No Deep Model:

The problem with empiricism in alternative practices

Barrett L. Dorko, P.T.

 Empiricists are not equipped to recognise the loud signals of alternative medicine as false. A deep model of the physical world is essential for choosing hypotheses to be tested and for learning from failures. Practices of alternative medicine that do not fit even at the far fringes of the model should not be tested in humans. Our decisions on which practices to test and which to adopt should be based on three things: empirical evidence; our deep model of the physical world; and our commitment to the wellbeing of our patients.

From (Alternative complementary) medicine: a cuckoo in the nest of empiricist reed warblers E, Beilinson and Leonard Leibovici British Medical Journal 1999; 319:1629–32. 

As I wade through the therapy literature these days or speak to students at my workshops, I often see a trend toward methods of management that do not include an understanding of physiologic processes familiar to me. I often wonder what I might have missed in school that would help me understand this, but, lately, I’ve come to understand that I hadn’t really missed anything, I just wasn’t seeing things from the same perspective. As it turns out, two things, empiricism and postmodern thought have overwhelmed therapy. 

The former is defined as “a theory that all knowledge originates in experience” and “a former school of medical practice founded on experience without the aid of science or theory.” To be empirical is not inherently a bad thing, and, in fact, is a necessary part of clinical practice. The danger lies in using empiricism while forgetting that what we can experience and observe does not represent the whole of any phenomenon. The “deep model” of the physical world includes the essentially invisible though well understood chemical alterations necessary for things like nerve conduction. Reducing this process to its essential characteristics is not easy, but it has led to every advance of care that I can think of. 

Consider this from Higher Superstition: T he Academic Left and its Quarrels with Science by Gross and Levitt (The Johns Hopkins University Press 1994): Even the simplest organism is an almost unimaginably complex system, whose fundamental chemical and physical processes reflect a heritage of several billion years of trial, error and modification. Yet a disease can, although certainly does not always, arise from a single, simple molecular difference buried deep within the continuous, labyrinthine reaction system that is the chemistry of life. 

Without a deep model to guide us, a model based upon years of careful scientific scrutiny, empiricism alone will often mislead and confuse us. It is obvious that those practicing “myofascial release” assign characteristics to the connective tissue that we know it doesn’t possess. When I say “we” I mean those of us who understand the deep model of collagen that includes its molecular structure.   

Again from Higher Superstition: If, as seems obvious, scientific and technical issues will become increasingly and urgently relevant to public policy in the decades ahead, how well will such matters be debated in this country? Obviously, we cannot hold high hopes. The historic record of American education in making the general public conversant with basic science has always been poor, except for a brief flurry of serious effort in the post-Sputnik era. Superstition, whether about astrology, ancient astronauts, or alien abductions, has always had easy and profitable going. Fringe medicine and outright quackery, long endemic in American culture, have taken on a new and ominous vigor, thanks in part to the dizzily rising costs and increasing impersonality of health care. The contrast between the incomparable virtuosity of professional American science and the general public disregard of scientific substance, whether from complacency or hostility, grows ever more pronounced. It is one of the great social paradoxes of history. 

I agree. And it doesn’t seem to me that this trend is heading in the hoped for direction. At least the direction I’m hoping for. 

Beilinson and Leibovici’s essay from the British Medical Journal is, well, thoroughly British, and therefore a bit difficult to follow. At least, from the perspective of someone educated at Ohio State. However, as I read it repeatedly I came to feel that its message was extremely important. I’d like to quote from it a bit more and add some commentary.  

Two conceptual frames are relevant to the present discussion. One (the empirical­social construct) is ill equipped to deal with the clamour of alternative medicine. Like the warbler, it ignores the absence of vital cues because of the loud signal. The other frame (the “deep model” empirical one) has deficiencies but is better protected against a loud false signal. Even firm empiricists, should use some of the protective mechanisms offered by the second frame. Both frames are defined by several assumptions. 

This reference to the warbler concerns the authors equating the loud voice of alternative medicine to the cuckoo chick deposited in the warbler’s nest. The mother warbler (purely an empiricist) is unable to recognize the essential difference between the loud chick and her own, and thus feeds the noisy one to the detriment of her family.  

The authors then explain the difference between what they call two “conceptual frames.”  The first is the combination of empiricism and social construct. Its tenants are as follows: 

Ø      Medicine is a social construct

Ø      Boundaries of medicine are defined in social terms

Ø      We firmly believe only in empirical proof

Ø      The origin of hypotheses does not matter

Ø      Our mission is (and always was) mainly to alleviate suffering 

Empirical proof (in the world of alternative medicine) is so powerful that we really do not care about the origin of the ideas we examine. The opposite is true: practices or hypotheses from everywhere are welcomed to be tested. It fits with our self-image of open mindedness and fairness. When the boundaries of medicine are shifting (and they shift because of social and political forces, by definition and belief) we will be able to use our empirical methodology to test what should or should not be adapted from the practices that are straddling the border now. 

The notion that medicine and, in fact, science itself is dominated by political and social forces is at the core of postmodernism. For more on this see my essay, What Went Wrong: Postmodern Thought and Physical Therapy Practice. The combination of empiricism alone with this is, in my opinion, a disaster. Beilinson and Leibovici sum up the alternative community’s attitude in this way: A combination of old­time patriarchalism and “scientific” hubris spawned a hardhearted and conceited breed in medicine. A bit of competition will do no harm. 

But when empiricism is combined with a deep model consistent with physical law and capable of passing the tests necessary to qualify it as a scientific theory (a higher standard than is commonly appreciated), it gains a power to help others and grow closer to the truth as time passes. 

The tenants of this second frame are:  

Ø      The scientific method changed the practice of medicine

Ø      Scientific medicine consists of empirical testing and a deep model

Ø      The deep model is essential for choosing hypotheses and learning from failures


Without a reasonable deep model in place we have no way of assessing failure or success. This leads quite naturally toward forms of intervention that “might” help, or, at best, “can’t hurt.” The charge that those of us unwilling to accept methods that have no rational explanation for their supposed effect are in some way heartless or less “spiritual” than alternative practitioners has always struck me as insulting and unfounded. Much could be said about practitioners who charge sick and vulnerable people for senseless care, but I’ll save that for another time. 

Beilinson and Leibovici are pretty clear in their next few words: 

(Alternative practitioners say) embrace empiricism, of course. In the long run, it is the only way to gain the legitimacy that you seek. But don't bother with “deep models.” Deep models are for snobs, oppressors, and wishful thinkers. The flat earth, phlogiston, bleeding, cupping, oppression of women, the Aryan Race—what are these but “deep theories”? 

They reply: 

(But) The core of scientific medicine is not empirical testing alone. Empirical proof (elicited using the best methodology) is very important. Standing alone, however, empirical proof might (and has) failed us. By definition, it is not protected from a small chance of error. Even with the best methodology, it is not easily guarded from inadvertent introduction of bias and from fraud. The building and falsification of bold hypotheses is at the core of the scientific method. However, there are hypotheses that cannot be accommodated even at the fringes: that livers of sacrificial beasts will predict the future; that a substance that causes complaints similar to the ones observed in a patient will, if diluted to an infinitesimal concentration, cure them.  

A deep model is necessary to choose the hypotheses we are going to test. Resources are limited. The moment we give up on the model, we should test everything. How do we choose what to test? There are thousands of practices, with a multitude of variations. The signals of the cuckoo chick ensure that it gets well fed by the reed warbler parents is easy to show that a way chosen from the framework of the “social construct” and does not use a deep model will soon encounter paradoxes and contradictions. 

Re­examining the deep model is the only way to use the failure of a trial or an experiment to move on. Failure should prompt a scrutiny of the model (the experiment was correct: how should we change the model, and which further experimentation will test it?); or a critical look at the experiment. After each unsuccessful trial in acupuncture there comes a flurry of letters saying that the study failed because the traditional method was not used, or that it was used; that it failed because needles were inserted too deep, or not deep enough, or twirled once too many times. We have no way of choosing from these explanations because we have no idea of how the treatment is supposed to work. If we have no idea, we must try all the alternatives. 

Scientific medicine was successful because it combined empirical proof with a deep model. This combination guards it, to a great extent, against chance, bias, and fraud. We will accept an empirical proof if it fits (even at the far, nebulous margins) the model of the physical world that we use. The model applies to the whole of the physical world, including our bodies. This model changes. A scientist can be defined a person who looks for explanations at the far fringes of the deep model and brings these fringes to the centre. 

There’s more, but I suppose you get the idea. The absence of a deep model that can be verified as reasonable without resorting to faith, metaphysics, religious or quasi-religious belief systems will, it seems, always separate alternative medicine from the scientific.  

I can see no way of reconciling the two as long as this is the case.