The problem of theory and practice in the medical profession 1

Scientific knowledge is a symbolic system consisting o f hypotheses, models and theories generated by means o f a paradigm-mediated interaction between a scientific community and a research domain. Such a knowledge generating paradigm consists o f already existing theories, as well as methodological and ontological beliefs or assumptions. In this article it is argued that the meaning ascribed to the central concepts o f medical science (such as patient, disease, causality and therapy) are fundamentally determined by the 19th century logical positivist scientific paradigm. The ontological and methodological implications o f the postmodern natural sciences (e.g. quantum physics) have not been applied to medical science. The 19th century ‘natural science paradigm ’ therefore acts as a meta­ theory fo r both medical science and medical practice. However, the theo­ retical knowledge system generated by medical science acts as the theory fo r the practice o f scientific clinical medicine which therefore functions with the same understanding o f the central concepts such as patient, disease and disease causality, therapy etc. The limitations o f this paradigmatic monism are illustrated by an analysis o f the medical and societal response to the AIDS epidemic and it is concluded that medical science and practice, because o f the complexity o f its research and practice domain, must accept in principle the possibility o f paradigmatic pluralism (as in the social sciences) or should attempt to develop a holistic paradigm that will cope more adequately with its fields o f research and practice.


Introduction
Scientific knowledge is a symbolic system generated by means o f a paradigmmediated interaction between a researcher or research group and an aspect of reality (the research domain).The paradigm consists of: There is no possibility o f a paradigm-free, direct access to 'reality'.The para digm functions very much as a world view for its adherents in that it has a regu lating function on their perception o f the world.It becomes institutionalized in social institutions such as university research departments, scientific societies, scientific journals, industrial research organizations, etc.As an institutionalized scientific world view it influences decisions by the community regarding the pro blems that will be investigated, how they will be investigated and w hat type o f solutions (i.e.'theories') will be considered to be 'scientific', i.e. w hat that com munity will consider to be valid scientific knowledge.
The paradigm therefore also has a fundamental influence on research funding either by scientific institutions themselves, or, through the political influence o f the institutions o f science, by government and other societal funding agencies.
Because 'scientific research' has become the accepted form o f rationality in W estern societies and the accepted basis for problem solving in all spheres, the scientific paradigm will have an influence and impact much w ider than the scien tific community.
The paradigm therefore has an influence not only on 'theoretical science' but also on its application in society by means o f technology, professional practice, poli tical decisions, etc. in so far as these claim to be based on 'scientific know ledge'.
In this essay I wish to explore the relationship between scientific know ledge and professional practice with specific reference to the medical profession.In order to do so I will

From neo-Aristotelianism to the natural science paradigm: an historical overview
A paradigm change involves changes not only in the theoretical beliefs ('know ledge') o f a scientific community, but more importantly in the methodological beliefs and ontological assumptions.This is what happened during the Enlighten ment when the neo-Aristotelian paradigm that dominated the intellectual life of the Middle Ages and formed the conceptual framework in terms o f which theorizing took place about the natural world (physical, biological), as well as the social and spiritual (religious) aspects o f life, was slowly replaced by w hat is now called the natural science paradigm.
In medical science this paradigm change w as signaled by the replacement o f the humoural theory o f disease by the cellular theory o f disease first comprehensively formulated by Virchov at the end o f the 19th century.
Humoural pathology postulated that the human body w as made up o f four humours namely white bile, black bile, blood and phlegm.Health w as a state o f harmony between the humours and disease a wrong mixture o f them.Therapy consisted o f intervention to restore the balance.The theory w as an application o f the theory o f the four elements (earth, water, air and fire) to the working o f the body.The function and structure o f the body was interpreted in terms o f the w ider cosmology -the body w as a mirror (or microcosm) o f the cosmos, the macrocosm.Verbrugh (1974:35) considers this theory to have been a major scientific achievement because a wide variety o f phenomena could be understood and controlled within its framework.This theory served medicine from Greek times to the 17th century.
According to the cellular theory o f disease, disease is located in malfunctioning at the cellular level in one or more organs.This is the basic theory o f disease with which scientific medicine still operates today.
The major difference between the neo-Aristotelian and the modem explanatory strategies lies in their understanding o f movement and o f causality.For Aristo Koers 59 (2) 1994:193-219 telians the beginning o f a process (the efficient cause) w as not as significant for understanding the present configuration as w as the final cause.The typically mo dem assumption is that something is explained once it can be explained in terms o f initial conditions and forces and factors that interact according to natural laws.Something has been explained once it has been described as a mechanism.
Because o f the explanatory and predictive pow er o f the developing natural science paradigm it w as applied to the biological or life world on the assumption that the structure o f the life world was intrinsically the same as that o f the physical world and that the general principles o f the physical sciences were adequate to access this aspect o f reality as well.This meant that both the physi cal and biological sciences began to operate with basically the same paradigm.
Although it originated as a paradigm to explain the physical world, it now became a natural science paradigm.The natural sciences therefore becam e characterized by a paradigmatic monism.
The natural science paradigm came to be regarded as defining the essence o f scientific rationality and even o f rationality as such.Scientists therefore began to apply this paradigm to the developing human and social sciences as well.
However, during the first part o f the 20th century the limitations o f this natural science paradigm with regard to the human world becam e clear and several other paradigms were developed (e.g.phenomenology and hermeneutics) by different scientific communities working within the human sciences, thus establishing the paradigmatic pluralism presently characterizing the human sciences.
1.1.4P a ra d ig m a tic monism vs. p ara d ig m a tic pluralism If the paradigmatic monism o f the natural sciences is taken as normative for scientific rationality, then the paradigmatic pluralism o f the social sciences often leads to the conclusion that the social sciences are still in a pre-paradigm atic phase o f development because o f the fact that there is no discipline in which there is a single dominant paradigm.
M outon and M arais (1988:150) argue that this conclusion is incorrect and is due to an application o f Kuhn's paradigm concept out o f context.Kuhn links the con cept o f a paradigm very strongly to that o f problem solving which is the major epistemic goal in the natural but not in the social sciences.They suggest that the term paradigm should therefore be understood more in a metaphorical sense when applied to the social sciences.There is no reason in principle for demanding pa radigmatic monism as a feature o f scientific rationality.The present monism cha racterizing the natural sciences is the result o f 'political activity' in the social institutions o f science, and is not a requirement o f 'scientific activity' as such.
At the beginning o f the 20th century, physics underwent a paradigm change in the development o f quantum physics and its related sciences.Toulmin refers to the natural sciences that operate within the paradigmatic framework o f quantum mechanics as 'postm odern sciences', while the natural sciences which operate within the Newtonian paradigm he calls 'm odem sciences' (Foss & Rothenberg, 1988:164).
The physical sciences therefore also essentially have a paradigmatic pluralism.
Although the postmodern insights are beginning to be incorporated in the frame work o f thinking o f the biological sciences, it has had little impact on medical science or practice.Systems theory seems to me to represent a possible alter native paradigm for the biological sciences.These developments have been set out diagrammatically in Figure 1.The gradual process by which medical science changed from the neo-Aristotelian humoural theory o f disease to the modem cellular theory o f disease, can be com pared to the gradual change from the Ptolemaic cosmology to the Copemican heliocentric theory.Foss and Rothenberg (1988) refer to this change as medici n e's first revolution.They suggest that medical science did not apply the impli cations o f the quantum revolution in physics and that it therefore continues to operate within the 'm odem paradigm '.They therefore explore the implications for medicine o f what they call 'the theory o f self-organising system s' (based on postmodern sciences) as an alternative paradigm referring to this as m edicine's second revolution.The life sciences are therefore entering a situation o f para digmatic pluralism as well.The human sciences have thus actually been leading the w ay in that they have been exploring the possibilities inherent in paradigmatic pluralism since the first quarter o f this century.
Koers 59 (2) 1994:193-219 It is now necessary to clarify the characteristics o f the natural science paradigm and then to show how the acceptance o f this paradigm by medicine affected the m ajor concepts in medical science.

T he natural science paradigm
As set out in section 1, a paradigm consists of: * Methodological beliefs (directives by which to approach enquiry) and * Ontological beliefs (directives by which to approach the world).

Methodological beliefs
The standard view o f science has changed with time following its first tentative formulation in the w ork o f Bacon.The Enlightenment ideal and understanding o f science as a method o f gaining knowledge o f reality as it exists objectively, cul minated in the logical positivism o f the W iener Kreis at the beginning o f this century.
Van Huysteen (1986:16) refers to logical positivistic as the first systematically formulated philosophy o f science and states that it is still accepted by many scien tists today.It definitely represents, I believe, the view o f science held by most medical scientists and medical practitioners working in the RSA today.I believe that it also represents the view o f science o f the éducated general public because it represents the underlying epistemology o f the academic school curriculum and thus influences the view o f scientific knowledge held by all school leavers.
The basic tenets o f the logical positivistic view on scientific methodology can be explained in terms o f the following concepts: empiricism, verification, the unity o f science, objectivity and scientific method.

Empiricism
A basic postulate o f the logical positivism is that all real know ledge is based on 'direct observation' which includes observation mediated through various forms o f instrumentation (Van Huysteen, 1986:17).It postulates that theories can be based on theory free observation and that this enables an understanding o f reality as it exists objectively i.e. independently o f the human observer.
Sallie M cFague (quoted by Van Huysteen, 1986:15) characterizes the logical po sitivism very succinctly as "... the assumption that it is possible to go directly from observation to theory without the critical use o f models".
Logical positivism thus believes that it is possible to interact with a research do main directly on the basis o f observation without the critical mediation o f a para digm.M ethodological beliefs are considered to be self-evident truths abstracted from the practice o f the most advanced o f the physical sciences namely physics.These beliefs guide the practice o f 'positivistic science' and thus fundamentally influence the type o f theories that can be generated and will be recognized as valid.
These methodological beliefs imply corresponding beliefs regarding the ultimate structure o f the research domain under investigation i.e. methodological beliefs imply ontological assumptions.

Principle o f verification
In The ontological assumption accompanying these methodological beliefs is that reality is structured according to mechanistic principles expressible in terms of linear causality.

Unity o f science
Implied in this empiricist view o f knowledge is that all sciences (i.e.scientific dis ciplines or 'vakw etenskappe') must have the same epistemological structure.A further implication is that all scientific knowledge must (eventually) be trans latable to the language o f the physical sciences and to that o f physics in particular (reductionism).

Objectivity
The term o b jec tiv ity plays an important role in the thinking o f natural scientists about their own practice.Within the logical positivistic conceptual framework, objectivity means knowledge obtained through the empirical method.It is thought to be free from any observer bias and a true (theory free) representation o f reality.Because it is 'theory free observation', scientific findings cannot be in fluenced by personal or cultural beliefs and values.
Reality is therefore assumed to be directly accessible through the scientific method which gives objective knowledge which is value-free, free o f observer bias and therefore impersonal, a-cultural, generally valid, universal and timeless.

Scientific method
Scientific rationality is thus identified with a specific method.This method was so successful, that proceeding in accord with it came to be identified with pro ceeding in accord with the precepts o f rational enquiry in general and theoretical science in particular (Foss & Rothenberg, 1988:45).
Objectivity is guaranteed by the application o f this specific method.Van Niekerk (1992:51) calls it "a stabilized procedure" which imposes a strict cognitive disci pline on the researcher and legitimates it as 'scientific'.
The method delimits w hat type o f questions can be asked and w hat types o f answers can be given.The cognitive discipline ensures intersubjective validity.
The method is used to generate scientific knowledge.It is also used in the pro cess o f verifying scientific theories.Important is the fact that the method is pu blic, i.e. it is open to public scrutiny because it is known to and (potentially) repeatable by others.
Van Niekerk (1992:50-51) characterizes the method which came to epitomize 'scientific rationality', by the terms experim entation, inductive log ic, verification and th eo ry form ation .I will now discuss each o f these in turn.

* Experimentation
The stabilized procedure by which scientific knowledge is generated is said to start with the performance o f strictly controlled, public and repeatable (stabilized) observations called experiments (Van Niekerk, 1992:50).B y 'repeatable experi m ent' is really meant laboratory experiments under strictly controlled circumstan ces in which essential variables are limited to measurable ones.In the experiment the scientist is thought to be a detached observer who has no influence on the in exorable logic o f the events unfolding before her (Foss & Rothenberg 1988: 45).

* Inductive logic
These experiments 'establish the facts' on which the theories will be based through the application o f the rules o f inductive logic by means o f which the particular judgm ents o f observational language are developed into generalized hy pothetical statements (Van Niekerk, 1992:51).Through the application o f induc tive logic the regularities brought to light during the experimentation are formu lated in the form o f generaliseable hypotheses.

* Verification
The next step is to try to prove that the provisionally formulated hypothesis is true.This is done by finding further experimentally controllable observations which will verify the claims o f the hypothesis.(Popper's principle o f falsification is not really a fundamental departure from the general approach.)

* Theory formation
Scientific theories are therefore considered to be constructed on the secure foundation o f objective, value-free facts obtained by factual direct observation.If the testing (by verification or falsification) o f the empirical generalizations are successful, a new scientific law has been established which can be integrated in a cumulative manner in the growth o f scientific knowledge.In this manner all nonempirical discourse can be eliminated from scientific discourse.Van Niekerk (1992:53) summarizes this 'stabilized process' as follows: So groei die liggaam van wetenskaplike kennis deur die akkumulasie van ware feitelike en teoretiese oordele, dit wil sê van oordele wat op 'n alge meen geldige wyse die objektiewe werklikheid met sekerheid ter sprake bring, en waarvan die waarheidsgehalte kontroleerbaar is deur die rekonstrueerbaarheid van die waamemings waarop die teorie berus, sowel as van die logiese argumente wat die afleiding van hipoteses en teorie motiveer.

Ontological assumptions
The ontological assumptions and the methodological beliefs o f a paradigm form a mutually interactive network.The methodological process must be directly rela ted to the ontological reality that it is assumed to be able to access.To a certain extent, the (assum ed) structure o f the ontological reality is established (created) by the methodological beliefs and vice versa.
The ontological assumptions o f the natural science paradigm obviously grew in sophistication over time.So e.g. the concept o f a machine which underlies the mechanistic world view o f the modem natural sciences, grew in sophistication as the sophistication o f technology developed more and more complex machines.
The basic concepts under which I wish to describe the ontological assumptions of the natural science paradigm are: dualism, materialism, reductionism, linear (uni directional) causality and determinism.

Dualism
D escartes' dualism w as o f fundamental importance in the development o f science as it dichotomized the material and immaterial worlds, permitting the scientist to concentrate on aspects o f nature more readily subject to experimentation and quantification (Foss & Rothenberg, 1988:44).It thus formed (and in essence still forms) the essential background to empiricism.Gradually the emphasis shifted to the material aspects o f reality so that the immaterial came to be seen as a deri vative o f the material.

M aterialism
The natural world, which is the domain o f investigation o f the natural sciences, is assumed to be a complex material whole, com posed o f assem bled atomic parts which, when interacting with sufficient complexity, can form all the physical and biological systems that w e encounter in the universe.This means that only funda mentally material explanations, i.e. explanations formulated in physicalist terms, will be considered as 'true' explanations.
Newtonian mechanics sanctioned the idea o f organized complexity, such as found in biological systems, as in principle reducible to the interaction o f its physical parts (Foss & Rothenberg, 1988:24).
The world view implied by the natural science paradigm presupposes what Foss and Rothenberg (1988:52) call 'a single fundamental material level o f reality'.Higher levels o f organization do not involve ontologically new entities beyond the fundamental level elements o f which the given entity is comprised.Primary sta tus is thus conferred on matter and the material world and secondary or epiphenomenal status to the biological and mental worlds (Foss & Rothenberg, 1988: 51).This leads to the following characteristic, namely that o f reductionism.

Reductionism
Reductionism is a complex concept.It indicates, as mentioned above, that com plex wholes exist only in terms o f the interactions o f their most fundamental parts.This has methodological implications as acceptable explanation now becom es the process by which a complex whole is analysed into its constituent parts and its behaviour explained in terms only o f the interaction o f those parts.
Reductionism thus implies that there is ideally a unified, physicalist language in whose vocabulary all events, physical, biological and mental are ultimately ex pressible.Only this type o f explanatory discourse is thus legitimated by the natu ral science paradigm.
Reductionism and materialism are thus intimately linked in the natural science paradigm -life, the human being and culture are, in the last analysis transient events, by-products o f physical organization because interacting physical particles governed by the laws o f m echanics and thermodynamics are the ultimate stuff o f the universe.As late as 1981 Hull (quoted by Foss & Rothenberg, 1988:51) sta ted: Today, both scientists and philosophers take ontological reduction for gran ted.Vitalism is dead.Organisms are 'nothing bu t' atoms, and that is that.

Unidirectional causality and determinism
In the explanatory framework o f the natural science paradigm, all events are explained in terms o f antecedent events organized in causal chains and networks which are characterizable in terms o f universal laws which make no reference to the causal efficaciousness o f future events or higher levels o f organization.Therefore all explanations must be determinate explanations.
Because o f the fixation on paradigmatic monism by the natural scientific com munity, this type o f explanation was extended to biological and social pheno mena.Contemporary medical reasoning in turn has borrowed its causal concepts from the other theoretical parts o f science in accordance with currently accepted explanatory goals (Maull, quoted by Foss & Rothenberg, 1988:55).

T he biom edical model
Contemporary medical reasoning has not only borrowed its concept o f causality from the natural sciences, but has, by identifying itself as a natural science, taken over the whole o f the natural science paradigm.Like the practitioners o f the social sciences, medical scientists and practitioners found themselves "standing in the shadow o f the sciences as aspirants to some more perfect kind o f knowledge which the natural sciences in particular represent" (Max W artofsky, quoted by Foss & Rothenberg, 1988:55).
The biomedical model is the model o f medical science and medical practice which follows from the acceptance by medical science o f the natural science pa radigm.Because the term 'science' here specifically refers to the natural scien ces, the term science o f medicine could thus be defined as the symbolic know ledge system concerning the mechanisms, diagnosis, treatment or prevention o f disease, developed according to the methodological directives and the ontological assumptions o f the natural science paradigm.
By thus opting for a natural science framework as its only conceptual framework, biology in general and medicine in particular, gave up its autonomy to develop and employ sui g en eris modes o f explanation, or for employing more than one explanatory framework as is possible within the human sciences.
The commitment to the natural science paradigm commits the medical scientist to a view o f the relevant research domain as one that is materialistic, deterministic, reductionistic and characterized by causal mechanisms, and the assumption that it is possible to exercise control through understanding the inherently predictable outcome o f various actions.
In addition, adopting the natural science methodological strategies commits the medical researcher to attach the same importance to conducting enquiry in a manner that is objective, a-cultural, a-temporal and impersonal in medicine as in physics or chemistry.
The biomedical model o f medicine (or biomedicine) is thus the result o f applying the natural science paradigm as the only paradigm for understanding the human being in health and disease.It locks medical science and medical practice into both a hierarchy o f natural sciences and into the methodological constraints o f the 'scientific method'.

Central concepts of biomedicine
The natural science paradigm thus has fundamental significance for the meaning attached to the central concepts o f medicine namely that o f patient, disease, di sease causation and therapy.I will now discuss each o f these in turn.

* The patient
The patient is understood as being essentially a biological organism which, in the words o f Descartes, is "so built up and composed o f nerves, muscles, veins, blood and skin, (that) though there were no mind in it at all, it would not cease to have the same functions" (quoted by Foss & Rothenberg, 1988:24).This is today still the basic model o f the human being underlying the teaching o f both the basic sciences and the clinical sciences as taught at medical schools in the W estern world.
D escartes' dualism sanctioned the idea o f the body as a machine that can be analysed independently o f the mind and o f the social and cultural context o f the person.This means that the diagnostic process abstracts the patient from his or her concrete existence and assumes that "the patient, conceptualized as a biolo gical organism, can be diagnosed in a context-free environment" (Foss & Rothen berg, 1988:61).Cartesian dualism was later reduced to a monistic materialism in which the body became primary, and psychological functions becam e secondary epiphenomena o f bodily processes.The psychological and social dimensions o f human existence becam e irrelevant to the science and practice o f medicine.

* Disease and disease causation
In biomedicine disease is conceptualized as something physically wrong in an organ or organs which can be described as a deviation from the norm o f m easur able biological parameters (functional or structural).D isease is thus a material entity and can be completely described in physicalist language.Similarly the cau ses o f disease are physical factors: the presence o f too much or too little o f a cri tical substance or the presence o f an intrinsically harmful agent.
Since disease is biology gone awry, and since the biological organism can best be understood by an understanding o f its constituent parts and their physical inter actions, there is no need to delve further than the physical mechanisms o f disease.
In this manner the ultimate level o f explanation o f disease becomes the level of molecular biology.To the extent that disease is strictly interpreted in physical terms, both research energies and research funding will be focused strictly in this direction.

* Therapy
Against this background it is clear that therapy will also be conceptualized in purely physical terms as physical intervention (chemical, electrical or surgical) that will com pensate for the surplus or deficiency o f the critical substance or will neutralize the pathogenic agent.
From this analysis it is clear that the adoption o f the natural science paradigm by medical science has had definitive consequences for the ascribable meanings of scientific m edicine's central concepts.

Institutionalized forms of medical science
Although scientific medicine or biomedicine thus locates itself within the discipli nary matrix o f the natural sciences, there is no specific scientific discipline called m edicin e which would be an equivalent to the disciplines o f physics and chemis try in the physical sciences, or to botany and zoology in the biological sciences.
The scientific activities o f scientific medicine take place within a large variety o f separate disciplines which include the physical sciences (e.g.medical physics, medical biochemistry) and biological sciences e.g.microbiology, anatomy, phy siology, pathology, genetics, immunology etc., as well as in the clinical disci plines, e.g.surgery, paediatrics etc.These clinical disciplines tend to refer to themselves as 'clinical sciences' especially within the academic environs o f uni versity faculties o f medicine.
In spite o f the fact that there are fundamental differences between the research methodologies as applied in the clinical situation (e.g.'controlled repeatable ex perim ents' are in principle not possible), clinical researchers see themselves es sentially as 'natural scientists' applying the natural scientific empirical method.
In actual fact their research process is much more closely allied to certain re search processes used in the social sciences in which sophisticated statistical procedures have to be used in order to detect 'patterns o f behaviour'.
The research may be done within the context o f academic departments in university faculties o f medicine, or in industrial laboratories e.g. the pharmaceu tical industry, but all these researchers would identify themselves as natural scien tists and their work as exemplifying the methodological directives o f the natural sciences.This is the basis o f their claim to being 'scientific' M ost o f these disci plines are laboratory based and are often referred to as the basic sciences -i.e.basic to clinical medicine, which is seen as 'an applied science'.
The term m ed ica l scien ce also includes scientific research done in clinical dis ciplines usually within the context o f university hospitals related to medical schools.This research consists mainly in applying the theories and knowledge generated by basic science research to the clinical situation in order to test the va lidity or applicability in the clinical situation.However, such research may gene rate its own type o f knowledge system especially regarding clinical manifes tations, diagnostic methods and criteria, therapeutic regimes, risk factor identifi cation, etc.
This research is usually done as an integral part o f clinical practice (hence patient fears o f being 'human guinea pigs').M any clinical researchers have their own la boratories attached to their wards, or work in the laboratories o f their basic science colleagues, while many 'basic scientists' are qualified medical practi tioners.The distinction between basic science research and clinical research is therefore not clear cut, but both see themselves as 'natural scientists'.
Although the practice o f clinical medicine involves persons, the research assump tion o f scientific medicine is that clinical research can be done as if the researcher is dealing with the same domain as is dealt with in the natural sciences.
Clinical researchers therefore go to extraordinary lengths to eliminate any refe rence to the human subject on which and by which the research is being done in order to obtain 'objectively valid' knowledge.
Such objectively valid knowledge is thought to be obtainable, e.g. by the use o f the so-called placebo controlled double blind cross over therapeutic trials which attempt not only to eliminate researcher and patient bias, but also the so-called placebo effect i.e. the biological effect o f interpersonal relationships and other non-physical factors.
This methodology is used to compare the therapeutic efficacy o f a new drug or therapeutic regime against that o f the presently accepted regimes and/or against that o f a 'placebo'.A placebo is a pharmacologically inactive substance which is administered as if it were a pharmacologically active agent.The 'placebo effect' measures the degree to which doctor-patient interaction affects the patients' response in terms o f subjective feelings o f well-being.This is especially impor tant where the patients' responses are an important criterion for assessing the agent.The ideal is to dispense completely with patient responses and only to measure objective indicators o f disease such as biochemical or histological changes.The assumption is that the body is a machine in which the 'personal do main' cannot affect the responses o f the machine.
The concept o f double b lin d means that neither the patient nor the doctor knows whether the patient is receiving the agent being tested or the placebo.C ro ss o v e r refers to the fact that half way through the trial those patients who were on place bo start receiving the agent, and vice versa.Those researchers who know which agent is being administered to whom, do not come into contact with the patients at all.
Medical science is thus practised within the context o f a large number o f disci plines which are considered to fall within the ambit o f the physical sciences, the biological sciences and the medical sciences.This scientific practice is guided by the natural science paradigm, which could thus be said to be the theory which guides the practice.The term th eory used in this sense obviously means some thing different to 'theory' as a symbolic knowledge system generated by scientific activity.It is really a meta-theory.
W e have now reached the point where we can analyse the relationship between theory and practice as it presents itself in the medical profession as well as the relationship between the natural science paradigm and the clinical practice of medicine.

Clinical practice as scientific practice
Clinical medicine as a system o f healing distinguishes itself from other systems of medicine on the basis that it, and it alone, is 'scientific'.Other systems o f medi cine such as homoeopathy, chiropractic, traditional medicine etc. are considered to be non-scientific, and therefore o f no practical importance in health care.
The basis for the claim to being a scientific practice is twofold: Clinical practice is viewed as the direct application o f 'scientific know ledge' (scientific being un derstood as natural science) and the clinical method is seen as resembling 'the scientific m ethod'.

* Clinical practice as the application o f scientific knowledge
C linical practice is considered to be scientific because it ap p lies the scientific kno w led g e (theo ries) g enerated by m edical science in its varying institutional for Koers 59(2) 1994:193-219 mats as set out in section 3. The term science is therefore understood as 'natural science' and the meaning given to the central terms o f medicine namely disease, therapy, patient, etc. will be the same as in medical science.
Certain aspects o f medical practice may be known to be based on very scanty or very doubtful 'scientific know ledge', but the ideal is to base clinical practice on 'sound scientific principles'.It is assumed that the domain o f clinical practice is identical (or sufficiently identical) with the 'research dom ain' o f 'medical science' (as described above) for this to be done.Clinical medicine is thus as sumed to be a direct application o f a body o f (discipline based) natural scientific knowledge.Natural science (in the form o f medical science) therefore supplies 'the theories', and clinical medicine applies them to patient care.

* The clinical method represents the scientific method
The interaction between the scientific doctor and the patient is regulated by a specific method known as 'the clinical m ethod'.According to this method the doctor first systematically interrogates the patient regarding his or her symptoms (the subjective experiences o f 'being ill' as well as other information regarding family history, work history, social habits that may have a bearing on the still-tobe-diagnosed disease).This process is known as 'history taking'.From this in formation the physician formulates a hypothesis regarding the possible disease or diseases that may cause the symptoms.
The history is followed by 'the physical examination' in order to find physical signs that will confirm or disconfirm the hypothesis.Physical signs are 'objective physical findings' e.g.enlargement o f the liver, signs o f anaemia, the presence o f abnormal heart sounds, etc.This is usually followed by special investigations o f e.g. the blood or urine, or Xray examination, etc.The aim is now finally to find objective evidence o f disease, i.e. o f abnormal physical processes.Without such objective evidence, there is no disease, regardless o f the patient's story.In these cases the concepts o f neurosis or psychosomatic complaints are invoked.The 'objective evidence' may be bio chemical or anatomical (X-rays, biopsies), immunological, etc.The disease label is the one that best links together the history, the physical examination and the special investigation.This disease label is o f vital importance, because it deter mines therapy.However, a disease is not simply a label, it is something that actually exists objectively.
From this description it can be seen that the clinical method presupposes the whole set o f assumptions and beliefs that were described above as the biomedical model.
The application o f the clinical method through which a diagnosis is established and a decision regarding therapy is made in the case o f an individual patient, is considered to be an exemplar o f (natural) scientific rationality, and is e.g.likened to the 'scientific process' o f data collection followed by hypothesis formulation and hypothesis testing.
The analytic process by which clinical information leads to the diagnosis is closely akin to the scientific method -the process whereby experimentation leads to the discovery o f new knowledge (Harvey et a i , 1984:2).
Although the actual position o f Harvey et al. (1984) is much more nuanced than that suggested by this quotation, there are two assumptions operative under the surface o f this model o f clinical reasoning.The first is that 'the scientific method' is an analytical one and secondly, the assumption that the clinical reasoning pro cess is (or can be) modeled on the 'scientific rationality' o f the research process.
In this case too the term scien tific is understood as n atu ral science.

T h e in stitu tio n a lize d fo rm s o f m ed ica l p ra ctice
The medical profession is institutionalized as a large variety o f clinical specialties such as surgery, paediatrics, internal medicine, obstetrics and gynaecology, fa mily medicine, psychiatry, oncology and radiotherapy, etc.However, the labo ratory specialties such as chemical pathology, virology, histopathology, medical microbiology, nuclear medicine, etc. as well as the community specialities such as community medicine and epidemiology, are considered to be part o f scientific medical practice.Although aspects o f the last two may be dealt with by 'pure scientists' they are mostly practised by medical graduates.
Each o f these forms o f practice has its own 'theory' in the sense that e.g. a psy chiatrist would not be able to handle the practice o f a surgeon -and by that I do not only refer to the specialized skills.The psychiatrist would be lost in the theory that informs the practice.But 'theory' here simply refers to the scienti fically generated knowledge base which is applied in the  Verbruch, 1974:46).But this educational process too pre supposes the whole framework o f biomedicine (Kriel & Friedman, 1990).

Theory and practice in the medical profession
The theory for the practice o f medicine is thus supplied by the science o f medicine.This theory is the output o f the enterprise o f medical science which is a paradigmatic interaction between researchers and their research domain.The paradigm in question is the natural science paradigm.
The natural science paradigm as applied to medical science is called biom edicine.
Biomedicine thus forms the meta-theory which regulates both the practice of medicine and the practice o f medical education.M edical education thus repro duces the biomedical paradigm.These relationships can be set out diagrammatically as in figure 2.
A final question needs to be addressed.W hat is the research domain o f medical science?
Researchers approaching ostensibly the same reality from the perspective o f dif ferent paradigms, 'see' different realities.The research domain is thus partly pos tulated or established or, at the very least structured, by the research paradigm.
So what is the research domain o f medical science?Is it health, or is it the ill person, or is it biological deviations from the norm called diseases?From what w as said above, it is clear that the research domain o f biomedicine is specifically biological deviations from the norm.It is not the whole field o f health and illness.This has important implications because the medical profession projects itself as the sole societal interpreter o f the field o f health and illness while in fact its theories (which form the basis for medical practice) have a very limited appli cation.Furthermore, its institutions guard the paradigmatic monism o f the natural science paradigm with all the scientific and political power at its disposal.

Conclusion
The standard view o f medical science as a natural science which shares the ontological and methodological beliefs o f the natural sciences as a group, thus fundamentally influences medical science and medical practice.M edical practice sees itself (in the form o f the social institution o f the medical profession) as applying the medical knowledge o f medical science.The output o f the research activity o f medical science is a theoretical symbolic system which at the same time acts as the 'theory' for the 'practice o f m edicine'.The monistic paradig matic framework o f medical science, therefore also forms the paradigmatic frame work (or meta-theory) for medical practice.
Schon (1987:1-40) has criticized what he calls the positivistic theory o f practice which assumes that practice is nothing more than the application o f 'scientific theories'.He points out that in professional practice the professional is not faced with clear cut problems as defined by the scientific paradigm, but by messy situations out o f which a problem must be constructed.Basically this means that the 'research domain' o f the natural sciences never is (or can be?) co-extensive with the reality within which practice takes place.If practice is approached solely from the perspective o f the natural science paradigm, then a significant area o f the real world is excluded from the interaction betw een the professional and her client.

B iom edicine and the A ID S epidem ic: a case study
In this section I wish to illustrate very briefly the limitations o f the biomedical model in dealing with the problem o f the AIDS epidemic.I will focus mainly on the implications o f the biomedical understanding o f the cause o f AIDS.For a more detailed discussion o f the relationship between the biomedical point o f view and AIDS, see Kriel (1991).

The biomedical view o f AIDS
In line with the biomedical definition o f disease, AIDS is seen exclusively as pathology in an organ system (the immune system) caused by a physical agent, the HI virus.All research efforts (and funding) are focused on finding a physical cure (e.g. a drug) or a physical form o f prevention (e.g. a vaccine or promotion o f condom use).
Sexual behaviour is seen as a risk factor -it is not seen as an essential aspect o f the disease or o f its causation.Very little attention is given to the nature o f sexual relations as a human phenomenon.Sexuality, in turn, is interpreted as being simply a biological phenomenon explicable in terms o f hormones and instinctual drives (whatever that means!).The social and cultural context (i.e. the human) context o f sexuality is considered to be a private matter falling outside the scope o f science.I contend that this biological model o f sexuality not only distorts our understanding o f human sexuality, but also that o f animal sexuality.
This model o f AIDS is not wrong, it is simply scientifically inadequate -using 'science' in its widest context as defined in the Introduction.But how does this inadequate model affect our approach to the epidemic? 1 will attempt to answer this question by focusing on the the aetiology (causation) o f AIDS.I will argue that AIDS is not caused by the HI virus, but by the HI virus within the context o f a specific pattern o f sexual behaviour that I will call 'open sexual relations'.
Because the cause o f the disease is not simply a physical factor, finding a phy sical cure or preventative strategy will therefore not break the epidemic.Finding a 'solution' to the epidemic requires a much more holistic approach than that of fered by biomedicine.

The cause o f AIDS
All sexually transmitted diseases (there are about 30 that afflict our species) have one characteristic in common, namely that they are dependent for their spread from person to person on the most intimate form o f contact between two individuals o f a species, namely sexual intercourse.The reason for this is that the causative organisms are very 'w eak' organisms that are extremely sensitive to any exposure to the environment.M odem technology has made it possible for some o f these organisms to spread from person to person through blood to blood contact, e.g. through the sharing o f infected needles by intravenous drug users, or through infected blood transfusions.However, these methods are essentially 'transmission artifacts' that can initiate an infection, but cannot sustain an epi demic.I will now demonstrate that an epidemic o f sexually transm itted diseases can only be maintained through the presence o f a specific type o f sexual behaviour within a community.Some sexually transmitted diseases, e.g.Hepa titis B, can spread by other means as well, but AIDS is for all practical purposes an obligatory sexually transmitted disease i.e. it is entirely dependent on sexual transmission in order to initiate and maintain an epidemic.

The cause o f AIDS: behaviour as causation
Both human and animal sexual behaviour is a very complex phenomenon.Most descriptive categories used for describing sexual behaviour have an ethical bias.
Terms such as m oral, im m oral, prom iscu ou s, fid e lity etc. are, I believe, not useful in dealing with an empirical issue.Terms such as hom o-or h eterosexual, celiba te, m onogam y and p o lyg a m y are empirically descriptive terms for the structure o f sexual relations and have been involved in the description o f the causality o f AIDS.So, e.g.AIDS was initially considered to be a 'homosexual disease'.It has now, however, become clear that AIDS has no essential rela tionship to homosexuality at all (see Saayman & Kriel, 1992:22).In fact, the epidemic is decreasing in many homosexual communities but exploding in the heterosexual community.Are there other descriptive categories available?
In animal communities living under natural circumstances, sexually transmitted diseases are a very rare occurrence.Sexual behaviour in animal communities is usually described as 'free' or as 'promiscuous'.This is a complete misperception o f animal sexuality.In nature sex is not only dangerous (many males lose their lives during the mating season and few survive without permanent scars!) but sexuality is very strictly controlled both hormonally and by species' specific mating behaviour patterns.Sex in nature is for gene selection, not for enjoyment!
The actual level o f sexual activity is very low.Outside the mating season there is no sexual activity and males and females often live in separate communities while most o f the mating season is spent on territorial delimitation and mate selection.But this alone cannot account for the low incidence o f sexually transmitted diseases in nature.By a closed sexual relationship is meant a relationship in which sexual intercourse is limited strictly to the partners involved in the relationship.Such a closed sexual community may be structured along monogamous lines, or it may be po lygamous (see Figure 3).It may be heterosexual or homosexual or celibate.The deciding factor is that the sexual partners remain absolutely loyal to one another, or remain celibate.M asturbation could therefore be seen as a form o f closed sexual relations.
The term open sexual relationship does not simply refer to multiple sexual part ners, but where one o f the partners in an otherwise stable relationship has sexual contact with a partner outside that relationship.A polygamous marriage is there fore not an open relationship.It only becomes so if one o f the partners (male or female) has an 'outside relationship'.The sexual relations in animal troops, herds or other natural groupings (e.g.o f chimpanzees), are often described as open be cause an oestrous female will mate with several o f the males in the troop.How ever, the sexual relations are limited to that specific troop, and is thus 'closed'.It is an extended form o f the human polygamous relationship.
In Figure 3 the boxes indicate that the infecting organism o f any sexually transmitted diseases (including the HI virus in the case o f AIDS) cannot enter into a closed relationship or, if it happens to have entered (e.g. via an infected blood transfusion), it cannot get out again into the community to initiate or maintain an epidemic.That is why it was stated earlier that transmission o f the HI virus by blood transfusions and infected syringes or needles are 'transmission artifacts'.They may initiate an isolated infection, but cannot maintain an epidemic in any community in which closed sexual relationships are the norm.

On cures, vaccines and condoms
By pleading that we must go beyond the biomedical view, I am not denying the importance o f seeking a cure or vaccine, but even from a biomedical perspective, there are immense ethical, technical, cultural and economic problem s facing re searchers and health w orkers in developing these strategies.W hile we w ait for biomedical research to deliver the goods the epidemic is escalating -and the most hopeful predictions talk about a decade or more before a 'breakthrough' can be expected.
But even if an effective and safe curative drug is discovered tomorrow, it will m ake no difference to the epidemic.It will o f course make a difference to the survival o f individual patients, but in terms o f the epidemic it represents essen tially a mopping up operation which does not affect the primary mechanism main taining the epidemic, namely the sexual behaviour patterns o f populations.It will be o f benefit to the rich in the rich nations o f the world, but for Africa it will be o f very little benefit at all.Confirmation o f this pessimistic view o f the value o f a curative drug is found in analysing the present worldwide epidemic o f syphilis and other sexually transmitted diseases.
For syphilis and gonorrhoea a cheap and very effective 'miracle cure' is available in the form o f penicillin, but the incidence graph for these diseases shows the same logarithmic increase as that o f AIDS.This is understandable because these diseases are dependent on the same behaviour patterns for their transmission as the HI virus.Therefore, unless w e find a means o f stopping transmission o f the virus, discovery o f a drug or other miracle cure will have no effect on the epide mic, especially in Africa, but also in other continents.
And what about a vaccine?Because the HI virus is limited to man, the possibility is there for an effective vaccine to eliminate it from the earth as was done (so the virologists assure us) in the case o f the smallpox virus.But I believe that the biological, social, cultural, economic and epidemiological context or ecology of the AIDS epidemic differs so vastly from that o f smallpox, that the same biomedi cal strategies will not work.
Apart from the economic and ethical problems in developing, testing and imple menting a vaccine, there is the problem o f antigenic drift which enables the virus to change its antigenic structure against which the vaccine acts.The possibility of the virus changing its antigenic profile increases with the rate o f transmission.Therefore, for a vaccine to be effectively deployed we need to decrease dras tically the rate o f transmission o f the virus -which brings us right back into the field o f human behaviour.
We are therefore in a classical Catch 22 situation: the biomedical strategies that are being proposed in order to break the epidemic (and millions o f Rand are poured into research efforts to actualize them), all require the epidemic to be bro ken for them to become effective.However, the obvious point in the causal net work at which the epidemic can be attacked, namely 'open sexual relations' is ig nored (or is considered to be an interim measure until a 'medical breakthrough' is made) in policy planning as well as research funding.This leaves condom s and programmes for promoting 'safer sex'.That condoms provide some protection against the transfer o f the vims, is undeniable.But con doms leak and burst, and unless they are o f very good quality (and thus very ex pensive) the virus can actually pass through the mesh.However, a much greater problem is the fact that the acceptance o f the use o f condoms depends on cultural factors that are specifically excluded by the biomedical model.Furthermore, the same attitudinal, social, cultural, economic and personal factors that underlie the open sexual behaviour patterns that maintain the epidemic, also militate against the use o f condoms.Condoms form a vital part o f the strategy, but again only in a situation in which the epidemic is under control -promoting condom usage will not bring the epidemic under control.
In homo sapiens the biological control o f sexuality has disappeared.Sexuality is stimulated by the mind and the biological controls o f sexuality have been replaced by social (cultural and religious) controls.These controls are by their very nature 'softer' control systems, and open sexual relations have probably been part o f mankind's history since the very beginning.In Proverbs 5:15 large sections deal with the dangers o f the adulteress and the wayward wife and the advice (to the males!) is: "Drink w ater from your own cisterns, running w ater from your own well." The problem o f theory and practice in the medical profession Human sexual behaviour is therefore more strongly influenced by socio-cultural and economic factors than by biological factors.Any breakdown in these spheres will also impact on sexual behaviour.
Although the major religions o f Africa, as well as the m ajor African cultural traditions, promote closed sexual relations, these controls were devastated by the impact o f colonialism and the postcolonial wars.W estern materialism and secu larism with the attendant ethical and cultural relativism have taken their toll.M igrant labour, which has formed an integral part o f apartheid but has also cha racterized labour practices in the rest o f Africa as well as Europe, has devastated family life and made closed sexual relations for practical purposes impossible for large sections o f the population.Religions cannot promote closed sexual rela tions without addressing these factors at the same time.
On the other hand there are cultural traditions that favour and support open sexual relations.These have equally to be addressed whether they are based in the libe ral sexual ethic or in the ethos o f patriarchal societies (Cullen, 1991).
We therefore need to go beyond negative approaches and seek to understand and address the root causes o f open sexual relations.In fact, in spite o f the near ob session with sexuality in the modem world, we have very little understanding of human sexuality except in terms o f an (inadequate) biological model.W e need to understand the essential nature o f human sexuality.From this understanding we can then address and confront the problems o f the relationship between sex and marriage, the problem o f sexuality and the single person, masturbation as a form o f sexual practice, the problem o f sex education o f teenagers etc.All o f this will require a thorough understanding o f sexuality as a specifically human phenom e non, its pow er and its controls.Various religions and cultural traditions will need to work together, because sexuality and AIDS know no cultural or religious boun daries.

*
the presently accepted theoretical beliefs o f the research community re garding the research domain; * their methodological beliefs regarding the appropriate m ethods that can access their research domain to provide valid knowledge (theories) o f the domain, and * their ontological assumptions regarding the domain structure.These ontolo gical assumptions may be consciously propagated by the research com munity or may be implied by the research methodology without the research community necessarily being conscious o f the implications (see M outon &M arais, 1988:125-151).

*
give a short overview o f the history o f the natural science paradigm (natural science paradigm); * explain what is meant by the natural science paradigm by setting out in broad outline the content o f the ontological assumptions and the metho dological beliefs; * explain how these assumptions and beliefs affected the central concepts of medical science, viz.that o f patient, disease, disease causality and therapy; * explain how the natural science paradigm and medical science have, through the specific content given to these concepts, affected medical practice and propose a model o f the relationship between medical science (theory) and clinical medicine (practice); * illustrate the argument with a case study o f the medical and societal re sponse to the AIDS epidemic within the context o f the biomedical model.

Figure 1 :
Figure 1: Paradigm changes in the natural and human sciences

Figure 2 :
Figure 2: The relationship between the natural science paradigm, scientific medicine, m edical practice and medical education to suggest that this phenomenon can be explained by the hypothesis that the function o f species' specific mating behaviour is not simply gene selec tion, but the setting up o f small temporary communities o f closed sexual relations and that it is the presence o f closed sexual relations that explains the absence o f sexually transmitted diseases in nature.If closed sexual relations do prevent the spread o f sexually transm itted diseases, there would be sound evolutionary reasons why the procreational dynamics in nature should favour the establishment o f closed sexual relations.The presence o f sexually transmitted diseases in a natural community would decrease its procreational potential (e.g. through sterility and abortion) so that natural selection would favour those communities whose mating behaviour tended towards closed sexual behaviour patterns which exclude the possibility of sexually transmitted diseases.The point o f the argument is that a prerequisite for the existence o f sexually transmitted diseases in a particular animal (or human) community is the presence o f open sexual relations, because closed sexual relations make it impossible for any specific sexually transmitted disease to propagate itself in the community, and provides social immunity to the spread o f the disease.So what is meant by closed sexual relations and how do they prevent the transmission o f sexually transmitted diseases?F ig u re 3: E xam ples o f closed sexual relations: (i) M onogam ous h etero sexual (ii) M onogam ous hom osexual (iii) Polygam ous (iv) T ro o p (v) C elib ate

Figure 4 :
Figure 4: Open hetero-and homosexual relations including a bisexual relationship.The open boxes indicate relatively perm anent relationships.

Figure 4
Figure 4 shows a situation o f open heterosexual and homosexual relationships, including a bisexual relationship.From Figure 4 three conclusions can be drawn.* One partner in many o f these relationships could believe that he or she is in volved in a closed relationship.* AIDS has no intrinsic relationship to homosexuality.The crucial factor is not whether sexual relations are homo-or heterosexual, but whether they are closed or open.* It shows that the HI virus can theoretically infect the whole community practising open sexual relations.The minimum causal equation for the AIDS epidemic is therefore: the presence o f the HI vim s + open sexual relationships.
Sexuality and AIDS are not phenomena that can be understood or managed purely as physical biological phenomena.The insight that closed sexual relations is an essential feature o f sexuality in nature and open sexual relations an essential aspect o f the causality o f AIDS has two very important implications: * We need to find ways o f helping individuals and communities establish closed sexual relations as the norm.This will require, amongst others, inten sive funding o f research into factors contributing to open sex u al relation s and into the most effective ways and means o f promoting c lo s e d sexu al re lations.At present, guided by the limited understanding o f causality provi But this common registration is based upon a common education.All m e dical practitioners go through the same undergraduate education.Following this basic education a further period o f specialization is required.It is success in the educational programme that allows registration.But the edu cational programme must be legitimated by the Council, so these two (edu cation and registration) are really aspects o f the same process.Only that which is taught at medical faculties is 'science' and only those who have undergone the social process o f training and examinations at such faculties, may participate ( *Registration with the M edical and Dental Council.It is this registration in a specific category o f practice that gives a medical practitioner the legal right to practise his or her specific branch o f medicine.* The virus is therefore not the (only) pathogenic (disease causing) agent in the production o f AIDS.It rather triggers action in something that is not a simple object (e.g. a body), but a pattern o f relations, an integrated circuit or an eco logical web o f relationships.Only in combination with way-of-life factors affec ting the host -in this case a way o f life characterized by open sexual relationscan the virus be said to cause the disease.It is therefore simply unscientific to maintain that the cause o f AIDS is the HI virus.The only scientifically valid statement about the cause o f AIDS is that it is caused by the HI virus in the presence o f open sexual relations.(Thisprinciple is true not only o f sexually transmitted diseases, but o f many categories o f disease.)Acommunity which practises closed sexual relations could be said to have social immunity against HIV infection.From this it is clear that the statement used in AIDS education campaigns 'Anyone can get A ID S' is simply not true, and is misinforming the public about the real causal relationships.