One of the strongest features (if not the strongest feature) of Marquis’s argument for the immorality of abortion is that it does not drag the debate into the territory of arguments over the status of an embryo or a fetus in terms of whether or not they count, at that stage, as human beings in any sense that gives them a status making them worth protecting. I think there are good reasons for saying that they do have such a status, but many proponents of liberal abortion laws do not. Reasoning solely in terms of the future of the fetus as Marquis does, therefore, makes those arguments rather moot.

However, in his case for the permissibility of the destruction of early embryos for the purpose of stem cell research (as opposed to stem cell research simpliciter, which does not require the destruction of embryos), Marquis does step into the territory of making judgements about the status of the entities in question, and when he does so, I think he goes wrong.

Marquis offers more than one argument in the paper that I am drawing on, however I am only going to address one argument, namely what I will call the argument from cell division. I will add my footnotes in [square brackets].

Marquis outlines the relatively simple “moral-principle objection” to the use of human embryonic stem cells (hESCs) as follows:

Opponents of hESC research argue that because intentionally to end an innocent human life is wrong and because intentionally to destroy a human embryo is intentionally to end an innocent human life, intentionally to destroy a human embryo is wrong. Because obtaining hESCs involves intentionally destroying a human embryo, obtaining hESCs for medical research is wrong. The view that it is wrong to end a human life in order to pursue the ends of medical research, no matter how valuable that research, is a moral principle taken for granted in medical research ethics. [Don Marquis, "The Moral-Principle Objection to Human Embryonic Stem Cell Research," Metaphilosophy 38:2-3 (2007), 191.]

I am going to assume here that hESC research always involves the destruction of a human embryo. Any hypothetical type of research that does not do this is beyond the scope of anything I say here. I think – and I daresay that anyone would think, on the face of it – that the above represents a very serious objection to the use of hESCs that requires a response from anyone who would defend the moral acceptability of embryonic stem cell research.

One trivial and implausible responses to this moral-principle objection are rejected immediately by Marquis, and rightly so, I think. On such response is the claim that we really don’t know when human life begins and it is just dogmatic to assert that it begins either at conception or so soon after conception that we can say with any certainty that hESC research destroys human life. As Marquis points out, “hESCs would be useless to researchers unless they were human and unless they were alive.” Still, we do need to qualify the moral-principle objection before it is true, says Marquis. After all, if we take some living cancer cells from a human being, they are human and they are alive, but very few people would object to destroying them. So while the moral-principle objection sounds plausible, it needs to be nuanced to avoid examples like cancer cells, or even just skin cells. [I think Marquis is simply wrong when he says that it's not a stretch of language to call living human cancer cells "a human life." I think it really is a stretch, since "a human life" intuitively sounds like a reference to a human that is alive, namely a living human being.]

What we really mean when we express the moral-principle objection is that it is wrong to kill “a human being” for the sake of medical research. Marquis refines this further still, to the claim that it is wrong to kill a being with a “future of value,” as he thinks this avoids cases where the previous version of the principle would not allow us to kill newborns born with no brain, or human beings in a persistent vegetative state. For myself I think this manoeuvre is unnecessary to achieve this end provided we make appropriate distinctions between killing a human being and allowing a human being to die. Dwelling here, however, would be a distraction from my disagreement with what follows in Marquis’s argument. For the purpose of this argument then, I will grant the claim that what makes killing in general wrong is that it takes away a future of value.

One way of arguing that early embryos are human beings is by using what Marquis calls the “trace back” strategy. I am a human being, and you can trace the history of this human being back through numerous phases: adult, adolescent, child, toddler, infant, fetus, embryo, and zygote. To destroy an embryo is to destroy a human being at one stage of its history, and to destroy the embryo that grew into the adult who is now typing these words would be to destroy a being with a future of value. At first sight, this claim has a lot going for it. Marquis proposes a few objections to the trace back strategy that he does not accept, and I will not discuss those here.

But Marquis does not ultimately accept that the trace back strategy is a good way of arguing against the morality of using hESCs in research, thereby destroying embryos.

The trace back strategy takes us right back to having once been a zygote. There are, in this view, one celled human beings (albeit for a very short period of time, before the first cell divides). This is the staging point for Marquis’s argument. For simplicity of reference, I will use the term “t1″ to refer to the period of time (however long it is) when the one-celled zygote exists, and “t2″ to refer to the time during which there are two cells, and before further division take places into four, eight, sixteen etc. (I would call those periods of time t4, t8, t16 etc.). Marquis’s claim is that there is a defeater for the belief that there is a human being at t1 that also exists at t2. The defeater arises in the form of the following puzzling observation:

Now consider tracing our individuality forward from this one-cell stage. This one-celled human being, this zygote, will split into two cells. Each of the two cells does not seem to be different in any important respect from the human being from which each originated. The situation here is like the fission of one amoeba into two amoebas, with (let us suppose) no space separating the amoebas. Therefore, if the zygote from which the two human cells originated is a human being, then each of these two cells must also be a human being. [Marquis, "The Moral-Principle Objection," 198-199.]

Two human beings? This sounds clearly wrong. What Marquis is doing, then, is showing that an argument that proves too much ends up proving nothing at all. Surely even a very morally conservative opponent of embryonic stem cell research would not accept this conclusion, yet, according to the above analysis, this is precisely where her reasoning leads. She should therefore, Marquis thinks, abandon her trace-back argument for the conclusion that we were once zygotes.

The argument is as follows:

1. At t1, a specific human being is identical with a specific one celled thing, a zygote.
2. At t2 there are two things that are identical with a human being at t1.
3. Therefore at t2 there are two human beings (that is, two things that are each identical with a specific human being at t1).

I want to suggest that this argument is not only unsound, but it does not even reach the point of being considered for soundness or unsoundness: It is logically invalid. That is, even if the premises are true, the conclusion does not logically follow. This is because I think the argument engages in equivocation over the word identical (and with the concept of identity). The word “identical” in the conclusion does not carry the same meaning as when it appears in premise 2).

Two concepts of identity

Imagine that you are standing with a detective in a dark room looking at a line up of possible offenders in a mugging to which you were a witness. You eye the four men standing on the other side of the glass, until your gaze fixes on the fourth man, Kevin Brown. He is the same height and build, he has the same facial features, same tattoos, you hear him cough, and he even has the same distinctive cough as the one you heard the offender with on the night of the mugging. That’s the guy, you think. But just then you notice that his hair is a bit shorter. he’s had it cut. He’s not exactly the same as he was that night. The detective notices you staring at him and asks you “does that guy look familiar?” “No,” you reply. “He’s a different man. That’s not the guy I saw on that night.” The man is released that night.

We all see what’s wrong with this sequence of events. The man behind the glass is not the same as he was before, but he’s the same man. The above scenario confuses two different concepts of identity. The first kind, the kind that the witness mistakenly adopts, is what I’ll call qualitative identity. The offender has a different quality when he stands in the police line up in that his hair is different. He therefore does not have identical features or qualities to those that he had on the night of the mugging. He is not qualitatively identical with his former self. The second type of identity is numerical identity. Thing 1 and thing 2 are numerically identical only if thing 1 is actually thing 2. Kevin Brown is the mugger because he is the same man who actually carried out the mugging, regardless of whether or not he still has all the same qualities that he had at the time.

To avoid one possible further confusion, let’s accept that when Kevin Brown carried out the muggings, he was both qualitatively and numerically identical with the mugger. However, the fact that these two types of identity are able to coincide does not mean they are the same. The property of having the features that Kevin Brown had on that fateful night is not the same as the property of being Kevin Brown. When we say that Kevin Brown is identical with the person who carried out the mugging, we are making a statement about who he is, and no matter what his features were at the time or are now, this identity statement would be true.

Let us turn back to the argument about the embryo at t1 and t2. What does premise 1) mean when it uses the term “identical”? Does it mean that at t1 a human being has the feature of having only one cell? Does it mean that at t1 a human being is a one celled thing? I think the proponent of the moral-principle objection could maintain either of these things, but I also think that in Marquis’s argument, premise 1) refers to numerical identity: At t2, a human being is a one celled thing.

But what then of premise 2? Whether or not the conclusion follows now depends crucially on the type of identity that is being referred to in the statement that “At t2 there are two things that are identical with a human being at t1.” Marquis’s argument has initial plausibility just because the two cells at t2 do in fact have the physical features of the one cell at t1. This, I expect we will immediately recognise, is qualitative identity. If my analysis of what drives Marquis’s argument is correct, then his argument really says this:

1. At t1, a specific human being is numerically identical with a specific one celled thing, a zygote.
2. At t2 there are two things that are qualitatively with a human being at t1.
3. Therefore at t2 there are two human beings (that is, two things that are each numerically identical with a specific human being at t1).

When spelled out this way, the invalidity of the argument is obvious. The property of being qualitatively identical with a human being is not the same as the quality of being numerically identical with a specific human being.

The trace back strategy is one that traces our numerical identity back through various stages in our existence to a point where we were a zygote. Marquis’s objection is only a good one if it actually shows that this is impossible to do. What he needs to do is draw on premises that the trace back strategist will accept, and show that they lead to a conclusion that the trace back strategist does not accept, thus showing that the premises should be abandoned. That is why I have construed premise 2) in terms of qualitative identity because doing so fulfils this duty of not inserting premises that ones opponent does not accept. Marquis could have made the argument valid by making premise 2) mean “At t2 there are two things that are numerically identical with a human being at t1.” But this is merely a case of the conclusion being used as a premise. Trace back strategists do not accept this, but instead trace our history through the stages of a one-celled human, a two-celled human that is numerically identical with the one-celled human, a four-celled human that is numerically identical with the two-celled human and so on.

A proponent of Marquis’s argument might begrudgingly accept the invalidity of the argument as I have presented it, but still find this a bit strange. How can there be two things that are qualitatively identical with a human being that are not themselves human beings? Two responses suggest themselves.

Firstly, although so far I have only commented on the equivocal nature of the argument, perhaps it is too generous to grant premise 2), that each of the cells at t2 is qualitatively identical with the cell at t1. Perhaps there are further qualities that I have not yet taken into account, such as the quality of being the first cell in a collection or the quality of not being joined to any other cell. Secondly (and building on this first response), I think that Marquis’s own “future of value” argument against abortion can be adapted for my purpose. Neither of the two cells at t2, considered on its own, has a future of being a human child or adult. But the thing made of two cells at t2 does have a future of being a human child or adult. But what, you might ask, if we take the thing that exists at t2 and split it in half without destroying the two cells? That is, what if we artificially induce something like twinning? Now each of the cells at t2 does have a human future. However, this does not show that if a zygote is a human then the two cells at t2 are two humans. All it shows is that those two cells are potentially two humans, and that potential can be realised if we intervene in a particular way.

This is not much of a concession. Let’s imagine that apples were produced, not individually on a tree, but by splitting just like cells. Let’s say I start an apple collection. Initially the collection only has one apple in it. Small beginnings! But before long, the apple splits and I now have two apples in my collection. To say that they are in my a collection is to say that they stand in a certain relationship with one another (and also with me, so long as I use the word “my”). Would you be persuaded by someone who insisted that I could not possibly have one collection, I absolutely must have two collections, because I now have two things that are each the same as my collection when it consisted of only one apple? I would not, but I would accept that I had what is potentially two collections, should I choose to separate them into different groups. [As a refresher on numerical vs. qualitative identity, the property of being made up of only one apple is a statement about qualitative identity, whereas the property of being Glenn's apple collection is a statement about numerical identity.]

What is true of two celled embryos and apple collections is also, on a physicalist account of human nature, true of adult human beings. We might think it very strange that an adult human being is potentially two human beings. But the fact that it is strange does not even get close to implying that it is not true. We already accept that this potential exists for many plants. Taking a cutting from one plant and planting it, allowing it to grow into a new plant on its own, is a common practice. The widely accepted possibility (setting aside the moral permissibility) of human cloning shows us that something similar is possible for human beings as well. What’s more, while it may currently (and maybe it will always be) physically impossible due to the risks and technological limitations we face, there is nothing theoretically impossible about splitting a human brain into two functioning halves capable of sustaining conscious existence in a body derived from the existing body of the original human owner of said brain. Therefore the fact that the two celled thing at t2 is potentially two human beings does not show that it is not, at t2, one human being. Remember, once the trace back strategist has given her account of why it is correct to say that a human being existed at conception, it is Marquis’s job to show that the account cannot be true.

I do not see that Marquis has fulfilled this duty as far as the argument from cell division is concerned.

One other argument that Marquis uses in this area of discussion, the argument from twinning, is addressed by Alexander Pruss. It is also discussed HERE by my friend Matthew Flannagan, who brought the argument from cell division to my attention.

Glenn Peoples

Similar Posts:

• No embryos were harmed in the development of this stem cell therapy
• Numerical identity does not require Qualitative identity
• Yay for stem cell research. But why bring Embryos into it?
• Embryonic Stem Cells no more?
• Nuts and Bolts 002: (numerical) Identity
• Episode 029: Is Abortion Immoral, and Should it be Illegal?