision of a all-or-nothing voluntary pooling of wills with something that allows differing levels of combination. In what follows I propose that we can do this by modifying Gilbert’s PST to create an analogue notion of wills becoming ‘entangled’ through levels of necessary identification with the plural subject. The variance in the three holiday cases will thus be explained by variance is the amount to which the individuals cannot help but identify with the collective, in turn this variance will be explained by how much of their own lives have been lived, and thus only fully make sense, within the scope of the collective agency.
The background to my claim that identification is important to understanding the variance in normative pain is an understanding of what it is about commitments of the will in general that binds us. What is wrong, one might ask, with agentive anarchism, i.e. doing whatever one feels the compulsion to do at any time and not feeling committed by ones plans and intentions? This is a tricky subject that is deserving of a longer treatment than I can give it here,25 but one proposal how this works that is helpful to us in understanding the possible variance of normative pain of breaking with commitments of the collective will, is that made by Michel Bratman.26 According to Bratman our seeing ourselves as being constrained by the commitments’ of our wills27 is necessary for us to be able to understand ourselves as agents who can govern our own lives. My claim here is that we can extend this claim to explain the normative power of commitments of the collective wills for couples in CRRs; i.e. that each lover must see themselves as constrained by the will of the we that they are part of in order to see that we as a collective agent.28
Bratman starts from the premise that there are a multitude of elements of our psychological make up that can be seen as causally responsible for the things that we do. However in order to see ourselves as unified agents able to act in the world (something we clearly have a fundamental reason to be able to do) we need to be able to have a “where-I-stand” constructed out of the psychic stew. The commitments of our wills (plans, intentions, goals) provide the scaffolding necessary to construct such a stand point. This is because seeing ourselves as necessarily constrained by the commitments of our wills is part of seeing “…ourselves as agents who persist over time, who begin, develop and then complete temporally extended activities and projects”29 (p.35). The commitments of our wills are thus the scaffolding around which we construct ourselves as singular unified agents.
How can we translate this to plural agents? Well just as singular agency requires a place where-I-stand, collective agency requires a place “where-we-stand”. The case is not however completely symmetrical between individuals and collectives. For individuals this sets up a fundamental normative requirement for one always needs to be able to see oneself as being able to govern-self for if we can’t see ourselves as individual agents there is no where left to retreat. It follows that there is no variability in importance of self-governance and thus that there is no variability in the normative pain of breaking with our own wills.30 For collective things are slightly different. We do always need to be able to see the collective we are part of as being able to govern-self if we want there to be a possibility of collective agency. However, to some extent we appear to have the option of retreating back to our own individual agency. What blocks this is that the fact that we have been part of on going collective agency, part of our understanding of ourselves will be bound up in being able to see the collective as acting. In as far as we have already lived part of our lives through the collective we can only continue to understand our contributory action as the kind of thing we set it out to be if we are able to see the collective as an agent capable of governing our collective actions, and because this requires its commitments of will to constrain, we must see them as doing so. Post-hoc reconceptualising our contributions is logically possible but seems to be a kind of inauthenticity for that is how, at the moment of our actions, we set them out to be.
The extent of the felt wrongness of rebellion against the will of the CRR is conditional on the extent to which our understanding of our own actions is entangled with the agency of the CRR. Given their ongoing entanglement with the CRR the socially situated autonomous individual is faced with a choice: (1) accept the normative constraint of the collective will or (2) abandon the possibility of collective self-governance.31
This difference between foundation of the power of the individual will and the collective will is, I believe, what explains the variability in the three holidaying couples. The variability comes from the fact that though we will always have part of ourselves bound up with the we the amount varies (with length of time of collective etc.) Acknowledging this requires a re-understanding of Gilbert by moving away from voluntary wills being pooled to wills becoming entangled over time perhaps in a sub-voluntary way, but this seems to fit better with our experience of love.
The superiority of this account over Gilbert’s is further illustrated by looking at what happens when a couple split up. Imagine that after Madeline scuppers their holiday plans and she and Patrick have returned home, she announces that she has feelings towards someone else and is considering leaving him. On Gilbert’s account part of what will make their splitting up painful, at least normatively painful, is that they are together bound up in a plural will. Patrick might then say to Madeline “you can’t leave me, we said we would be together forever!”, appealing not to his own hurt feelings but to the violation of the collective they together form. The question we can ask is why should they care about destroying this plural will? Here Nozick says something that I think is a kernel of an account like that I give above, he says that “A willingness to trade up [i.e. to find a new partner], to destroy the very we you largely identify with, would then be a willingness to destroy your self in the form of your own extended self”32 This seems to be right as a phenomenological description of what it feels like to break up the question is how do we model each individuals identifying with the collective we?
On Gilbert’s account this identification is a necessary result of the fact that they are bound up with a joint commitment to the plural subject-hood. This commitment being joint can only be rescinded jointly and thus each individual is in a sense normatively compelled to continue to identify with the we unless the other releases them – and given this – if they seek anyway to switch to another (in the full sense of abandoning the former not just having an additional relationship on the side – though that in itself will carry its own problems) then they cannot help but feel themselves to be destroying themselves. For Westlund such hard set commitments do not fit with her view of love. She claims that the commitment of each is a personal commitment – thus it would seem that, though there may be a great deal of other pains involved in breaking up, the feeling of destroying an extended part of oneself need not be one of them – for one can escape it by merely personally rescinding ones commitment to the plural subject.
On the view I am proposing Gilbert’s account is along the right lines, and gets closer to the experience than Westlunds, but it must be amended to reflect the fact that becoming bound up with the collective will is not an all-or-nothing voluntary act but rather happens as each individual come to only be able to understand more and more of what they do in terms of existence of the collective agent. When Patrick says to Madeline “you can’t leave me, we said we would be together forever!”, she may reply “We! There is no we” – in doing so she will be giving up the ability to understand part our her own self, but given that they have been drifting apart, given that they have more and more started to live separate lives, this though still a violation of part of her history, will be less painful.