440 pages in Swedish
The problems adressed in this thesis are whether it ever - and, if so, under which conditions - is morally right to undergo or perform an abortion, and if there are cases in which it would be wrong not to undergo or perform an abortion.
Ordinary medical and legal definitions of the term "abortion" are found inadequate for a moral discussion. Abortion is often defined as an intended premature termination of pregnancy (sometimes "premature" is defined with reference to some specific point in time - e.g., 12 weeks after conception - which is set to ensure that only non viable foetuses can be aborted). What makes abortion controversial from a moral point of view is its status as an instance of killing, but these definitions either do not imply the death of the aborted foetus, or do not include all terminations of pregnancy which occur before the start of any natural process of birth and which lead to the death of the foetus. At the same time there are good reason for excluding pure accidents, e.g. unsuccesful caesarean sections, from the moral discussion. Two proposals, the one restricting the use of "abortion" to those terminations which are made with the intention of killing the foetus, the other restricting it to terminations made with the knowledge (or well-founded belief) of the foetus' death, are found either to beg controversial moral questions discussed in the debate on abortion, to include morally uninteresting terminations of pregnancy resulting in the death of the foetus, or to exclude such terminations which are interesting from a moral point of view. The definition of "abortion" adopted in the study instead includes those (and only those) intended terminations of pregnancy which occur before the start of any natural process of birth, which are not carried out with the intention of saving the life or health of the foetus, and which result in the death of the foetus.
The discussion proceeds in two stages. The first stage is a critical review of the international debate on the moral status of abortion mainly among philosophers and theologians. The review is carried out with reference to five different positions on the abortion-issue:
2) Moderate Positions: abortion is right in the same type of cases which are accepted by conservatives and in some types of cases which are not accepted by them (e.g., when the pregnancy is the result of rape, or before some point in the foetus' development), but there are still some types of cases in which abortion is wrong, even if it is competently performed and requested by the woman.
3) The Liberal Position: most voluntary and competently performed abortions which are or can be performed in practice are right. (There are several versions of the liberal position. But this formulation seems to me to be the most plausible and one which is also relevant for practical deliberation and action.)
4) The Radical Position: most voluntary and competently performed abortions which are or can be performed in practice are right, there are cases in which it would be wrong not to have an abortion, but it is not right to perform an abortion against the will of the pregnant woman.
5) The Extreme Position: most voluntary and competently performed abortions which are or can be performed in practice are right, there are cases in which it would be wrong not to have an abortion, and there are cases in which it is right to perform an abortion against the will of the pregnant woman.
The review starts with a close look on the argument for the conservative position put forward by the catholic church. The premise of this argument is that it is wrong to cause the death of the foetus if (i) the foetus is a human being, and (ii) the causing of the foetus' death is carried out with the intention of bringing about this result. Granting the truth of (i), (ii) is mainly thought to permit at least some abortions necessary to save the woman's life. It is argued that (ii) may indeed permit considerably more than that, but at the same time it seems extremely difficult to draw any specific conclusions from (ii). The moral relevance of (ii) is questioned. Reference to the intentions of some agent may be relevant for judgements concerning the moral character of this person, but it is not relevant for the moral status of his actions.
Clause (i) is related to a principle embraced by all proponents of the conservative position, namely the principle that (with rare exceptions) it is wrong to cause the death of human beings. This principle has been interpreted in four different ways. According to the theological interpretation what matters is if the foetus is in possession of a soul. Granting that there are souls of this kind it is argued that the foetus cannot be in possession of a soul as long as twinning may occur (i.e. before the conclusion of the implantation). The supposition of this argument is a view shared by many theologians that the soul is an individable substance, and that God has not put two souls in place in every case where twinning occurs. Granting the common religious view that the soul continues to exist after death, it is also argued that the possession of a soul does not in general make it wrong to kill its possessor. This instead seems to depend on what kind of existence the soul would enjoy after the death of its body. The argument that the wrongness of killing ensouled creatures can be derived from a supposed prohibition of God is rejected as either irrelevant or morally unacceptable.
According to the biological interpretation, what matters is that the foetus belongs to the species Homo Sapiens. Although the argument for this being the case usually put forward by conservatives is invalid, it is concluded that the foetus indeed belongs to this species from the moment of conception. Taking two arguments put forward by Michael Tooley as a starting point it is however argued that the basic moral principle presupposed in this interpretation of (i) is untenable. Firstly, there seems to be cases in which it may be permissible to kill a being which belongs to Homo Sapiens (e.g. if the being suffers an irreversable and total loss of consciousness). Secondly, there seem to be no reasons for and some reasons against incorporating the concept of a biological species into basic moral principles. I conclude that (i) is irrelevant as an argument for the wrongness of abortion under this interpretation.
The mentalistic interpretation instead focuses on the mental capacities of the foetus. Since it seems obvious that human foetuses cannot have any such capacities before they have started to develop a brain, it is concluded that this interpretation makes (i) false. It is however noted that the lack of mental capacity is compatible with the presence of a potentiality for mental life.
This potentiality is the focus of the interpretation from potentiality. The presence in a foetus, F, of a potentiality for mental life is found to be dependent on the truth of the following (counterfactual) conditional: If F is not (had not been) aborted, F will develop (would have developed) into a being with some mental capacity. Since many foetuses do not have this kind of future even if they are not killed it is concluded that reference to potentiality cannot support the conservative position. The possibility of invoking further conditions in addition to abortion in the analysis of potentiality for mental life is rejected on the ground that this would result in moral principles which not even conservatives would be willing to accept. It is also argued that the presence in a foetus of a potentiality for mental life does not in it self make it wrong to kill it. The main weakness of this thought is that it does not take into consideration the content of that mental life which the continued life of the foetus would result in. Especially, if the foetus is deformed in a way that would result in a life full of misery and agony and short of any compensating advantages, it is hard to see why it cannot be the case that it may be right to kill it.
One argument for a moderate position starts off from the same kind of premises as the conservative arguments. According to this argument abortion is permissible if the pregnancy is the result of rape, even though the foetus is a human being and it is in general wrong to kill such beings. The reason is that since, in the case of rape, the pregnancy is not the result of any voluntary action of the woman, she is not responsible for the presence of the foetus in her body. Three different interpretations of this thought are considered. One of these is found to include all rape-dependent pregnancies, but also a host of other pregnancies. The two others fail to include all possible pregnancies resulting from rape, but at the same time include other types of pregnancies.
The thought that whether or not abortion is permissible depends on whether the woman is responsible for her pregnancy (or the foetus) is developed further in an argument often refered to by feminists. According to this argument, the foetus may have a right to life, but this does not make it wrong to kill it. The reason is that the woman has a right to decide over her body which includes a right to abortion (if she so wishes). Only if the foetus as well has a right to the woman's body may abortion be wrong. The way in which the foetus may acquire such a right is, according to this argument, through the woman being responsible for her pregnancy and the foetus, e.g., it seems that proponents of this argument would claim that if a woman becomes pregnant without being forced to have intercourse, with some knowledge of the fact that intercourse may result in pregnancy and without using any reasonably reliable contraceptive, she is responsible for the foetus and may not abort it. I claim that this view reflects a much too simple view of rights and their relative strength. It seems that the woman's responsibility for her pregnancy has nothing to do with how much of a right the foetus has to the continued use of her body. Assuming that the foetus can have such a right, its strength should instead be seen as dependent on the interest of the foetus to continue to exist and the woman's interest in removing it from her body. This also means that the foetus may have a right to the woman's body even if she is not responsible for its presence there (e.g., because of lack of knowledge). If the foetus has such a right it has a right to life (in a stronger sense than assumed in the argument).
This conclusion shifts the focus to the question of whether or not the foetus has a right to life. According to one argument it has such a right when it has developed a capacity for consciousness. Therefore, it is said, abortion is only permissible before the foetus has acquired such a capacity. I agree that capacity for consciousness is necessary for the holding of a right to life, but deny that it is sufficient. While capacity for consciousness indeed seems necessary for the foetus to be able to have an interest in its own continued life, it is not sufficient for the foetus to have this ability. It is further argued that the foetus cannot have a capacity for consciousness, and thus no right to life, before the start of the development of its brain.
Another argument for the wrongness of aborting foetuses with a capacity for consciousness begins instead from rule-utilitarian premises. According to this argument, the system of moral rules which it would be best to accept in a society contains a rule generally prohibiting the killing of beings with a capacity for consciousness. The utility of such a rule is explained by it reflecting a fruitful borderline between killing (which is normally wrong) and non-reproduction (which is normally permissible). The need for such a line is in turn explained by the usual differences in side-effects between these two types of action. However, since abortion, from a utilitarian point of view, may be considered as one form of non-reproduction the need for such a line does not motivate the line drawn in the argument. This is instead done through an appeal to the need for a fairly simple and coherent system. The consciousness-borderline should be seen as just one application of the simple rule that it is in gereral wrong to inflict harm upon others. Since there is still need for excluding non-reproduction from the class of prohibited acts, this rule must however be interpreted in a very special way which implies that we cannot harm someone if a conscious being whose situation is worsened, compared to some alternative state the agent could have brought about, has never existed. Such a rule indeed does the job of motivating a borderline between the killing of beings with and without a capacity for consciousness. But at the same time its inclusion into the system of rules makes it probable that this system is after all not the best one to accept in a society. This is shown by the implications which, as pointed out by Derek Parfit, follow from such a rule in certain cases. Among other things, there is probably nothing wrong in totally destroying the environment for future generations, according to the rule in question. And even if people accepting the rule will not always follow it in such cases, these implications seem to me so grave as to show that the best system of moral rules cannot include this rule, it is sufficient to suppose that they sometimes succeed in following it.
Even if the prospects seem meagre for finding a specific point in pregnancy at which abortion becomes generally wrong, it has been claimed that the reasons for the wrongness of abortion gradually become stronger as the pregnancy continues. Such a gradualistic view of the moral status of abortion has been advanced on two different levels. On the one level it has been argued that, even if the foetus lacks a right to life, the risk for detructive effects on desirable human attitudes (such as loving and caring for our children) as a result of abortion increases if the abortion is performed in later stages of pregnancy. It is of course hard to find any evidence for this being the case, as a matter of fact there seem to be some evidence against it, but the risks mentioned cannot be disregarded as the argument seems to have at least some relevance for the abortion issue. The other level instead concerns the intrinsic moral status of the foetus. According to gradualism at this level, the foetus may have a right to life which in earlier stages of pregnancy is very weak as a reason for the wrongness of abortion, but this right gradually becomes stronger as the foetus develops. I claim that even if the property in virtue of which the foetus acquires a right to life (e.g. some mental capacity) develops and grows gradually as pregnancy continues, the strength of the right is not determined by the amount of this property present in the foetus. The strength or moral force of a right instead varies with the interest of the right-holder in keeping that to which she has a right, and such variations are in no clear way tied to the right-holder's degree of (mental and/or physical) development. The only way in which the foetus, if it apparently lacks the property needed for a right to life, still may have such a right is if it is a part of an individual which exists over time, some other part of which has the property in question, e.g., if a foetus, if it continued to exist, would belong to the same person (existing over time) as a latter (piece of this) person with a right to life and an interest in its own continued existence, it is not unreasonable to claim that this foetus has a right to life of some strength. This strength does however not seem to be directly dependent upon the foetus' degree of development. If the interests of the future individual the foetus would become if not aborted are also the interests of the foetus, and if it is this fact which gives it a right to life, the strength of this right depends only on the strength of these interests. Whether or not the foetus even can have rights in virtue of possible future interests is however not discussed at this stage, but pursued below.
The last argument for a moderate position begins from act-utilitarian premises. At the heart of this argument is the view that if we disregard effects on people other than the foetus and the being it would become if not aborted, it is wrong to abort a foetus which would result in a life worth living. The reason is that abortion in this case results in less intrinsic value (i.e. happiness) than would have been produced if the pregnancy had been carried to term. Most abortions thus carry a moral burden - in most cases compensating factors are needed if abortion shall be right. By the same token it is equally wrong not to conceive such a foetus (e.g., by the use of contraceptives). This claim is however softened in the argument by a reference to the disutility of creating new happy children when we instead could make existing unhappy children happy. Surprisingly enough this line of reasoning has however not been applied to abortion. This can be made through a comparision between the alternative of carry a pregnancy to term and care for the resulting child (thus making it happy), and that of undergoing an abortion and transfering the resources that would have made the aborted child happy to an already existing child, whose life, because of lack of such resources, is worth not living. By the principle that resources normally create more utility if directed to people who are worse off, it seems plausible that the second alternative (called abortion & transfer) would be the best one, and that it therefore can be right to perform abortion & transfer even if the foetus would result in a worthwhile life and none of the possibly compensating factors normally discussed are present. This argument, dubbed the argument from transfer, has one peculiarity which is discussed more in stage two of the study: even if it is right to perform abortion & transfer, since most women who perform abortion do not follow it up with donation, it may still be wrong to perform abortion. But, on the other hand, if abortion & transfer is performed this is right even though abortion is performed. It therefore seems that the argument from transfer can justify abortion, even though it cannot do so regardless of how the abortion is followed up.
The weakest formulation of the liberal position (L4) is that the fact that abortion kills the foetus does not supply any reason for the wrongness of abortion as long as we disregard the possibility of harm to other beings. The truth of L4 is assumed in all stronger versions of the liberal position (e.g. the one formulated above). The argument put forward in support of L4 (the so called personhood argument) is that the foetus lacks a right to life, and that people therefore cannot have any obligations not to kill it which are based on how this would affect the foetus or the being it can become. The most elaborated view of this right put forward in the abortion debate implies that in order for the foetus to have a right to life, it must be capable of having an interest in its own continued existence. The personhood argument denies that foetuses meet this condition. This denial is dependent on two different basic claims.
The interests relevant for the right to life are analyzed in terms of the possible satisfaction of desires. The first basic claim is that these desires are limited to those which can be understood as propositional attitudes. I deny this claim. Desires understood as behavioural dispositions, whose actualization is tied to some mental event which does not involve any representational or propositional aspect, may very well be considered relevant to the question if a certain being possesses a right to life. Foetuses who have attained a capacity for consciousness may have such desires not to be killed and thus seem capable of having an interest in their own continued existence.
The second basic claim is that a foetus cannot at a certain moment have an interest in its own continued existence which is dependent on this continued existence being necessary for the satisfaction of some desire which this foetus would have at some other time if it continued to exist. The reason for this is that a foetus existing at a certain point in time cannot have desires at other times. Of course, the foetus may develop into a being with important desires, but, the argument goes, since the foetus cannot be a part of a bearer of desires which exists over time, these desires would not belong the foetus. This claim is supported by a theory of personal identity over time, which, among other things, demands memory-links between two persons existing at different times in order for them to belong to the same person over time. I argue that if this demand is to be acceptable it must be explicated in a way which permits that there may exist memory-links between a foetus and e.g. the child it can give rise to. Firstly, even if no one seem to remember "their" time as foetuses, there may be overlapping chains of memories reaching back to some foetal experience. Secondly, it may be that many people have either such chains or direct memories of some foetal mental state of which they are wholly or partially unconscious. Even if the demand is softened in this way it is too strong however, at least in a theory of personal identity with important implications for who can have a right to life. There are of course other conditions that have to be met if a foetus shall be able to have desires in the future (or past), but these conditions seem to permit that foetuses in later stages of pregnancy may have future desires.
It is also argued that the claim that the foetus lacks a right to life cannot support L4. The fact that other people cannot have obligations not to kill the foetus which are solely based on how this would affect the foetus or the being it can become, is compatible with there being other ways in which such effects may in themselves constitute a reason for the wrongness of abortion. e.g., if the foetus can become a person with a fairly good life the killing of it means a loss of intrinsic value. The possible claim that only peoples rights and obligations can have moral relevance is rejected - there are also other things of moral importance.
The radical position has been defended from a slightly restricted act-utilitarian basis. The most important argument for the liberal component of the radical position is that there should be as few unwanted children as possible. I accept this claim with some qualifications, but at the same time find it irrelevant. Since there are good reasons to believe that many unwanted pregnancies, if not cut off, would result in wanted and loved children, the claim cannot justify an abortion because it is wanted by the woman. At the same time, also this utilitarian argument disregards the kind of reason for abortion found in the argument from transfer. Furthermore, several arguments for the wrongness of abortion based on the effects on society and people's attitudes in general are found to have weak plausibility and relevance. On the whole then, the liberal part of the radical position may after all be supported by act-utilitarianism. The same moral outlook also provides reasons for the wrongness of omitting abortion in several cases, e.g. when the foetus is very severely deformed in a way which makes a future good life for the born person impossible. In many of these cases what is wrong is, however, not strictly to omit abortion, but to omit a sequence of acts which involve a new child being concieved and born in the aborted foetus' place. Such a duty to exchange the foetus for a new child is very plausible if the foetus is less severely deformed, if the woman wants an abortion because she is very young or because of rape or incest. The consequences of abortion for the woman are found to have relevance in that they support that abortion should be performed as early in pregnancy as possible.
However, it seems very hard to maintain this support for the first two claims of the radical position without being forced to accept an extreme position. Of course, it is wrong to force a woman to have an abortion. But if an unwanted abortion would save the foetus from a future full of agony and without any advantages, it may be right to perform the abortion if this can be done without the woman or anyone else (besides the one performing it) noticing. It is argued that there may very well be some cases in practice where these conditions are met.
A promising possibility of excluding the extreme position is provided by the principle of respect for autonomy which is built into the argument put forward in support of the radical position as a restriction on the act-utilitarian principle used in this argument. According to this principle, as long as we disregard effects on other people and apart from extreme situations, it is wrong to treat someone against her will, even for her own sake. Reference to autonomy does not provide any further reason for the wrongness of voluntary and competently performed abortions, but it may be used as an argument against involuntary abortions. I argue, however, that since a plausible principle of respect for autonomy must permit infringements of people's autonomy when this is necessary to prevent serious harm to other people, it cannot be used to reject the kind of extreme position which seems to be supported by act-utilitarianism. The reason is that respecting the woman's autonomy in these cases would mean to condemn the foetus to nothing but terrible torture in what is most properly described as a grotesquely lengthy death-struggle.
Stage two of the study is devoted to developing a tenable act-utilitarian argument on the morality of abortion. To do this it is necessary to formulate and defend a more precise act-utilitarian principle. This is done through a discussion of three crucial questions in normative ethics, which during the review have shown themselves to have strong relevance for the implications of act-utilitarianism on the abortion issue. The discussion results in a broadly hedonistic act-utilitarian principle called BHU, which does not disregard or discount the value or disvalue that may be created through the adding of new people to a population, and which permits the possibility of moral dilemmas where, whatever the agent does, she acts wrongly.
Three objections to hedonism are considered. The first is based on preference-utilitarianism, but seems to have very weak relevance for the abortion issue. The second objection proceeds from a qualitative hedonism of the kind often ascribed to John Stuart Mill. The objection is rejected on the ground that a qualitative hedonism presupposes, first, a clear categorisation of different kinds of pleasures and pains and, second, a hierarchy of these expressing their relative quality. I argue that none of these presupositions can be defended in a non-arbitrary and acceptable way. The third objection is that intrinsic value must also be ascribed to the existence of innocent human beings. This argument is rejected on two grounds. Firstly, innocence, even if relevant for deontological rules for self-defence, seems irrelevant for the ascription of value to states of affairs. Secondly, it is claimed that the idea that human beings are intrinsically valuable is ruled out by the arguments put forward against the basic principle in the conservative argument.
Another objection to BHU is based on an argument from the so called repugnant conclusion. According to to this argument BHU implies that, other things being equal, a state of the world with very many people whose lives are barely worth living is better than a state with fewer people living better lives. Several proposed modifications of BHU to avoid this implication are rejected, mostly on the ground that they have been shown by Derek Parfit to have even more unacceptable implications. It is also argued that since the reactions of people to the repugnant conclusion are very unstable, they do not constitute a good argument against BHU.
The last objection to BHU takes the argument from transfer as its starting point. The conclusion of this argument (that abortion & transfer is right and that abortion is wrong) can be questioned on the ground that it involves comparisons of irrelevant alternatives. I claim, however, that in order to support such an argument one has to presuppose a view of utilitarian comparisions of alternatives which is untenable. Either it yields conclusions which, from an utilitarian point of view, are morally unacceptable, or it is unfruitful because of its inability to answer pressing moral questions. Another objection to the argument from transfer (and thus BHU) is that an acceptable moral theory cannot permit moral dilemmas where the agent cannot avoid acting wrongly. I claim that this, on the contrary, is acceptable, at least as long as the dilemma is not dependent on the fact that the moral theory implies norms with which it is impossible to act in accordance.
With BHU or any other utilitarian moral theory as basis, it is not possible to find good reasons for any specific position regarding the moral status of abortion. This is due to complete lack of knowledge concerning several important factors, severe uncertainties concerning the knowledge of many other relevant factors, and even more severe uncertainties as to how different relevant factors should be weighted in relation to each other. However, utilitarianism is not unique in its inability to provide good reasons for solutions to practical moral problems. Most other moral theories pay some attention to the consequences of acts and their alternatives, which means that they share some of the severe application-problems of utilitarianism. Those theories which focus on the intentions of the agent also have to face the underestimated problem of determining which these intentions are. So, on the whole, I conclude that the problem discussed in this study is insolvable in practice, and probably will remain so.
It is however possible to argue for some tentative conclusions on the basis of BHU - conclusions which may be seen as a first step towards a well-founded solution to the problem of the moral status of abortion which in practice we will never reach. These conclusions are based on the identification of factors of which we have some (let alone uncertain) knowledge and which are directly or indirectly relevant for how BHU will judge an abortion. Some of the factors are whether the foetus has some severe defect which in practice guarantees that the child it can become would lead a life worth not living, whether the abortion is safely performed without high risks for the woman or the doctor being punished or for the woman being ruined, whether the pregnancy is unwanted in combination with some factor which makes it reasonable to believe that also the child would become unwanted, whether the abortion is performed early or late in pregnancy.
The conclusions reached are these: 1) It is very difficult to justify abortion in countries where it is illegal and/or most women lack the option of having a competently performed abortion at a reasonable cost. 2) It is often possible to justify an alternative which involves the performance of abortion in countries where abortion is legal and pregnant women have the option of having a competently performed abortion at a reasonable cost. 3) In most cases of those mentioned in 2), where the alternative involving abortion is abortion & transfer, the woman is trapped in a moral dilemma, where she cannot avoid acting wrongly even if she does what she ought to do. 4) In some of the cases of those mentioned in 2) it would be wrong not to perform the alternative involving abortion. 5) Especially, it is in general wrong to omit (safe and legal) abortion when the foetus is so deformed that it would develop into an individual whose life would be worth not living. In most cases it is also wrong to abstain from abortion if this means that the abortion is performed later. 6) In some very rare cases of those mentioned in 4) it may be permissible to perform abortion against the will of the pregnant woman. The supposition is that she or anobody else (apart from the agent) never discovers this. So, in practice we have no means of identifying these cases.
Back to the bibliography, please!