Skip to content

Jeremy Hunt, Abortion, and the Sorites Paradox

Controversy erupted last week when the UK Health Secretary, Jeremy Hunt, said that the law on abortion should be changed. But before looking at what he said, can I ask what you think? Here are some options:


I’m not sure that the results will be visible immediately on the blog, but I promise to post them in a couple of days if they’re not.
However, the main reason I asked this question in this way was to illustrate the apparent arbitrariness of the various options that are often debated. Okay, so I chose the questions, but most of these reflect dates that are those that people argue over. 14 days is when the ‘primitive streak’, the line of cells that will become the spinal cord, appears. 12 and 24 weeks equate (approximately) to the first and second trimesters. 12 weeks is the law in much of Europe. 24 weeks is also the approximate date when fetuses become viable. 28 weeks was previously the law in the UK. While even asking the question ‘throughout pregnancy’ is bound to raise a few hackles, previous surveys have suggested that some people give this as an answer (interestingly, it turns out that there is currently no relevant date in Canada, though this is more to do with a constitutional stalemate than any outlying Canadian morality). Jeremy Hunt’s own view was that the current UK law (of 24 weeks) should be reduced to 12 weeks.

Of course, these neat, round figures cannot have particular moral significance in themselves. It therefore occurs to me that there is a striking parallel between abortion dates and what is known as Sorites Paradox. For those unfamiliar with the paradox, let me briefly explain: the paradox was originally a riddle attributed to the Megarian logician Eubulides of Miletus. The riddle, or paradox, is to ask how many grains of wheat make a heap (soros is the Greek for ‘heap’). Clearly 1, or 2 grains, do not make a heap. By contrast, a thousand (or larger number) grains of wheat do make a heap. The difficulty arises in trying to define the point at which the wheat becomes a heap.

There seems to be an obvious parallel with abortion in that settling on, say, a 12 or 24 week period seems a very arbitrary way of reconciling the divergence of views. (It would be interesting to see what numbers people would settle on if they were asked to express their views in, say, days. Would they settle on 75, 100, 150, or 200 days (10.7, 14.3, 21.4, and 28.6 weeks respectively)?)

My own feeling about the Sorites Paradox is that relies on an erroneous belief that our concept of a heap can be reduced to, or equated with, a simple number. The belief is erroneous because empirical evidence suggests that many of our decisions depend on a wide, and complicated, range of factors. Many of these we are probably consciously unaware of, but could be shown empirically to affect our conclusion nonetheless. In the case of the paradox, some possible candidates for affecting our conclusion might be what the heap was of (wheat, sand, boxes), or whether there was an incentive for us to conclude one way or the other.

Abortion is a fundamental moral issue, not merely a riddle. Nonetheless, just as it would not be a very satisfactory way to resolve the Sorites Paradox by arbitrarily deeming a heap to be more than 100, or more than 1,000 grains of wheat, I find it unsatisfactory that the resolution of such a fundamentally important moral issue such as abortion is fought on such arbitrary terrain of regular sounding weeks.

Granted, there are provisions, such as in the UK law, that provide for exceptions to the primary 24 week period. However, these address fairly extreme circumstances, such as risks of death, grave permanent injury to the physical or mental health of the mother, or substantial risks of serious physical or mental handicap to the child.

My suggestion would be that we find a more sophisticated means of taking into account the various factors that affect our views as to when abortion is permissible. Ought not factors such as rape, physical or mental consequences for the mother, inability to access medical advice, inability to provide for the child, or serious disability to be taken into account as much as how much time has passed? Though there is a paradox in that most people believe that the more time has passed by, the more unacceptable abortion becomes, other factors are more tractable.

Share on

14 Comment on this post

  1. I believe the Sorites Paradox analogy is more appropriate when applied to what is assumed in the post: when something becomes a person; conception, after 12 weeks, 24 weeks?

    1. Dear Emilian, thanks for the response! Yes, I agree with you, as an analogy, the Sorites Paradox seems most apt if the comparison is with the question of when an embryo becomes a person. This is, as you refer to in your post, a commonly adopted way of asking the question. If I assumed that this was the way that the question ought to be posed in my post, this was not intended. I think that such an assumption suffers from a similar problem to that which I wanted to highlight in my post: the assumption that there is a point somewhere along the timeframe when acceptable becomes unacceptable and that this point can be discerned. However, whether this question is posed in what seems like a more concrete fashion (when an embryo becomes a person) or a seemingly more abstract fashion (when abortion is permissible / impermissible), I still think that the Sorites Paradox is an appropriate analogy. It’s just that the latter is slightly less obvious.

      1. Hi Paul,

        It’s nice to interact again. I wasn’t claiming that the analogy did not hold, just that the embryo/person case is more striking, and you’re right the permissibility/impermissibility less obvious. I was wondering if we can find some analog which is also normative. Consider cases when a lie is thought permissible via the less bad consequences. The time frames are prima facie as you say arbitrary. But it might also be the case that your survey tests some intuitions that we have on permissibility. Nevertheless, you’re major point is about the plurality of factors which I admit that it puts pressure on the standard way of dealing with abortion permissibility.

        1. Hi Emilian, lovely to hear from you again. Abortion is an area where minds much greater than mine have spent a lot of time thinking up analogues. Think of Thomson’s violinist argument! I would be interested to know whether people facing difficult questions such as this find these analogies helpful. Perhaps Planned Parenthood or Marie Stopes have counselors who advise people facing difficult questions to think of their embryo like a famous violinist…

          More seriously, some of these more familiar analogies clearly are very persuasive. Yet I doubt that any of them are sufficient to provide a definitive answer to such complicated questions. I personally feel that analogies are better conceived of as arguments that people use to convey their position to others.

  2. Based on the survey results so far, it seems I’m the only one who came to the same conclusion as you: I picked “Other” and recommended circumstantial based decision).

    I love the reference to the problem of the heap of sand. I was unaware it had a specific name, but always attributed it to the Stoics, who thought of it as a “problem of vagueness” – some ideas, concepts, etc. are inherently vague and there is nothing we can do about it! (The ancients came up with some pretty unsatisfying resolutions, such as refusing to answer what you BEFORE you are unsure so as not to be lead into falsehood.. lame!)

    For the analogy to follow, however, we have to discuss the relevant subject (which you unfortunately didn’t get around to in your post) and [a] figure out what the relevant subject(s) is, then [b] determine if it is necessarily vague, as the heap of sand was thought to be.

    The answer to [a] is not straight forward at all. For some, the only relevant factor will be “when does the soul enter the body”, or in secular terms, “when does the fetus become a human being”. For others, it may be a matter of determining that point in which has consciousness or susceptibility to pain. In such cases, the problem of vagueness could easily be argued to apply (sorry Abrahamic faiths, but “moment of conception” based on scripture doesn’t count as an argument).

    For others still, it may be further removed from such specifics: (Ie. at the point where killing the fetus become immoral – where “immoral” could be based on communitarian values). In these broader terms, vagueness doesn’t have to be an issue at all. For the utilitarian, vagueness will have to be accounted for (the uncertainty of whether the fetus can feel pain) and decision can be come to based on the situation. For the libertarian, well… I find them philosophically and morally inept, but i’m sure someone could come to their defense…

    1. Dear Dyami, thanks for the comments! I’m glad to see that the results are visible once you answer the question. To tell you the truth, I’m a bit mystified as to how to interpret the results. I had not intended that any particular result be equated with supporting the argument that I made in the post, but I suppose you’re right, that ‘other’ is probably the closest option. If so, with only 5% of the votes cast so far, I’m a bit of a failure. That said, the most surprising statistic is the number of people who seem to have voted for ‘throughout pregnancy’. I put this in as a bit of an afterthought, but in so far the vast majority of votes (almost 40%) have chosen this option. This could mean (a) people are interpreting this as the answer that most closely reflects my argument (unlikely I think as I suspect most people vote before they read – who could resist clicking at the outset), (b) this is what many people really think (I find this improbable), or (c) people who choose this option are not taking it seriously. Unfortunately the system does not allow terribly complicated analysis to be carried out on the responses. It would be interesting to hear what people who have selected this option give for their choice.

    2. In relation to the vagueness point, I’m not convinced that this is the way to answer the question. I think there are two obstacles to us answering the question in this way. The first is the number of inputs to the reasoning that we are unaware of, the second is the way our brain works.

      Extremely briefly, as to the first, I believe there are a huge number of pieces of information that will affect an individual’s judgement as to whether abortion in specific or general circumstances. We do not yet know what all these are, but they could be identified by careful empirical research. One example that appears quite likely is images of one’s own foetus. It seems to be the case that once somebody has seen their own foetus on a scan, they are much less likely to choose an abortion. This, I suspect, is the (to me) creepy reason behind the insistence in many US states that a woman has to seen an image of her foetus before deciding to undergo an abortion. There are undoubtably many others.

      The second aspect is how these many different pieces of information are taken into account by the brain. The brain is made up of billions of very simple neurons and many more connections between them. (Connectionist theories try to explain how these simple cells can handle extremely complex tasks.) Somehow, all these different inputs are taken into account by the brain to come up with a very simple conclusion: either yes abortion is permissible, or no it is not. The mistake (I think) is to assume that this complex process is completely accessible to consciousness.

      I believe that if you could somehow control all these variables for an individual, then you could probably identify quite closely the contours of the circumstances in which that individual found abortion to be acceptable / unacceptable. In that way, the concept would not be inherently vague, just practically vague. ‘Trolley experiments’ where the different variables are carefully controlled are the type of method that might be promising.

      In relation to the identification of the correct subject point, I do not think that this is likely to be feasible for the same reasons as above. Ie, the question that people pose, such as ‘when does an embryo become a person’ are too high level to reflect the two factors identified above. Ie, they are not accurate reflections of the underlying thought processes, but (I believe) an imperfect way of conveying something relevant to other discussants.

      1. “In relation to the identification of the correct subject point, I do not think that this is likely to be feasible for the same reasons as above. Ie, the question that people pose, such as ‘when does an embryo become a person’ are too high level to reflect the two factors identified above. Ie, they are not accurate reflections of the underlying thought processes, but (I believe) an imperfect way of conveying something relevant to other discussants.”

        — To the contrary, you appear to endorse what you consider the relevant subject(s), to wit, the “underlying thought processes”, what “[effects] individuals’ judgement” and identifying “the circumstances in which that individual found abortion to be acceptable / unacceptable” – in short, descriptive brain-states.

        If that is the relevant subject, you appear to argue that the vagueness problem dissolves. (Problem of vagueness definitely exists in blogosphere, so let me know if I’m butchering your ideas lol)

        -In direct response to this, I’d say I am less optimistic re the explanatory power of neuroscience… but perhaps this isn’t the time to RANT about that. So let us assume H1 = hypothetical situation where there is no descriptive vagueness)

        The next obvious question is: on what moral standard do we evaluate H1? This is the truly difficult question, and it is here that the vagueness problem may creep back in. No doubt, some neuroscientist/philosophers/psychologists are working hard at getting beyond these issues. But from what I’ve read, none are fully satisfying (though usually informative and always interesting – case in point, S. Harris, O. Flanagan, P. Churchill). While the ontology does demand more proof and progress (which research will surely provide), the philosophical arguments and prescriptive elements are equally if not more in want than the descriptive deficiencies..

        1. Dear Dyami, I think a little butchering of my ideas is very necessary. The only question is whether we end up with steak or sausages…

          I should clarify: I don’t think the vagueness problem can be feasibly overcome by neuroscience. There will always be some vagueness. My thought was that neuroscience (or just philosophical dialogue) could help to minimise this vagueness. Ie, if the debate is until what number of months do we consider abortion permissible, or alternatively, at what point does an embryo become a person, then there seems to be a lot of vagueness. If we try and break it down a bit more into a more complex question that allows an answer that depends on a wider range of variables, then perhaps we start to reduce some of this vagueness.

          I should also clarify (that I agree with you): neuroscience that makes the factors that we rely upon more explicit only helps to clarify the question, it doesn’t do much to answer the question directly. I’m not sure that there is an explicit moral standard to evaluate moral questions. Take, say, consequentialism and detontological theories. As standards to assess moral behaviour they seem pretty inadequate. Even the most logical or coherent theory invariably conflicts with our intuitions. Kantians who insist that lying is wrong even to prevent a murder have a very clear principle, but it seems pretty obvious that clarity is bought at a cost. By contrast, arguments based on consequences or arguments based on principles can be very powerful. I suspect that resolving conflicts over moral questions is more about concentrating on getting the process right, rather than finding an elusive external standard by which to assess moral questions. Hence my suggestion that making the debate over the permissibility of abortion more sensitive to wider factors than a simple scale may be a way to reconcile different positions in a more satisfactory way.

          1. Well said. Now we have to remove all those practical problems and we can continue getting closer to a consensus.

            I proffer this timeline, based on a my interpretation of this weeks horoscope, however, I remained behind the veil of ignorance so what follows is completely unbiased and perfect:

            I think we can now come to agree that the correct answer is 72 1/2 days for affluent women/societies, 211 days for impoverished women/societies, 13 days post-birth for extraordinary survival circumstances in remote village societies, 0 days in any of Abraham’s homes, 0 days in the Kingdom of Ends, 0 days for future Ubermensch women/society of ideal conditions and self-realization, and the maximum number of days for the maximum number of babies in the utilitarian society where there are no individuals. Of course we will have to add 1 day to each category on leap years to keep our stellar phases in sync. Australia, Newfoundland and Iceland are all exempt from this calculus.

            Cheers,
            D

  3. I wonder if the results are skewed because, based on the available options, the survey could be seen as a loaded question (that is, it assumes that there IS a maximum date). You could argue that by putting ‘other’, you technically escape this…

    I bet if you made one of the options read: “Depends on the situation” or “there is no maximum date for an abortion to be permissible” then the results would read different… BUT.. now that I think of it, it is likely that you could ask a misleading question merely to reflect the current discourse in the public sphere, only to later reveal our confused understanding of the problem?

    1. Dear Dyami, I totally agree with you that the way the question is posed will affect the answers. It has been shown repeatedly that the order (high to low v low to high) will affect the result, as will random numbers presented immediately before hand (in one experiment, participants were shown a roulette wheel land on either 10 or 65, then asked to estimate what percentage of African countries were members of the UN. Those shown 10 estimated on average 25%, but those shown 65 estimated on average 45%).

      I therefore suspect that the way the above question was asked will have affected the answers. They were made to be presented in a random order which is better than a given order, but still problematic. If people had been asked simply to enter a number, in days or weeks, or months, or use a sliding scale, I suspect the results will have been different again! I hope this goes some way to illustrating some of the points I was hoping to make in my original post!

      1. I do wonder whether Jeremy Hunt actually thinks the limit should be set at 12 weeks, or whether he was simply trying to shift the ‘Overton window’ of public opinion towards accepting Maria Miller’s 20-week limit by making it appear more acceptable in comparison.

        In reference to the ordering of options, I wonder whether it would be more effective if the 12 or 20 week-limit were suggested first. Intuitively I think if Hunt had made the 12-week claim first, the public would have been more accepting of Miller’s later 20-week claim than compared to the other way round (i.e what actually happened).

  4. I want to relate to the ‘Sorites Paradox’.
    The human mind has an intuitive attitude for logic symmetry, i.e. if the heap has 1,000 grains then 1,000 grains is a heap.
    However, this logic fails to work in many cases.
    When seeing a heap (for example) a multitude of factors are taken into account. No one needs to count the grains to decide if it’s a heap or not.
    I believe that this wrong attitude toward symmetry leads to many fallacies.

    As for the discussion at hand, I think that any rule / law / decision should take into account the multitude of reason and thus cannot be reduced to single number.

Comments are closed.