Category: On ErikaRS’ Mind

First, the caveat: This is neither a critique nor a defense of J.K. Rowling’s recent commentary on Twitter (link, link) or the subsequent essay where she expanded on her initial statements.

Rather, this is an exercise in exploring applying the principle of charity to see if we can extract any value from a argument which, on its surface, seems to provide no value or anything to think about.

I’m going to mostly focus on the second twitter thread. The first tweet tried to communicate a subtle point through sarcasm and was, in my opinion, a complete failure. The essay expands on the second Twitter thread but wasn’t published when I started thinking about this. 😀

So let’s start with Rowling’s words: “If sex isn’t real, there’s no same-sex attraction. If sex isn’t real, the lived reality of women globally is erased. I know and love trans people, but erasing the concept of sex removes the ability of many to meaningfully discuss their lives. It isn’t hate to speak the truth.”

I, of course, have no more insight into what Rowling “really” meant than anyone else. But if I interpret her words through the lens of trying to find the nugget of truth, I see her as making a two part argument.

1. Being biologically female entails a lived experience that is distinct from the experience of gender.

2. The effect of sex and gender can only be understood through the lens of intersectionality. Thus, cis women experience forms of discrimination which are not reducible to sex or gender alone (so do the other intersections, of course).

Let’s tackle the first part first. Does biological sex affect a person’s lived experience? From my perspective as someone with a uterus, the answer is most assuredly yes. There’s the lived experience of menstruation and the whole constellations of concerns around pregnancy (preventing it, achieving it, experiencing it). Those with female bodies who do not experience these things in what is considered a medically normal way still have their bodies judged by these standards.

Then there are the social implications of these lived physical experiences. Debates about birth control and abortion and fertility treatments, the way society judges pregnancy, these all illustrate that having a biologically female body does have significant impact on a person’s lived experience. Enough for biological sex to be worth considering an identity that is separable from gender.

What about the second part? Can we understand biological sex in a non-intersectional way? Can we say that the impact of biological sex can be analyzed independently of the impact of gender? I think this is a harder question to answer, in part because there is no single intersectional experience (kind of the point).

But overall, I am inclined to say that we must understand the interaction of gender and biological sex in an intersectional way. The stories shared by transfolk show that their experience of biological sex is often highly influenced by how it does not match their gender. And in my personal experience, sex and gender are highly tangled, even as I see how they are separate factors in my life.

Bringing all this back to Rowling’s comment, I think I land on saying that erasing the concept of sex does erase something important from the lived experience of individuals. However, I don’t think that transfolks and their allies are trying to do that. Rather, I think they acknowledge that sex has meaning but also want society to more broadly acknowledge that sex and gender are separable concepts. Understanding exactly how that separation works is a complex and subtle space where, I think, our understanding is still evolving.

Originally published on Medium on May 13, 2020.

One of the most important skills to cultivate if you want to be an effective leader across groups with multiple goals and perspectives (i.e. any human group) is to learn to separate observations from the narrative you build around those observations. That doesn’t mean forgoing your own narrative; you will have one. Rather, it means understanding that your narrative is just one of many that can be assembled from the observations at hand. (And that’s before you start to account for the fact that your observations generally account for a subset of all relevant observations.)

The separation is important for many reasons. One, it allows you to better listen to the concerns and narratives of those around you. Second, it gives you more flexibility and creativity to try on different narratives and see where they do and do not have better predictive and explanatory power. Finally, when you are not attached to your narrative you are better able to change it when it is no longer consistent with the observations.

I’ve heard it said that theodicy — the problem of how an all powerful God allows pain, suffering, evil — is a modern problem. (Modern in the sense of a Christian tradition going back a couple thousand years.) It’s interesting to think on why this may be.

I suspect that, in part, it’s because in the pre-modern era, humanity’s lot didn’t really get any better. Thus, the world was as the world was. If there were things that were bad, it was not implausible to think that it was somehow necessary.

As we entered modernity, humanity started to improve things. We started seeing that things could be better. They could be better through the effort of us mere humans. Thus, it makes more salient the fact that an all powerful God didn’t make things better, even in the small ways that humanity was able to. And so for those embedded in the modern mindset, theodicy becomes a critical problem. One critical enough to be at the root of much of the Western version of atheism.

I suspect a postmodern attitude tends back toward finding it less of an issue. Not because of any assumption that the world as it is has the best possible form. Rather, I suspect it’s because a postmodern approach to theodicy sees omnipotence as merely a lens for describing what deity might be. Saying God is omnipotent is a way to describe useful properties of people’s theological beliefs, but it need not be taken literally as all powerful, in the sense that we imagine power.

Originally published on Medium on January 8, 2020.

One of the scenarios that drives me crazy as an owner of a complex and complicated tech stack is this: we want to be able to do thing X. Someone investigates thing X and decides (rightly!) that it’s not worth it for now but that it would be great to do someday. Then we completely drop that exploration and move onto the next thing.

What we should do is capture the core elements that are directional in an architectural north star for our stack. Then as we are making decisions in the relevant space, we can use whether or not it moves us in the direction of the north star as one of our guiding criteria. This should be a living document. It should be brief. It should talk about the high level capabilities we want to have, not how we would achieve them.

The piece that is missing when we drop explorations is the recognition that even if we’re not going somewhere right now, there is still value in knowing where we want to go. I believe that you can get a lot more value out of evolving systems than we currently do.

That said, I’m the manager. So if folks don’t know how to do this or that they should, it’s my fault ¯\_(ツ)_/¯

Originally published on Medium on November 21, 2019.

Don’t underestimate the power of closing open threads of thought and action. Things that aren’t closed out take up more attention than they should. You see them, and you have to decide “What was this? Why do I care? Is this done? What’s the next action?” And since these items are often left in an ambiguous state, this will require reprocessing part or all of the information time and again.

Particularly insidious cases are the not quite done task and the currently unactionable task. The first are things where the primary goal has been achieved, but there’s something left that would be nice to do. For bugs this often looks like a clean up task. For email it often looks like something that the email reminds you of but which is not actually the direct task the email thread was meant to achieve.

The second category includes things that will be actionable at a particular time or when a particular prerequisite is done. It also includes the even trickier case of things that you _could_ take action on, but the limiting factor is time or interest (“I ought to…”).

You probably have more of these open threads than you think. Emails in your inbox, open bugs, unresolved doc comments, open tabs, meeting notes, unanswered chats. Essentially, every channel of communication is a source of open threads.

This is not a new observation, and neither is the solution for handling it. Close the threads, and find some way of capturing the niggling remainders—the things that kept you from closing it out—in a more actionable form. Create a new bug with the remaining work and the appropriate priority, and close up the old one. Integrate the content in the open docs thread into the doc, and close it. Add the action the email reminds you of to a to-do list, either your active one or your aspirational one, and archive it. Or snooze the email to a future date, whether a specific one, or an arbitrary one sometime in the future (although snoozing often requires reprocessing the information when it pops back into your inbox, because it doesn’t allow you to snooze with a note; but at least you do it less often). For managers, often closing a thread means finding the right owner and communicating both transfer of ownership and expectations.

This takes time, but it is better to spend even a minute or two processing something now than have it take up 30 seconds again and again and again, perhaps multiple times per day for many days. Even more than the directly attributable time is the gain from decreased context switching. Future you will thank current you.

Originally published on Medium on September 13, 2019.

Yesterday we had curriculum night at our daughter’s elementary school. All-in-all, I liked what I heard. It sounds like the academic programs are strong, the enrichment programs were varied, and there is a real emphasis on community building and social and emotional intelligence.A lot of this is because of the activities of a very strong and, critically, a very well funded PTSA. The PTSA even funds teachers aids to improve the adult to child ratio in the classrooms.

This is where I get to the part I didn’t like. All the talk of what the PTSA did with donated money emphasized how much inequity there is in the public school system. A lot of what makes our elementary great comes from the ability of many of the parents in the area to make significant donations to the school. What should be the baseline quality of education becomes something we get because we can afford to live where we do. I love that we have great public schools, but the emphasis on all the great things our school can do because of parental donations almost makes me feel like a private school would be a more honest choice — at least then everyone is on the same page about being able to trade financial privilege for a better education.

Note that I’m not saying we should restrict parental donations and involvement. Schools are a crucial part of community building and parental involvement is a crucial part of educational outcomes. Plus, local communities should have the freedom and flexibility to customize their schools to fit their local needs and values.

However, parental funding and involvement shouldn’t be necessary to provide a good baseline educational experience. The funding for that experience should be fully funded by public dollars.

Originally published on Medium on June 26, 2019.

Comparative advantage is an important concept in economics. It explains why, in general, it is advantageous for folks to cooperate even if some people are better at nearly everything. The gist of the idea is that there is often an opportunity cost to doing an activity (you aren’t doing something else that is potentially more valuable to you). Thus, even if you can do something better than someone else, they may be able to do it at a lower opportunity cost, and it may be worth your while to pay them to do it instead.

This becomes particularly interesting when thinking about the distribution of resources in society. Under absolute advantage thinking, if I do something that produces value X and you do something that produces value Y, then I deserve X worth of society’s value and you deserve Y. We can trade to adjust the exact content of the value, but the proportion of the value is directly related to the value produced by an individual. An alternate way about thinking about the distribution of resources in society is that they should be more equal or based on need. As Marx said, “From each according to his ability, to each according to his needs.” For the purpose of this essay, I will assume we are aiming for proportional distribution but will question the common thinking about what proportional means.

Thinking about comparative advantage provides yet another way of thinking about distribution. Before I get into the details, a caveat: the explanation I am about to go into uses a counterfactual world. In reality, we don’t have access to the counterfactual world, so what follows will not be an explanation of how to actually execute this alternate distribution scheme. What I present here is a principle for evaluating options, not a policy in its own right.

With that caveat, how does comparative advantage give us another option when it comes to distributing the value produced by society? Imagine two worlds. In the first, I produce X and you produce Y, where X is greater than Y. In the second world, I produce X and you produce Y, but I am only able to produce X because we trade in a way that frees me up from doing an activity that only produces Y value.

In the first world, if we want distribution to be proportional to production, you get value Y and I get value X. The total value produced by that two person society is the sum of our independent production decisions.

In the second world, if we had acted as independent agents, then the total value created would have been 2Y. It is only because we cooperated that I was able to produce X. Your comparative advantage allowed us to produce a value X+Y for society. Thus, we both deserve credit for the excess value created by cooperation (X-Y). The portion of society’s value that I deserve is something less than X and yours is something more than Y. This difference is not a wealth transfer. It is a fair distribution of the value created by choosing to collaborate.

Once you start thinking about things this way, the idea of comparative advantage and its impact on distribution and allocation of credit starts appearing everywhere!

(This builds on but does not directly reflect some ideas developed in Ryan Muldoon’s Social Contract Theory for a Diverse World: Beyond Tolerance.)

Originally published on Medium on June 11, 2019.

I have a narrow definition of cultural appropriation. (Short version: it’s only cultural appropriation if the thing being appropriated has symbolic value that needs to be earned by those within the culture it is taken from.) Although this definition has flaws, I believe that the definition of cultural appropriation needs to be narrowed somehow from common usage if it is to avoid condemning all cultural exchange. Cultural exchange is the lifeblood of cultural evolution, and it has been the source of much good.

Still, with such a narrow definition of cultural appropriation, it is useful to think about the broader set of questionable cultural borrowings. For example, copying clothing that is strongly associated with a particular culture may not be appropriation, but it may imply a level of affinity with that culture which members of that culture would not consider appropriate. For cases like these, where the cultural item borrowed does not have granted symbolic value but is perceived as closely tied to the originating culture, I propose the term cultural infringement.

The term cultural infringement is intentionally meant to invoke the idea of copyright. Copyright is a delicate balance between control and usage. While copyright cannot be directly translated to cultural infringement, it does have some useful lessons to teach us. One is the idea of fair use: there may be contexts in which it is appropriate to use the artifacts of a culture even if members of that culture might disapprove. Criticism is one such use.

Another idea we can learn from copyright is that borrowing can often be generative. See, for example, the rich world of fan fiction. Much of it is terrible, but the best fanfic helps the reader learn something new about the work. Similarly, cultural borrowing can often be beneficial and lead to mutual enrichment both for the source culture and the destination culture. For copyright, we aim to balance control and generation by having copyright expire. That does not apply to borrowing culture, but the observation that borrowing culture can be generative still stands.

As the expiration of copyright illustrates, cultural infringement cannot be literally treated as copyright. Most of the time, there is not a recognized owner of a culture, and members of the culture may perceive acts of infringement differently. However, thinking of the trade-offs of copyright can lead us to think about the trade-offs, good and bad, that come from non-appropriating borrowing from other cultures. It helps us think about how a cultural has certain claims to its own artifacts but also how giving too much control impedes positive change.

In this framework, whether a particular instance of cultural borrowing is generative or infringing depends largely on how the borrowing is perceived by the members of the source culture. To understand this, we need to listen and learn from the perspectives of those in the culture, both those who feel the borrowing is inappropriate and those who feel it is reasonable. If cultures are going to realize the benefits of borrowing from each other, taking time to learn a bit more about what is being borrowed seems like a small price to pay.

Originally published on April 26, 2019.

Why take time to try to really understand what someone is saying? It increases your credibility.

Let’s start by clarifying what I mean by really understanding what someone is saying. In my mind, the key property of such understanding is that when you summarize their statement, either explicitly or implicitly via the assumptions in your reply, the other person agrees that what you said matches what they said. An understanding summary reflects the content of a statement in a way that the original speaker considers accurate. It does not say what you think they “really” meant. And it means that when you’re unsure then you ask questions instead of making assumptions.

What about credibility? Why does understanding this way increase credibility? Let’s look at what happens when you don’t really understand what someone is saying. When that happens — especially when it happens repeatedly between the same pair of people — your lack of real understanding starts to make the other person question your assessment of behavior and information in general. “If,” they think, “this person gets it so wrong on these statements of mine, why should I believe that they are accurately representing anything?” By not listening, you are casting doubt on your basic ability to present reliable data.

In my experience the only way you can get someone to listen to your response is to first show you can listen to them. It’s far far from a guarantee. Still, a possibility is better than an impossibility.

Originally published on Medium on January 3, 2019.

This post is about my views on the morality of abortion. I strongly support access to safe and legal abortion. I also support decreasing the need for abortion by improving access to contraceptives and improving the social and economic conditions that lead to women aborting pregnancies with regret. Since some may find it relevant, know that I had an abortion (of a much-wanted baby who had serious medical issues).

Abortion is complex. I have yet to see a claim that it is morally simple that did not ignore myriad issues. A person’s conclusions can be simple. Yet those conclusions, if well founded, come from a moral reasoning process that must deeply contemplate some of the fundamental issues we face as ethical beings.

Abortion is complex.It creates a direct conflict between two individuals: the person carrying a fetus[1] and the person that fetus could become (or is, depending on your philosophy of personhood). When a pregnancy is unwanted, one person will become a slave for nine months or another person will die. And yes, that sentence was constructed to elicit anger. Slavery is a strong and loaded term. Yet, in this age when pregnancy can be safely aborted, what is a forced pregnancy if not involuntary and inescapable control of a person? And whether or not one’s philosophical belief is that a fetus is a person at the time of an abortion, most aborted fetuses would otherwise have become a person with time.

Abortion is complex. Thus, with this essay, I am not aiming to change anyone’s position on the issue. My goal is to give a brief outline of the dominant factors that influence my perspective on the moral factors involved with abortion.

What follows is heavily influenced by

• Jeff McMahan’s The Ethics of Killing. In this long, dense book, McMahan discusses personhood, death, killing, autonomy, self-defense, and many more topics while developing his Time-relative Interest Account of death. You can read my shallow-yet-detailed summary.
• Alison Gopnik’s The Philosophical Baby and Lise Eliot’s What’s Going on in There discuss how the brain develops in the fetus, infant, and young child. Both books strongly influenced my view of personhood. In particular, they reveal the artificiality of many of the lines drawn in both popular and philosophical explorations of personhood.
• Judith Jarvis Thomson’s essay “A Defense of Abortion”, a.k.a., Thomson’s violist, has also been deeply influential. In particular, it influenced my views on where abortion stands on the spectrum of moral requirements (prohibited, permitted, encouraged, required).
• The many discussions I have had with folks from all across the spectrum of beliefs about abortion over the years.

[1] I don’t say mother here for two reasons. One is that not everyone who carries a baby identifies as a woman and a mother. Second is that, regardless of the debate on gender identity, calling someone with an unwanted pregnancy a mother is already making a moral judgment about pregnancy, and the purpose of this post is to analyze the morality of abortion.

What are we?

Unless unknown intelligences are reading this, we are homo sapiens, human beings, creatures commonly called people. Yet usually, when we talk about what it means to be a person, we are not talking about a genetic identity but rather a functional one. To inaccurately summarize volumes upon volumes of philosophical thinking (of which I have only read a tiny tiny fraction), we are conscious beings. We have an awareness of the past and the future, we make predictions, we reflect and, above all, we each do this from the perspective of an “I”, whatever that may be.

Yet this “I” does not spring into being fully formed. No matter how we define it, a fetus is not a fully developed human being. It is a work in progress. The development process is continuous, with few bright lines marking progress. Even the brightest line that exists — birth — is only bright from the perspective of the person carrying the fetus; developmentally, a fetus one day before birth and a baby one day after are very similar[2].

• Belief: Pregnancy is a continuous process during which structural, functional, and psychological continuity start out weak and become gradually stronger over time. Birth does not significantly impact structural, functional, or psychological continuity.

Note that in healthy adults, structural and functional continuity are effectively identical. However, when thinking about what McMahan calls the margins of life, it is useful to separate the two. Structural continuity is the continuity of the physical matter of a person: cells splitting and dying yet still part of one human organism. Functional continuity is the working of the systems of that organism.

Although I label it a belief, continuous development is really a fact of fetal development. From here on out, my stated beliefs will be philosophical. I will not attempt to rigorously defend them. I will merely sketch the reasons these particular premises appeal to me and attempt to trace some of the consequences.

As noted above, I relate personhood to consciousness. Another common position is to relate it to the presence of a soul — some part of the self that exists independent of a physical body and mind. As an atheist, I do not believe in souls.

• Belief: The idea of a soul — an aspect of one’s self that is independent of one’s existence as an embodied mind — is irrelevant to the morality of abortion.

It is worth sitting with this premise a moment longer. Many people believe that abortion is immoral even from an early stage because a soul confers personhood well before any embodied sense of consciousness could possibly be present. This is a defensible position to take (although ensoulment has its own philosophical complexities even among those who believe in souls). However, this is a fundamentally religious belief and asking others to believe abortion is immoral based on an argument that depends on people having a soul is, in essence, forcing religious beliefs upon others.

Embodied minds

If it is not a soul that matters, what does matter? I believe that we are essentially embodied minds. More concretely,

• Belief: Our personhood relies on our structural, functional, and psychological continuity with psychological continuity being the most important.

Structural and functional continuity are matters of anatomic and physiological definition. Psychological continuity comes from our maintenance of strongly connected and overlapping sequences of psychological states. There is a lot packed into this one. Probably ten or more hours of reading from The Ethics of Killing summarized into one imprecise sentence. This premise appeals to my intuitions that we remain the same person even if our psychological continuity decreases (e.g., because of a cognitive degenerative disease).

It also allows for my intuition that a physically continuous self that does not share psychological continuity is, in some sense, less me than one that does. If my psychological continuity were to diverge from my physical continuity (e.g., a sci-fi body swap), I imagine feeling a greater sense of concern for the psychological continuous version of me (unless I expected to get the previous body back). This has what some philosophers, including McMahan, consider to be disturbing implications since it implies that “I” can split into two separate identities — and perhaps be merged from multiple identities. I do not consider the uniqueness of “I” to be a requirement.

The rolling rock of consciousness

Often the gold standard for psychological continuity has to do with knowing I am me, for various formulations: self-awareness, egoistic concern, awareness of death, etc. I diverge from this; fetal development has influenced my views. I see consciousness as the culmination of a continuous process of increasing continuity rather than a minimum bar. My best guess is that consciousness arises from our existence as complicated feedback loops. Thus, I put the minimum bar for personhood at the point where that feedback loop is kicked off.

To use an analogy, I am looking for the point where the rock starts rolling down a hill rather than where it reaches the bottom. A rock that is stable is only a tiny bit different than one that has just started to roll, yet the divergence will, in the course of normal events, result in a dramatically different fate.

Given that most fetuses (although not most embryos) become people if left to develop, one could say that pregnancy is when the rock begins to roll. Yet given the weight I put on psychological continuity, I prefer a point that is psychological rather than merely physical. This is my current best attempt at pinning down what that is.

• Belief: What makes us people is our ability to learn by forming, testing, and updating increasingly elaborate models of the world.

Studies of prenatal development have shown that this process starts in the womb. Language is the best example of this since language is so clearly a learned phenomenon. At birth (and, more recent studies have indicated, before birth), babies can differentiate phonemes from their native language. Clearly, babies have a long way to go from here to understanding “me” (the evolving interaction between an infant and a mirror is quite entertaining), but it is this same process of model building that eventually gets there.

This places my view of personhood much earlier than the more common consciousness-based views. Those views place personhood anywhere from months to years after birth. It can be as late as 4–7 years old if awareness of death is taken as the point at which one is fully aware of oneself as an independent self. For philosophers, the challenge this raises is that it implies that if late abortions are morally permissible, then infanticide may be too. And that aside, as a parent, my problem with the self-awareness criteria is that interacting with a baby and watching them learn shows that even a newborn is categorically different from an animal.

When do we start creating and updating models? Structurally, we know a lot about fetal brain development (a super technical article; a more approachable article). It is less clear what this means cognitively. We do know that it is late in the second trimester that the developmental focus in the brain shifts from the brain stem to the cerebellum and cerebral cortex. This is the absolute minimum level of structure likely necessary for learning and counterfactual development. Thus, the end of the second trimester is an early (but not impossible) lower bound for considering the fetus to have achieved personhood. To sum that up more conversationally,

• Belief: To the best of our knowledge, sometime around the third trimester (for a normally developing pregnancy), an abortion may cause a person to die.

Note that in this view, it does not matter that during the early parts of the process the fetus has a lower level of psychological continuity than various animals that we never attribute consciousness to. What matters is the process, not the current product.

[2] Birth does kickstart some interesting development. This becomes interesting when assessing development milestones for a baby born early. Sometimes gestational age is a better marker of milestones, sometimes birth age is, and sometimes it is in between.

Why is death bad?

Now we’re done, right? At the beginning of the third trimester, an abortion causes a person to die. Killing people is bad. Abortions before then, not after. kthxbye

Remember, though, that this is a continuous process. The line is blurry, not bright. Earlier in development, the fetus is still on its way to becoming a person. And even after birth, a newborn is at the beginning of a long process that involves ever-increasing levels of mental sophistication.

Add to this the sometimes-hard-to-admit fact that we do not consider all deaths equally bad. I do not mean this in the sense that those still living may care about particular deaths to differing degrees. Rather, the difference I am referring to is how bad death is for the person who died.

• Belief: The most important (but not only) reason death is bad is because of the loss of the future to the person who died.

There are two intuitions behind this belief. One is that the external regret for a person’s death is incidental. That one person was beloved in their community and another a hermit in the woods is, in the personal sense, irrelevant when we consider their deaths. The other intuition connects the death back to the idea of personhood. Since what matters to me is structural, functional, and psychological continuity, the badness of death arises from the destruction of a self — the disruption of continuity.

But what is that loss? Let us approach it in pieces.

• Belief: The value of a lost future can vary from death to death.

This is one of those statements that is more easily intuited than justified (with all the dangers that entails). Imagine two deaths: a young adult just starting to feel like they are able to realize their dreams and an older adult who feels they have lived a long and fulfilling life. While the death of either would be sad (especially depending on the cause of death), it is deeply intuitive to me that the death of the older adult who feels they have fulfilled their dreams is less regrettable than that of the young adult about to grasp life fully.

Yet this intuition quickly becomes uncomfortable once we start trying to categorize which types of death are worse than others. Is the death of an older adult who has lived a fulfilling life more or less regrettable than one who has lived an unhappy life which has no prospects of changing (as far as we know)? I do not want to be the one to make that call. Yet I think our discomfort points at an important idea. Perhaps the focus on whether a life is good or bad is a red herring. The important part is that the young adult loses a life that they rationally expected to have and the older adult, whether fulfilled or disappointed, expects less future. More precisely,

• Belief: Since we cannot predict how much future goodness or badness any individual would have, the value of the loss of a future depends on how much reasonably expected psychological continuity that future contains relative to the present self.

If earlier premises made large philosophical jumps, this jump must be described as enormous. In fact, many of the 560 pages in The Ethics of Killing are spent developing the basis for something similar to this as a reasonable claim (my credit to McMahan on this one is huge). Since this post is meant to summarize where I currently stand on the morality of abortion, we’re going to skip all that. 🙂

I will make a note on the caveat of a “reasonably expected” future self. This is my way of cutting through the philosophical tangle that McMahan spends a lot of time on. We cannot know a person’s future. It may be that if someone had not died painlessly today, they would have died an incredibly painful death in two weeks, in which cases their earlier death could be seen as less of a loss — even a relative good. However, when we are evaluating loss, we are generally doing it relative to expectations, not relative to an oracle’s prediction of alternate futures. Our expectations are all we realistically have.

Relating this back to the fetus, while a fetus, in general, loses more good (and more bad) by death than those who die later, the strength of their connection to their future self is non-existent for most of the pregnancy and, at best, incredibly weak for months and years after birth. They have no expectations of psychological continuity. Thus, while their loss is greater in absolute terms, the fetus experiences no loss.

This distinction between absolute loss and experienced loss brings up an issue worth thinking about. It implies that death is less bad for an infant than it is for an adult since an infant will not experience the loss in the same way. Since we are only considering the aspect of badness from death that relates to the personal experience of loss, I do not consider this problematic.

The discussion above lays out that abortion does not cause a bad death for the fetus because the fetus does not experience death in the way that someone with stronger psychological continuity would experience it. However, it does not imply that there are no other reasons to regret the loss of a fetus. Rather, it says that if we consider the loss of a fetus bad, it is because of the other reasons death is bad.

So far…

What I have laid out so far implies that, in my view, killing a fetus is more than killing an animal because it would, under the course of natural development, likely become a person. It is, however, less than killing a person until at least the third-trimester since the structural requirements for the beginnings of psychological continuity are not present. I have also laid out that even after the beginnings of psychological continuity, death for a fetus is less bad for the fetus itself than death for a more developed person because that psychological continuity is so weak and, therefore, strictly less bad overall than the death of a more developed person.

What this does not establish, however, is whether or not killing a fetus is morally permissible.

Moral Expectations

Now that we are talking about actions, it is necessary to understand the different moral expectations we may put on actions. In my readings, it has been most common to break down moral actions into three categories:

• Morally necessary actions are actions that it is immoral to not perform.
• Morally permissible actions are neither forbidden nor required.
• Morally forbidden actions are actions that, if taken, are immoral.

This common breakdown fits a whole spectrum into the realm of morally permissible. Some morally permissible actions may be worthy of praise. They make the world a better place even though they are not required. Other actions may be worthy of negative judgment. They make the world a worse place even if they are not forbidden. And some actions are truly morally neutral. Taking a page[3] from Gregory Mellema’s The Expectations of Morality, we can split the morally permissible into three categories:

• Morally encouraged actions are not morally required but are worthy of praise.
• Morally neutral actions are worthy of neither praise nor judgment.
• Morally discouraged actions are not morally forbidden but they are worthy of blame or criticism.

It is useful to make these categories explicit. In practice, a relatively small set of actions are morally necessary or forbidden. But that does not mean that most actions are morally equivalent. There are differences, and we need a way to label them. (As an aside, I am intentionally ignoring whether morality is an issue just of action or if it is also a function of belief. Since we are discussing acts, this should not make a difference.)

[3] Literally just taking the page. I haven’t read this, but this was the only reference I found for the phrase “morally discouraged”. That does not mean the concept is novel to Mellema’s work, just that the same phrasing that occurred to me is the one that he uses. And although I have not read this yet, I did buy the book 🙂

Killing

Some people believe that killing people is always bad. Under that assumption, abortion is wrong as soon as a fetus has any claim to personhood. I, like many, believe that killing is morally permissible under some circumstances. When is killing morally permissible? Continuing as before, this is not a full-fledged philosophical discussion of the morality of killing. These are the intuitions that influence my views.

Before discussing abortion, I want to discuss some situations in which at least some people believe that killing is not morally forbidden. None of these are perfect analogies for abortion and they should not be taken as such. Rather, their purpose is to illustrate a spectrum of feelings we have on killing.

Defense

The obvious case of permissible killing is killing in defense, either of oneself or of others (especially others unable to defend themselves). What makes for allowable killing in defense is an area of debate.

• Belief: Killing in defense is morally permissible if (1) there is an immediate threat (not a potential threat) and (2) there is real uncertainty as to whether or not the threat can be defused without harming the one creating the threat.

What it means to create a threat is non-obvious. I personally hold to a strong form of defense: it is permissible to defend yourself even if the person creating the threat is not morally responsible for the threat. For example, if an explosion were to turn a person into a projectile coming at you and you could somehow deflect them by doing harm to them (e.g., hitting them with another projectile), I consider this a sufficient justification. Thus, for me (unlike for McMahan in The Ethics of Killing) permissible defense does not require that there be a redistribution of justice.

• Belief: There is no requirement that the person who creates the threat be morally responsible for the threat.

That said, I do believe that the greater the degree to which the one creating the threat is responsible for it, the more justifiable defense becomes. It is merely morally permissible to kill someone who is threatening you because they have been involuntarily made into a projectile. It is, perhaps, morally encouraged or even morally required for someone who is capable to defend against someone who is actively presenting a credible threat to a group of defenseless innocents.

There is also a dimension of proportionality when it comes to the moral expectations around killing in defense.

• Belief: If a threat can be diffused without killing (even better, without violence), then that is a morally preferable outcome. If a threat is not credible, killing is not permissible.

Proportionality also helps us think about what threats justify killing: the threat has to endanger that which is so precious to us that it is proportional (but not necessarily equal) to life itself. A threat to lives is clearly one such case. I believe that significant threats to one’s well being or autonomy (think kidnapping, not taking the snack you want) are sufficient justification.

Failure to save

Another category of killing is letting people die. In this discussion, I am referring to situations where one had a concrete chance to save another and chose not to: think about watching someone drown, not people dying of hunger on the other side of the world. Put another way, what is the morality of failing to be a Good Samaritan? It is a matter of debate whether this should even be called killing.

• Belief: Being a Good Samaritan is not morally necessary. It is morally encouraged.

Since the non-samaritan did not cause the victim to need help and since there is always some cost of helping (including a chance of harm to oneself), there is no strong basis for a moral obligation here. However, I believe this case illustrates an important boundary condition. Being a Good Samaritan is morally encouraged. The strength of that encouragement increases the lower the cost/risk to the Good Samaritan. To borrow the language of mathematics, moral necessity is the limit as the cost goes to 0 of this moral decision.

Defense and failing to save are quite different. Still, they have an important similarity.

• Belief: It is morally acceptable to take one’s own interests into account even in situations that may end in harm to another. This is true even if allowing harm to oneself would lessen harm to others.

Both examples have elements of proportionality too. For defense, the larger the threat, the more harmful the response can be. For the Good Samaritan, the higher the cost, the lower the obligation to help. These sound different, but both get at the same idea: the more harm that would come to oneself, the more one can justify harm to others.

Rejecting compulsory saving

Now consider one of the most famous discussions of harm in the context of abortion: Judith Jarvis Thomson’s essay “A Defense of Abortion”. If you haven’t read it before, it is worth the time. The brief version: suppose a woman wakes up and finds herself attached to a famous violinist such that her body provides life support for theirs. If she detaches, the violinist will die. Is the woman morally obliged to stay attached to the violinist?

Before discussing the relevant points from that essay, it is worth noting the common criticisms of this as an analogy for abortion (as discussed in The Ethics of Killing). One objection is that although the woman did not consent to pregnancy merely by having sex (a ridiculous claim), she does bear some responsibility for the pregnancy (in the same way that a person who accidentally bumps another off a dock bears responsibility for that person needing aid). Another objection is that the biological relationship between the woman and the fetus has moral significance even without consent. A third common objection asserts that there is a moral difference between letting someone die by withdrawing support and killing them.

For this discussion, those are not relevant. I am less interested in the violinist as a direct analogy for abortion than as a way to continue looking at intuitions about the morality of killing. The violinist provides an intermediate point between defense and being a Good Samaritan. Being forcefully attached to the violinist constitutes a threat to the woman’s freedom, so detaching could be seen as a defensive action. Yet the woman is also making a choice about whether or not to choose someone in need, making this a Good Samaritan case too, albeit one with a very high cost. In fact, it might be fair to generalize Thomson’s violinist as a case of an involuntary Good Samaritan.

Since I have already concluded that being a Good Samaritan is morally encouraged, not morally required, neither the woman nor anyone else would have an obligation to become attached just because the violinist’s life is in danger. However, since she already is attached, is she morally required to stay attached? It is here that the comparison between the voluntary and involuntary Good Samaritans becomes interesting. In the case of the voluntary Good Samaritan, if nothing has changed — if the risks and timeline remain what they were when the Good Samaritan accepted the cost of aiding the person in need — then it seems morally necessary that they continue to provide that aid.

• Belief: Having chosen to provide aid under certain assumptions, it is morally necessary to continue providing that aid as long as those assumptions remain true.

The involuntary Good Samaritan, however, never agreed to take on the costs. Not only that but, in another disanalogy with the abortion case, she has suffered an injustice, not just bad luck. Some agent is responsible for her plight, even if we assume that violinist is innocent. The injustice factor certainly seems sufficient to revoke any moral obligation she has to continue providing aid. Is the lack of agreement sufficient on its own? I believe so. What makes sustaining aid morally necessary for the voluntary Good Samaritan is not that they have started paying the costs of aid; it is that they agreed to provide the aid. The voluntary Good Samaritan would be just as morally obliged to provide aid if they had committed but had not yet started. The involuntary Good Samaritan has not agreed and so is not morally obliged to give or continue aid.

• Belief: It is the agreement to provide aid that creates a moral obligation, not the action of providing the aid. Without an agreement, there is no obligation.

Since it will kill the violinist if the woman detaches, it would be reasonable to say that the woman is morally encouraged to remain attached. That moral encouragement is much closer to the neutral point of the spectrum than the necessary point.

Accidental creation of risk

It is valuable to consider one more example: the accidental creation of lethal risk. Suppose I, despite taking reasonable care not to bump into anyone, knock someone who cannot swim off of a pier into deep water. Further, suppose I have the skill to save them. Do I have a moral obligation to do so? This is similar to the Good Samaritan case. Thus, I am, at the very least, morally encouraged to rescue the non-swimmer. As with the Good Samaritan, the degree of moral encouragement increases as the cost of the rescue decreases. The difference is that in this case, the threat was created by my action. This responsibility, even if innocent in intention, increases the strength of moral encouragement.

• Belief: Having responsibility for creating a threat (even justifiably accidental responsibility) increases the strength of moral encouragement to mitigate the threat.

The greater the responsibility for the threat, the greater the moral encouragement. In cases where the person who can prevent the harm created the threat, the point of moral obligation will come before the cost of helping goes to zero.

• Belief: The cost which one is morally obliged to take on increases the more one is responsible for the threat.

Thus, a threat created innocently becomes an obligation if the cost to rescue is small but not if it is large. A threat created through gross negligence becomes an obligation even at a reasonably high cost. And if a threat was intentionally created (not in defense) one is morally obliged to prevent it at any cost (although someone who intentionally created a threat is unlikely to care).

Abortion

“Finally!” the reader thinks.

It is worth taking a moment to think about how each of the examples is and is not similar to abortion.

If a pregnancy is unwanted, then the fetus is taking control of another person’s body. Thus, abortion could be seen as a defense of one’s bodily autonomy. The threat is immediate (the pregnancy is occurring). The threat cannot be diffused without harming the one creating the threat (given current medical technology, at least). Note that, given my belief in a strong right of self-defense, the fact that the fetus is innocent does not reduce a person’s right to defend against the threat of pregnancy. Thus, the key remaining question is whether or not the defense (killing the fetus) is proportional to the threat (9 months of a loss of bodily autonomy with a non-trivial chance of long-term medical issues or death).

Since the person carrying the pregnancy did not choose to be pregnant, this case has some resemblance to the involuntary Good Samaritan case (this should not be surprising, given its origins). In so far as abortion is analogous to involuntary Good Samaritan cases, there is no moral obligation to carry the pregnancy. However, unlike in the discussion above, the risk did not exist independent of the person carrying the pregnancy. There was no external agent that set up this situation where the only way for the woman to retain her autonomy was to kill another. The exception, as I suspect many readers are thinking, is if the pregnancy was a result of rape. Thus, there is a sound basis on which to believe that pregnancies resulting from rape have different moral obligations than pregnancies resulting from voluntary sex (a point that I have been unsure of in the past, so I appreciate this distinction).

Accidental creation of risk is a stronger analogy for a pregnancy resulting from voluntary sex. The risk to the fetus only exists because the person is pregnant. Thus, the person who is carrying an unwanted pregnancy should consider themselves to be morally encouraged to prevent the death of the fetus, and they should be willing to take on a higher cost than they would to save some other person’s pregnancy that they had the ability to save. However, none of this discussion settles how much cost a person should be willing to take on. Because the loss of bodily autonomy is so fundamental and personal (even if only for a limited time), concrete guidelines are hard.

There is something worth calling out here that may make some people uncomfortable. Another belief in my analysis of the accidental creation of risk is that the greater the responsibility of the risk creator, the greater the cost they are obliged to take on. This means that, in this framework, someone who knows how reproduction works and had sex without birth control when they were potentially fertile has a greater moral obligation than a someone who used birth control properly and got pregnant anyway. The first person may even have a greater obligation to carry a pregnancy than an adult ignorant about reproduction, although that depends on your philosophy of moral obligations around obtaining accurate knowledge. I am okay with these distinctions because we are discussing the morality of abortion. These distinctions would be legally abhorrent because they push up against the idea of equal protection. For the purpose of this discussion, we must remember that two acts can be equally legal (or illegal) and unequally moral.

There is one more factor that the above discussion did not take into account. All of these discussions of killing assume that the one being killed is a person. As I noted earlier, I do not believe that an early fetus has any claim to personhood, although the line separating non-personhood from personhood is fuzzy, not bright. Pulling this all together, you now have enough background to understand where I stand on abortion.

• Belief: Abortion is morally permissible since a pregnant person cannot regain their bodily autonomy and be rid of a pregnancy without killing the fetus. However, because the person carrying the pregnancy usually created the risk to the fetus, abortion is morally discouraged. As the fetus moves along the continuum from non-personhood to personhood, that moral discouragement goes from weakly discouraged (nearly neutral) to strongly discouraged to morally forbidden once a safe birth[4] could be induced instead. Moral discouragement further strengthens if the pregnant person had made an earlier decision to voluntarily carry the pregnancy and there is no material change since the time that decision was made.

Some people will likely disagree when I say even an early abortion is (very weakly) morally discouraged. It is worth noting, however, that taking a morally discouraged action does not make one a bad person, especially if it is a situation where the cost of not taking that action is high. Littering is morally discouraged, but someone who littered in the course of getting out of the way of an oncoming truck was still making the best overall choice.

• Belief: Despite being morally discouraged as an isolated decision, an abortion may still be the best overall choice for a person carrying an unwanted pregnancy to make. The decision needs to be assessed in the full context of the costs it was being traded off against.

There is one more important consideration above. By committing to a pregnancy, a person becomes morally obliged to see it through. However, this changes if the information changes. As those of you who read my abortion story know, this point is very important to me since I had a late abortion after finding out that my baby’s brain was not growing.

• Belief: If something significant changes, either external context or (especially) internally to the fetus or the person carrying the pregnancy, then the moral decision can be reconsidered.

[4] What constitutes a safe birth is itself a matter which could be debated at length. Premature birth imposes real risks on a baby. I would say that, as a practical matter, having decided to carry an unwanted pregnancy, a person should carry it to term or as long as medically advised for her and the fetus. Morally, the person is only obliged to carry the pregnancy until the fetus would be fine once birthed. Practically, with current technology, that means a full term pregnancy.

A personal note: abortions for medical reasons

I intentionally stayed abstract about the specific reasons a person might choose to get an abortion. However, since I chose my own abortion when learning that my baby had a major (but non-lethal) medical issue, I want to briefly address this category of abortions. Only about 1% of abortions happen after 20 weeks and about 90% happen in the first trimester (source). However, many of those late abortions are because of medical reasons.

If the medical concern is for the health of the person carrying the pregnancy health, justifying the abortion is easy. If their health is threatened by a pregnancy, then they have a strong defense rationale for the abortion which overrides the earlier commitment to carrying the pregnancy.

But that does not explain why it’s ok for parents to choose to abort a child with significant but non-lethal medical problems. This is a hard problem. I believe that it is ethical to choose to abort a pregnancy for fetal health reasons. Still, despite having made this choice, I still struggle to clearly articulate the ethics.

Here are some of the factors that float in my mind: There is an idea that some lives are worth living and some are not. There is an idea that generally a psychologically developed person can only make that decision for themselves. There is an idea that a developing person is unable to make some decisions and that it is a guardian’s role to make the decisions they are incapable of making. There is an idea that since the fetus is not yet a person (or is very weakly a person), the death is not bad for them and so the decision to abort is more akin to them never having come into existence than it is akin to killing a person. But each of these needs a good deal more discussion to be fully fleshed out into an ethical view of parental responsibility with respect to medical decisions.