Vous êtes sur la page 1sur 16

J Med Humanit (2006) 27:151166 DOI 10.

1007/s10912-006-9015-8 ORIGINAL PAPER

From Folklore to Fact: The Rhetorical History of Breastfeeding and Immunity, 19501997
Amy Koerber

Published online: 21 July 2006 C Springer Science+Business Media, Inc. 2006

Abstract This article examines the recent construction of human milks immune-protective qualities as scientic fact, demonstrating that long-standing controversies about human milks immune-protective effects have not been resolved by a particular scientic discovery. Rather, experts consensus on how to respond to this uncertainty has been transformed, and this transformation has had as much to do with a change in the metaphor that governs interpretation of evidence about immune protection as it has with discovering new evidence about either human milk or the antibodies in it. Keywords Medical discourse . History of medicine . Breastfeeding . Immune system . Metaphor . Rhetoric of science A 1988 immunology article cites an 1892 German study as the rst to report that human milk affords some kind of immune protection.1 A 1974 pediatrics article cites studies published in 1922, 1934, 1935, 1958 and 1961, all of which report clinical results suggesting that breastfed babies fared better immunologically than their bottle-fed peers.2 However, despite this apparent cause-effect relationship, throughout much of the twentieth century, the idea that breastfeeding provides immune protection was perceived as mere speculation rather than scientic fact. As the introduction to a 1979 book titled Immunology of Breast Milk explains, . . . both medical and veterinary scientists have long recognized the potential importance of immunity acquired after birth from the mother, although in clinical practice the arguments have often been accused of being wives tales and folklore (Tomasi, 1979, p. xiii).3 By contrast, in todays medical discourse, the immune-protective qualities of human milk are touted as a scientically proven benet that cannot be provided through

A. Koerber ( ) Technical Communication & Rhetoric, English Department, Texas Tech University, Box 43091, Lubbock, TX 79382 e-mail: amy.koerber@ttu.edu
1 2 3

Hanson et al, Antiviral and Antibacterial Factors in Human Milk. Gerrard, Breast-feeding: Second Thoughts, 757764. Tomasi, Introduction, xiii. Springer


J Med Humanit (2006) 27:151166

bottle-feeding. For instance, the American Academy of Pediatrics current policy statement on infant feeding cites numerous recent studies to support its claim that breastfeeding protects against childhood diseases including diarrhea, lower respiratory infection, otitis media, bacteremia, bacterial meningitis, botulism, urinary tract infection, and necrotizing enterocolitis. It also cites studies that show a possible protective effect against several adult diseases, many of which are currently understood as related to the immune system, including insulin-dependent diabetes mellitus, Crohns disease, ulcerative colitis, lymphoma, allergic diseases, and other chronic digestive diseases.4 This 1997 document, widely publicized for its declaration that breastfeeding should be treated as the infant-feeding norm, attributes its position to considerable advances that have occurred in recent years in the scientic knowledge of the benets of breastfeeding, in the mechanisms underlying these benets, and in the practice of breastfeeding (italics added).5 In this article, I investigate what lies behind this recent shift from perceiving human milks immune benets as anecdotal observation, or wives tales and folklore, to perceiving such benets as veried scientic fact. Specically, I examine the scientic arguments about infant feeding and immunity that unfolded in key medical journals in the last half of the twentieth century, adopting a critical perspective informed by recent observations about the immune system as a scientic and cultural phenomenon.6 As Donna Haraway observes, in the last quarter of the twentieth century, the immune system became a potent and polymorphous object of belief, knowledge, and practice in scientic and popular understandings of bodies and selves.7 Emily Martin traces the origins of the phenomenon that Haraway discusses to 1954 when gamma globulin (an immune component of human blood) rst became widely available to researchers, and there was a corresponding explosion in scientic knowledge about the immune system. According to Martin, the eld of immunology continued to be invigorated by these ndings throughout the 1960s and 1970s, when a series of scientic breakthroughs further expanded scientic understanding of antibody function. The model that emerged through these discoveries, according to Martin, began to increasingly idealize antibodies for their exible specicity. Scientists who studied the immune system knew about the wide variety of antibodies before this, and they knew that the body produced antibodies specic to different antigens, but the new model that emerged in the 1970s caused these traits to be further understood and emphasized. The belief entailed in this new model is that the body produces large numbers of antibodies to a wide variety of antigens, and when it encounters a particular antigen, it selects from this wide variety, and the antibodies specic to this antigen multiply.8 Based on these fundamental changes in scientic understanding of antibodies, and drawing on recent developments in cybernetics and systems thinking, Martin characterizes

4 5 6

American Academy of Pediatrics, Breastfeeding and the Use of Human Milk, 1035. Ibid., 1035.

The articles rhetorical analysis is based on a case study of 59 articles from a wide range of pediatrics and immunology journals. To select these articles, I started by locating the key texts on infant feeding and immunity that the 1997 policy statement cites to justify its claims about the immune benets of breastfeeding. From there, I worked backwards, attempting to locate any text that was cited as making groundbreaking claims about infant feeding and immunity. Although my research has led me to encounter a few much older publications, the majority of the articles that the chapter considers have been published between 1940 and the present.
7 8

Haraway, The Biopolitics of Postmodern Bodies: Constitutions of Self in Immune-system Discourse, 211.

Martin, Flexible bodies: Tracking immunity in the American culturefrom the days of polio to the age of AIDS, 3637. Springer

J Med Humanit (2006) 27:151166


late twentieth-century medicine and U.S. popular culture as moving from a hierarchical machine to a complex-systems metaphor of the immune system. In Martins interpretation, the complex-systems metaphor entails a fundamentally different way of thinking about the bodys ability to ght disease. Rather than viewing the immune system as located exclusively in the bloodstream and functioning to keep disease-causing agents out of the body, Martin says, the immune system in this new view is an integral part of a body capable of ghting disease at all of the bodys surfacesa body that actively relates to the world, that actively selects from a cornucopia of continually produced new antibodies that keep the body healthy and enable it to meet every new challenge.9 A complex system, as Martin describes it, consists of multiple feedback loops, so that instead of a centralized point of control, there are multiple points of control. Martin characterizes such systems as far less predictable than hierarchical-machine systems in that complex systems are extremely sensitive to uctuation and change in response to outside stimuli received by one or more parts of the system.10 I argue that the shift Martin identies in late twentieth-century perceptions of the immune system has been an important factor in the recent construction of human milks immuneprotective qualities as scientic fact. As I demonstrate, there has not been a particular scientic breakthrough or event that suddenly produced evidence so convincing that it ended the long-standing question of whether or not human milks immune-protective effects could be scientically veried. Rather, what has occurred is a transformation in expert consensus about how to respond to this uncertainty. This shift, I demonstrate, has had as much to do with a change in the metaphor that governs interpretation of evidence about immune protection as it has with discovering new evidence about either human milk or the antibodies in it. To develop this argument, I proceed chronologically, illustrating how the hierarchical-machine metaphor shaped scientists interpretation of data about infant feeding and immunity in mid twentieth-century research, but how such interpretations gradually shifted throughout the last half of the twentieth century as the new complex-systems metaphor began to take hold. Although the complex-systems metaphor has played a central role in this process of fact construction, my analysis suggests, metaphors do not emerge fully formed at a precise moment in scientic research; rather, they constantly shift in response to new evidence, at the same time that they actively shape how scientists interpret such evidence at any given moment. Thus, my analysis tells a story of scientic progress that is neither about rhetoric nor about science, but about the complex intersections between them. In this story, the discovery of new evidence is intertwined with the rhetoric surrounding it in such a way that neither makes sense apart from the other. Specically, although it had long appeared that breastfeeding enhanced immunity, this cause-effect relationship was not treated as scientic fact until the dominant metaphor of immune protection could account for an immune system operating in parts of the body other than the bloodstream. Infant feeding and immunity in the mid twentieth century The mid twentieth century in the United States has been characterized as an era of increasing faith in scientic solutions to problems, not only in the area of infant feeding, but in other aspects of life as well.11 Because the newly available infant formulas at the time were seen

Ibid., 37. Ibid., 119. Rosenberg, No Other Gods: On Science and American Social Thought. Springer

10 11


J Med Humanit (2006) 27:151166

as the fruits of scientic and technological progress, many mid twentieth-century physicians were eager to embrace bottle-feeding. However, they also desired the kind of scientic evidence that would nally settle the infant-feeding question that physicians had faced since the turn of the century when bottle-feeding began to rise in popularity and articial formulas proliferated.12 , 13 A 1958 review article by pediatrics researcher Bo Vahlquist, The Transfer of Antibodies From Mother to Offspring, provides a useful starting place for examining where research on infant feeding and immunity stood in relation to the mid twentieth-century thirst for scientic answers. Vahlquists review article is signicant because it seems to have offered a denitive answer to one of the many questions that researchers faced in this period: the question of whether the antibodies in human milk could be transferred from the mother to the nursing infant. Vahlquist acknowledges that in many non-human species, newborns receive antibodies through their mothers colostrum and milk. However, he asserts that this is not the case in humans and he rmly refutes other researchers claims to the contrary: The essential features of antibody supply in many animal species differ from those of man, and disregarding this fact has led to many erroneous conclusions. A pertinent example is the false evaluation of the signicance of colostrum and milk as vehicles for antibody supply to the human newborn.14 The review article proceeds by compiling an impressive amount of evidence demonstrating that antibodies from mothers milk can hardly ever be detected in more than trace quantities in the bloodstreams of nursing infants and, therefore, that human milk does not pass any signicant amount of antibodies to the nursing infant. Vahlquists claim that human milk could not afford immune protection because the antibodies it contained were not showing up in the bloodstreams of nursing infants is consistent with the hierarchical-machine metaphor of the immune system that dominated scientic thinking in this period. Specically, according to Martin, the hierarchical-machine metaphor depicted antibodies as residing and enacting immune protection primarily in gamma globulin, a component of human blood, and it did not account for the possibility of immune protection operating in other bodily systems or uids.15 Although, as will become clear in the next section of this essay, Vahlquists conclusions held a great deal of sway in the pediatrics community for many years after they were published, his conclusions did not represent the only possible interpretation of the evidence he presented. In fact, prior to Vahlquists 1958 article, some researchers had already begun to look for a form of immune protection that human milk might offer through a means other than the bloodstream. However, because the dominant metaphor in 1950s research dictated that immune protection occurred primarily in the bloodstream, and the antibodies from mothers milk were not appearing in the bloodstreams of nursing infants, such researchers largely failed to make convincing scientic arguments. Because metaphor is intimately connected to knowledge construction, metaphor theorist Ken Baake argues, when scientists encounter evidence that is not consistent with the dominant metaphor, they must accommodate the available language to make it correspond with the

12 13

Apple, Mothers and Medicine: A Social History of Infant Feeding, 18901950, 1617.

Wolf, Dont Kill Your Baby: Public Health and the Decline of Breastfeeding in the 19th and 20th Centuries, 74101.
14 15

Vahlquist, The Transfer of Antibodies from Mother to Offspring, 305. Martin, 31. Springer

J Med Humanit (2006) 27:151166


new evidence as closely as possible.16 In attempt to accommodate their ndings to the dominant metaphor, researchers described such ndings in terms that now seem vague, as an element that seemed to kill specic viruses and bacteria through direct contact, either in laboratory settings or on surfaces of the body such as the respiratory and digestive tracts. For instance, a 1950 article reports that 80 percent of the human milk specimens in its study contained varying amounts of neutralizing substance for the virus of poliomyelitis.17 Based on these ndings, the article urges the medical community to consider the role that breast feeding beyond the third month, or the drinking of antipoliomyelitic cows milk, might play in preventing or modifying the infection in human beings.18 Reecting the fact that researchers in this period understood the immune system in a way that did not encompass this type of anti-toxic effect on bodily surfaces other than the bloodstream, the author is careful to stipulate that the antipoliomyelitic substance in milk was not antibody.19 Similarly, a 1955 article concludes that Milk (human, cow, goat) and human colostrum have a strong destructive factor against virulent and non-virulent mutants of leptospira.20 Taking a slightly different approach, but still reecting the idea of an unknown substance in human milk that enacts immune protection directly on bodily surfaces other than the bloodstream, a 1953 article reports that the intestinal ora of breastfed infants consist mostly of Lactobacillus bidus, and it suggests that a particular biochemical substance in human milk promotes growth of this disease-ghting organism.21 These researchers efforts to accommodate their ndings to the dominant metaphor at the time were apparently not very successful. Interestingly, Vahlquists review article cites several such studies, which he characterizes as suggesting a local protective effect. However, he quickly dismisses the possibility of such an alternative form of protection, stating simply, . . . the results of Sabins experiments to show such a local effect in the digestive tract have not been unequivocal and moving on to state more evidence against the idea that human milk affords immune protection.22

Resisting Vahlquist: Unsettling the mid twentieth-century stasis Vahlquists review article was signicant not because it presented new evidence but because it presented large amounts of data accumulated over several decades in an argument that satised the medical communitys desire for denitive answers to the infant feeding and immunity question. However, in establishing such a strong and well-accepted argument, Vahlquist also provided the target at which anyone who wanted to argue for breastfeedings immune benets would aim for years to come. Researchers efforts to hit this targetthat is, to resist the truths that Vahlquist establishedare my focus in this section. In rhetorical terms, I demonstrate, Vahlquists article played an important role in establishing a stasis that dened infant-feeding research throughout the 1960s and 70s and was eventually unsettled in

16 17 18 19 20

Baake, Metaphor and Knowledge: The Challenges of Writing Science, 114. Sabin, Antipoliomyelitic Substance in Milk of Human Beings and Certain Cows, 867. Ibid., 867. Ibid., 867.

Kirschner & Maguire, Antileptospiral Effect of Milk, 564. Leptospira is a bacteria that can be passed from cattle to humans, leading to outbreaks of a disease called leptospirosis.
21 22

Gyorgi, A Hitherto Unrecognized Biochemical Difference between Human Milk and Cows Milk, 98. Vahlquist, 321. Springer


J Med Humanit (2006) 27:151166

the 1980s. Stasis is, dened simply, the place where rhetoric begins, an explicit or implicit disagreement or conict.23 In science, stasis can be understood as the unresolved issues or tensions that scientists agree are at the heart of their quest for knowledge at any given moment, as a gap in certiable knowledge or a point . . . that must be settled somehow before any further progress can be made in discussing the overall problem.24 As John Gage observes, the situations of conict of knowledge that constitute points of stasis are the mutual invention of the writer and the audience.25 In other words, a stasis does not exist entirely independent of the rhetors moves; rather, it is, in part, shaped by these moves. Thus, to say that Vahlquists text dened a stasis is to say that it articulated one of the burning questions that his contemporaries were asking, but also that it established the point of stoppage subsequent researchers would encounter when they made arguments in response to this question. Because of the stasis that Vahlquists article helped to solidify, in some ways, his article made it much more difcult for later researchers to convince the pediatrics community that the antibodies in human milk were providing signicant immune protection. However, as I demonstrate, researchers did not stop looking for evidence of immune protection altogether; they simply had to intensify the effort to nd a different kind of immune protection than that which had dominated understandings of the phenomenon up until this time. Using various approaches, new ndings published throughout the 1960s and 70s began to converge around the speculation that Vahlquists article had mentioned but swiftly dismissed: the possibility that milk antibodies exercise local, rather than systemic, immune protection by resisting antigens directly on the surfaces of the digestive tract. The situation these researchers faced was made easier by some ndings that immunologist Lars Hanson reported in 1961 and 1962. Interestingly, perhaps because Hanson was an immunologist rather than a pediatrician, his are some of the few post-1958 publications I consider in this article that do not cite Vahlquist. In any case, Hansons article identies a type of antibody in human milk that he suggests might function differently than the serum antibodies that medical researchers were accustomed to studying. In some ways, this 1961 article does not sound too much different from the 1950s articles I discussed earlier, which hypothesized the existence of a local immune protection operating on bodily surfaces other than the bloodstream. Like the authors of these earlier articles, Hanson uses currently available terminology to accommodate his ndings to the dominant metaphor at the time, which dictated that gamma globulin was the primary vehicle for immune protection: he calls the antibodies he identies gamma-globulins of milk. However, as an immunologist, Hanson goes further than previous pediatrics researchers in speculating about the differences between milk and serum antibodies and the mechanisms underlying such differences. Although his language is imprecise and speculative, Hanson suggests that milk antibodies might be serum antibodies to some extent modied in the mammary gland and that the difference between the antibodies in serum and milk could be important in relation to the absorption of immune globulins from the gut of the breast-fed child.26 Although, according to Martin, immunologists did not fully embrace the complex-systems metaphor until the 1970s, Hansons suggestions anticipate the

23 24 25 26

Carter, Stasis and Kairos: Principles of Social Construction in Classical Rhetoric, 99. Prelli, A Rhetoric of Science: Inventing Scientic Discourse, 145. Gage, An Adequate Epistemology for Composition: Classical and Modern Perspectives, 16667.

Hanson, Comparative Immunological Studies of the Immune Globulins of Human Milk and of Blood Serum, 262. Springer

J Med Humanit (2006) 27:151166


unpredictable, multiple, and non-linear way of thinking about the immune system that would accompany this new metaphor. In a 1962 follow-up to this article, Hanson and collaborator Bengt G. Johansson add further credence to these speculations, reporting the results of their successful attempt to isolate human milk antibodies [i]n order to examine more closely the relationship between the immune globulins of milk and of blood serum. . ..27 In particular, the researchers isolate a milk antibody that they call milk 2A -globulinthe existence of which was rst postulated in the 1961 article. Although this milk antibody was still not understood as precisely as it would be later, the 1962 article is important because it uses an imaging technique called immune electrophoresis to conrm what was merely suspected in the 1961 article: that antibodies unique to human milk do exist and that they behave somewhat differently than serum antibodies. As Bruno Latour and Steve Woolgar observe, the type of isolation and identication accomplished by images such as those of Hanson and Johansson is an important step in the production of scientic truth because once a hypothesized substance has been identied and isolated, researchers can stop devoting energy to disputing its existence and nature, and they can instead use it as a tool. Once this happens, the substance provides one less concern, or one less source of noise, for the researcher.28 Thus, isolation and identication of a substance can contribute to the construction of scientic facts by causing researchers speculations to converge around a belief or set of beliefs that is increasingly certain, rather than continue to move in various directions. Latour and Woolgars insights usefully illuminate the trajectory that research on breastfeeding and immunity followed throughout the 1960s and 70s. After the electrophoretic images that Hanson and Johansson produced in 1962, researchers nally had laboratory evidence for an immune component in human milk, rather than just clinical evidence, and consequently, they were able to develop more precise scientic arguments in regard to this component. In the work that such researchers accomplished with the new tool available to them, we see how an emerging metaphor can generate knowledge by causing scientists to see old evidence in productive new ways. Specically, after these breakthroughs in immunology, human milk gradually became associated with the complex-systems metaphor of the immune system that, in Martins interpretation of history, was beginning to emerge in the 1960s and became more fully formed in the 1970s. As this association developed, researchers gained access to a new vocabulary, and along with this vocabulary came a new framework for interpretations of evidence related to the question of human milks immune protection. Even though the researchers who challenged Vahlquist continued to cite his article, they were able to grant his claims a much narrower scope than Vahlquist himself had done, and with the more precise scientic terminology that became available as the new complex-systems metaphor emerged, they were eventually able to unsettle the stasis that Vahlquists article had established. Thus, the power of his broadsweeping claim against human milks immune-protective qualities diminished throughout the next couple of decades as pediatrics researchers came to restate it in increasingly narrow terms and to place greater emphasis on the local, rather than systemic, immune protection offered by human milk.

27 Hanson and Johansson, Immunological Characterization of Chromatographically Separated Protein Fractions from Human Colostrum, 65. 28

Latour and Woolgar, Laboratory Life: The Construction of Scientic Facts, 110. Springer


J Med Humanit (2006) 27:151166

We begin to see this effect in an article titled Bacterial and Viral Coproantibodies in Breast-Fed Infants.29 Like many other pediatrics texts in this period, the article cites Vahlquists 1958 article in its opening paragraph to support the following synopsis of previous research: The existence of bacterial and viral antibodies in human milk has been known for some time. However, since most investigators have found that little parenteral absorption occurs in the human being, these antibodies generally have been considered unimportant in the immune system of the newborn infant.30 The article then sets out to resist this previously accepted fact about the role of antibodies in human milk, citing Hanson and Johanssons 1962 publication to support the claim that human milk might offer some degree of immunity to intestinal infections during early infancy.31 In search of evidence for this kind of protection, the authors of this article conducted a study that looked for coproantibodies, which means antibodies in the stools, rather than in the blood, of newborn breast-fed infants. They nd evidence of such antibodies, and they conrm that such antibodies come from mothers milk because titers of antibody in infant feces correlated well with those in the mothers milk.32 Furthermore, they rule out serum as a direct source of fecal antibody by noting little correlation between antibody levels in mother/infant serum and those in infant stool. As further evidence for their claim, they note that such antibody activity was not found in the stools of bottle-fed infants. As we see in the articles Comment section, one reason why these authors more denitive arguments became possible was that, as human milk came to be associated with the newly emerging idea of an alternative, supplementary immune system, it also came to be afliated with a metaphor that could compete with the hierarchical-machine metaphor that had governed Vahlquists claims. Specically, reecting the move toward what Martin calls the complex-systems approach to thinking about the immune system, in its Comment section, the article on coproantibodies echoes Hansons earlier speculations about the existence of an additional, previously undiscovered immune system that operates in a self-regulating manner in response to specic stimuli that the intestinal cells encounter: Antibody appears earlier in intestinal mucus than in serum, suggesting that the intestine has an immune system of its own with cells which may produce antibody following local antigenic stimulation . . . .33 The momentum toward such an alternative, supplementary immune system continues to build in research conducted throughout the 1970s. For instance, a 1971 article that cites both Vahlquist and Hanson, as well as the above-mentioned article on coproantibodies, studied 99 infants, including 33 infected with neonatal meningitis or urinary tract infections, and 66 uninfected controls. The study concludes that those infants in the study who had consumed the most breastmilk were the least likely to be infected. The authors connect this nding to the emerging idea of human milk as a source of local protection: These antibodies [in human milk] are probably not absorbed, but their presence in colostrum and their ability to pass through the gut with well-retained antibody activity would accord with a role in local

29 30 31 32 33

Kenny, Boesman, and Michaels, Bacterial and Viral Coproantibodies in Breast-fed Infants. Ibid., 202. Ibid., 202. Ibid., 205. Ibid., 209. Springer

J Med Humanit (2006) 27:151166


defence against E. coli.34 Next, they repeat the speculation that multiple systems might work together to ght infection in the newborn, again echoing the move toward associating breastfeeding with the complex-systems metaphor of the immune system: Thus there might be a defence against gram-negative septicaemia at several levelsone local . . . the other humoral . . . . Obviously other defence systems are present.35 Interestingly, this 1971 article cites Vahlquist as the single source for its claim that These antibodies [in human milk] play little if any, role in mans humoral defence, but it then goes on to position its hypothesis relative to Vahlquists stance: but since some of them are passed along the intestinal tract with activity intact they may have local effects.36 This same article later cites Vahlquist again as a source for the following statement: These antibodies are probably not absorbed, and it again contrasts its own ndings to Vahlquists established fact: but their presence in colostrum and their ability to pass through the gut with well-retained antibody activity would accord with a role in local defence against E. coli.37 Articles throughout the 1970s continue to produce further evidence for the assertion that human milk affords local protection and, more important, that this local protection is an integral part of an immune system consisting of multiple components that work together to ght infection on multiple fronts. An article that includes Hanson as one of its co-authors provides a good indication of the status of beliefs about infant feeding and immunity at the beginning of the 1970s. Published in 1972, the article reects back on the signicance of Hanson and Johanssons 1962 ndings, and it explicitly connects these ndings to the new understanding of breastfeeding and immunity that the pediatrics community would later adopt. The article begins by noting that newborns in the 1970s were being exposed to more disinfectants and antibiotics and less breastfeeding than ever before. The authors then observe, The fact that this trend has continued side by side with a situation where infections have played a diminishing role in neonatal morbidity and mortality has created the impression that such changes have been benecial, or at least not harmful, to the infant.38 The article then provides evidence of the long-standing inuence of Vahlquists article in perpetuating doubt about human milks immune benets, conrming that the 1962 isolation and identication of human milk antibodies played a central role in enabling researchers to overcome Vahlquists truth claims: The possible role of the low consumption of breast milk has attracted little attention, and historically the fact that the antibodies of milk are not absorbed from the gut in the human (Vahlquist, 1958) is perhaps partly responsible for this. With the appreciation of secretory IgA as the main immunoglobulin in breast milk, and also as an important immune factor for epithelial surfaces, a new situation has developed. This, as well as the recognition of several other antibacterial factors in breast milk, has reopened discussion of the value of breast milk.39

34 35 36 37 38 39

Winberg and Wessner, Does Breast Milk Protect Against Septicaemia in the Newborn, 1094. Ibid., 1094. Ibid., 1091. Ibid., 1094. Hanson & Winberg, Breast Milk and Defence Against Infection in the Newborn, 845. Ibid., 845. Springer


J Med Humanit (2006) 27:151166

However, the articles conclusion is still speculative about breastmilks immune protection, and it ends with a call for more research.40 Thus, a full decade after Hanson and Johanssons 1962 identication and isolation of milk antibodies, researchers still perceived a need for more scientic research and rhetorical work to convince scientic and medical audiences that the 1962 evidence was relevant to the infant-feeding and immunity question. Although such researchers were using the important new evidence from immunology to strengthen the argument that human milk provided local protection, they could not refute Vahlquists truth claims all at once because the complex-systems metaphor of the immune system had not fully entered into these researchers interpretive framework yet. Shortly after Hanson and Winbergs 1972 article, however, a handful of publications start to assume a more denitive stance. This more denitive stance is clearly connected to the fact that the complex-systems metaphor for the immune system was gaining ground as the eld of immunology continued to explode. For instance, a Commentary published in Pediatrics says the following in regard to the challenge that recent immunological research was posing to Vahlquists 1958 review article: [Vahlquists] review of the passage of immunoglobulins from mother to fetus, written before the recent explosion of information on the structure and function of immunoglobulins, stated that the fetus derived all of his immunologic protection transplacentally and that, though some immunoglobulin was present in colostrum, this was of no clinical consequence.41 Gerrard then proceeds to counter Vahlquists well-established truth, and to urge the pediatrics community to join him in rethinking long-standing medical beliefs about infant feeding and immunity: Most of us in the practice of pediatrics, as in so many other areas, have taken a permissive attitude toward breast-feeding. This attitude is understandable, since we presumed that the function of breast milk was little more than the provision of nourishment. We now know that breast milk also provides effective protection, more effective than antibiotics, against certain common enteric pathogens, and that it can also be expected to provide relative freedom in infancy from allergic disease, a growing problem of modern feeding habits.42 Gerrards remarks indicate that the stasis Vahlquists 1958 review article had played such an important role in establishing was, by the mid 1970s, well on its way to being unsettled as the new knowledge resulting from the rapid expansion of immunology as a discipline continued to make its way into pediatrics researchers understanding of infant feeding and immunity. Three years after Gerrards commentary was published, an immunology article cites Hanson and Winbergs 1972 article to support the following claim: Human colostrum and milk has now been shown to possess a wide variety of host-resistance factors. This 1977 article seems to dismiss Vahlquists ndings altogether. Interestingly, the authors do not cite him, and their article contains the following sentence: Although in many mammalian species, the absorption of immunoglobulin in the intestine is well established and often
40 41 42

Ibid., 847. Gerrard, 757. Ibid., 763. Springer

J Med Humanit (2006) 27:151166


constitutes the sole source of passive immunity in the neonatal period [the authors cite several studies here], the data on intestinal absorption of immunoglobulin in man are too fragmentary and conicting43 The article then concludes that breast feeding during the 1st week after the onset of lactation may be crucial in providing the infant with immunologically active ingredients in a concentrated form at a time when the mechanisms of mucosal defense are less than optimal.44 Suggesting that the stasis established by Vahlquists 1958 article had been altered even further by the 1980s, a 1986 pediatrics article begins by citing Vahlquist as the source of the following claims, which the authors feel condent in portraying as outdated: The presence of immunoglobulins in breast milk was initially thought to have no signicance for infants. It was believed that all immunologic protection in the human neonate was acquired transplacentally.45 The article then goes on to cite several recent studies that refute Vahlquist by demonstrating the signicance of the unique kind of antibody human milk contains, and it concludes, It is now generally accepted that human milk affords the infant primarily enteric protection, whereas systemic immunity is enhanced by transplacental passage of immunoglobulins.46 The article also states what is emerging as the new view of breastmilk: Because articial replication of these species-specic immunity factors is not possible currently, breast milk remains unique when compared with prepared formulas.47 This same article also emphasizes the centrality of the new understanding of immunity to the emergence of the new view of breastfeeding: The ability of breast milk to confer passive immunity has withstood the test of time well and remains the leading impetus to encourage breast milk feedings.48

Conclusion Beginning with Emily Martins observation that the complex-systems metaphor for the immune system emerged gradually in immunology research throughout the last half of the twentieth century, this article has traced how this metaphor has enabled the construction of new scientic facts about breastfeeding and immunity in the realm of pediatrics. As I have demonstrated, the only piece of evidence that can be considered new in this story of scientic progress is Hanson and Johanssons 1962 electrophoretic images of the antibodies in human milk. However, this piece of evidence did not single-handedly or instantaneously change researchers beliefs; it took more than a decade after Hanson and Johanssons images for researchers to start talking about human milks immune benets as scientic fact. As my analysis has suggested, neither the new metaphor nor the new evidence emerged fully formed at a precise moment. Rather, the relationship between the two has been always in uxthe metaphor shifting in response to new evidence, but also shaping how scientists interpret such evidence at any given moment. In fact, the term metaphor derives from the Greek words epiphora and phora, which both imply movement, and Such movement
43 44 45 46 47 48

Ogra, Weintraub, & Ogra, Immunologic Aspects of Human Colostrum and Milk, 247. Ibid., 248. Lemons, Lemons, & Lemons, Breast-feeding the Premature Infant, 11213. Ibid., 112. Ibid., 113. Ibid., 119. Springer


J Med Humanit (2006) 27:151166

suggests that truth can never be xed in language, but must be constantly renegotiated (constructed) as new metaphors appear.49 Bearing out this complex relationship between new evidence and the metaphors that govern its interpretation, I have demonstrated, new scientic evidence about human milks immune-protective qualities has been so closely intertwined with changing metaphors of the immune system that it is impossible to pinpoint an exact moment when the truth changed or a place where science ends and rhetoric begins in this narrative of scientic progress. To understand the signicance of Hanson and Johanssons 1962 evidence in unsettling the stasis that Vahlquists 1958 article established, we have to look beyond the evidence itself and consider the larger context that surrounded it. This enlarged perspective reveals that the images were an important piece of evidence not only within the narrow connes of infant-feeding science, but also because they coincided with a larger explosion of scientic knowledge about the immune system in general. With this explosion of knowledge, the eld of immunology became an increasingly important object of scientic study throughout the time frame this article has considered: the number of researchers in immunology was 10,000 in 1970, compared to only 3300 in 1950 and 110 in 1930.50 Although it would be difcult to obtain an exact count of how many of the 10,000 immunology researchers in 1970 were specically focused on human milk, the following paragraph from the Introduction to a 1979 book entitled Immunology of Breast Milk suggests that human milk denitely occupied a prominent position in the spotlight cast by the explosion of knowledge about the immune system that began in the 1950s and gained increasing force throughout the 1960s and 70s: Although the predominance of IgA in human colostrum and milk had been described, the importance of this observation was not fully appreciated until the discovery that IgA was a predominant immunoglobulin class in a variety of mucosal secretions in addition to breast milk. This set the stage for the explosion of knowledge concerning mucosal immunity that has accumulated over the past 15 years. . ..51 Thus, it seems, Hanson and Johanssons evidence was important because it caused human milk to become associated with the newly emerging metaphor of the immune system that was causing so much excitement for immunology researchers. As Hanson et al. say in a 1988 article that reects back on the recent history of developments in understanding human milks immune protection, Only when the special characteristics of sIgA antibodies in milk and other external body uids were discovered was it realized that these antibodies were to function on mucosal membranes, not in tissues.52 The authors summarize these special characteristics as follows: The main function of SigA antibodies is to bind antigens. The most important function in host defense is presumably to bind bacteria preventing them from attaching to mucosal membranes, an important rst step in the initiation of most infections. The antibodies in milk against bacterial adhesins, such as pili or mbria, may mediate such anti-adherence activity.53

49 50 51 52 53

Baake, 62. Jerne, Towards a Network Theory of the Immune System. Tomasi, xiii. Hanson et al., 142. Ibid., 142. Springer

J Med Humanit (2006) 27:151166


Hanson et als description of how antibodies in human milk function clearly reects the phenomenon Martin is referring to when she says that biomedical discourse has moved toward a model of antibodies as exibly specic. The language that Hanson and his co-authors use also reects the shift to what Martin calls a complex systems metaphor of the human body, and the authors embrace the new metaphor with more certainty than any of the previous texts I have examined. For instance, in summarizing recent advances in understanding the immune functions of breastmilk, these authors refer to research that demonstrated the presence of an enteromammaric pathway of lymphoid cells homing to the mammary glands after antigenic exposure in the Peyers patches in the intestine.54 In other words, through this enteromammaric pathway, antigens to which the infants intestine is exposed cause the mothers mammary glands to produce antibodies specically designed to ght against them. Additionally, the authors point to a study that suggests the existence of a bronchomammaric pathway that functions in a similar way to produce and deliver to the infant antibodies against antigens that affect the respiratory system. Thus, we can conclude, the rise of the complex-systems metaphor allowed researchers in the 1970s to gradually reject Vahlquists well-established version of the truth. One reason this happened is that human milk suddenly became afliated with a different sphere of research immunology rather than pediatricsand this new sphere of research was being invigorated in productive new directions by the rise of a new metaphor. In rhetorical terms, we can conclude, the movement facilitated by the metaphor unsettled the stasis (which, recall from above, can be translated as stoppage point or lack of movement) that had circumscribed infantfeeding debates for several decades prior to this. Specically, my analysis has indicated, when human milk became afliated with the emerging metaphor of the complex system, pediatrics researchers were able to make more convincing arguments about it. This afliation forced medicine to adopt a more expansive view of human milk: rather than just a feeding substance whose nutritional qualities could be reproduced in an articial substance, it came to be more highly respected as an object of scientic research.

Implications and unanswered questions As Paul Starr observes in his history of U.S. medicine, infant feeding is an area in which it could be argued that medical progress has done more harm than good. From Starrs perspective, the recent explosion of medical knowledge about breastfeeding has, paradoxically, produced scientic progress only by pushing medical attitudes back to what they were more than a century ago, before commercially produced formulas were widely available.55 The detailed picture this article has provided of the scientic events that Starr criticizes is important for several reasons. First, from the perspective of medical practice guidelines, the shift in beliefs that has resulted from increased understanding of human milks immune qualities has given valuable support to the small minority of people in the medical community who had long believed in the advantages of breastfeeding over bottle-feeding. As I argue elsewhere,56 mid twentiethcentury medical discourse was always able to justify ambivalence toward breastfeeding by
54 55 56

Ibid., 142. Starr, The Social Transformation of American Medicine.

The Battle of the Breast: Communicating U.S. Infant-Feeding Norms, 19402000. Book manuscript in progress. Springer


J Med Humanit (2006) 27:151166

claiming that there was a lack of certain knowledge: there were no facts to prove human milks superiority, so medical discourse would not take a denitive stance in support of breastfeeding, and instead perpetuated the idea that formula is, or eventually could be, an equivalent substitute. The increased medical understanding of how antibodies in human milk operate has helped breastfeeding advocates argue convincingly that human milk cannot be replaced by any articial substitute. It poses a serious challenge to the long-standing assumption that the food womens bodies could provide to their infants was less than perfect, and therefore, could be imitated and even improved upon through scientic research. The new medical perception of breastfeeding also resonates with popular images of health and wellness, which, Martin observes, are coming to be increasingly understood as dependent on a well-functioning immune system: The underlying common regard for the immune system in our culture allows it to become a kind of currency in which healthdegrees of it can be measured and compared among different people and populations.57 In fact, Martin mentions in passing that breastfeeding is one of several examples of phenomena [that] have been given new interpretations and understandings as the immune system has become more deeply entrenched in our collective understandings of the body and diseases that threaten it.58 Second, this articles historical analysis has important implications for recent efforts to develop a feminist discourse of breastfeeding.59 Working from various disciplines, such researchers have recently made great strides toward reversing the long-standing tendency to devote less feminist attention to breastfeeding than to other aspects of womens reproductive health.60 This articles case study contributes to such feminist efforts by examining breast milk as a unique example on which contemporary views of immunity have played out. When human milk became recognized as part of the immune system, it was no longer seen as something unique to womens bodies, but instead began to be seen as central to the human immune system, which has become a subject of intense scientic fascination in recent years. From a feminist perspective, it might not be just a coincidence that after this shift in perception occurred, breastfeeding suddenly gained more respect in the medical community. On the other hand, feminist scholars would be wrong to think that the prestige breastfeeding has gained through afliation with immune-systems knowledge is enough to make it a truly integral part of medicine or the culture at large. Certainly, the American Academy of Pediatrics 1997 policy statement stipulating that breastfeeding is now the infant-feeding norm pays lip service to this ideal. This statement received wide media coverage and enjoyed a high prole in the medical community. It is also cited in other politically inuential documents such as the federal governments Healthy People 2010 objectives, which call for increasing the percentage of breastfed newborns to 75 percent by 2010,61 and the Department of Health and Human Services Blueprint for Action on Breastfeeding.62

57 58 59 60

Martin, 183. Ibid., 183. Hausman, Mothers Milk: Breastfeeding Controversies in American Culture, 193.

See, for example, Blum, At the Breast: Ideologies of Breastfeeding and Motherhood in the Contemporary United States; Carter, Feminism, Breasts, and Breast-feeding; Galtry, Extending the Bright Line: Feminism, Breastfeeding, and the Workplace in the United States; Hausman; Law, The Politics of Breastfeeding: Assessing Risk, Dividing Labor.
61 62

U.S. Department of Health and Human Services, Healthy People 2010.

U.S. Department of Health and Human Services, Ofce on Womens Health, HHS Blueprint for Action on Breastfeeding. Springer

J Med Humanit (2006) 27:151166


However, the 1997 statement and all of the attention to breastfeeding it has sparked have not necessarily made breastfeeding a more normal activity for women, in any sense of the word. Rather, I would argue, it has led to a strange and contradictory situation for breastfeeding women in the U.S. today. Thanks to the increased status of breastfeeding as an object of health knowledge, women experiencing breastfeeding difculties have more access to expert assistance than ever before. In fact, such women can call a 1-800 number sponsored by the federal government to receive support and answers to their questions.63 However, the same women can also expect impolite stares or even a request to leave if they attempt to carry out this normal activity in a public place.64 Furthermore, as Bernice Hausman asserts, medical practice is not necessarily keeping pace with the ofcial discourse of the AAPs 1997 statement. Specically, Hausman argues, bottle-feeding continues to be the normative model for infant feeding, in terms of how infant feeding is approached in routine pediatric practice.65 Thus, from a feminist perspective, a key question that remains unanswered in this article is why there is not stronger support for breastfeeding in medical practice (as opposed to medical discourse) and in society at large? In the absence of such support, breastfeeding will continue to be an activity at which individual women are expected to succeed or fail, rather than a truly normal part of life. If the scientic knowledge about human milks immune benets continues to accumulate, this will increase the pressure for women to succeed at breastfeeding, but will not remove the many obstacles they face.

American Academy of Pediatrics, Task Force on Infant Feeding Practices (1984) Report of the Task Force on the Assessment of the Scientic Evidence Relating to Infant-Feeding Practices and Infant Health. Pediatrics 74(Supplement): 579762 American Academy of Pediatrics, Work Group on Breastfeeding (1997) Breastfeeding and the Use of Human Milk. Pediatrics 100: 10351039 Apple RD (1987) Mothers and Medicine: A Social History of Infant Feeding, 18901950. University of Wisconsin Press, Madison Baake K (2003) Metaphor and Knowledge: The Challenges of Writing Science. State University of New York Press, Albany Blum LM (1999) At the Breast: Ideologies of Breastfeeding and Motherhood in the Contemporary United States. Beacon Press, Boston Carter M (1988) Stasis and Kairos: Principles of Social Construction in Classical Rhetoric. Rhetoric Review 7(1): 97112 Carter P (1995) Feminism, Breasts, and Breast-feeding. St. Martins Press, New York Gage JT (1984) An Adequate Epistemology for Composition: Classical and Modern Perspectives. In Connors RJ, Ede LS, Lunsford AA (eds.) Essays on Classical Rhetoric and Modern Discourse. Southern Illinois University Press, Carbondale, pp 152169 Galtry J (2000) Extending the Bright Line: Feminism, Breastfeeding, and the Workplace in the United States. Gender and Society 14: 295317 Gerrard JW (1974) Breast-feeding: Second Thoughts. Pediatrics 54(6): 757764

63 The Information, Referral and Breastfeeding Helpline is sponsored by the National Womens Health Information Center. It can be reached at 1-800-994-WOMAN. For more information, see http://www.4woman.gov. 64 On November 10, 2003, a Burger King employee at an Orem, Utah, restaurant asked a customer who was breastfeeding her baby to leave. The employees request was reportedly a response to the remarks of another customer who claimed to feel uncomfortable with the mothers public breastfeeding. Burger King later issued an ofcial apology. See Burger King apologizes for the full story. 65

Hausman, 23. Springer


J Med Humanit (2006) 27:151166

Gyorgi P (1953) A Hitherto Unrecognized Biochemical Difference between Human Milk and Cows Milk. Pediatrics 11: 98 Hanson LA (1961) Comparative Immunological Studies of the Immune Globulins of Human Milk and of Blood Serum. International Archives on Allergy 18: 241267 Hanson LA, et al (1988) Antiviral and Antibacterial Factors in Human Milk. In Nestle Nutrition Workshop Series 15, Biology of Human Milk. Raven Press, New York, NY, pp 141157 Hanson LA, Johansson BG (1962) Immunological Characterization of Chromatographically Separated Protein Fractions from Human Colostrum. International Archives on Allergy 20: 6579 Hanson LA, Winberg J (1972) Breast Milk and Defence Against Infection in the Newborn. Archives of Disease in Childhood 47: 845848 Haraway D (1991) The Biopolitics of Postmodern Bodies: Constitutions of Self in Immune-system Discourse. In Simians, Cyborgs, and Women: The Reinvention of Nature. Routledge, New York, pp 203230 Hausman BL (2003) Mothers Milk: Breastfeeding Controversies in American Culture. Routledge, New York Jerne NK (1974) Towards a Network Theory of the Immune System. Annales dImmunologie 125: 373389 Kenny JF, Boesman MI, Michaels RH (1967) Bacterial and Viral Coproantibodies in Breast-fed Infants. Pediatrics 39.2: 202213 Kirschner L, Maguire T (1955). Antileptospiral Effect of Milk. New Zealand Medical Journal 54: 560564 Latour B, Woolgar S (1986) Laboratory life: The Construction of Scientic Facts. Princeton University Press, Princeton, NJ Law J (2000) The Politics of Breastfeeding: Assessing Risk, Dividing Labor. Signs: Journal of Women in Culture and Society 25: 407450 Lemons PM, Lemons S, Lemons JA (1986) Breast-feeding the Premature Infant. Perinatal Nutrition 13(1): 111122 Martin E (1994) Flexible Bodies: Tracking Immunity in the American Culturefrom the Days of Polio to the Age of AIDS. Beacon Press, Boston Ogra SS, Weintraub D, Ogra PL (1977) Immunologic Aspects of Human Colostrum and Milk. The Journal of Immunology 119(1): 245248 Prelli LJ (1989) A Rhetoric of Science: Inventing Scientic Discourse. University: South Carolina Press, Columbia Rosenberg CE (1976) No Other Gods: On Science and American Social Thought. Johns Hopkins University Press, Baltimore Sabin AB (1950) Antipoliomyelitic Substance in Milk of Human Beings and Certain Cows. American Journal of Diseases in Children 80: 866867 Starr P (1979) The Social Transformation of American Medicine. Basic Books, New York Tomasi TB Jr (1979) Introduction. In Ogra PL, Dayton DH (eds.) Immunology of Breast Milk. Raven Press, New York U.S. Department of Health and Human Services, Ofce on Womens Health (2000) HHS Blueprint for Action on Breastfeeding. Government Printing Ofce, Washington, D.C. U.S. Department of Health and Human Services (2000). Healthy People 2010. Government Printing Ofce, Washington, D.C. Vahlquist B (1958) The Transfer of Antibodies from Mother to Offspring. Advances in Pediatrics 10: 305338 Winberg J, Wessner G (May 29, 1971) Does Breast Milk Protect Against Septicaemia in the Newborn? The Lancet : 10911094 Wolf JH (2001) Dont Kill Your Baby: Public Health and the Decline of Breastfeeding in the 19th and 20th Centuries. Ohio State University Press, Columbus