The Lay Referral System

The Problem of Professional Power

Eliot Freidson's early paper, “Client Control and Medical Practice”— published in 1960 (Freidson 1960) and based on the data he collected interviewing medical patients about their experiences with their diseases and with the system of professional medical practice most of them were involved in — made an early and prescient statement of a problem which preoccupied him continuously from then on: how can the professional tendency to monopolize power in medical settings be controlled?

In his analysis of the situation of patient and doctor, the professional's quasi-monopoly on knowledge about disease and its treatment created a crucial disparity in power between the two. If you know why I'm sick, and I don't, if you know what to do so that I'll get better and I don't — then the power lies in your hands. I have to do what you want, whatever it is and whatever I think about it, to achieve my goal of getting rid of my disease. Freidson saw that power corrupted and that doctors' prescriptions for their patients sometimes reflected interests beyond the simple desire to effect a cure, interests having to do with their obligations to colleagues, hospitals, and other people and organizations in the medical arena. Thus, fighting any tendency for complete power to lie in the profession's hands would preserve some freedom of action for people with disease, who were in danger of being pushed around by professionals in pursuit of their own interests.

The distinction between lay and professional was crucial to his analysis. Lay people associated with one another in the context of communal life, a life in which they took account of the full range of activities they engaged in together and the interests they shared. Professionals associated with other professionals on the basis of their shared knowledge and the common interests that arose from their similar social and institutional situations, but not with anyone outside their professional community.

In Freidson's view, patients can best protect themselves from medical monopoly and the resulting possible abuses of power under certain conditions. Two conditions in particular affect people's ability to preserve some independence of action. 1) The members of every community of whatever kind (but he had in mind especially residential communities) share some cultural understandings about disease, treatment, and cure. When the culture of the lay (i.e., non-medical professional) community with respect to questions about how and why people get sick differs substantially from that of the professional community, lay community members find support for doing something other than what doctors tell them to do. When the two cultures resemble each other closely — when doctors and lay people share the same ideas about causes and cures of disease, and especially when that means that lay people accept medical definitions unquestioningly — their margin of interpretive freedom and freedom of action is reduced.

2) He introduced the term “consultant,” borrowed from medical practice, to name the people a sick person turns to for advice and help. In the professional medical community, one doctor calls in another with specialized expertise to assist in the diagnosis and treatment of a case (to “consult”). Freidson generalized the term to refer to anyone a sick person called on for advice or treatment, whatever their professional status. When the number of lay people sick people consult (“lay consultants”) is high (how to establish what constitutes “high” is a serious question) they have a far better chance of preserving their independence from professional controls.

He combined these two conditions — high or low congruence between lay and professional communities' definitions of disease and treatment, and high or low number of lay consultants — in what was then the relatively novel procedure (introduced by Lazarsfeld) of producing a four-celled table, an exhaustive listing of the logically possible combinations of two dichotomous variables. He focused on two of the resulting possibilities. The more or less self-sufficient ethnic communities medical anthropologists studied exemplifed low congruence between professional and community culture and, simultaneously, the presence of a large number of lay consultants. In such situations, sick people are relatively free of medical definitions and treatments of their ailments and have access to a lot of what medical people call “folk practitioners,” often admitted to have a certain practical knowledge but disparaged as having no real scientific results to base treatment on. In these situations, diagnostic authority relies on inexplicable gifts, which some have and some don't, and on your knowledge of your own body and its ways. In the U.S., his own and others' researches had shown, this situation was most commonly found in the lower classes.

In the other situation he focused on, professional and lay community culture are highly congruent and there are few lay consultants, a situation he thought characterized the experience of people of the North American professional class. Such people accept the ideas about disease and treatment current in the professional medical community, and have relatively few lay consultants to offer them alternative explanations and recommendations. When people in this situation become ill they have only their own private understandings to rely on. Once they have exhausted that resourece, they find themselves dealing with medical professionals.

The Career of an Illness

Freidson made the operation of these influences clear by considering the career of an illness. When people first feel that something is wrong with their bodies, and consider the possibility that they may be “sick,”they search for interpretations of what they are feeling, and turn first to their own experiences and understandings: “It's probably an allergy that's making me sneeze and cough”; “my stomach upset is probably due to something I ate that didn't agree with me”; “this headache is the result of that extra drink I had last night.” These interpretations, common in almost any American community, suffice to explain a large number of the transient incidents of malaise that affect us all. The headache and stomach go away; the sneezing continues but I don't develop a fever and the newspaper reports that the pollen count is high this spring. All of which confirm my initial self-diagnosis.

But suppose the symptoms don't go away and the pollen count is reported to be exceptionally low. The next stop for most people is their circle of relatives, friends, and fellow workers. If I report my symptoms to these people and they all tell me that they, or someone they know, has had the same thing and it goes away after a few days, I once again stop worrying: I have “something that's going around” and what is affecting so many people probably is not serious enough to worry about — unless my symptoms don't go away as other peoples' have.

Many people find, in their circle of intimates, people who have (usually by virtue of their own medical histories) more extensive experience to draw on and more detailed interpretations of symptoms and possible courses of action to be taken. Perhaps they are a little hypochondriacal and have had occasion to consult a variety of sources of possible knowledge. My own father was such a person (I often accused him of being an amateur doctor). He read such medical literature as he could find (most of it, to be sure, contained in the Reader's Digest), and collected information from doctors he knew and from a pharmacist friend who also gave him pills that would not have been otherwise available to him. He used his knowledge and pill collection (augmented with over the counter medications) to “treat” friends and relatives, who were ready to take his advice.

Freidson gave a more systematic account of this process of seeking help:

This casual exploring of diagnoses, when it is drawn out and not stopped early by the cessation of symptoms or by resort to a physician, typically takes the form of referrals through a hierarchy of authority. Discussion of symptoms and their remedies is referral as much as prescription — referral to some other layman who himself had and cured the same symptoms, to someone who was once a nurse and therefore knows about such things, to a druggist who once fixed someone up with a wonderful brown tonic, and, of course, to a marvelous doctor who treated the very same thing successfully.

Indeed, the whole process of seeking help involves a network of potential consultants, from the intimate and informal confines of the nuclear family through successively more select, distant, and atuhoritative laymen, until the “professional” is reached. This network of consultants, which is part of the structure of the local lay community and which imposes form on the seeking of help, might be called the “lay referral structure.” Taken together with the cultural understandings involved in the process, we may speak of it as the “lay referral system.” (Freidson 1970, 377)

If people's diseases continue to bother them, and their ordinary resources of home remedies, self-medication, and consultation with their normal lay consultants does not produce results, they may finally turn to the professional system. This is where the lay and professional systems touch: “the local physician may be seen as the 'hinge' between a local lay system and an 'outside' professional system.” (p. 376)

The crux of the argument is that, when patients and prospective patients rely on the lay referral system, physicians and the results of their work get judged by lay people using lay systems of thought and evaluation. Professional power comes into play when patients leave the lay system behind and enter a world organized and run by medical people, in which professional understandings and judgments dominate. Physicians, while they may be sensitive to patient evaluations when they think about getting new patients from the community, are (especially, of course, specialists) much more sensitive to the judgments of their fellow professionals. That's why they're powerful. They don't care, and needn't care, what patients think as long as their colleagues think well of them.

This is an elegant argument, the terms clear, the links between them clear and intuitively believable. It is, of course, too good to be true, and in the latter part of the paper Freidson indicates that empirical reality exhibits all kinds of mixtures, with some medical practitioners being very much dependent of patient judgments while others are practically immune.

Lay Referrals as a General Phenomenon

The distinction between lay and professional understandings of an area of professional practice makes sense in many areas besides the medical arena Freidson focused on. I'll consider two examples, recreational drug use and learning to use a computer, to see how Freidson's framework can push us to new discoveries about other kinds of social behavior.

Recreational Drug Use

Users of recreational drugs seldom have conventional scientific knowledge available to them about the drugs they take. Many drugs have not been studied at all by research scientists. Often the drugs are new, invented by chemists working outside the conventional settings in which pharmaceutical research usually goes on. Even in the case of drugs which have been studied by chemists, neurophysiologists, and pharmacologists, the information is often quite skimpy and, in any case, only available in specialized scientific journals and couched in a highly technical and not easily understood language. The information is skimpy, certainly from a recreational user's point of view, because conventional research in this area aims to answer a very limited number of questions and not the ones users are most interested in. Aside from a conventional biological assay of a drug, research on their drugs of choice most often looks for harmful effects: damage to the cardiovascular system, to the lungs, or of a psychological sort (psychotic episodes, or dasmage to brain tissue and functioning due to long-term use). These questions concern the agencies which most often fund the research,which are connected to the federal government and particularly to law enforcement agencies and to legislators or government administrators who hope to use the findings to further an agenda of discouraging drug use.

Users look for answers to a different set of questions. They want to know if the drug will get them high and, if so, for how long, and what the nature of the experience will be; if it has any other effects that they would prefer not to experience, and how those can be avoided, minimized, or dealt with when they do occur; if there are any effects to be concerned about associated with use over a long period of time. In short, they want to know what it will do for them, as well as what it will do to them, and are prepared to make their own judgments about whether these results are desirable or what balance of desirable and undesirable effects they are prepared to accept. Conventional scientific analyses of drug effects seldom address these questions.

Users do find answers to these question, but not in scientific journals and probably not from a local physician either. But they do find them, and that prompts us to wonder where the knowledge they acquire comes from and what it is based on. An evenhanded consideration of this topic, in the style of Freidson's evenhanded consideration of lay and professional medical knowledge providers, raises provocative questions.

We might say that knowledge comes from research. People have questions and they search for answers. If they don't find an answer they take to be authoritative to settle their question, they do research to create the necessary knowledge. I don't use the word “research”ironically. It may not look like the research reported in Nature or Science or JAMA, but it shares many features with research of that kind: it is based on data collected by someone and on the collation and analysis of that data by the same person or someone else, according to some set of shared standards.

What makes the case of recreational drug use interesting is that it demonstrates a very different variety of scientific research, which produces, through procedures different from conventional science, knowledge which is perfectly adequate for the purposes of the people who use it. Drug user research, if I can call it that, produces masses of data, based on experiments conducted by people who take the drug. Everyone who smokes marijuana, everyone who takes LSD, everyone who takes amphetamines — all these people are doing research on the effects of those drugs. They take them and they note the results, because they are taking them in order to experience something and they want to see if they are having the experience they expected and wanted. They carry the results with them, as a memory, and on occasions when the subject of that drug and its effects comes up, they may contribute their data, their observations.

How do these individual “scientists” pool their observations to produce collectively a body of knowledge that can be taken as authoritative by the people who use the drug? They have no journals to consult, the value of whose articles is guaranteed by a system of peer review. Instead, the observers communicate their observations through conversation and other informal means, and the guarantee of the validity and value of what they report lies in its congruence with what others report. If a new drug appears in a community, and hundreds of people take it and observe its effects, and they are in a position to interact, it's likely (not sure, but likely) that they will develop a body of shared understandings about just those things users care about which conventional science ignores: how high it gets you and what kind of high it is, what kinds of unpleasant other effects may be encountered, and what you can do when they occur. This is the solid core of meaning which can be attributed to the otherwise trite and misleading expression, “drug culture.”

Some of these participants in the development of this kind of collective knowledge play a role similar to the lay consultants Freidson described. They are the people you go to when you have a question, a sympton, an experience, you don’t understand and want some reassurance or advice about. The difference between this situation and the common medical arrangements (lay and professional) Freidson describes is that in the case of recreational drugs no other authoritative source, widely accepted among users, exists. It is the situation he describes in which the difference between the culture of the lay community and that of the relevant professional community (in this case, conventional drug researchers) is so great that lay people do not trust or make use of any resources from the other community. (This case is explored more fully in Becker 1967, 1973.)

Computer Gurus

Many, perhaps most, people who use computers have very little understanding of how the machines they depend on actually work. Business people know just enough to start up “their program,” whether it is for bookkeeping or medical appointments or library withdrawals, and do what the screens tell them to do. A lot of people have figured out how to get and answer e-mail, and how to consult the internet for information they want. But if something goes wrong, if they press the accustomed key and it “doesn't work,” they are at a loss. They then engage in something like the search in their own experience Friedson described in the medical case. They rely on folklore or on common sense, which is of very little use in the totally artificial and arbitrary, though logical, environment of computing. If they are lucky, they have someone to turn to who knows more than they do and can tell them what they did wrong.

An example: a friend of mine had typed his way into a crash; the computer responded to no command he knew how to give. So he called the friend who had advised him on buying the computer and had set up the programs he used for writing and for e-mail. The friend came, like an emergency physician, asked some questions, fiddled with the keyboard, and fixed whatever was wrong. My friend wanted to know how this disaster had occurred, and his savior explained that computers were dumb, literal machines which did whatever you told them to. In this case, he explained, my friend had given it two contradictory commands and the dumb machine couldn't decide which one to obey. My friend got testy and said: “It knew what I meant!” The machine didn't but his savior did.

These relations of guruship (to coin an ugly word) go quite far up the ladder of knowledge. My wife and I act as consultants (or gurus) to quite a few people who use far less of a computer's possibilities than we do, and who are lost when they venture beyond simple word-processing, net surfing, and e-mail. We advise them on what computer to buy and what software to use, how to solve simple hangups and do things they didn't know were possible. But each of us had to learn what we knew from someone and each of us quickly reaches our limits.

I was an early computer user (early, that is, for someone who did not routinely engage in statistical work), beginning to use the Apple II sometime in the 1980s, and heavily influenced by my guru, a colleague named Andy Gordon, who had had extensive experience programming mainframe computers and was a serious proselytizer for the not yet popular cause of personal computing. He recruited me easily, taught me with great patience the somewhat arcane procedures characteristic of those early days, and even persuaded me to try my hand at programming: under his guidance, I created, after many weeks and hundreds (probably thousands) of mistakes, a small program that opened a file of Census data and used it to create a population pyramid, essentially two bar charts back to back, which displayed the age and sex composition of a population. Andy was a model guru. He saw what I didn't understand, nudged me along so that I discovered what I needed to know for myself. The experience left me with no desire to write any more programs, but with a rudimentary understanding of what was making the programs I used work. And that was good enough to make me a guru in turn to people who understood less than I did, and who got hung up on “elementary” points.

After Andy and I both moved to the University of Washington in Seattle, I acquired a new guru (and I think he served in that capacity for Andy as well), the head of social science computing there, Fred Nick, a genial man with a deceptively simple manner which hid an encyclopedic knowledge of what was available in hardware and software, what could go wrong with it, how you could fix what might go wrong, and who you could consult in the rare case where he didn't have the answers. (His knowledge went so far as to allow him to tell you that something you wanted to do couldn't be done, not now anyway.)

When Dianne Hagaman (my wife) wanted to construct a visual work that required doing something that the complex software called Macromedia Director did not have available as an already programmed option, she had to take classes in that software. But that wasn't enough and she finally found that creating the effect she wanted would require more study than she (being a photographer with no ambition to become a programmer) wanted to do. So she hired a guru, a programmer who was expert with that software and had in fact worked on its development. But this guru, though he knew a lot, could not solve the problem either, and appealed to a community of programmers more expert than he was — his gurus — who gave him the clues and suggestions that eventually led to a solution.

Gurus tell you what you need to know when you need to know it. They are usually part of your own community, friends, and friends of friends, who will take on this responsibility. Such services have been somewhat regularized now, with the advent of experts for hire, who come to your house or office, set up your equuipment (avoiding the intense anxiety that accompanies that procedure for people who aren't experienced in the troubles it invariably produces), and tell you what to do when things go wrong. This has been further regularized in the Genius Bar that is an important feature of the Apple Store.

“Guru” is not the name of a social type. It describes a social relationship between a more knowledgeable person and one who knows less. The person you call when you don't know what to do with a computer problem is your “computer guru.” Your guru may not know a lot about computers, but knows more than you do, and can solve the problem you have experienced. The person who is guru to you probably has a guru of his or her own, who solves the problems your guru can't solve for himself. (Shapiro 1988 contains much case material relevant to this discussion; see also Barley and Kunda 2004.)

Some Comparative Thoughts

Freidson's paper on lay referral systems is, ultimately, concerned with the power professionals exercise by virtue of a real, perhaps imagined, in any case claimed, monopoly of knowledge about some esoteric field of activity. We do what doctors tell us to to do because we think they know more than we do and thus can help us achieve our goals (in the medical case, the goal of health) when we would otherwise be helpless. But we are not in a position to judge what they do for us, and they may not always be as competent as they proclaim themselves and may at times act to further interests of their own which don't coincide with ours.

He described the medical case, in which professionals are licensed by the state and so empowered to exercise a monopoly over activities in their arena of supposed expertise. There's a clear line between professionals and lay people and, once you've crossed that line, you are in the hands of the professional monopoly. His research suggests that a countervailing force, which protects us from this kind of exploitation, is the existence of a system of lay opinion which tells us when professionals act against our interests.

The cases of recreational drugs and of computer expertise show us that the relation between lay and professional is not always defined that way, and taking medical practice as the archetypal situation misleads us. In the case of recreational drugs, professional experts do exist, but their expertise is very limited, extending only to the “bad effects” of the drugs they investigate. Having no knowledge that is helpful to interested users of these drugs, they cannot function as knowledgeable experts. Users rely instead on a network of amateur investigators like themselves, whose pooled knowledge, based on personal experiences and distributed through an alternate system of communication, replaces that of the experts. In this case, lay people evade the control of knowledgeable and “legitimate” experts by creating an alternate source of knowledge based in their shared experience.

The state does not take sufficient interest in the area of computer knowledge to licence experts and monitor their training and activity. Instead of a clear division between licensed professionals and amateurs, knowledge is distributed along a chain of increasingly experienced and competent experts, who can function as gurus for those less knowledgeable than them but who turn to more knowledgeable people to serve as their gurus. Everyone can be a guru and everyone can have a guru. Knowledge is widely distributed but the degree of knowledge any person has is variable.

Freidson's analysis of the lay referral system thus opens up a large vista of research possibilities, focused on questions like the following. How is knowledge that is not “common knowledge,” more or less known to all members of a community (or at least all members of the same age and sex groups), controlled? Who needs it? How do people who need that knowledge get access to it? What is the role of the state in controlling the accumulation of and access to such knowledge? What are the consequences for the ultimate users of this information?


Barley, Stephen R., and Gideon Kunda. 2004. Gurus, Hired Guns, and Warm Bodies: Itinerant Experts in a Knowledge Economy. Princeton, NJ: Princeton University Press.

Becker, Howard S. 1967. “History, Culture and Subjective Experience: An Exploration of the Social Bases of Drug-Induced Experiences.” Journal of Health and Social Behavior 8:163-176.

—. 1973. “Consciousness, Power and Drug Effects.” Society 10:67-76.

Freidson, Eliot. 1960. “Client Control and Medical Practice.” American Journal of Sociology 65:374-82.

Shapiro, Reva. 1998. “Analytical portraits of home computer users: the negotiation of Innovation.” Ph. D. Thesis. Sociology. University of California-San Francisco.