New paper on morality and polarization

Moral and political polarization is a popular (and important) issue these days, for obvious reasons: it is increasing dramatically, making it ever-more-difficult for people on different moral and political sides to get along and cooperate. Walter Sinnott-Armstrong has suggested that learning to argue better can help.

In this new paper, I argue the problem of polarization and solution to it may both be much deeper: polarization is probably caused by how philosophers and laypeople commonly think about morality itself--and changing how we think about morality may be an important step toward a better, less polarized future. Here's the paper's abstract:

This article argues that philosophers and laypeople commonly conceptualize moral truths or justified moral beliefs as discoverable through intuition, argument, or some other purely cognitive or affective process. It then contends that three empirically well-supported theories all predict that this ‘Discovery Model’ of morality plays a substantial role in causing social polarization. The same three theories are then used to argue that an alternative ‘Negotiation Model’ of morality—according to which moral truths are not discovered but instead created by actively negotiating compromises—promises to reduce polarization by fostering a progressive willingness to ‘work across the aisle’ to settle moral issues cooperatively. This article then examines potential methods for normatively evaluating polarization, arguing there are prima facie reasons to favor the Negotiation Model over the Discovery Model based on their hypothesized effects on polarization. Finally, I outline avenues for further empirical and philosophical research. 

Comments

[Derek Bowman]

Thanks for sharing, Marcus. I enjoyed reading the paper, and as the paper went on you anticipated and responded to man of my initial objections.

But in the end I'm still not sure who the parties to the Negotiation Model are meant to be and what, exactly, they are negotiating about. Within the Discovery Model I can understand advocates of competing comprehensive moral views negotiating a political settlement on how to live together. There the relevant priorities are given, in part, by each party's considered moral judgments. This may be hard to do in practice - e.g. in the abortion case - but at least I know what we're negotiating about and what is at stake for each side.

But how does this work for the Negotiation Model? Presumably those who already think the fetus is a full moral person with corresponding rights and dignity are not part of the negotiation in this model (nor are those who think the fetus has no moral status), since they've already reached a moral judgment. Are the parties then those who have a stronger preference for the welfare of fetuses vs those who have a stronger preference for the welfare of pregnant women? Do they have any prior basis for these preferences, or are they simply treated as bare preferences?

[Marcus Arvan]

Hi Derek: I'm glad you enjoyed the paper, and thanks for the questions!

The long answer to your questions will be in my new book, which should be out late next year. The book provides a better case for Rightness as Fairness (and the version of the Negotiation Model I defend) based on behavioral neuroscience. It also revises my account, as well as how I understand negotiation, in some important ways. So, if you are interested in much longer and more detailed answers to your questions, I'd suggest checking out the book when it appears!

Here, though, is a very short (and oversimplified) answer. On my account, rational agents should be motivated by an ideal of coercion-avoidance & minimization (Negative Fairness), and a positive ideal mutual assistance (Positive Fairness). I argue these two regulative ideals are ones that *can* be discovered through rational thought, but that another requirement can also be discovered through rational reflection: a Principle of Fair Negotiation to seek compromise agreements with others on how to settle any conflicts the first two ideals generate (both within and between each other).

Here is what this means in practice (as an overly simplified first pass). On my account, the coercion-avoidance ideal (The Principle of Negative Fairness) requires people to be motivated by an ideal of avoiding coercing other sentient beings. Since fetuses are sentient beings (at least at a fairly early stage of development), and women are sentient beings, the Principle of Negative Fairness requires every person to want to *avoid* and *minimize* the extent to which women and fetuses are coerced. The Principle of Positive Fairness in turn requires every person to want to *help* women (e.g. in family planning and making their own health decisions) but also help sentient fetuses (whose lives are a stake).

Alas, as we all know, these moral ideals generate conflicts. First, the best way to avoid coercing women (unrestricted right to abortion) is arguably the *worst* way to prevent fetuses from being coerced; but the best way to ensure that fetuses are not coerced (depriving women of a right to choose) is the *worst* way to avoid coercing women. Second, the best way to help women (viz. Positive Fairness) may be the worst way to help fetuses; and vice versa. According to the Negotiation Model I defend, people motivated by the above ideals--but who might disagree on how to resolve those conflicts--then have a duty to negotiate a compromise, and after the compromise is generated, identify moral rightness *with* the compromise unless and until a new compromise is negotiated (more on this below).

Here, then, is what my version of the Negotiation Model requires, say, in the case of abortion. It requires everyone to the discussion to *agree* that women's autonomy and sentient fetal life both matter. People who don't think women's autonomy matters, or who don't think sentient fetal life matters, should be excluded from the discussion and negotiation as unreasonable. This is a pretty strong requirement--but I think there is a good case to be made for it (just like I think there is a good case for not negotiating with Nazis).

Anyway, because the Negotiation Model as I defend it holds there is no objective fact of the matter to discover of what morality requires beyond this--beyond recognizing *that* women and fetuses both matter, and that we must fairly negotiate comprises to the conflicts that coercion-avoidance and mutual assistant generate--the model allows that in negotiation people should be able to assert things like this: "Women's autonomy matters morally" and "Abortion is a serious moral issue because sentient fetal life is at stake." What the model prohibits is people in discussion and negotiation--and in their own moral-belief formation--assenting to things like this *before* fair negotiation occurs: "Abortion is wrong! It violates the fetuses' right to life", and also, "Any restrictions on the right to choose is wrong--as it violates women's rights!" On the Negotiation Model, people *can* try to advance their preference (viz. "I really want to ensure that women retain control over their bodies", and alternatively, "I really, really want to protect fetal life"). However, relative to those *desires*, the model says that people should *not* assert a moral discovery (e.g. that abortion is wrong, etc.), but instead seek to fairly negotiate a compromise: one that gives both sides *enough* of what they want to arrive at some kind of temporary agreement--one that, if one side is not entirely satisfied, they can renegotiate later.

The question then, of course, is how negotiation is supposed to work. The story here is a very long one, and in my new book manuscript I argue that Rightness as Fairness ultimately leads to a version of the 'Nonideal Original Position' I defend in my forthcoming JAPA paper (https://philpapers.org/rec/ARVNJA-2 )--which I've argued generates social and political principles to *govern* actual negotiation in an unfair world (e.g. by compensating for unfair bargaining power), a set of principles which (I plan to argue) enable us to create *fairer* and *fairer* answers to moral questions over time, through repeated iterations of social negotiation (following principles of nonideal justice), iterations which we should identify as "fairer and fairer" insofar as following principles of nonideal justice generate a greater approximation of equal bargaining power among members of society.

Notice, finally, that something like this underlies a lot of social activism today: namely, that justice is a matter of progressively equalizing bargaining power (away from domination), and that the more bargaining power is equalized (across races, genders, etc.), the *fairer* our answers to moral questions (e.g. consent, autonomy, etc.) progressively become. One deep part of my philosophical project is to spell out a coherent picture of how roughly this picture is correct, but it must be governed by specific discursive rules (including seeking compromise and overlapping consensus) that are not widely accepted in activist circles. But unfortunately, I must leave things at that for now. For, as you can see, there is a ton to discuss here.

Hopefully my book in progress will make some good sense of these matters, even if (as books tend to be) it does so very imperfectly. I guess we'll have to wait and see! ;)

[Untenured Ethicist]

Is there a connection between the Discovery Model and *increasing* polarization? Are Americans more inclined to make judgments in accordance with the Discovery Model than we were in 1994?

I would be careful about excluding people from moral discussion on the basis of their supposedly unreasonable views. We do not agree about what moral views are reasonable any more than we agree about what moral views are true.

Some people think it unreasonable to deny that bees and lobsters are sentient and that their experiences of pain matter. Others think it unreasonable to assert that bees and lobsters have experiences of pain that should matter to us. I would not exclude either group of people from moral discussion. Nor would I exclude people who think that morally significant human life begins at birth.

[Marcus Arvan]

Untenured ethicist: Thanks for your comment and feedback. I discuss the first issue (increasing polarization) in the paper. There are reasons to believe that the internet and social media dramatically increase the Discovery Model’s polarizing effects—so that would explain increasing polarization since 1994. I understand your concerns about excluding people from conversation, and argue in my 2016 book that we should err on the side of inclusiveness in conversation and negotiation. The point of the model, though, is that we do need to draw the line somewhere. And this seems like common sense: we shouldn’t compromise with Nazis, slaveowners, racists, etc. The question is where to draw the line—and this is an area I’m still thinking about.

[Derek Bowman]

Marcus,

Thanks for the detailed response. It seems that, while the Negotiation Model may be an alternative to the Discovery Model with respect to some discrete moral questions, it only works against a substantial background of 'Discovered' norms. So it can't be an alternative with respect to the moral domain in general.

This makes the account much more plausible than it would otherwise be - as you say we have to draw the line somewhere, and some things shouldn't be up for negotiation. But this also casts doubt on the ability of the Negotiation Model to solve the problem of polarization. At best it can hope to do so on some specific issues, among those who already share a very hefty overlap of moral commitments. And it seems to exclude at the outset of a number of parties to some of the most acute polarized disputes we have.

[Marcus Arvan]

Hi Derek: That is mostly correct - but I am pretty clear about that in presenting the model!

As I defend it, there are several regulative ideas we can discover through reason (coercion-minimization, mutual assistance, and fair bargaining). The crucial point is that the model *denies* that other first-order moral questions (about abortion, gun control, etc.)--you know the main things that divide us!--can be discovered.

Another way to put this is: the model holds that we can *discover* that we should negotiate fairly (constrained by a number of regulative ideals)--but all other moral questions must be settled by negotiation.

I think that is a pretty distinctive view--one that, if people came to progressively believe it and follow the conversational norms I defend, would significantly mitigate polarization.The key, though, is for it to actually get through to people--that is, for people to accept the model and *change* their commitments to conform to it.

As Untenured Ethicist's comment indicates, changing how people think and speak about morality will be difficult to change. For example, Untenured ethicist writes, "Nor would I exclude [from conversation] people who think that morally significant human life begins at birth." The very point of the Negotiation Model is that people need to stop thinking that there *are* facts to discover like that "morally significant human life begins at birth." The point is that it will be when we stop believing these kinds of things--and tolerating this kind of conversational move (asserting a 'discovery' that fetuses don't matter until they born!) that we really stand any change of preventing the kind of polarization that has torn the world apart since...well, forever.

The fundamental changes to our entire conceptual scheme the Negotiation Model recommends--to our deeply-entrenched moral practices, conversational behavior, and moral-belief formation--may be very, very difficult to achieve. Still, for all that, I think the model may well be right. And when has anything important in this world been easy to achieve? (I also think the model has very important implications for programming ethical A.I. so that robots don't kill and enslave us all, so that's another thing!: https://philpapers.org/rec/ARVMTS ).

[Derek Bowman]

Marcus,

The way you point in the paper - as in your most recent comment - makes it seem like the 'discovered' regulative ideals provide some background constraints, but that most of the substantive moral content comes out of negotiation.

But your response to my initial comment seems to show the regulative ideals doing most of the important moral work. You claim the Negotiation Model precludes thinking that there are any discoverable moral facts of the sort that "morally significant human life begins at birth."

But in fact your model of negotiation on the abortion question appears to rest on the discovered answer that sentience is the mark of morally significant human life. The only thing up for negotiation is the best way to respond to that moral significance. But there seems to be no room to negotiation with those who would either deny the importance of coercion-avoidance for sentient fetal life, or for those who would insist on the importance of coercion-avoidance for pre-sentient life.

[Marcus Arvan]

Hi Derek: thanks for following up. I think I've been equivocating a bit in this conversation for two reasons.

First, we need to distinguish the Negotiation Model per se from my own favored version of it. Although the version I have defended (Rightness as Fairness) holds that we can indeed discover pretty strong constraints on negotiation (including the moral relevance of sentience), this is just my own favored view--whereas the present paper is about the Negotiation Model in general (which includes other people's theories, like Muldoon's, Habermas', etc.). So that's one thing.

Second, I'm revising my second book right now and revisiting these very issues--in ways that are tempting me to change my minds on some things here (including a fairly different account of how to circumscribe the limits of negotiation that might be more permissive).

Anyway, I agree that that this this is the most difficult part of the Negotiation Model to get right. Some proponents of the model (Muldoon) defend little to no constraints (if I recall correctly)--and that leads to serious problems (or so I and some of Muldoon's critics think): namely, leaving too much on the table open for negotiation (including evil views). It seems like the model needs *some* constraints on negotiation, and so the trick is to get those discovered constraints right. As you and Untenured Prof note, my account in RF may have been too restrictive--and this conversation (as well as some stuff I'm working on in draft) have me toying with a somewhat different approach. We'll see how it goes. But I hope this helps you understand why my comments have sort of been equivocating on this stuff (the long and short of it is: this is still stuff I'm working on figuring out better!).

[Marcus Arvan]

An interesting new study just published related to these issues: https://link.springer.com/article/10.1007%2Fs11109-019-09526-z