Mad Science: Psychiatry is not science

“The discrepancy between the views held by the public and many professionals on the one hand and the actual evidence on the other hand brought the authors of this book together. Our goal was to understand and describe how psychiatric research and science of the past half century have shaped the public understanding of expressions such as ‘mental illness’ or ‘mental disorder’ and of the effectiveness of psychiatric treatment. Much of the psychiatric research that has fueled the expansion of the mental health enterprise has not contributed to a science of madness. Instead, it has fueled mad science, which rests on unverified concepts, the invention of new forms of coercion, unremitting disease mongering, the widespread use of treatments with poorly tested and misleading claims of effectiveness, and rampant conflicts of interest that have completely blurred science and marketing. This is the ‘madness’ of American psychiatry, and of psychiatry in much of the world.” (KL 81)¹

“…We will demonstrate that the touted achievements of psychiatry in the past half century – keeping disturbed people out of psychiatric hospitals for extended periods, developing a novel and easily applied diagnostic approach embodied in the modern Diagnostic and Statistical Manual of Mental Disorders (DSM), and using ‘safe and effective’ drugs as the first-line intervention for every ill and misfortune – are little more than a recycled mishmash of coercion of the mad and misbehaving, the mystification of the process of labeling people, and medical-sounding justifications for people’s desires to use, and professionals’ desires to give, psychoactive chemicals.” (KL 189)

This will be a two-part post about the new book Mad Science: Psychiatric Coercion, Diagnosis, and Drugs by Stuart A. Kirk, Tomi Gomory, and David Cohen.



This part will summarize the main argument of the book – that contemporary psychiatry is pseudoscience – and the next will discuss some of the problems with the authors' alternative understanding of mental health.

Mad Science doesn’t provide a wholly original perspective on the failings of contemporary psychiatry, but it does offer a solid, up-to-date, readable overview of the critical arguments and evidence.² The book is loosely tied together, as the title with its double meaning suggests, by its focus on “mad science” – on how, in psychiatry, pseudoscience is taken for real science. The authors show mad science to be an assault on science as much as (and connected to) an assault on human freedom and development. This focus should make the book particularly interesting to more science-oriented readers.³

The authors drive home repeatedly that the categories of the DSM (which, “with its list of disorders and symptoms is the foundational scientific cornerstone on which psychiatry stands,” KL 3143) have no scientific validity. Given the persistence and effectiveness of rhetoric that relies on terms like “overdiagnosis,” “diagnostic expansion,” and “overdrugging” to suggest that there exists a scientific core at the center of psychiatry, I think it’s worth quoting several statements from the book:

• “This was the Kraepelinian goal: to discover biological markers (hopefully causal) to confirm the existence of the diseases in those in whom diseases were hypothesized to exist and the inexistence of such markers in others.

This hasn’t happened. There are no known biological markers for any category.” (KL 4255)

• “…no significant improvements on any indicator of any major mental disorder, including depression, schizophrenia, and bipolar disorder have been demonstrated in studies that have been conducted relatively free of industry design or funding.” (KL 836)

• “Despite heroic or, depending on who is doing the judging, desperate and enormously expensive research efforts over many decades, no genes or reliable pathophysiology that maps schizophrenia or any other ‘mental disease’ has been found.” (KL 1036)

• “Neither the many theories nor the implied causes of madness have been scientifically validated, perhaps because mad science rests on hundreds of constantly shifting diagnostic categories of ‘mental illness’, which have little in common.” (KL 1049)

• “There are no biological markers, or pathophysiology for any of the diagnoses listed in the DSM.” (KL 4093)

• “Describing a set of behaviors and labeling them as pathological symptoms never establishes the validity of an illness.” (KL 4133)

• “DSM offers behavioral diagnostic criteria as if they confirm the existence of a valid disorder, when the criteria merely describe what is claimed a priori to be an illness. Descriptive diagnosis is a tautology that distracts observers from recognizing that DSM offers no indicators that establish the validity of any psychiatric illness, although they may typically point to distresses, worries, or misbehaviors.” (KL 4133)

•“No independent tests can verify any underlying mental disorder; the very symptoms used to construct the alleged mental disorder confirm the presence of the disease.” (KL 4330)

As the authors note, the fact that the core concept(s) of biopsychiatry lack scientific validity is broadly - if often deceptively – acknowledged by the biggest names in the field. The psychiatric researchers they discuss,

who have committed their careers to developing this biological understanding… state flatly that nothing biological has been reliably associated with any DSM diagnosis that aids to make the diagnosis or to predict how someone will respond to drug treatment. There are no biomarkers for psychiatric disorders – no biological signs that can be used reliably to measure the presence, change, improvement, or worsening of the condition that one might deem ‘pathological’. To be sure, every month, investigators propose new biological measures in the literature as candidate biomarkers, but none survives for long. (KL 6248)

The authors quote DSM-IV task force head Allen Frances’ stark admission to Gary Greenberg (2011): “There is no definition of a mental disorder. It’s bullshit. I mean, you just can’t define it” (KL 297). They note that the FDA, which continues to approve psychiatric drugs for the “treatment” of alleged mental illnesses, has acknowledged that these so-called illnesses are not demonstrated pathologies and that it relies on psychiatry’s descriptive diagnoses. Just a few months ago, near the time of the book’s publication, came the “startling” public admissions from Insel and Kupfer that the so-called illnesses in the DSM have no scientific validity.⁴

As they argue, the basic premise of biopsychiatry and its diagnostic manual fails fundamentally in establishing itself scientifically:

To state that mental illness is a valid concept (that it truly identifies a phenomenon of nature), means that some body of evidence has been amassed according to the guidelines of a specific theory, and then has survived rigorous tests devised upon the notion that the specific theory might be false. Then and only then might we entertain with some confidence that behaviors now defined as symptoms of mental disorders might ‘truly’ be the manifestations of diseases or brain disorders. This has not occurred. (KL 4106)

The modern DSM purports to provide guidance in identifying valid instances of mental illness. The manual explicitly claims that it is ‘atheoretical’, terminology that was used to imply that the categories did not contain any inferences about the ‘causes’ of the disease, no theory about what it is or why it occurred. That’s why it is known as ‘descriptive diagnosis’. This is about all the manual could officially claim, because there is no convincing scientific evidence concerning what causes people to feel or act in ways that are uncomfortable to themselves or bothersome to others. (KL 4304)

The scientific invalidity at the core of biopsychiatry has enormous implications – for clinical advances, for diagnosis, for epidemiological and drug research, and for science more generally.

This basic invalidity has resulted in scientific sterility. The authors state plainly: “No modern (or older) edition of DSM has produced scientific breakthroughs” (KL 4907). I can’t think of a more damning indictment of an alleged field of scientific study.

The lack of validity of the concept of “mental illness” and of the specific constructs called mental illnesses or disorders expectedly makes a farce of psychiatric diagnosis. Early in the book, the authors describe what a mature psychiatry would look like if its basic concepts were scientifically valid. Among its features, “The techniques of diagnosing mental illness would be more accurate, and valid, than methods used previously. Diagnosis would rest on biological markers rather than conversation as the ‘biological basis of the more than 300 types of mental illnesses would have been substantiated or disconfirmed” (KL 55). You don’t have to be a strict Popperian to appreciate that if the concepts were valid, several decades in some biomarkers would very likely have been identified. As described above, though, they haven’t.⁵

This lends a bizarre cast to efforts to “refine” diagnoses:

There is no established method of standards that DSM uses in deciding how many criteria should be required for any disorder. This is because no scientific links or any other indicators exist that psychiatric researchers themselves consider important between the diagnostic criteria and any specific biological pathology (or anything biological). Without that scientific foundation, the decision about the number of criteria is essentially a practical and public relations problem. (KL 4235)

…[M]ost of the struggles over DSM-5 are, in fact, disputes over where the arbitrary boundaries should be drawn between phenomena that have not been validated as disease categories. (KL 4304)

The lack of validity of the core concepts means that disagreements about diagnoses and their resolution are necessarily political rather than scientific:

What is most striking about all the controversies over DSM-5 and all its prior editions is that the conflicts are not really about science or empirical evidence about the reliability of psychiatric diagnoses. The evidence suggests that reliability has worsened over time, undermining the validity of the classification system. In a truly scientific endeavor, this would send the DSM-5 task force back to the drawing board. True science does not require publication deadlines, press releases, public relations specialists, reviews by the public nor by patients or their loved ones. Science stands or falls on the empirical outcome of rigorous tests of claims, regardless of popular opinion. (KL 4828)⁶

The image that the spokespeople for psychiatry and their supporters want to project – of an advancing science identifying underlying pathophysiology and targeting treatments to specific mental disorders - is an illusion:

In fact, [in] the opening chapters of their book, A Research Agenda for DSM-5, Kupfer and his associates (2002) admit that research has not been able to document the validity of the DSM classification system (Kupfer et al., 2002). They note the following: that research has not discovered common etiologies for the major DSM disorder categories; that not one laboratory marker has been discovered for any DSM-defined syndrome; that studies have found disorder categories overlap, undermining the belief that these disorders have distinct causes; that there are high degrees of short-term instability for many disorders; that there is lack of specific treatments: the same treatment, drugs, or psychotherapy are used for many supposedly different disorders; and that even studies of twins have contradicted the DSM assumption that different disorders have different underlying genetic bases. (KL 4907)

As detailed in the book, in the later twentieth century psychiatrists wanted to develop a system of reliable diagnoses. Reliability, of course, is something of an adjunct to validity, but reliable measures could potentially be seen as indicators of possible validity, which would still have to be established. As the authors show in some detail, though, while improved diagnostic reliability has been claimed for psychiatry, it hasn’t in fact been achieved. Even the DSM-III wasn’t the advance in reliable diagnosis it’s widely believed to be, and the measures of reliability for the various diagnoses have remained stable – and none too impressive - or declined over the years. (The creators of DSM-5, as I reported a while back and as the Mad Science authors note, attempted to justify using measures with recognized crappy kappa values by…redefining “crappy.”) As with the other problems, this failure to develop increasingly stable categories and reliable diagnoses follows naturally from the fundamental invalidity of the model - it’s exactly what we’d expect to see.

The lack of valid concepts at the heart of the enterprise also wreaks havoc with research. Studies determining the prevalence of various disorders or mental illness as a whole across populations or over time are impossible when those conditions aren’t scientifically valid. “Having valid categories of disorder matters fundamentally in epidemiological research,” the authors point out. “…If the diagnostic criteria do not or, as we have argued, cannot validly identify who is disordered, epidemiological research based on them is largely meaningless, except as a sociopolitical tool for psychiatric expansionism” (KL 4880). The same goes for drug testing. “Most telling” about the failure of any of the regularly appearing biomarker candidates is that not one is “used in any clinical trial testing a drug for the treatment of any psychiatric condition” (KL 6248).

Biopsychiatry’s continued faith in the model with nothing to show for it leads the authors to view it not as a challenging or even failed scientific endeavor but at its core something else:

The quest of biological psychiatry, adopted and aided by DSM, is to claim a brain disease for every human trouble. From Kraepelin to NIMH’s Decade of the Brain, the perennial promise is that brain diseases will be discovered. Despite decades of failure to confirm that misbehaviors and emotional turmoil are caused by disordered brains – a search floated on the massive stream of funding from the federal government and the drug industry – the effort has never lost momentum and even enthusiasm, as if the key to human misery will be discovered. The failure for the enterprise to deliver on its promises has not discouraged efforts to support it, which suggests that it is at its core a moral crusade. (KL 4975)

They have this to say about Insel and NIMH:

…while Insel correctly admits the lack of progress in understanding and treating madness and even in alleviating any of its burdens on patients and their families, he fails to consider that the failure may result from madness not being, for the most part, a biomedical problem. This simple hypothesis may explain why, after a hundred years of pursuing the neuroscience hypothesis, so little (actually, nothing) has been discovered of any direct relevance to clinical psychiatric practice. That Insel does not even entertain this hypothesis in print – even if only to dismiss it – suggests that the NIMH is not a scientific institute. (KL 695)

The latest episode, in which the NIMH and APA came clean about the lack of scientific validity (if not about the lack of reliability) of their diagnoses, just offered more evidence that this is a faith-based field. It appears that nothing will lead them to conclude that they’re mistaken. Even as they continue to sell their manual and prescribe (or forcibly administer) their drugs in the present, their rhetoric is about the fantastic new discoveries that will come in the future and retroactively justify their model. As I’ve argued, this is not science. I don’t think it needs pointing out that it’s dangerous to have people in positions of political and legal power making decisions on the basis of this sort of unshakeable faith.

The authors describe Insel’s acknowledgment in the recent past – very late in coming - that the “chemical imbalance” notion is bogus: “We note Insel’s assertion that the long-held view of mental disorders as conditions characterized by ‘loss of cells’ should now give way to one of conditions characterized by ‘disruptions in the circuitry map’. In essence, Insel states that the dominant neurobiological hypothesis of mental disorder has been abandoned” (KL 6287). Insel has been saying something similar for the past several years, despite the fact that he’s been consistently unable even to define a neural pathway. What’s shocking is that this hasn’t been news – the abandonment of the central “hypothesis” at the core of the model is noted and everything goes on as usual. “Will the 75 percent of Americans who believe in ‘chemical imbalances’ be informed that circuitry has dethroned chemistry?” the authors ask. “Does it matter to anyone?” (KL 6287).

A good part of the explanation for that continued support or acquiescence can be found at the intersection of moral crusade and immoral crusade. Kirk, Gomory, and Cohen, amongst many others, describe the perverse and corrupting influence of drug companies on psychiatric “science”:

In our view, what has occurred and what is occurring is the systematic manipulation of the scientific process to reach predetermined conclusions, which means, the paralysis of science as an unbiased enterprise to advance knowledge or even as a pragmatic enterprise to improve the quality of human life. (KL 7040)

The virtually complete blurring of marketing and science has paralyzed science – or what has passed as science. Conflicts of interest permeate the FDA and its advisory committees, scientific journals, and the scientific literature. (KL 6975)

On the one hand, the drug industry’s marketing efforts and its massive infusions of money to support psychiatric activities, and on the other hand psychiatry’s enthusiastic acceptance of the partnership, have completely subsumed psychiatry as a satellite branch of the multinational pharmaceutical industry. (KL 6962)

Where they’ve failed at science they’ve succeeded brilliantly at selling the model and the image of themselves as science-based healers and purveyors of targeted and effective treatments. This successful global campaign has led to a frustrating situation in which not only do internal and external critics of psychiatry have to fight an uphill battle, but even the most frank admissions by the leaders in the field are met with indifference. We’ve arrived at a comical reality in which people and organizations with the halos of scientific authority the pharmaceutical companies have spent billions to purchase (like those clunky gold dinner-plate halos of the International Gothic style paid for by the wealthy medieval commissioners of works of art) can use the most obvious language of pseudoscience or even publicly acknowledge that their model has no scientific foundation without really being called on it, without any significant diminution of their power or authority. This has to change.

In the second post on the book, I’ll discuss the problems with the alternative understanding of psychological problems and mental health that the authors propose. This alternative model isn’t central or necessary to the book, which focuses on debunking psychiatric pseudoscience. But talking about alternative understandings is crucial both for the success of debunking efforts and for building a better world more generally.

¹All Kindle locations are approximations, due to the fact that the Kindle app for Windows 8 is garbage and they’ve made it impossible to “downgrade” to a functional version.

²In addition to the sections on the biopsychiatric model of mental illness, the emergence of the diagnostic categories of the DSM, and the development and use of psychiatric drugs, the book contains chapters on psychiatric history and historiography and on ACT (Assertive Community Treatment). These are among the more interesting and original aspects of the book, and the discussions do fit with the general argument about the misuse of science, but they could have been better integrated with the other chapters (the ACT chapter would maybe work better towards the end rather than at the beginning).

³This is not to say, of course, that other books critical of psychiatry lack this sort of scientific critique, but that this issue is not as central as it is in Mad Science. In this sense, the book can be usefully compared with Ben Goldacre’s Bad Science and Bad Pharma. The difference is that while Goldacre considers psychiatry one medical specialty amongst others, subject to the same manipulations as others though with the same scientific potential, these authors show the ways in which psychiatry is special – not a temporarily corrupted medical science but a fundamentally corrupt and invalid model.

⁴I put “startling” in quotation marks for two reasons: first, because for those following the situation these admissions were nothing new (see Table 2 in Brett Deacon’s article and similar quotations stretching far back in time in Anatomy of an Epidemic and other works); second, because they’ve so far fallen largely on deaf ears.

⁵Even if they were, I should note, this wouldn’t be proof of a genetic or biological cause. Experiences of course can lead to changes in people’s brains.

⁶There are many arguments that can be made about this assertion if it’s read as a general statement about science. It’s better and more accurate, though, to appreciate the context and specific meaning of the statement. The authors aren’t saying that science is in no way a publicly oriented or political endeavor, but that psychiatry has no scientific basis on which to resolve disagreements.

The Path of the gods, it has many steps.

No, I haven’t converted to polytheism. The title is a reference to the Sentiero degli dei, a hiking trail that hugs the Amalfi coast, overlooking the Mediterranean. I walked/climbed the stretch between Praiano and Positano last month, and thought I would share some pictures and information.

Before I get into details, though, I should say that it was a remarkable experience and I would do it again in a second. If and when I return to the area, I’ll make sure not to miss it. Looking at my pictures, I’m wishing I could transport myself back there.…

But I think having more knowledge ahead of time can only enrich the experience. I’d done a little – in retrospect, not enough – internet investigation before I left the US, and went into it with some impressions that turned out to be inaccurate. Three subjects that could use some attention are: the difficulty of the trail, its suitability for people afraid of heights, and the quality of the signage.

(A few general notes, on which people seem to agree: First, it’s best to walk toward Positano rather than away from it. Second, start early, before the sun is high in the sky. We were fortunate to be sheltered by some misty clouds that diffused the sunlight and made the climb more interesting, but you shouldn’t go if there’s any threat of rain. If you’re taking the bus andor ferry to your starting point, plan accordingly and find the schedule that will get you there as early as possible. Wear sunscreen, and bring more. Bring plenty of water. Bring some cash. Bring your camera – you’ll want it.)

So, first, the level of difficulty. Maybe the most important element is the amount of climbing. The path itself is high, close to the top of the cliffs, so you have to reach it somehow. I saw several references to the 1700-step staircase at the Positano end, but I guess I hadn’t taken enough note of the discussions about starting the path in Praiano as opposed to Agerola. Apparently, if you begin in Bomerano (a neighborhood in Agerola), the bus drops you off high on the cliffs so you don’t have to climb up to reach the path itself. In contrast, where we entered in Praiano, the bus leaves you relatively low on the cliffside and you have to climb 1900 steps, to many of which the term “step” is loosely applied.

You won’t really see much of this portion in people’s pictures, because it’s extremely difficult and vertigo-inducing to take pictures as you’re climbing. Here are a couple from places we stopped to catch our breath.

After we’d climbed probably more than a thousand steps, we came upon a little church/monastery with a cute snack bar and a bathroom.

The man there was quite nice, and we also met a family of hearty Belgians who were just leaving as we arrived. Catching our breath, we asked hopefully, “There isn’t much more climbing after this, right?” “I don’t want to get your hopes down,” the man said gently, “but yes, there is.” Resigned to our future after hearing that, we set off climbing again, and yes, there was quite a bit more. Once we’d accepted it, though, and knowing we were making progress toward the top, the climb became more pleasant and the views increasingly amazing.

So, the Praiano climb is an important element in assessing the trail’s difficulty, and could explain to some extent the widely varying accounts. The descriptions I’d read before I left for Italy were largely ones like this (from recent posts on TripAdvisor):

“I had read many reviews about this walk hoping to find out if it would be suitable for someone with a bad knee and hip could walk it and I found no difficulty at all.”

“There is quite a bit of up and down but it is very achievable and no where near as hard as some reviews have suggested. My step mother (60) who has low to average fitness managed the walk and 1600 steps down [in Positano] and has now a great sense of achievement.”

“This is an easy walk on good paths, the one or two rougher patches are very short not difficult so long as sensible footwear is worn.”

“Agree with previous posters that this isn't really a ‘hike’. It's more of a stroll along an extremely beautiful path.”

If I’d dug deeper into the older posts on TripAdvisor, I would have found ones more like this, several pages in:

“The walk is quite demanding as the path is often no more than rough gravel and steps made from granite rocks. Wilst the path roughly stays at 600 metres or so above sea level, it does go up and down alot. We are both fit and active and would not describe this as a modest stroll.”

“This is a wonderful hike with awesome vistas along the entire route. Be advised that it is somewhat challenging and rocky at points.”

“Although a hard walk and very high at times, the walk was absolutely fantastic.”

“I consider myself in good shape (I run and workout several times a week) but there were a couple of times I was huffing and puffing. The hike is strenuous, but doable, with breaks, water and snacks.”

As I said, I probably hadn’t paid enough attention to the fact that most if not all of the people describing the walk as easier had begun in Agerola and not Praiano as we did. And that makes a difference – climbing almost 2000 stone steps carved into a cliff adds considerably to the overall difficulty. From what I’ve come to understand, the section that goes from Agerola to Praiano is comparatively leisurely.

However, even as a description of the part of the path from above Praiano to Nocelle/Positano, the first set of statements I think give a false impression. Very few sections could honestly be characterized as an easy walk or stroll, or suitable for people who aren’t physically fit or have knee or hip problems. Any trail where you have to use your hands in parts and keep your focus in finding your footing amongst rocks is pretty much by definition not a stroll. At one point we turned and took pictures of the path behind us:

Not a stroll. (I had to laugh as I was looking through my pictures, and then when I did a search online to see others’. None of them really show the narrow, rocky, ascending/descending sections, leaving the impression that the path is wide and flat throughout. I eventually realized that this is because it’s generally only in the easier sections that it’s advisable – or possible, really – to take pictures.)

Which brings me to the fear of heights. Here are a couple of descriptions from TripAdvisor:

“For the most part, the trail is wide and not as precarious as the overall steep slopes would seem to suggest.”

“People mention that there are some bits that aren't great for vertigo sufferers which worried me a bit but I saw nothing that was an issue. I think it would only be a problem if you have serious vertigo rather than a dislike of heights which is what I have.”

I can’t agree with this. Again, this is probably in some part related to the Praiano climb, since it included some of the narrowest sections beside steep drops. Climbing there, even for me and my companion who aren’t especially afraid of heights, often involved not turning around. I did a couple of times, and I imagine that if I were afraid of heights panic might have set in. (I would not recommend descending those steps in Praiano to anyone, really, but especially not to people with any fear of heights at all. It’s one thing to have the cliff at your back, but quite another to be facing the drop while propelled forward and downward.) But even setting this portion aside, the trail itself is often narrow (see the picture just above) and you’re often separated from the edge by a rickety excuse for a fence or nothing. This is in my view the more useful assessment for the acrophobes:

“If you are like me and do not like heights or being close to the edge of some extremely high and sheer drop offs on somewhat treacherous terrain, either go shopping, wear blinders or just figure on dealing with it. I managed to wear out a pair of leather gloves by crabbing across rocks and clutching at rock walls and there were several times where I really thought panic seemed to be the only way to a merciful end of the fear. I am however delighted that I didn't give in as this is truly a spectacular hike. I do believe though that a career in mountain climbing is not in my future. ;-)”

Finally, signage. Once again, although there are several people providing accurate information, there are some strange claims being made on TripAdvisor and elsewhere:

“The trail is easy to follow and there are signs that guide you along the way, just follow the red and white marks left on certain landmarks, such as rocks.”

“We found the path to be very well marked…”

By no stretch of the imagination is the path well marked. You often have to search the rocks around you for a splash of paint that looks like an arrow, and one of the funniest parts to me was that you would have to feel your way and pretty much guess for a long stretch, only then to come to a tiny, inconspicuous sign when you no longer needed it. The consensus seems to be that the entrance to the path in Agerola is well marked, and I can’t confirm or refute that, but neither the entrance nor anything else in Praiano is, and the stairs down to Positano are extremely hard to find.

Despite some initial surprises and confusion, we had a great experience. But I think everyone should go into it with as much information as possible, so here are my thoughts and suggestions:

- If I have the opportunity to do it again – and I hope I do soon – I would probably try it from Agerola. If you’re staying in Positano, though, it seems like this can be a bit of a pain, since you have to take a bus or ferry to Amalfi and then a connecting bus to Agerola.* Several people reported that the buses can be crowded, and the whole process can take more time than you might want to spend before you even get on the trail itself. (A few people mentioned hiring a private driver to take them, but the idea of spending 60 euros or more to get to a free hiking trail seems strange to me.) I’ll reiterate that the experience is well worth these minor difficulties.

But I would also probably do the Praiano stairs again, now with the full knowledge of what I’m in for. I would recommend the Praiano climb to people who a) want a serious workout,** b) want a sense of accomplishment, c) are physically fit, and d) have no or minimal issues with heights. (As I said above, I wouldn’t recommend descending in Praiano to anyone, really.)

- The path is moderately difficult in parts, but absolutely worth it as far as I’m concerned. The climb has probably to some extent colored my impression of the path as a whole, making it seem slightly more difficult than it would be for those beginning high up in Agerola, but still people should know that reports that it’s an easy stroll aren’t correct. We had possibly the best conditions: it was fairly early in the morning when we began; we had some light, misty clouds; and it wasn’t an extremely hot day, especially for mid-July. Even under these conditions, there were sections that were more difficult and strenuous. I don’t think you have to be in perfect shape to walk it, but don’t expect an easy stroll.

- For people who are slightly afraid of heights, I wouldn’t say “Don’t do it.” You should absolutely do it if you want to challenge or push yourself. But you should go into it with the knowledge that it might be a challenge. Claims that the trail doesn’t contain sections that would be stressful to people with some fear of heights aren’t correct. But if you want to try to confront and reckon with such fears, I can’t think of a more beautiful place to do it.

I hope this is useful information. Please feel free to ask questions or share your own impressions or advice in the comments.

*Of course, if you’re already at or near that end of the path you would want to begin there. Wherever you plan to begin, if you’re taking the bus you should probably have written on a card (or be able to say in understandable Italian) “Sentiero degli dei” and the name of the exact stop you need to get off, and show/tell this to the driver. It doesn’t hurt to ask the other passengers, either. Even if the driver doesn’t know what you’re talking about, you might well run into others going to the path – you’ll know them by their clothes - or locals who can show you where to enter. Speaking of locals, I was pleasantly surprised to discover that the hikers weren’t all foreign. We met several Italian people, and later learned from some local people that they’ve walked it many times.

**As one person said: “We started in Praiano, hiked up 2000 stairs, then trekked over the mountains back to Positano. There are other approaches to the path - I was told this is the best one if you want exercise (which I did after 2 weeks of Italian dinners!).”