A Journal of Philosophy, Applied to the Real World

Seeing, Feeling, Doing: Mandatory Ultrasound Laws, Empathy and Abortion

Monash University


In recent years, a number of US states have adopted laws that require pregnant women to have an ultrasound examination, and be shown images of their foetus, prior to undergoing a pregnancy termination. In this paper, I examine one of the basic presumptions of these laws: that seeing one’s foetus changes the ways in which one might act in regard to it, particularly in terms of the decision to terminate the pregnancy or not. I argue that mandatory ultrasound laws compel women into a position of moral spectatorship and require them to recognise the foetus as a being for whom they are responsible, particularly through empathic responses to ultrasound images. The approach I propose extends the project of a bioethics of the image and highlights the need for a critical analysis of the political mobilization of empathy in discussions of abortion.


Obstetric ultrasound is today one of the most widely used of prenatal care technologies in the world, such that almost all pregnant women who interact with medical care in pregnancy will encounter it in some way (especially in resource-rich contexts but increasingly also in resource-poor settings). This is often in the form of prenatal testing for conditions such as Down Syndrome, spina bifida and other morphological anomalies. Even in pregnancies in which prenatal testing is not undertaken, women may have ultrasound images taken in the course of the pregnancy to examine factors such as foetal location, placental function, or increasingly, to identify risks factors for preeclampsia, amongst other conditions.

Within this context of international routinization, ultrasound technology sits at the crux of a “prenatal paradox” (Taylor 1997). Obstetric ultrasound contributes in fundamental ways to decisions about abortion on the basis of foetal characteristics such as malformations or genitalia. It makes such decisions possible by revealing foetal characteristics to prospective parents. At the same time, ultrasound technology has long been implicated in anti-abortion politics and the legal restriction of access to abortion services. This takes various forms, including the use of ultrasound images and video on pro-life websites, or as evidence in legislative proceedings that seek to restrict access to abortion services, or the legal requirement that women access ultrasound services prior to undertaking an abortion.

In recent years, a number of states in USA have adopted laws that require a woman seeking an abortion to undergo ultrasounds prior to accessing abortion services. Such ultrasounds are not required as part of the abortion procedure as such, but attempt to ensure that women see, or have the opportunity to see, their foetus prior to undertaking, or deciding to undertake, an abortion. These so-called mandatory ultrasound laws are often cast as a matter of information provision and consent. However, this is misleading. Instead, mandatory ultrasound laws seek to precipitate and mobilise a particular relationship between the pregnant woman and the foetus she carries, with the aim of influencing her decision-making and action in regards to that foetus. As Carol Sanger (2017, 109) puts it, mandatory ultrasound laws aim to “produce a confrontation” between the pregnant woman and foetus, compelling a woman to recognize the life she would end in abortion and aiming to deter her from that. In essence, the legal requirement to look is based on the idea that seeing one’s foetus changes how one might feel about it, and further, how one ought to act in regards to it. In short, the idea is that seeing the foetus has ethical effects.

While these laws have been widely criticized by legal scholars, the ethical implications of this harnessing of foetal images to legal access to abortion has received little attention in bioethics. Moreover, of the literature on mandatory ultrasound laws, very little considers in detail the underlying premise of these laws, that is, that seeing one’s foetus changes how one might act in relation to it—an idea, I argue, that is based more on the feelings precipitated by the image of one’s foetus rather than any informational value it might have (or be purported to have).

In order to elaborate this idea, I build on the case made by Paul Lauritzen that a ‘visual bioethics’ is required to understand the ways that foetal images such as those produced through obstetric ultrasound have persuasive force, and are mobilised within antiabortion campaigns as “visual arguments”. One difficulty with Lauretzin’s approach is that it says very little about why images of the foetus might have persuasive force; I remedy this by turning to discussions of moral spectatorship and empathy. In a brief discussion in The Ethics of Care and Empathy (2007), Michael Slote suggests that ultrasound images may affect the empathic relation that can be maintained with the foetus, and consequently, their moral status. I extend Slote’s comments about empathy and abortion to consider the feelings involved in seeing the foetus and its connection to moral decision-making. Slote’s comments, then, provide a starting point for a more extensive reflection on the relationship between visibility, emotion and ethics as enacted in obstetric ultrasound.

My discussion proceeds in three parts, the first two of which are broadly motivated by a concern with the connection between seeing and feeling. In the first section, I provide a brief outline of various ways that obstetric ultrasound has historically and contemporarily been linked to abortion politics, because of its capacity of make the foetus visible. One aspect of this is the notion of maternal bonding, to which ultrasound has been seen as contributing. In section two, I focus on the concept of empathy as a means of elaborating the ethical significance of the seeing-feeling nexus, specifically by taking up Slote’s comments on empathy and abortion to discuss three points. First, I discuss the question of how we ought to characterise the emotional relationship with the foetus that is at least in part made possible by imaging technologies such as obstetric ultrasound. Second, I discuss the way in which ultrasound necessarily mediates the relationship that the viewer maintains with a foetus in utero and consider whether this matters for an ethics of empathy, such as that proposed by Slote. Finally, I consider the question of how ultrasound images are interpreted, especially given the background context of medical and social norms. I suggest here that the social and medical framing of an ultrasound image makes a difference to the empathic relationship that may come into play. In the third part of the paper, I consider the connection between feeling and doing, and the way that mandatory ultrasound laws mobilise this for political purposes. I argue in this section that such laws attempt to mandate a specific emotional relationship between the pregnant woman and the foetus she carries. Moreover, understood in its political context, it becomes clear that doing so comes at the expense of empathy with the pregnant woman herself, as she engages in the decision-making process around abortion.

Before moving to these discussions, let me be clear about the aim of this paper, and particularly what it is not. Most obviously, I do not aim to provide a detailed discussion of the varieties of mandatory ultrasound statutes currently in existence, nor of legal challenges to them (eg. Denbow 2015; Robertson 2011; Sanger 2008; Sanger 2017; Smith 2013). Nor do I aim to engage in long-standing debates in bioethics around the ethics of abortion, or at least not directly. My interest is related to these, but also somewhat orthogonal to discussions of the moral status of the foetus, at least insofar as these revolve around claims about intrinsic characteristics of the foetus (see Mills 2014). Instead, what I am interested in is the entwinement of foetal imaging technology and law in framing the ethics of abortion, and specifically the supposition that seeing the foetus ought to or does change one’s moral responses to it. In this, the paper aims to extend what might be called the bioethics of images,1 by offering a more fulsome account of the connections between seeing, feeling and doing than has otherwise been proffered in bioethics.

1. Obstetric ultrasound, abortion politics and maternal bonding

According to historians Malcom Nicholson and John E. E. Fleming (2013), obstetric ultrasound technology has been a major driver of the medicalization of pregnancy and childcare since its invention in the 1950s; today, obstetric ultrasound is routinized as a form of prenatal testing in many countries. In this routinization, it has also been at the heart of a “prenatal paradox”, in which knowledge provided by ultrasound may lead to abortion (such as when foetal anomalies are revealed by it), while at the same time ultrasound images have been extensively used in anti-abortion politics, in attempts to foster ‘pro-life’ sentiment and restrict access to abortion services (Taylor 1997). As such, obstetric ultrasound plays a complex and significant role in the technological mediation of the ethics and politics of abortion.

This linking of obstetric ultrasound to anti-abortion sentiments and politics is in itself not new. For instance, the obstetrician credited with driving its use in obstetrics, Ian Donald, was strongly opposed to abortion and used ultrasound images to attempt to convince unmarried pregnant women in Scotland during the 1960s not to terminate their pregnancies. He also showed a real-time film of a foetus at an anti-abortion rally in Milan in 1979, in the midst of Italian abortion reform debates, and was invited to an audience with Pope John Paul II—something that Donald himself saw as a crowning glory of his career (Nicholson 2000). In more recent years, ultrasound has been increasingly harnessed to the legal regulation of pregnant women and their access to abortion services, and features heavily in legislative and public debates.

In the United Kingdom, for instance, the capacity of ultrasound to make the foetus visible featured in debates over a proposal to reduce the gestational limit for abortions for non-medical reasons to 18-weeks. In a series of editorial and opinion pieces, leading obstetrician Stuart Campbell, who was a driving force in the routinization of obstetric ultrasound for prenatal testing and the later use of 3 and 4-dimensional ultrasound, made much of the emotional impact of seeing the foetus in utero. Campbell writes,

There is something deeply moving about the image of a baby cocooned inside the womb…Advanced scanning means we have a window on the secret life of foetuses. At 11 weeks, we can see them yawn, and even take steps. At 22 weeks, they begin to open their eyes. Between 20 and 24 weeks we watch as they seem to cry, smile and frown…When I see a foetus that can smile at me, I know absolutely that we should not tear it from the womb. (Campbell 2008).

Campbell’s slippage from the claim that a foetus may “seems to cry, smile or frown” to the idea that it can “smile at me” points to the difficult matter of interpretation and whether, or to what extent, actions on the part of a foetus can be seen as expressions of emotions or feelings (Mills 2014). Further, though, Campbell links the construal of a direct relationship in which the foetus smiles at the viewer to a moral conclusion, one that makes abortion impermissible at this stage of gestation.

A similar harnessing of ultrasound images to moral conclusions has been advanced in the USA, where anti-abortion groups have pushed various legal measures to restrict access to abortion services in recent years. One of these campaigns has sought to introduce laws requiring women seeking an abortion to undertake an obstetric ultrasound examination, not as an aspect of the procedure, but prior to it. Currently, some twenty-six states regulate the provision of ultrasound to pregnant women seeking abortions, with various statutes in place (Guttmacher Institute 2018). The most onerous of these laws, in Texas, Louisiana, and Wisconsin, require that the abortion provider perform an ultrasound for each abortion sought, and both display the image for a woman to see, and describe it to her. Other states require that providers give the woman the opportunity to see the images, and, in some, hear a detailed description of it, before performing an abortion. Others require providers who already perform an ultrasound as part of the preparation for an abortion to provide women with an opportunity to view the image. The weakest forms of the law require providers to offer to perform an ultrasound.

The key justification offered for such legislation has been that ultrasound examination ensures that women are properly informed before they make an irreversible decision about their health and their pregnancy. A supposed effect of this is to protect their reproductive autonomy and strengthen their capacity to resist coercion. This reasoning is typified in the Texas Woman’s Right to Know Act, which requires that women be provided with an information booklet and undergo an ultrasound examination, including a verbal explanation of the foetus’s development and morphology, at least 24 hours prior to undergoing an abortion. This bill initially appears to be motivated by the principles of informed choice and reproductive autonomy. For example, the recently revised information pamphlet that must be provided to women seeking an abortion opens with the statement “You need good information in order to make important decisions about your pregnancy and your life. You have the right to make these decisions freely. No one else should make them for you” (Texas DHS 2016, 1). The pamphlet then goes on to provide information about the risks of abortion, detailed information about foetal development including colour pictures, and information about financial and other services to assist in raising a child.2

This justification in terms of informed choice and reproductive autonomy has been widely criticized, including by academic commentators on the laws. For instance, numerous commentators have pointed out that the ultrasounds do not provide information that is medically necessary to consent to an abortion procedure. Others challenge the link drawn between the informational aspects of ultrasound examinations and reproductive autonomy. For instance, James Rocha (2012) argues that the laws do not enhance autonomy but limit it by overriding a woman’s prerogative to determine “how she wants emotion to be inserted into the process” of deliberation (2012, 49). Going further, Jennifer Denbow (2015) argues that such uses of ideas of autonomy are actually reshaping what autonomy means culturally and politically; she writes “autonomy is being conceived in terms of risk minimization and a medicalized, though legally controlled, notion of informed consent”. Further, she points out that this conception of autonomy aligns with the notion that women are being coerced by medical practitioners and sexual partners to undertake abortions. The idea, then, is that providing additional information about the foetus will enable women to resist that coercion, and make their own autonomous decision in light of increased knowledge about the foetus.

Useful and important as such critiques are, they still tend to give too much credence to the claim that ultrasound examination prior to gaining access to abortion services is in fact about information provision. In this, an opportunity to critique the epistemology of ultrasound and foetal imaging is missed. For instance, as Julie Palmer (2009) has pointed out in an extended reflection on Campbell’s commentaries on ultrasound and abortion that I mentioned above, the public mobilization of ultrasound images often entails a conflation of seeing with knowing. This, she claims, is evident in the idea that ultrasound allows one to ‘face the facts’ of abortion, for instance, which belies the necessity of interpretation in order to make sense of what one sees. In this, ultrasound is cast as a “moral speculum” (Mills 2014, 93), insofar as it seems to allow the viewer to peer inside and see/know what kind of moral being one is dealing with. Interestingly, this conflation also underpins Paul Lauritzen’s (2008) claim that the interplay of words and pictures in the use of ultrasound images and other visual media in abortion debates constitute “visually mediated arguments”. He builds on this to claim that the characteristics of argumentation, such as consistency and factual accuracy, can be used to assess the contribution that images make to debates on abortion. Contrary to the supposition of this approach, it is often extremely difficult to say with any clarity just what argument an image might be advancing in and of itself. In fact, this approach returns attention to the narrative that frames an image and its interpretation, such that the image is merely supplemental to that narrative.

What is obscured here is the emotional or affective impact an ultrasound image may have. To be sure, the emotional aspect of images is often mentioned, but is rarely analysed in any depth. For instance, Lauritzen suggests that uses of ultrasound images may lead to “emotional manipulation”, while Rocha acknowledges that the requirement to view the ultrasound image injects emotion into the decision-making process (Lauritzen 2008; Rocha 2012). However, there is no articulation of what emotions might be involved, or of the moral significance of this dimension of seeing an image. In order to bring this into focus, it is first worth stating that the informational justification for mandatory ultrasound is in fact fundamentally misleading, since the provision of information could be achieved in ways other than requiring women to undergo ultrasound examinations. For this, the provision of a pamphlet (such as that also mandated in the Texas Woman’s Right to Know Act) providing medically accurate information about foetal development and the abortion procedure should be sufficient.3 Further, if it is the case that ultrasound images provide some specific information that other images may not, such images could also be included in an informational pamphlet. What becomes clear at this point, is that the requirement that a woman seeking an abortion has an ultrasound (and views the images thereby produced) relies on and seeks to mobilize a crucial extra element. This is the idea that seeing one’s own foetus (not just any foetus) has a significant effect on how one feels about it, and further, how one might act in relation to it (Sanger 2008; 2015).

This notion that seeing one’s own foetus makes a difference to how one will act toward it aligns with the theory of maternal bonding, to which ultrasound has been harnessed for some time. Although the idea of maternal bonding has a longer history, it has been explicitly linked to ultrasound since 1983, when two obstetricians published a letter to the editor in the New England Journal of Medicine, claiming that ultrasound has the capacity to enhance maternal bonding and potentially forestall decisions to abort. In this letter, the authors, Fletcher and Evans (1983), discuss two cases in which they argue 2D ultrasound is implicated in maternal bonding in the first trimester, whereas traditionally, maternal bonding was understood as restricted to later stages of pregnancy, after quickening. The authors cite the two women as saying of their foetus after viewing ultrasound images “I feel that it is human. It belongs to me. I couldn’t have an abortion now”, and “I am going all the way with the baby. I believe it is human” (Fletcher and Evans 1983).

It is worth noting the contexts in which these comments are made: the first woman was a victim of domestic violence and her pregnancy was discovered in the course of x-rays to determine damage to her abdomen. The second woman was involved in a trial of early steroidal intervention to suppress foetal androgens in cases of congenital adrenal hyperplasia, and she had to make a decision about whether to undertake early intervention or wait till the mid-trimester amniocentesis to determine the sex of the foetus and consider termination in the event that it was female. These are not the standard contexts of women considering abortion in the first trimester and one imagines the intensely emotional and even traumatic occasions of the ultrasound may well have heightened its perceived effects. Further, it is important to note the caution of the authors: while they do suggest that ultrasound may have an effect on maternal bonding in the first trimester, they also suggest this is only in the case of an already strongly wanted pregnancy, and maternal ambivalence toward the foetus may only be resolved with later ultrasounds, in the mid-trimester. Regardless of these caveats, though, this letter has been credited with (inadvertently) sparking the anti-abortion lobby’s interest in ultrasound (Taylor 2002).

The notion that ultrasound precipitates or enhances bonding seems to have been strengthened by 3 and 4D ultrasound, which presents a much more vivid and ‘realistic’ image of the foetus than is obtained in grainy black and white 2D images. For instance, Campbell claimed in 2002 that while the clinical advantages of 4D ultrasound were disputable, the real gains lay in “parental behaviour and foetal behaviour” (Campbell 2002, 2). The latter is because 3 and 4D ultrasound makes it possible to observe small movements on the part of the foetus such as blinking. Of parental behaviour, Campbell argues that these images provoke significant emotional responses from parents, from which he concludes that the “natural desire of parents to see and know and love their baby before birth” ought to be recognized by obstetric ultrasound services. In 2006, he went further to suggest that augmentation of the bonding relationship by 4D ultrasound might be used to leverage positive parental behavioural changes – by implication, primarily on the part of the mother (Campbell 2006). In this opinion piece, Campbell claims that it is now widely accepted that ultrasound scanning at 12 and 20 weeks are “the main factors involved in initiating this bonding process” (Campbell 2006, 27). He again urges obstetricians to take more note of this aspect of ultrasound scanning, particularly since there is evidence that this may have health benefits, for example, women may reduce their alcohol consumption during pregnancy following scanning. This construal of the parental-foetal relationship gives enormous significance to ultrasound (and hence, medicine) insofar as it positions this technology as the principal instigator and fundamental mediator of the parents’ affective relation to the foetus or “baby”.

As Janelle S. Taylor (2008, 3) points out, while bonding theory was institutionalized in neonatal care in the early 1980s, the linking of bonding theory with ultrasound required a new supposition: that bonding can occur through spectatorship independently of the embodied experience of gestation. Moreover, obstetric ultrasound effectively renders the pregnant woman herself as a spectator. As feminist scholars have often argued, ultrasound images contribute to an understanding of the foetus as a being separate from the maternal body; in effect, ultrasound makes the foetus an other. Further, rendering the pregnant woman as a spectator obscures the privileged embodied relationship that she has to her foetus, placing her in a position more like that of other viewers of foetal images. As Iris Marion Young (2005, 61) argues, for instance “the pregnant woman’s experience of that image is just the same as anyone else’s who views it”. In the context of mandatory ultrasound, Young’s conclusion that this experience is the same as anyone else’s is overstated. For one, no-one else is required by law to view images of a foetus. Nor is anyone else required by law to offer up their body in order to produce the image. Nevertheless, pregnant women (are required to) participate in a visual experience of spectatorship, one which is understood to generate a specific emotional response to the foetus.

Interestingly, the notion of spectatorship has been discussed in an ethical register by film theorist, Lisa Cartwright (2008). She proposes the concept of moral spectatorship to elucidate the way that representations interpellate viewers or spectators in particular ways, and, according to her, this occurs specifically through empathy. For Cartwright, empathy means “the reflexive experience of awareness of the thoughts, emotions . . . or concerns of an other or others” (Cartwright 2008, 23). In this, Cartwright understands empathy not as a matter of “feeling like” the other, but rather, of “feeling for” him or her (33-34). Further, the significance of empathy is that it may prompt a sense of responsibility for others. Cartwright writes “spectators may also ‘feel themselves into’ those they can imagine not as themselves but as theirs, or rather, as their responsibility.” (Cartwright 2008, 235-6). What is important in this analysis, then, is the claim that an image or representation can precipitate the assumption of moral responsibility on the part of the spectator, specifically through an empathic response to the image. This claim shapes my discussion in the following section.

To summarize so far, we have seen that ultrasound images of foetuses have frequently been linked to abortion politics, and the idea that seeing the foetus will change maternal behaviour is not unique to mandatory ultrasound laws. In fact, a connection between ultrasound images and maternal bonding has been postulated for some time, an idea that relies upon the further supposition that spectatorship can itself precipitate a response of bonding. The idea of bonding here is obscure, but the notion of moral spectatorship may help make some of the stakes of the legal mandate to see one’s own foetus clearer. Specifically, we can say that mandatory ultrasound laws, especially those that make viewing the ultrasound image compulsory, compel pregnant women seeking an abortion to take on the position of a moral spectator; in doing so, they are asked if not forced to take responsibility for their foetus through an empathic response to them. To make this plausible as a general description of the work of mandatory ultrasound laws, then, I need to discuss the notion of empathy for the foetus in more detail. To do this, I turn to some brief comments on empathy and abortion made by moral philosopher, Michael Slote, in his book The Ethics of Care and Empathy (2007).

2. Empathy and Abortion

Slote is recognized as one of the leading contemporary theorists of moral sentimentalism, and he has done much to refocus attention on the concept of empathy in moral philosophy. His general project is to give contemporary credence to the sentimentalist claims of David Hume and Adam Smith, through reference to the developmental psychology of C. D. Batson and Martin Hoffman. In this, he elaborates an ambitious agenda of providing both ethical and meta-ethical reasons for centring on the concept of empathy in normative philosophy. My task here is not to engage with or assess this project as a whole; rather, I am interested in a brief set of comments that Slote made about the value of the concept of empathy in clarifying the ethics of abortion in his book, The Ethics of Care and Empathy (hereafter ECE). In this section of the paper, I reflect on Slote’s comments to amplify some aspects of his claims and draw out the connections between visibility, emotion and ethics that he suggests but does not elaborate. I focus on three issues: how the gestational age of a foetus may impact on empathy; the significance of the mediation of empathy by technology; and the way that social and medical norms frame the interpretation of an image and hence its capacity to generate empathy.

Slote’s project in ECE is to develop a comprehensive account of care ethics that makes the concept of empathy foundational to the practice and normative implications of caring. In this, he defines empathy as “having the feelings of another (involuntarily) aroused in ourselves, as when we see another person in pain” (Slote 2007, 13). Empathy is thus distinguished from sympathy, which entails feeling for another, but not feeling their emotions per se.4 In this, empathy is morally motivating in a way that Slote contends sympathy is not. He claims that empathy is crucial to altruism and moral motivation, such that “empathy makes a difference to how much we care about the fate of others in various different situations” (Slote 2007, 15). Because of this, in his view, feminist ethics of care needs the concept of empathy to elucidate why we might care for others. Beyond explaining the motivation for caring, the concept of empathy can also help extend care ethics into a fully-fledged moral theory—one that competes with, rather than merely complements, so-called justice based approaches. This is because empathy can provide a plausible criterion for moral evaluation, allowing us to determine the morally right and wrong. As he puts it “empathy can be used to clarify moral issues about what is intuitively better or more acceptable and what is intuitively worse or less acceptable” (Slote 2007, 19).

Slote attempts to demonstrate these points through a brief and, he admits, very preliminary discussion of the ethics of abortion, wherein our intuitive responses to the foetus guide moral evaluation. Slote claims that his use of the concept of empathy is almost unique in approaches to abortion, and preceded only by an article by John T. Noonan, a Catholic appellate court judge. In an article entitled “Responding to Persons” (1973), Noonan rejects forms of theorizing about abortion that do not focus on persons, a category that he supposes the developing embryo and foetus ought to be included within. This inclusion requires making the foetus visible and perceptible, and rests on recognition of the foetus as like us in some fundamental sense. In short, recognition of the embryo or foetus as like us is not based on physiological aspects of the foetus per se, but on comprehension of its experience. In this, empathy plays a crucial role. Noonan claims that while we have little direct access to the experience of being a foetus, since we don’t remember it “empathy may supply for memory” (Noonan 1973, 303); as such, it is this empathic identification with the foetus that allows it to be included in the “family of man” (Noonan 1973, 305). The moral consequence that Noonan draws from this is that empathic recognition of the personhood and correlative right to life of the developing foetus prohibits abortion.

Slote extends on Noonan’s claims to show that the concept of empathy may in fact yield an ethics of abortion more in line with pro-choice commitments, by identifying two complexities that Noonan ignores. These relate to first, foetal development and gestational age, and second, the difference between the late-term foetus and the neonate. I address each of these issues in turn; I also address a third issue that Slote does not mention, that of the ways in which an ultrasound image is perceived and interpreted within a given set of medical and social norms that shape our capacity for empathy.

2.1. Empathy, other minds and the foetus

Slote emphasizes the issue of foetal development and gestational age to argue that the concept of empathy supports a conclusion that abortions undertaken earlier in a pregnancy are morally acceptable in a way that later abortions may not be. He suggests that when made visible, early embryos often appear “alien”, and “look more like fish or salamanders or (at least) non-human, lower animals than like human beings”. As such, “we naturally tend to empathise more with [foetuses at] later stages than with the earlier” (Slote 2007, 18), and insofar as empathy is an evaluative criterion, this suggests that a moral distinction can be made between earlier and later abortions. This characterisation of a natural tendency to empathize more with foetuses later in gestation provokes a number of questions about the kind of beings with which it is possible to empathise, or, in other words, about the limits of empathy. Given that Slote defines empathy as “feeling what the other feels” (and not simply feeling for the other), there is a question about what a foetus feels, if it feels anything at all, as well as one of how we know what a foetus feels. I will address the second issue later, but here, the key issue is whether a foetus feels things of the kind that it is possible to empathize with. A first step in addressing this question is to consider whether a foetus feels anything.

In Slote’s and other approaches, empathy is thought of as a means of interaction with other minds, or more specifically, with the feelings and experiences of a being with a mind. As such, the presupposition is that the being that one empathizes with (the ‘target’) has a mind, though that does not necessarily entail that it be the mind of a person. The question, then, is whether a foetus has a mind—or, to put it in other words, is a human foetus (generally understood) the kind of being that is sufficiently sentient as to have experiences and feelings with which one could empathize? While there are several layers to this question, one way to address it is through considering the basic development of the foetal brain. As it is, the field of foetal neurodevelopment has itself become embroiled in abortion controversies through debates on foetal pain. Central to this is the issue of at what point the neural pathways that transmit sensorimotor information are sufficiently developed to give rise to experiences of pain, with some researchers placing this as early as 15 weeks gestation and others at around 26 weeks gestation and later.5 Even given this divergence, though, there is a broad consensus that such apparatuses are not in place within the 1st trimester of gestation – the time during which most abortions take place.

Given this, it seems reasonable to suggest that the foetus (or embryo) with a gestational age of less than 12 weeks does not have a mind, and is therefore not a sentient being with feelings and experiences. If that is the case, then it would seem that—insofar as empathy presupposes another mind—it is not possible to empathize with it. In short, at least according to Slote’s definition of empathy, it is conceptually impossible to empathize with an embryo or early-term foetus in utero prior to the second trimester, since there are simply no feelings for another to feel. This is not to say that it would be misplaced to care about an embryo or early-term foetus in some way. For instance, one may have concern for its safety now with a view to the wellbeing of the future child it may reasonably be expected to become (in the absence of abortion or miscarriage). But, in this case, Slote’s understanding of empathy will not help to articulate that care. This suggests that Slote’s particular construal of empathy may limit its usefulness for thinking about ultrasound and empathy, and a less stringent definition in terms of “feeling for” may be more appropriate. Note, though, that Slote’s point nevertheless resonates with evidence that some women required to have ultrasound examinations prior to abortion actually express positive feelings about having seen the foetus, particularly relief that it is not more recognisably human. I return to this point later.

The recognisability of the foetus also relates to the question of the limits of our empathic imagination. We can see from the discussion above that both Noonan and Slote rely on the presumption that empathy is facilitated by similarity or likeness, such that the more something is ‘like us’, the more able to empathize we are, and the more dissimilar it is from us, the more empathy is vitiated. In this, though, likeness can be understood in different ways, for example, as a matter of morphology or as a matter of behaviour. While a human embryo (prior to about 10 weeks gestation) is not morphologically similar, it is incontrovertible that at developmental stages later than this a human foetus does look like ‘us’ (by which I mean postnatal human beings), that is, like a very small human being. Despite these apparent similarities, a human foetus is significantly different from us as well: foetuses live in a different world to us and are in some ways, a different kind of beings. Foetuses do not breathe air, they do not eat or ingest food, and their basic habitat is significantly different to that of postnatal human beings—they live and flourish in a fluid-filled sac that would be the death of the rest of us. Moreover, they necessarily live inside the body of another.

Because of this, significant difficulties emerge in the interpretation of foetal behaviours. A foetus can seem to display behaviours expressively similar to ours—for example, moving away as an expression of pain, smiling as an expression of happiness and so on. However, it remains unclear whether movements and facial expressions on the part of the foetus are expressive of a conscious (emotional) response to the world or are simply reflex actions. In short, poking at the belly of a pregnant woman may produce foetal movement, but this does not necessarily signify that the foetus consciously experiences the poke as discomforting and moves away because of that.

A further difficulty relates to the way in which perception of morphology and interpretation of behaviour interact. The question is, does the interpretation of behaviours as expressions of certain feelings or emotions as akin to ours, and with which we may empathize, follow from perception of morphological similarity in revealed in ultrasound images? And, does the interpretation of certain behaviours as having the same expressive function as ours moves us to attribute (moral) similarity to the foetus? In other words, are we moved to interpret behaviours like ours as expressing the same emotion because the morphological similarities mean we already think of the foetus as a person? Or does the interpretation of behaviour as expressing the same emotion as it does for us lead us to attribute personhood to a morphologically similar though also clearly different being? What is at issue here is whether empathy is a matter of comprehension of the other, or of projection onto the other. Also at issue is the extent to which the capacity for empathy with a foetus is implicated in the performative attribution of moral status to the foetus, or the recognition of that status (see Mills 2014).

As a final point in this section, the problem of other minds and the lack thereof for establishing an empathic relation with a foetus might be avoided altogether through a recent phenomenological argument that empathy is not circuited through minds, but is instead a form of immediate bodily intuition. The argument here is that the basically Cartesian premise of most debates on empathy, which casts empathy as a mediation between ourselves and the mind of another, is mistaken. Instead, empathy should be understood as “fundamentally a pre-Cartesian experience that does not differentiate between the body and the mind of another” (Taipale 2015). On the face of it, this suggests a way that the foetus might be understood as a being that one can empathize with, insofar as it is capable of bodily expression. This would at least be suggested in regards to the foetus late in pregnancy, though prior to the development of the nervous system this would not be the case. Even so, one might suggest that this understanding still requires a mind—it is simply that the mind is no longer separated from the body and its expression in the manner of the Cartesian approach. Further, it might also be objected that, in the case of the foetus, there is no opportunity for immediate bodily intuition—except, perhaps, on the part of the pregnant woman—since there is no bodily apprehension of the foetus without technological mediation. This leads us to the second complexity that Slote introduces.

2.2 Empathy, perception and technology

Extending on the developmental difference thesis further, Slote asks whether there might be a similar difference between a foetus later in pregnancy and a newborn. Such a difference might arise, he suggests, because ultrasound and other technological visualisations of the foetus are indirect modes of perception: they are less immediate than contact with a newborn.

As this indicates, then, Slote gives some significance to immediate bodily apprehension in empathy insofar as he differentiates between the newborn and the foetus on the basis that the former are available to us through immediate perception whereas the latter are not; our perception of the foetus is necessarily technologically mediated through, for instance, obstetric ultrasound. For Slote, this difference between direct and indirect modes of perception matters for moral status, but unfortunately, he offers little explanation for this.

In a subsequent discussion of the moral significance of distance, Slote goes on to reiterate the claim that perceptual immediacy can make a difference to moral obligation. For him, this means that our obligations to those not immediately in front of us are relaxed or diminished. Even so, this discussion does not elaborate what perceptual immediacy amounts to, and nor does Slote return to the initial claim that the indirect perception of the foetus via imaging technologies such as ultrasound differentiates it morally from the newborn. In short, he fails to illuminate the role that technology plays in shaping perception and therefore empathy, and in particular, the ethical significance of this mediation of perception. Here, then, I want to reflect briefly on this point that technological mediation of perception impacts on moral status, insofar as it shapes empathy.

Although the question of the technological mediation of perception has had little play in bioethics, it has been taken up in philosophy of technology. For instance, Don Ihde (1979) has argued that technology changes our hermeneutic relation to reality. Building on this, Peter-Paul Verbeek has explicitly attended to the technological mediation of morality. Verbeek’s general project is to elucidate the ways in which ‘things’ shape our moral lives. He argues that technologies and things give shape to our lives and precipitate certain moral actions; further, things often have embedded within them certain moral claims and they mediate our moral lives in fundamental ways. One of the technologies that Verbeek discusses is obstetric ultrasound. He argues that ultrasound “establishes a hermeneutic relation between the unborn and the people watching it” and effectively translates the materiality of the foetus into a particular representation. In the process, it constitutes the foetus as “an individual person…made present as a separate living being rather than forming a unity with its mother, in whose body it is growing” (Verbeek 2011, 24). Further, ultrasound examination transforms prospective parents into decision-making moral agents in regards to a foetus. Verbeek’s analysis is not altogether novel insofar as feminist scholars have often pointed to the ways in which ultrasound is implicated in the constitution of personhood. However, what it helps to focus attention on is the way in which ethical relationships are themselves fundamentally mediated by technology.

Several points can be made about this mediation here. First, and most fundamentally, ultrasound constitutes the foetus as an other, a necessary prerequisite for the possibility of empathy. In other words, in presenting the foetus as separate from the maternal body, ultrasound imaging renders the foetus as a being with whom it is possible to empathize in a way that it would not be possible if the foetus is conceived of as part of a woman’s body. In this, then, ultrasound makes empathy with a foetus possible in a fundamental way, by making it appear as an other being. In addition to this, though, imaging technologies such as ultrasound are the means of access for interpreting foetal behaviours as expressions of emotions with which we might empathize. In other words, ultrasound and other foetal imagining technologies operate as the medium through which it becomes possible to observe and interpret foetal behaviour. They are the means by which we ‘know’ a foetus and what it might be ‘feeling’.

At this point, the double edge of the technological mediation of perception in Slote’s account becomes clearer. On the one hand, imaging technologies such as ultrasound are necessary to establish the possibility of empathic relations with a foetus, insofar as they represent the foetus as another being. They are also crucial to establishing empathy in that they are the primary means by which we can see behavioural indicators that might be interpreted as expressions of foetal emotion. On the other hand, though, the very fact that ultrasound has this role ensures that the empathic relation established with the foetus is secondary to that possible with the newborn. By virtue of this dependency on technology, the empathic relation with the foetus is inherently mediated in way that the haptic, bodily apprehension of the newborn is not. Thus, ultrasound both makes empathy with the foetus possible, and at the same time, ensures that the empathic relation thereby established is necessarily and fundamentally inferior to the direct relation possible with the newborn. It might be said that Slote’s reasoning exemplifies the logic of what Jacques Derrida (and others) identify as a “metaphysics of presence”, in which the unmediated encounter is taken as the primary reference point and the mediated encounter is derivative of and inferior to that.

Even so, Slote’s distinction between immediate and mediated perception brings into focus questions about the ways in which images may elicit empathy, which is itself complicated in regards to ultrasound. The value of ultrasound lies in its capacity to see the foetus, but what is seen in the ultrasound image (whether a still or a moving one) thereby produced is not simply the thing itself. Ultrasound is not “a window to the womb” as is popularly supposed, but rather, a complex technology that produces an image as an artefact of sound waves. The ultrasound image is in a strict sense a re-presentation of a foetus, produced through a synaesthesic ‘translation’ of sound into sight. Further, the image thus produced often requires interpretation to make any sense of it—it is often only when the sonographer or technician tells the viewer how to read the image that a foetus even becomes visible or recognisable in the image. This is perhaps less true of 3 and 4-dimensional ultrasound images, in which the addition of the plane of depth and in the latter, time, produces an image that is more susceptible to non-specialist interpretation. But, these images nevertheless require interpretation, as, in fact, all images do. Recognising the necessity of interpretation involved in the representational dimension of the ultrasound image brings to the fore the fact that the image and its interpretation is always framed in particular ways. In short, the technological mediation of the visibility of the foetus frames the encounter with the foetus in specific ways.

2.3. Empathy, norms and framing

It is a virtual truism of feminist analyses of ultrasound imaging that such images frame the foetus and its relationship to the woman gestating it in specific ways, especially to occlude the bodily presence of the pregnant woman and present the foetus as a separate being (Petchesky 1987; Stabile 1998; Hartouni 1997; Franklin 1991; Lupton 2013). This point rests on a critical understanding of what appears in an image, as well as what is left out. But the interpretation of images is also shaped by a broader notion of framing, insofar as the context of interpretation will also lend authority to some interpretive strategies and not others. As Judith Butler argues in regards to photography, images have a “transitive affectivity,” which means that “[t]hey do not merely portray or represent—they relay affect” (Butler 2009, 68). Further, this harnessing of affect occurs in conjunction with the transmission or “iteration” of social norms that regulate the appearance of the socially recognizable. In other words, the body that ‘appears’ in an image is necessarily understood in the context of social norms that render some bodies more socially acceptable than others. If this is correct, then it raises a question about how such framing by norms impacts on the interpretation and reception of ultrasound images and makes some empathic relations possible while closing off others.

In regards to ultrasound images more broadly, there are (at least) two interconnected sets of norms that shape their reception and interpretation. The first and most obvious of these is medical norms. As I have discussed in more detail elsewhere, as a medical technology, ultrasound has been central to the formation and application of norms in obstetric care (Mills 2016). In particular, ultrasound technology has enabled the formation of norms of foetal development, thereby making possible the identification of foetal anomalies and/or abnormalities.6 This is in fact one of the central uses of obstetric ultrasound today. While ultrasound was initially used in obstetrics to track foetal growth rates and estimate gestational age, the routine screens throughout pregnancy today are often seen primarily as a means of detecting foetal abnormalities. Consequently, at least in the context of the legal permissibility of termination of pregnancy, it also makes possible decisions about foetuses that do not develop in accordance with norms. In this, it contributes to what Eva Kittay (2006) identifies as a “desire for the normal”.

Arguably, the desire for the normal signals the point of intersection of such medical norms and the second set of norms that shape the interpretation and reception of ultrasound images, that is, social norms. Social norms can broadly be understood as the informal or customary rules and principles that guide behaviour in a given social grouping. However, social norms do not simply regulate behaviour; rather, as accretions of power, they also operate as an inescapable lens through which forms of embodiment are rendered socially acceptable or not. In her account of the impact of social norms on embodiment, Judith Butler (1993) argues that the social legibility of different bodies is unavoidably shaped by the operation of regulatory norms. Butler was primarily concerned with questions of gender and sexuality, but her view of the regulatory effects of norms on the social legibility and acceptability of different forms of embodiment has also been extended into matters of race and disability. The upshot of this is that forms of embodiment that deviate from social conceptions of normal bodies are rendered illegitimate or unintelligible, as fundamentally undesirable.

In terms of the interpretation of ultrasound images, the point is that the interpretation and affective response to such images is itself entangled in and shaped by the operation of social and medical norms. In short, the framing by norms means that any empathic relation with the foetus represented therein is itself shaped by norms both social and medical, and further, that the recognition of any particular foetus as an appropriate object of empathic relations is, to some extent, dependent on those norms. In philosophical discussions of empathy, the ways in which norms work upon a capacity to empathize is typically analysed under the rubric of bias. In this literature, it is broadly recognized that the capacity to empathize with the feelings of another are highly susceptible to bias. One issue in this susceptibility is the way in which moral approval and disapproval is subject to the modifying effects of social norms. But social norms also shape who one can empathize with, that is, who can recognizably be an appropriate target of empathy. With this in mind, it becomes apparent that empathy for the foetus is not straightforward, as the interpretation in a context of bias generated by social and medical norms means that the foetal image may interrupt empathic bonding as much as it precipitates it. We can identify two ways in which norms operate to interrupt and direct empathy in regards to the foetus.

First, it is worth considering one of the structuring oppositions within discussions of ultrasound, which I touched on in passing earlier. Recall that in the letter to the editor regarding maternal bonding, the women were quoted as saying “I feel that it is human” or “I believe it is human”. This designation of the foetus as human indicates that ultrasound has the capacity to bring the foetus into the sphere of empathic targeting through enabling or producing its recognition as human. However, as Campbell notes, upon seeing 3D ultrasounds in the first and early second trimester, many women regard the foetus as “alien” (Campbell 2006, 243). This is echoed in Slote’s claim that the early foetus often strikes us as alien or animal-like, such that we “naturally empathize more with later stages than with the earlier” (Slote, 2007,18). This suggests that seeing the foetus at the wrong stage of gestation might actually interrupt any bond that may have begun to emerge for the mother or parents – her baby is now rendered alien, and hence not an appropriate target for empathy. Interestingly, as I discuss further in the following section, this may mean that mandatory ultrasound viewing prior to the termination of a pregnancy has the opposite of the effect intended. As John Robertson (2011, 351) notes, rather than causing women to abandon plans for a termination, seeing the early foetus or embryo may in fact reassure women that a termination is morally acceptable.

Second, this interruption can also happen at later stages of development insofar as the interpretation of ultrasound images is saturated with norms that differentiate between the normal and the abnormal. What is interesting is how the medical norms of foetal development mesh with social norms that come to define pregnancies worth continuing. Moreover, these norms have themselves become fully integrated into the solicitation or, conversely, the suppression, of empathy. In effect, the diagnosis of anomaly can set a foetus outside the boundaries of empathy, breaking the maternal bond and rendering the pregnancy undesirable. As Catriona Mackenzie and Jackie Leach Scully (2007) argue, embodied experience may place significant limits on the capacity to empathically engage with the lived experience of differential embodiment, and this failure may help to account for high rates of abortion following the diagnosis of foetal anomalies or abnormalities. In the context of social norms that underwrite a differential valuation of bodily forms, it requires a particular imaginative effort to foresee a ‘life of value’ for the future child and person that a foetus with abnormalities may become.

As these points indicate, while the concept of empathy may be useful in articulating the general emotive effect of ultrasound images on decisions about abortion, we ought not assume that the response of one woman to an ultrasound image of the foetus she carries will necessarily be the same as that of another. Women seek abortions for a variety of reasons. Some do not wish to become a parent at all, others not now, or in the circumstances in which they become pregnant. Some women may wish to be pregnant, and to become a parent, but not with this sexual partner. Others want to be pregnant now, but seek an abortion following the diagnosis of abnormality. Given this, in the following section, I explore the disparities in women’s responses to ultrasound images in more detail.

3. Empathy and moral action

So far, I have discussed some of the issues that arise in relation to the connection between seeing the foetus and feeling for it. What remains to be discussed though, is the connection between feeling for the foetus, and any (moral) action in regards to it. While I am not going to be able to explore this issue in great depth here, it is necessary to address several points. First, there is an underlying theoretical claim being made that there is a connection between empathy and moral motivation. I discuss this claim here, and particularly whether the mediated nature of the empathic relation established through ultrasound has implications for the strength of this connection. Second, throughout the paper I have noted claims about how ultrasound might change maternal behaviour, but I have not discussed any empirical studies of this claim. I do that here, including studies of the efficacy of mandatory ultrasound laws in preventing abortion. Following this, I turn to discussing the legal compulsion of empathy and make two points about mandatory ultrasound laws; first, I note a gendered misalignment in their operation, and second, highlight the way in which the compulsion of empathy for the foetus has come at the expense of empathy for pregnant women.

The recent revival of interest in the notion of empathy—both within philosophy and more broadly in psychology and popular literature—means that there is now a well-established debate about the role that empathy plays in moral action. As I mentioned earlier, Slote argues that it plays a foundational role, not only motivating moral action but also providing criteria for moral approval or disapproval. Others, including other moral sentimentalists, argue that empathy does not have any significant moral role—it neither provides a foundation nor a motivation for moral action. For instance, Jesse Prinz (2011) has argued that empathy does not fulfil the roles attributed to it, and further, is in fact a moral ‘liability’. Prinz and other critics of empathy point out that as a moral motivation or foundation for moral approval, empathy is significantly misleading and fails to track moral judgements. This is because empathy is subject to various distorting effects, including bias and partiality.

For the purposes of this paper, it is not necessary to establish that empathy has a moral worth in general; what is important to consider is whether empathy has a moral motivating effect at all, and further, whether it has that effect in the context of gestation and the maternal-foetal relationship. Interestingly, the critical points made by opponents of empathy themselves lend strength to the idea that empathy is morally motivating, even if that motivation is not sufficiently wide to make empathy a good candidate for yielding moral action in general. For instance, critics of empathy point to a phenomenon that psychologists call the ‘identifiable victim effect’. Experiments on this effect show that knowing identifying features of the victim of misfortune produces a more significantly caring response, in terms of charitable donations and so on. Paul Bloom (2016) links this effect to empathy and goes on to argue that this shows that empathy is a poor moral guide. Perhaps; but it also shows that insofar as knowing the identity of a moral patient yields empathy, then empathy seems to have moral effects in relation to that identified target (even if not others).

This is significant in the context of gestation and abortion, because it might be argued that mandatory ultrasound laws are precisely an attempt to mobilize the identifiable victim effect through making the foetus visible – what is at issue is no longer a foetus in general, but this one foetus. After all, such laws do not require that women adopt anti-abortion beliefs and work to prevent the abortion of all foetuses, only that a woman responds to her own foetus, this particular being. In effect, ultrasound makes a foetus appear as a singular, embodied being with whom the pregnant woman as spectator bears an ethical relationship (Mills 2011). This raises the more specific question, then, of whether seeing the foetus in the context of deciding on abortion actually makes a difference to the decisional outcome. In other words, do women change their minds about abortion after seeing the foetus?

Throughout the period of the international routinisation of obstetric ultrasound since the 1980s, there have been numerous empirical studies of the extent to which ultrasound images precipitate or strengthen parental, and especially maternal, bonding. Given the methodological and other variations between these, it is difficult to make general statements about the results of them.7 However, it seems that viewing ultrasound images of it may have some measurable effect on emotional bonding to a foetus for prospective parents (or at least prospective mothers, since many such studies only recruit pregnant women). Recent studies have also investigated the degree to which maternal attachment is impacted by factors such as the gestational age of the foetus, and whether the ultrasound image is generated using 2-dimensional or 3 or 4-dimensional ultrasound, with contradictory conclusions drawn (Ji, Pretorius et al. 2005; Sedgmen, Mcmahon et al. 2006; Atluru, Appleton et al. 2012; De Jong-Pleij, Ribbert et al. 2013). Sedgmen et al (2006), reported that viewing an ultrasound image may reduce maternal consumption of alcohol (Sedgmen, Mcmahon et al. 2006), though it did not have an appreciable impact on other health behaviours. It should be noted that all of these studies were undertaken with women who were seeking to continue their pregnancies—they were ‘wanted’ pregnancies.

This array of evidence has led both proponents and opponents of mandatory ultrasound laws to assume that viewing an ultrasound image in the context of seeking an abortion for an unwanted pregnancy will impact on a woman’s decision to continue with or forgo pregnancy termination. There is, however, little evidence that seeing the foetus in the context of mandatory ultrasound prior to termination does have such straightforward effects. One qualitative study of women’s views on ultrasound in abortion care found that while both these views received some evidential support, mandatory ultrasound also had some unexpected results. In this study, Kimport et.al (2012) interviewed a sample of women receiving abortion care in two conservative states in the USA that have mandatory ultrasound statutes in place. They found that some women were dissuaded from having a termination, when they were already feeling ambivalent about that decision or undecided about their course of action. Conversely, as argued by pro-choice activists, some women found that viewing an ultrasound had a heightened emotional toll, but did not dissuade them from their prior decision to have a termination. Unexpectedly, women also reported that seeing the foetus made them feel better about having a termination, since it made the whole process of pregnancy termination “more real” or because it helped to concretize their responsibility in the process and “honor the complexity of her choice” (Kimport et al. 2012, e516). They conclude that in the context of an unwanted pregnancy, the effect of ultrasound viewing can have differential effects for women, including “improving their overall experience of abortion and satisfaction with their decision” (Kimport et al. 2012, e517).

Interestingly, the findings of this small qualitative study were reinforced in a larger quantitative study, involving more than 15,000 patients at a Planned Parenthood clinic in Los Angeles, California. In this, Gatter et.al (2014) collected data from women who chose to view an ultrasound image as part of their pre-abortion care, since the Californian statute requires that an ultrasound be offered, but does not mandate viewing. The conclusions of this study were that voluntary viewing of an ultrasound image in the context of abortion care has had little effect in terms of decision-making: 98.4% of women who chose to see their foetus continued to termination, compared to 99% of women who did not see their foetus beforehand. Of these, women with “low decisional certainty” (ie, who were confused, conflicted or did not want to abort), had slightly lower rates of continuing to termination after seeing the foetus (95.2%). As the authors conclude, viewing an ultrasound was only of significance for women who were already uncertain about the decision to terminate. And furthermore, this effect was strongly related to gestational age, such that women at 17-19 weeks gestation were 20 times as likely to continue the pregnancy compared to women at less than 9-weeks gestation (Gatter et al. 2014, 84).

Finally, a very recent mixed method study by Upadhyay et al. (2017) has examined the effects of ultrasound viewing in Wisconsin, a state that mandates that abortion providers present and describe the ultrasound image to the pregnant woman seeking a termination, regardless of whether she wants to view the image. Even in the context where seeing the foetus was legally mandated, the authors report similar results to those above. They conclude from quantitative data that Wisconsin’s mandatory ultrasound law was associated with a statistically significant but small increased likelihood of women continuing a pregnancy. However, the authors suggest that this needs to be considered in context and that factors other than viewing the foetus – such as having to pay out-of-pocket for abortion services—may explain this increase (Upadhyay et al. 2017, 18). From qualitative data, they show that most women reported that seeing the foetus had no impact on their decision; for a small number (8), it had an impact, but for most of these (5) the impact was positive—it helped solidify their decision to abort. Two women reported in interviews that seeing ultrasound images contributed to their decision to continue the pregnancy, both of whom reported significant ambivalence and uncertainty about seeking a termination. One reported that her boyfriend wanted her to abort, though she did not want to; the other reported that seeing her foetus on ultrasound at 16 weeks gestation resolved her uncertainty and made the decision to continue the pregnancy clear to her (Upadhyay et al. 2017, 17).

Two points from these empirical studies can be highlighted. First, it is worth noting that gestational age appears to play a role in the extent to which viewing one’s foetus generates morally significant attachment to it. Seeing the foetus at later gestational stages reportedly had more impact on women’s decision-making, and especially the decision to continue the pregnancy, than seeing the foetus at earlier stages of gestation, wherein the emotive responses were more mixed. This seems to synchronize with Slote’s developmental thesis, in which an image of a foetus carries more weight in terms of moral decision-making the more the foetus appears ‘like us’. The second point to highlight is that seeing the foetus appears to have most impact in a decisional context of maternal ambivalence. In other words, when a woman is less certain about the decision to terminate a pregnancy, seeing the foetus may have some impact, though for women with high decisional certainty, it has little impact, or, indeed, may have a positive effect of reassurance.

This context of maternal ambivalence points to an interesting feature of mandatory ultrasound laws and the justifications offered for them. As Denbow (2015) discusses in detail, mandatory ultrasound laws are cast as a means of protecting the informed choice of women against the influence of men who would have them abort. This means they work with two somewhat contradictory ideas. On the one hand, they posit that viewing an ultrasound can help precipitate bonding in circumstances where it is absent or insufficiently developed. On the other hand, though, they posit that women are naturally bonded with their foetuses, and, further, that natural bond needs to be protected against the coercive intrusions of others, such as abortion providers and male sexual partners. In relation to the second of these, the suggestion is that the influence of men diminishes women’s capacity to provide properly informed consent for terminations of pregnancy, since they coerce women into abortions they otherwise do not want. If we take this claim at face value, however, what becomes apparent is that there is a gendered misalignment in mandatory ultrasound laws. For if it is the case that men are coercing women into having abortions against their will, it is not obviously pregnant women who need to be shown ultrasound images of foetuses so much as their male sexual partners. This is not a recommendation, not least because in many circumstances, it would be unworkably impractical. However, it does make apparent the way in which such laws fit within the (biopolitical) regulation of the reproductive behaviour of women (Rodrigues 2014). Moreover, it also highlights the moral burden that is being placed on women: they are caught in the bind of being simultaneously a threat to the foetus and its potential rescuer.

The double edge of this positioning also contributes to the final point that I wish to make here, which pertains to the political distribution of empathy in the context of mandatory ultrasound laws. In the previous section, I made the point that ultrasound works to constitute the foetus as other, which is a necessary prerequisite for empathy. Further, I argued that the directionality of empathy is shaped by social norms, such that the question of who can be recognized as an appropriate target of empathy is itself a question of social power. Arguably, one effect of the mobilization of ultrasound images within the anti-abortion campaigns has been to direct empathy toward the foetus at the expense of empathy for the pregnant woman herself. In the unwanted pregnancy, two beings increasingly vie for empathic concern, and insofar as the pregnant woman is cast as threatening, she is seen as less deserving of that concern. In essence, insofar as ultrasound contributes to the formation of the two beings model of pregnancy, through the constitution of the foetus as other, it has meant that empathy has either been directed toward the foetus or the pregnant woman. And because women are seen as causing moral harm to another, they are no longer the appropriate target for empathic concern. In short, moral norms can get in the way of empathy.

4. Conclusion

This paper is guided by two aims—first, to extend the project of a bioethics of images through reflection on ultrasound imaging; and second, to critique the mobilization of such images in efforts to reduce access to abortion, specifically in mandatory ultrasound laws in the USA. While there is a growing critical literature on such laws, there has been little philosophical engagement with them, and few attempts to unpack the basic presumption of such laws: that seeing one’s foetus changes the ways in which one might act in regards to it, including in terms of the (ethical) decision about whether to allow it to live or not. However, in my view, this presumption raises significant questions, specifically about the relation between visibility, emotion and ethics. Throughout the paper, I argue that mandatory ultrasound laws compel women into a position of moral spectatorship and require them to recognise the foetus as a being for whom they are responsible, particularly through empathic responses to ultrasound images. In an effort to clarify the emotive effects of ultrasound images, I focus on the concept of empathy, to show that empathy with a foetus is complexly related to gestational age. I also point out that empathy is technologically mediated, and shaped by the intersection of medical and social norms. In the final section of the paper, I consider the relationship between empathy and moral action, specifically through empirical studies of the extent to which seeing the foetus impacts upon decisions about termination of pregnancy. I also highlight two critical points: first, mandatory ultrasound laws suffer from a gendered misalignment insofar as the presumption is that men coerce women into abortions – in this case, if the aim is to prevent abortions, it is men who apparently require the bonding experience supposedly precipitated by ultrasound images. Second, seen in political context, mandatory ultrasound laws attempt to foster empathy for the foetus, and they do so at the expense of empathy for pregnant women.


Atluru, A., Appleton, K. and Plavsic, S. K. (2012). “Maternal-Fetal Bonding: Ultrasound Imaging’s Role in Enhancing This Important Relationship.” Donald School J Ultrasound Obstet Gynecol 6(4): 408-411.

Batson, C. D. (2009). “These things called empathy: eight related by distinct phenomena.” The Social Neuroscience of Empathy. J. Decety and W. Ickes. Cambridge, MIT Press: 3-16.

Bloom, P. (2016). Against Empathy: The case for rational compassion. HarperCollins.

Butler, J. (1993). Bodies That Matter: On the Discursive Limits of ‘Sex’. New York, Routledge.

—(2009). Frames of War: When Is Life Grievable? London, Verso.

Campbell, S. (2002). “4D, or Not 4D: That Is the Question.” Ultrasound in Obstetrics and Gynaecology 19.

—(2006). “4D and Prenatal Bonding: Still More Questions Than Answers.” Ultrasound in Obstetrics and Gynecology 27: 243-244.

Cartwright, L. (2008). Moral Spectatorship: Technologies of Voice and Affect in Postwar Representations of the Child. Durham, Duke University Press.

De Jong-Pleij, E. A., Ribbert, L. S., Pistorius, L. R., Tromp, E., Mulder, E. J. and Bilardo, C. M. (2013). “Three-Dimensional Ultrasound and Maternal Bonding, a Third Trimester Study and a Review.” Prenatal Diagnosis 33(1): 81-88.

Denbow, J. (2015). Governed through Choice: Atutonomy, Technology, and the Politics of Reproduction. New York, New York University.

Derbyshire, S. W. G. (2010). “Foetal pain?” Best Practice and Research: Clinical Obstetrics and Gynaecology 24(5): 647-655.

—(2015). Fetal pain. Prenatal and Preimplantation Diagnosis: The Burden of Choice. Springer International Publishing: 119-130.

Fletcher, J. C. and Evans, M. I. (1983). “Maternal Bonding in Early Fetal Ultrasound Examinations.” New England Journal Medicine 308(7): 392-393.

Franklin, S. (1991). Fetal Fascinations: New Dimensions to the Medical-Scientific Construction of Fetal Personhood. Off-Centre: Feminism and Cultural Studies. Eds. Franklin, S., Lury, C. and Stacey, J. London, Harper Collins Academic: 200-201.

Gatter, M., Kimport, K., Foster, D. G., Weitz, T. A. and Upadhyay, U. D. (2014). “Relationship between Ultrasound Viewing and Proceeding to Abortion.” Obstetric Gynecology 123(1): 81-87.

Guttmacher Institute. (2018). “Requirements for Ultrasound.” Available from: https://www.guttmacher.org/state-policy/explore/requirements-ultrasound [accessed 3 June, 2018].

Hartouni, V. (1997). Cultural Conceptions: On Reproductive Technologies and the Remaking of Life. Minneapolis, University of Minnesota Press.

Ihde, D. (1979). Technics and Praxis. Dordrecht, D. Reidel Publishing Company.

Ji, E. K., Pretorius, D. H., Newton, R., Uyan, K., Hull, A. D., Hollenbach, K. and Nelson, T. R. (2005). “Effects of Ultrasound on Maternal-Fetal Bonding: A Comparison of Two- and Three-Dimensional Imaging.” Ultrasound Obstet Gynecol 25(5): 473-477.

Kimport, K., Preskill, F., Cockrill, K. and Weitz, T. A. (2012). “Women’s Perspectives on Ultrasound Viewing in the Abortion Care Context.” Womens Health Issues 22(6): e513-517.

Kittay, E. F. (2006). Thoughts on the Desire for Normality. Surgically Shaping Children: Technology, Ethics and the Pursuit of Normality. Ed. Parens, E. Baltimore, The Johns Hopkins Press: 90-110.

Lauritzen, P. (2008). “Visual Bioethics.” Am J Bioeth 8(12): 50-56.

Lupton, D. (2013). The Social World of the Unborn. Houndsmill, Palgrave Macmillan.

Mackenzie, C. and Scully, J.L. (2007) “Moral imagination, disability and embodiment.” Journal of Applied Philosophy 24(4): 335-351.

Mills, C. (2011) Futures of Reproduction: Bioethics and Biopolitics. Springer: Dordrecht.

—(2014). “Making Fetal Persons: Fetal Homicide, Ultrasound and the Normative Significance of Birth.” philoSOPHIA: A journal of continental feminism 4(1): 88-107.

—(2015). “The case of the missing hand: Gender, disability and bodily norms in selective termination.” Hypatia: A Journal of Feminist Philosophy 30(1): 82-96.

—(2016). Resisting Biopolitics, Resisting Freedom: Prenatal Testing and Choice. Resisting Biopolitics: Philosophical, Political and Performative Strategies. Eds. Wilmer, S. and Zukauskaite, A. London, Routledge: 109-122.

Nicholson, M. (2000). “Ian Donald—Diagnostician and Moralist.” Available from http://www.rcpe.ac.uk/heritage/ian-donald-diagnostician-and-moralist [Accessed 18 December 2018]]

Nicholson, M. and J. E. E. Fleming (2013). Imaging and Imagining the Fetus: The development of obstetric ultrasound. Baltimore, The Johns Hopkins University Press.

Noonan, J. T. J. (1973). “Responding to Persons: Methods of Moral Argumentation in Debate over Abortion.” Theology Digest 21(4): 291-307.

Palmer, J. (2009). “Seeing and Knowing: Ultrasound Images in the Contemporary Abortion Debate.” Feminist Theory 10(2): 173-189.

Petchesky, R. (1987). “Fetal Images: The Power of the Visual Culture in the Politics of Reproduction.” Feminist Studies 13.

Prinz, J. (2011). “Against Empathy.” The Southern Journal of Philosophy 49: 214-233.

Roberts, J. (2012). The visualised foetus: a cultural and political analysis of ultrasound imagery. Farnham, Ashgate.

Robertson, J. A. (2011). “Abortion and Technology: Sonograms, Fetal Pain, Viability and Early Prenatal Diagnosis.” Journal of Constitutional Law 14(2): 327-390.

Rocha, J. (2012). “Autonomous Abortions: The Inhibiting of Women’s Autonomy through Legal Ultrasound Requirements.” Kennedy Institute of Ethics Journal 22(1): 35-58.

Rodrigues, S. (2014). “A woman’s “right to know”?: Forced ultrasound measures an an intervention of biopower.” International Journal of Feminist Approaches to Bioethics 7(1): 51-73.

Sanger, C. (2008). “Seeing and believing: Mandatory ultrasound and the path to protected choice.” UCLA Law Review 56: 351-408.

—(2017). About Abortion: Terminating pregnancy in twenty-first Century America. Belknap Press.

Sedgmen, B., Mcmahon, C., Cairns, D., Benzie, R. J. and Woodfield, R. L. (2006). “The Impact of Two-Dimensional Versus Three-Dimensional Ultrasound Exposure on Maternal-Fetal Attachment and Maternal Health Behavior in Pregnancy.” Ultrasound Obstet Gynecol 27(3): 245-251.

Sekulic, S., K. G. Bukurov, M. Cvijanovic, A. Kopitovic, D. Ilic, D. Petrovic, I. Capo, I. Pericin-Starcevic, O. Christ and A. Topalidou (2016). “Appearance of fetal pain could be associated with maturation of the mesodiencephalic structures.” Journal of Pain Research 9: 1031-1038.

Slote, M. (2007). The Ethics of Care and Empathy. New York, Routledge.

—(2013). Moral Sentimentalism. Oxford University Press.

Smith, C. D. (2013-2014). “Mandatory ultrasound statues and the First Amendment, Shifting the constitutional perspective.” Cardozo Journal of Law and Gender 20: 855-884.

Stabile, C. (1998). Shooting the Mother: Fetal Photography and the Politics of Disappearance. The Visible Woman: Imaging Technologies, Gender and Science. Eds. Treichler, P. A., Cartwright, L. and Penley, C. New York and London, New York University Press: 171-197.

Taipale, J. (2015). “Beyond Cartesianism: Body-Perception and the Immediacy of Empathy.” Continental Philosophy Review 48(2): 161-178.

Taylor, J. S. (1997). Image of Contradiction: Obstetrical Ultrasound in American Culture. Reproducing Reproduction. Eds. Franklin, S. and Ragone, H., University of Pennsylvania Press: 15-45.

—(2002). “The Public Life of the Fetal Sonogram and the Work of the Sonographer.” Journal of Diagnostic Medical Sonography 18(6): 367-379.

—(2008). The Public Life of the Fetal Sonogram: Technology, Consumption and the Politics of Reproduction. New Brunswick, Rutgers University Press.

Upadhyay, U. D., Kimport, K., Belusa, E. K. O., Johns, N. E., Laube, D. W. and Roberts, S. C. M. (2017). “Evaluating the Impact of a Mandatory Pre-Abortion Ultrasound Viewing Law: A Mixed Methods Study.” PLoS One 12(7): e0178871.

Verbeek, P.-P. (2011). Moralizing Technology: Understanding and Desigining the Morality of Things. Chicago, Chicago University Press.

Young, I. M. (2005). On Female Body Experience: “Throwing Like a Girl” and Other Essays. New York, Oxford University Press.

1.This formulation is both more focused and broader than that a ‘visual bioethics’ and allows for analysis of both the production and reception of images. Further, its neutrality in relation to sense perception may allow greater room for analysis of the affective and emotional aspects of the reception of images.

2.Notably, the images of the foetus included in the 2003 version of the pamphlet were taken from Lennart Nillson’s classic book, A Child is Born. This is interesting because Nillson’s images are not representations of the foetus designed for medical information provision, but highly stylised works of art, often involving foetal remains. For an informative discussion of Nillson’s images, see Lupton, D. (2013).

3.The information pamphlet provided by the state of Texas mentioned previously continues to include discredited claims linking abortion to increased risk of breast cancer.

4.Slote’s definition of empathy is diametrically opposed to that proposed by Cartwright noted above. In literature on empathy, this definitional difference is not unusual. See especially Batson, C. D. (2009). Also see my earlier discussion of Peter Goldie’s conception of sympathy and ultrasound, in Mills, C. (2011). Slote later revises his understanding of empathy into what he calls a 3rd person perspectival account. (Slote 2013).

5 For opposed positions on fetal pain, see for example, Derbyshire (2010) and Sekulic et al (2016). For a useful overview of the issue, see Derbyshire (2015).

6.See Mills, C. (2015) as to why both the terms ‘anomaly’ and ‘abnormality’ are appropriate here.

7.For a critical discussion of recent studies in ultrasound bonding, see Roberts (2012).