I wish to begin by thanking the seven scholars who kindly consented to reading and commenting on my paper. The especially broad range of suggestions and criticisms will prove highly useful in my ongoing study of medical slang and points to future lines of research in the area of language and health.

Donald Pollock highlights identifying medical slang for health care institutions in addition to that used for patients and other physicians. Since this pertains to my comments on the work by Gordon, I should take advantage of the opportunity to say how important his study on hospital slang in California was for my own interpretation. I observed that Gordon concentrated on medical slang for patients and not (additionally) for health care institutions themselves. This should be taken more as an observation than a criticism. As Pollock notes, either there may have been little or no medical slang for health care institutions in the USA in the early 1980s, or Gordon chose not to focus on the issue, i.e., he did not necessarily overlook it. Although I disagree with some of Gordon's conclusions (see below, in response to Trostle), I agree with him on two central premises, without which the rest of the analysis becomes fruitless for either his paper or mine: that medical slang exists as a linguistic entity amenable to study, and that it has a relevant bearing on the medical ethos.

Regarding slang and changes in health care institutions, although I did not interview any physicians working in the United States, I have received a number of pertinent jokes circulating on the Internet, especially jests concerning health maintenance organizations and the conditions imposed on access to treatment from them. Interestingly, although the original authors are anonymous, the jokes have been sent or forwarded by American physicians. I am more familiar with the health field in Brazil, but an interesting aspect in the United States (or for a comparative study) might be to investigate whether changes are occurring in physician discourse that reflect what I understand to be the radical adoption of cost-effectiveness analysis in choice of treatment (with all its potential benefits for various segments of society).

James Trostle, also in regard to Gordon's study, is correct in pointing out that comatose patients can actually 'claim' treatment in the literal sense of 'requiring' rather than 'demanding' it. Still, it is interesting that Gordon chose the polysemous 'claim' rather than the less ambiguous 'require' when he classified comatose patients together with those described by California hospital slang as whiners: crocks, cricks, bell-ringers, etc. In this sense, perhaps unintentionally, the original meaning of claim, clamare 'to cry out' tends to intrude figuratively on the term as used.

Without intending to beleaguer Gordon's study, but to respond to Trostle again, I found the study methodologically meticulous, but emphasize a difference in our conclusions. Gordon concludes that hospital slang serves primarily to promote rapport among the health care staff and explicitly plays down any callousness or social bias it may connote. I agree that the phatic function of hospital slang can serve as one means, among others, for promoting rapport among health personnel, but contend that this is only half the story. I suggest that, far from either encouraging or censuring medical slang, we accept that some of it may be, for lack of a better term, "politically incorrect" (Johnson, 1998) and attempt to understand what it means for medical practice. Such an understanding might ultimately foster more humane health care, benefiting both patients and physicians. Although the California and Carioca tropes for patients bear noteworthy similarities, the divergent conclusions may also reflect differences in medical practice in the United States and Brazil, the 15-year time lag between the two studies, and/or my personal bias, having worked in public health care facilities in Brazil.

Trostle helpfully requests a definition of certain terms. I use chiasmus as defined by Ducrot & Todorov (1979/83:277): "the repetition and simultaneously inversion of the relationship between two words in the course of a sentence". A variation on this pattern is used in several sayings concerning the medical fields, of which I gave only one or two examples because of limited space, e.g., "The clinician knows everything and solves nothing, while the surgeon knows nothing and solves everything." I found the form relevant to the discussion of acquisition of medical knowledge, since this form of inter-specialization jab is repeated by med students and residents during their training and provides a sort of opposition or cross-over between the various fields at a time when they are pondering over their future field of specialization. (Coincidentally, chiasmus is also a neuroanatomical term, referring to the decussation or crossing of two neural tracts.)

I also use paronomasia as defined by Ducrot & Todorov (1979/83:278): "the juxtaposition of words that have the same sound but different meanings", e.g., in my article, the case of the expert medical opinion, parecer, referred to ironically as parece ser ('it appears to be...'). This is one way puns are formed, the type Freud (1905:31) called klangwitz, or 'sound-jokes' (which he considered inferior to more elaborate word play). Another type of trope occurs when the two terms are not just phonetically similar, as in paronomasia, but where one and the same word is used in a figurative sense within the same context (a form of antanaclasis ­ Ducrot & Todorov, 1979/83:277), e.g., the play on the verb drenar, 'to drain' a patient. I call attention to this distinction because it highlights Trostle's other question about a confusing sentence concerning the connotative usage of slang. A layperson catching the shift of sound in a paronomastic pun might guess the meaning, or at least infer that some joking is going on. But it would be difficult for him to discover that to 'drain' a given patient is to transfer him to the doctor's private office. Since the words are not just similar, but identical, the situation is perhaps one where connotation is most heavily context-dependent.

I use catachresis not in the standard sense of misuse or strained use of words, but concurring with Black (1962:32-33), as "the use of a word in some new sense in order to remedy a gap in the vocabulary; catachresis is the putting of new senses into old words". As Black points out, when "catachresis serves a genuine need, the new sense introduced will quickly become part of the literal sense."

My article opens and closes with a metaphorical exercise posing a conundrum: do all metaphors 'die', i.e., tend inexorably to catachresis? (By the way, the opening actually happened, while the ending is fictional.) Catachresis is exemplified by clavicle, coined by an ancient anatomist to plug a gap in the lexicon (since there was no name for that bone), who at that very moment 'put a new sense' into the old meaning of 'little key'. Clavicle soon became a dead metaphor and for centuries has required an etymological exercise to unveil it. To the extent that what is signified is seen as 'natural', the metaphor dies. Live metaphors, on the other hand, are those which do not merely 'plug gaps', but which continue to express double or complex meaning and provoke surprise, amusement, discomfort, disagreement.

Trostle suggests an important proviso to my approach: I have encompassed under the term 'medical slang' what actually constitutes not a distinct register, but a collection of terms culled from different styles of medical discourse. Future research should focus on what Labor (1972:186) proposes as a superordinate/subordinate hierarchy amongst styles. In the case of medical discourse, this might place formal scientific, deontological, and clinical discourse at the former pole and hospital-locker-room and out-of-patient-earshot conversation at the latter. What I have classified thus far as slang would be found primarily towards the subordinate pole, but interspersed elsewhere along the gamut. For example, medical conferences and clinical case reviews could be expected to feature scientific, sometimes hypercorrect speech, but might also be heavily punctuated with jokes, not only for didactic purposes, but also since the participants would perceive themselves as being among peers. Situational shift between different speech styles in the operating room was described nicely but indirectly by one of the anesthesiologists I interviewed, and it would be important to directly record this and other physician discourse styles (an important reference is Mishler, 1984). In addition to consideration for observational bias, more complex ethical implications are obviously involved in such direct recording of medical procedures, with due respect for physician, patient, and family consent.

Three of the discussants highlight concepts particularly relevant to slang during medical training. Pollock refers to what he calls physicians' 'training tales', and here I am reminded of having obtained metaphorically rich responses from my interviewees by asking them to describe their first shift on ward duty or whether they had ever felt afraid while providing care, following suggestions by Labov (1972: 180-83) on types of questions likely to elicit shifts in speech style. Quoting Good, Knauth points out that learning medicine is like 'learning a foreign language' and suggests situating slang within this more overall context of medical discourse. And Deslandes suggests that the derogatory terms often used for med students are part of the humiliation they experience in the 'rite of passage' during training.

Deslandes and Carrara question how the terms mulambo and mulambulatório help to create exclusion, or how the semantic and social levels interact. When a physician refers to given patients as mulambos rather than by their proper names or as 'the patient' or metonymically as an organ or disease entity, and when the same physician states that the patients will be treated in a mulambulatório, based on a performative analysis of this speech act (Searle, 1979:75-116), the physician has performed both 'expression' (expressing his feelings and attitudes) and 'assertion' (telling others how things are, in this case the patients and the treatment facility). The term mulambo is not a random utterance from one individual in society to describe another. A physician's assertions are invested by the state and society with scientific and moral authority, and the patient confirms this authority (although not necessarily uncritically) by seeking treatment from him. The speech act does not have to be a necessary or sufficient condition in order to 'help create' (i.e., serve as a cofactor for, or contribute to) patient exclusion or discrimination. Such speech acts would only fail to help create exclusion if they were totally divorced from the accompanying medical acts. If we could extend this Debate, and hopefully we can in another forum, I would ask Deslandes and Carrara to explain how such total estrangement might be achieved between speech acts and other acts. We might even productively consider, based on Searle (1979:85-93) that truth assumptions about the metaphor's literal meaning need not hold in order for it to function as such. That is, suppose that a mulambo does not 'really' have the properties we associate with a 'rag' discarded, inert, silent but rather is active and vocal. This lack of 'truth' (i.e., unequivocal denotation) in the literal meaning of the term would not alter what the speaker means by the metaphorical utterance.

It can also be argued, as Carrara suggests, that the physician is doing his best to treat the patient under adverse conditions, and that, as both he and Deslandes suggest, the term mulambulatório may be a critique of the physician's 'raggedy' working conditions. Both discriminatory assertion and exposé of medical working conditions may operate in this same metaphor. While the physician is apparently attempting to place a distance between himself and this context, his extension of the term mulambo to mulambulatório ends up naming the entire process, encompassing the line of low-income patients, the clinic itself, and the heath care staff, including, most importantly, himself.

Carrara also notes correctly that racial bias cannot be inferred from the African origin of the term mulambo, although, coincidentally, it labels low-income or homeless Brazilian public outpatients amongst whom people of African genetic origin are heavily overrepresented. He points out properly that Brazilian Portuguese has many other etymologically Bantu words, and that, besides, people do not walk around with an etymological dictionary in their heads, and generally speak without thinking twice about where the words came from. For example, we could easily praise a skilled Brazilian surgeon by calling him a bamba, and etymologically we would be comparing him to a m'bamba, or Bantu expert or official (an absurd comparison, since Brazilians are obviously not Africans, and in addition, for historical reasons, there are precious few black surgeons in the country). But Carrara's objection raises the important point of whether racial differentiation in health conditions and health care is measurable in Brazil and the relationship this might bring to bear on a derogatory term for certain patients. In this sense his discomfort focuses attention on what is not said (or cannot be scientifically tested) about race and health in the country. Under the terms of a Presidential decree (Ministério da Justiça, 1995), a Working Group of experts was convened three years ago by the Ministry of Health to discuss and propose measures concerning health conditions and health care among the black population. The Group published a report with this aim, one of whose recommendations reads:

"Promotion of knowledge concerning the relationship between health care professionals and black patients, who are subject to a negative social stereotype, in order to identify inadequate forms of behavior and, consequently, the adoption of educational measures to correct them" (Ministério da Saúde, 1996:17).

I doubt whether there is any more than the Working Group members' own expert opinions to support this recommendation. But the main reason for this lack of evidence is suggested by the report itself (Ministério da Saúde, 1996:17), which also recommends:

"Inclusion of the item 'color' or racial identification on patient records and [administrative forms] in the Unified National Health System...so as to allow for an epidemiological analysis of racial/ethnic groups..."

In 1998, two years after the report was published, the Ministry of Health has still failed to act on the above recommendation, so that 'color' is only included on patient records in a few scattered cities, and not nationwide. An opportunity continues to be lost to provide valuable information (at zero cost) to the Unified National Health System. Omission of information on color/race has serious implications at both the individual clinical and epidemiological levels, for diseases where African genotype plays a determinant role (e.g., sickle-cell disease) or where there is research ongoing elsewhere in the world concerning race as a potential cofactor in differential expression of cardiovascular and other diseases. In addition, racial identification could provide important information on differential access to (or discrimination in) health care and for redressing inequality, if it should exist, through specific public policies. Lack of inclusion of 'color' on health care forms produces a sort of colorblind alley, where racial inequality is Presidentially decreeable but scientifically unsayable. Carrara notes correctly that it would be possible to test the specific point of physician bias by clocking examination times for white and black patients, but it should also be noted that current patient records would be useless for such a study, for the reasons discussed above.

Sérgio Carrara has corrected me on the origin of pitiático: not petit mal, but 'pithiatism', "a morbid condition curable by suggestion", a concept related to hysteria and attributed to Babinski (1857-1932) (Cunha, 1982:610). The outdated term has survived as a catch-all for patients perceived as malingering or overly vocal rather than those actually presenting with hysterical neurosis.

Ana Maria Canesqui makes several valuable methodological and bibliographical suggestions aimed at more extensive and in-depth contextualization of medical metaphor. My article discusses some forty tropes (amongst others excluded for lack of space) relating to a range of aspects in medical experience, classified tentatively into three thematic areas as discussed previously and compared with the Brazilian literature on the respective issues (e.g., Minayo on epidemic social violence, Machado on medical professionalism, etc.). Canesqui's comments, directed to medical professionalism and defense of physicians' corporatist interests, relate mainly to the third of these, physicians' relations to the health care system. Her suggestions are especially relevant to how medical slang contrasts with scientific discourse (as expressing physicians' exclusive command of "esoteric and abstract knowledge") and deontological discourse, governing the profession in its entirety and specificity. She is correct that differences in gender, generation, and specialization can be reflected in medical discourse and require further investigation. Anesthetists do indeed have special characteristics, two of which are that they 'control' the patient's vital functions during interventionist procedures and that, as compared to other specialized fields in Brazil, they have succeeded in negotiating relatively favorable working conditions with health maintenance organizations. In addition, I only interviewed physicians in Rio de Janeiro, so extrapolation to Brazilian medical practice as a whole is necessarily speculative. A more complete picture of the health field would also benefit from linguistic research with other actors, especially other health professionals and patients. With regard to the latter, Duarte (1986) provides a fascinating reading of working-class Brazilians' description of illness.

Maria Elizabeth Uchoa appears to have agreed with the thrust of my interpretation and has provided me with a clear synthesis of my own paper, for which I am very grateful.

Responding to Carrara's objection to the paper having been written in English: I wrote it originally in Portuguese and translated it into English at the request of Cadernos de Saúde Pública as part of the journal's larger editorial policy to increase the volume of its work reaching the non-Portuguese-speaking community. Translating the material out of Portuguese posed several problems, as Carrara notes, since it hinges so extensively on culturally impregnated tropes. But this merely compounded two previous problems. First, even in the original language, live metaphor cannot be simply 'translated' into its roots. As Ricoeur notes, metaphor produces a 'surplus of meaning' making it more than the sum of its parts. Second, and prior to interpreting the material, medical slang had to be transposed from a spoken to a written medium, with an unavoidable loss of intonation and other prosodic traits. So even before tackling my interpretation of medical slang per se, the reader is forced to make this triple leap of faith. Hopefully the resulting debate will have compensated for the effort.

I particularly thank Carrara for referring me to my original interest in the theme of medical metaphor. While doing written and simultaneous translation of medical topics in general, I noticed that plays on words were almost always 'untranslatable' (to the point of defying the laws of physics in the case of simultaneous translation, i.e., two ideas attempting to occupy the same space at the same time), but that such tropes were rich in meaning in a way that literal speech and catachresis were not. Therefore, it was the challenge of translation that originally motivated me to write this paper.

The comments by Nancy Flowers, although sent by her on August 5, for some reason only reached me today (Nov. 19), when I am told that this issue of Cadernos is about to go to press. I am happy that her review will be included, but must apologize for my hasty acknowledgment.

Flowers focuses precisely on what I have asserted to be the most relevant aspect of medical slang, that it expresses changes occurring in the health care system. I am particularly grateful to her for including references I had not accessed previously and which I am anxious to read. Her reference to Coombs and his conclusions regarding medical slang at different stages in physicians' careers are consistent with my observations of jokes and sayings during medical training, although I did not raise the specific point of trends across generations. And the reference to Konner touched a personal note, since I also studied medicine in my thirties.

References

DUARTE, L. F., 1986. Da Vida Nervosa nas Classes Trabalhadoras Urbanas. Rio de Janeiro: Jorge Zahar.

JOHNSON, E., 1998. Political correctness. In: Encyclopedia of Applied Ethics (R. Chadwick, ed.), pp. 565-578, San Diego: Academic Press.

LABOV, W., 1972. The study of language in its social context. In: Sociolinguistics (J. B. Pride & J. Holmes, eds.), pp. 283-307, London: Penguin Books.

MINISTÉRIO DA JUSTIÇA, 1995. Decreto Presidencial de 20 de novembro de 1995. Coleção de Leis da República Federativa do Brasil, 187:5124-5126. Brasília: Ministério da Justiça.

MINISTÉRIO DA SAÚDE (MS), 1996. Mesa Redonda sobre a Saúde da População Negra: Relatório Final. Brasília: MS.

MISHLER, E. G., 1984. The Discourse of Medicine: Dialectics of Medical Interviews. Norwood: Ablex Publishing Company.

SEARLE, J. R., 1979. Expression and Meaning: Studies in the Theory of Speech Acts. Cambridge: Cambridge University Press.