Looks like Autism and neurodevelopmental disorders are going to be re-classified in the new DSM-V


Some of the work I did for my dissertation dealt with “nosology“, the categorization and classification of symptoms, signs, syndomes, and diseases. I took a class in neuropsychology with David Tucker, an excellent teacher and clinician who got my interest in this subject going. Clinical neuropsychologists confront the problem of how complex an individual’s experience is, and diagnostic criteria may not capture this very well.

A minor theme of my dissertation was the particular issue of knowledge representation for cardiac “body knowledge” or “body cognition” disorders compared to autism. Psychiatrists, neurologists, pediatricians, psychologists, and other clinicians wrestle with how different one autistic patient is compared to another. The new classifications for autistic spectrum disorder coming out in the 2013 DSM-V will re-work how autism is defined, hopefully leading to better diagnoses. I write about this issue for DailyRX:

“Much discussion has centered on exactly who should be considered autistic, based on which diagnostic rules doctors should use. Diaglogue among clinicians, scientists, and patient advocates has focused on the proposed reworked definitions to be published by the American Psychiatric Association’s Fifth Edition of the Diagnostic and Statistical Manual of Mental Disorders in mid-2013.

Currently, the 4th edition of the DSM categorizes autism, Asperger’s disorder, childhood disintegrative disorder, and “pervasive developmental disorder not otherwise specified” as separate conditions.

If the proposed changes are indeed ratified and published, the larger category of “autistic spectrum disorder” will be used to categorize individual experience and behavior, ranging from mild to severe impaired functionality.”



A study where the brain’s complexity is nested in a lovely, lucid, elegant layer of simplicity


Most of the time when you study neuroscience you get the sense that every scale you look at, every system and subsystem you examine, every mechanism you investigate, is amazingly intricate and complicated. Turtles all the way down, so to speak. Just browse neurodynamics guru Walter Freeman’s free content if you don’t believe me.

I have on my table Eric Kandel and colleagues formidable Principles of Neural Science. You could actually get strong lifting this hefty tome. While over 1,000 pages it does not cover body cognition or mechanisms of interoception or neurophenomenology or body knowledge disorders or neurodynamics as I might prefer. There is just too much for one book to cover. Biology is like that. You could spend a year reading it and the field would have advanced…

And yet, and yet. Sometimes you find a study where all the complexity is nested in a lovely, lucid, elegant layer of simplicity. The baroque, spiraling layers of structure and process are folded into a rational, understandable, even beautiful, architecture:

A recent study by Van J. Wedeen of the Department of Radiology at Massachusetts General Hospital and colleagues reported surprising results from imaging brain fibers.

The data shows the brain fibers are aligned into an unexpectedly simple grid-like structure, rather similar to intersecting streets.

Can conscious experience affect neural states via “downward causation”?


My students have been asking questions about how the chemical processes in the brain are related to emotions.

This is forcing me to really think about how neurophenomenology should relate to dualism, monism, panpsychism, reductionism, elimitivism, and other stances regarding conscious states and neural states.

In a broad sense I favor Merleau-Ponty’s notion of la corporeal, which conceives of that physiological body that is observed by science as in some fundamental sense the same as my body, my flesh, embodied me, the foundation of my experience.

What about the specific, in principle falsifiable question of whether conscious experience can affect neural states via “downward causation”? I found a few resources to address this question.

I first encountered this idea in an Omni interview with the eminent neuroscientist Roger Sperry:

“In wrestling with the split-brain problem, I realized that this kind of interaction with objects requires that consciousness have a causal impact on brain activity. Consciousness can be viewed as a higher emergent entity that supercedes the sum of it’s right and left brain awareness”

Once you have an interesting idea like that to ponder, you are primed for more! For what’s it worth, Dave Demaris, coming from a neurodynamics, computational neuroscience and systems engineering background, tells me he finds nothing controversial about downward causation from mind to brain. I suspect Francisco Varela was ahead of all of us on this, I am still very slowly working my way through his massive output. Someone needs to get Walter Freeman to get his views on this documented (for what’s it worth, Walter told me in 2001 his general views on mind and brain are closer to Varela’s than people seem to think).

I found thoughtful, provocative writing from the clinical psychologist Brian Kohler at http://www.isps-us.org/koehler/neurophenomenology.htm:

“Neurophenomenology relies on two key concepts: emergence and embodiment. Emergence extends and enriches the notion of natural causation, without violating the supposed causal closure of physics. Emergence entails both upwards and downwards causation. Embodiment provides the tools for criss-crossing the ‘explanatory gap’ between first-person phenomenology and third-person neuroscience. This is not closing the gap via reductionism, rather it is a way of moving productively from the one domain to the other by way of a third mediating domain, ie, dynamical systems.Varela and Thompson (2003) noted:“Given that the coupled dynamics of brain, body, and environment exhibit self-organization and emergent properties at multiple levels, and that emergence involves both upwards and downwards causation, it seems legitimate to infer that downwards causation may occur at multiple levels in these systems, including that of…cognitive acts in relation to local neural activity” (p.276).

These authors cited the idea of J. A. S. Kelso who wrote: “Mind itself is a spatiotemporal pattern that molds the metastable dynamic patterns of the brain.” Walter Freeman described consciousness as an order parameter and state-variable operator in the brain that mediates relations among various neurons. According to Freeman, mind is not epiphenomena, rather, it plays a crucial role in intentional behavior-it is the task of the neurodynamicist to define and measure what that role is. Another good example, and clinically useful, of ‘downward causation,’ is recent research on human epileptic activity. There is evidence that subjects can voluntarily affect the conditions leading to the initiation and course of seizure activity ( see Francisco Varela & Evan Thompson’s “Neural synchrony and the unity of mind: a neurophenomenological perspective” in Axel Cleermans’ edited volume “The Unity of Consciousness: Binding, Integration, and Dissociation” published in 2003 by Oxford University Press).

Epileptogenic zones are embedded in a complex network of other neural regions that actively participate in mental life. These networks are multiple and distributed over a large scale. The global level of integration ( the result of ‘upwards causation’ ) may produce ‘downwards’ effects, acting eventually upon the local level of the epileptogenic zones. Recent studies by Varela and colleagues have demonstrated that there are deterministic temporal patterns within the apparent random fluctuations of human epileptic activity, and that these patterns can be modified during cognitive tasks (Le Van Quyen et al 1997). Varela and Thompson (2003) concluded: “ …the act of perception on the part of the patient contributes in a highly specific manner, via the phase synchrony of its associated neural assembly…to pulling the epileptic activities towards particular unstable periodic orbits. Thus downwards causation need be no metaphysical will-o’-the wisp, but can be an empirically tangible issue” (p. 277). “

I successfully defended my dissertation and am getting my Ph.D


I can’t tell you how much work this whole thing has been!

Note to anyone hiring postdocs or profs: I identify as a cognitive scientist, not a psychologist. The University of Texas is notating transcripts to reflect their approval of my completion of a program of work via the Ad-Hoc Interdisciplinary PhD, and labeling it as Medical Cognitive Science.

The title is “”Modeling the clinical predictivity of palpitation symptom reports: mapping body cognition onto cardiac and neurophysiological measurements.”

This dissertation models the relationship between symptoms of heart rhythm fluctuations and cardiac measurements in order to better identify the probabilities of either a primarily organic or psychosomatic cause, and to better understand cognition of the internal body. The medical system needs to distinguish patients with actual cardiac problems from those who are misperceiving benign heart rhythms due to psychosomatic conditions. Cognitive neuroscience needs models showing how the brain processes sensations of palpitations. Psychologists and philosophers want data and analyses that address longstanding controversies about the validity of introspective methods. I therefore undertake a series of measurements to model how well patient descriptions of heartbeat fluctuations correspond to cardiac arrhythmias.

First, I employ a formula for Bayesian inference and an initial probability for disease. The presence of particular phrases in symptom reports is shown to modify the probability that a patient has a clinically significant heart rhythm disorder. A second measure of body knowledge accuracy uses a corpus of one hundred symptom reports to estimate the positive predictive value for arrhythmias contained in language about palpitations. This produces a metric representing average predictivity for cardiac arrhythmias in a population. A third effort investigates the percentage of patients with palpitations report actually diagnosed with arrhythmias by examining data from a series of studies.

The major finding suggests that phenomenological reports about heartbeats are as or are more predictive of clinically significant arrhythmias than non-introspection-based data sources. This calculation can help clinicians who must diagnose an organic or psychosomatic etiology. Secondly, examining a corpus of reports for how well they predict the presence of cardiac rhythm disorders yielded a mean positive predictive value of 0.491. Thirdly, I reviewed studies of palpitations reporters, half of which showed between 15% and 26% of patients had significant or serious arrhythmias. In addition, evidence is presented that psychosomatic-based palpitation reports are likely due to cognitive filtering and processing of cardiac afferents by brainstem, thalamic, and cortical neurons. A framework is proposed to model these results, integrating neurophysiological, cognitive, and clinical levels of explanation. Strategies for developing therapies for patients suffering from identifiably psychosomatic-based palpitations are outlined.

The challenge of building a clinical neurophenomenology of palpitations and cardiodyamics


from spiritualnetworks.com

Heart disease is a leading threat to health, and people worry when they feel changes

in their heartbeat. Should clinicians trust descriptions of these palpitations? How should

clinicians and scientists model personal, phenomenological statements about what is

happening inside the body of a subject or patient? In an increasingly standardized,

scientific, and objective world of medicine, what role is there for a doctor’s intuitions and

instincts about a patient’s bodily sensations?

These are not simple questions, as they attempt to straddle a fundamental duality

between patients understood as embodied persons in an existential context of health and

disease, and humans understood as systems, as bio-machines modeled by science.

Clinicians collect measurements and interpret data about that category of object known as

a human body, and must compare this externalized, ostensibly objective, techno-scientific

knowledge to their patients’ description of what their bodies feel like. Like the Roman

god Janus, the healer faces two worlds. As modern medicine becomes more involved

with science, integrating these two domains requires ever-more flexibility,

thoughtfulness, and careful techniques for acquiring and modeling data. As complex as

these practical necessities are, science peers even deeper, into the meaning of the often

enigmatic gap that can exist between patient descriptions of the heart speeding up or

missing beats, and the lack of corresponding measurement of heart electrical output as

measured through electrocardiograms (ECG). Medicine needs good approaches for

distinguishing palpitations of psychosomatic origin from those with cardiac etiology, as

well as general guidelines for the trustworthiness of patient-reported data about their

bodily sensations. Science needs to understand what mechanisms in the brain, body, and

mind explain both accurate and inaccurate palpitations reports.

Knowledge of, and theories about, fluid dynamics, hematology, processing of cardiac

state by the peripheral nervous system, receptor activation, hormone binding, protein

signaling, up-regulation and down-regulation of genes, and models of perfusion support

sophisticated models of cardiodynamics. Yet the heart can be thought of in rather more

intuitive terms, as a pump made of muscle that moves oxygen-poor blood to the lungs,

and newly oxygenated blood to the rest of the body. Electrocardiograms show the

rhythms of this pumping as sometimes more regular and periodic and other times less so.

But how much personal knowledge do patients have about what the heart is doing?

Personal knowledge of the body is a problem for mechanistic science. While cardiac

periodicity is an object of scientific measurement, and therefore clinically and

epistemologically privileged for scientists constructing explanations, the personal

experience and phenomenological knowledge of the body may be considered merely

subjective opinion or anecdotal.

People introspecting about their interior sensations sometimes report to their doctor

that their heart is racing, pounding, or skipping beats. In some instances, such data are

compared to that from publically available sources such as ECG, and a diagnosis of

cardiac etiology is made, but in other cases doctors believe the patient is  psychosomatically

cognizing benign heart rhythms as dangerous. When measurements of cardiodynamics do

not correspond well to unwelcome sensations of altered heartbeats, how should medicine

and science understand the discrepancy? This work addresses this problem directly,

by modeling the probabilities that a patient’s experience corresponds to

a medically important heart rhythm disorder. For the patient, feeling a change in

the rhythm or intensity of this fundamental aspect of ongoing embodied existence can

be very worrisome. When the cause is psychosomatic, medicine categorizes it as unexplained,

and cognitive neuroscience faces an explanatory challenge. Somewhere between the cardiac nerves,

brainstem, thalamus,and cortical regions, normal heart rhythms are processed as abnormal and

threatening, but why?

A true understanding of such a gap between personal bodily feelings and cardiac

measurement requires an implicit or explicit mapping of cardiographic, radiological, and

other data onto a description from the patient about what is going on inside their body, or

vice versa. This is not the sort of problem that cognitive science has heretofore usually

focused on, but the field of medical cognitive science can apply ideas from neuroscience

to come up with an explanation. Current evidence (Damasio, 2010) suggests a role for

multiple areas in the peripheral and central nervous systems that process cardiac rhythm

signals, which are cognized into feelings of skipping beats and other abnormal rhythms

(Barsky, 2000).

Such theoretical problems aside, clinicians must apply complex psychological,

anatomical, neurophysiological, and etiological concepts to interpret their patients’

symptom reports. What patients have to say about what is happening in their bodies must

be taken seriously, but not necessarily believed. Traditionally, a doctor might have had

some intuition about the reliability of a patient’s description of their heart fluttering or

racing and would consider the possibility that emotions, stress, and existential or

psychological issues partially or mostly explain the diagnosis. Yet the demands placed on

modern clinicians increasingly constrict the time they may spend listening to the patient,

making it harder for them to get a rich description of the proper existential context

framing the presenting complaint. As such, the need for quickly ascertaining the

probability that palpitation symptoms have a cardiac or psychosomatic etiology becomes


What good is patient phenomenology in this new world of evidence-based medicine?

To determine this, I shall focus in particular on comparing the predictive utility of patient

palpitation reports for cardiac arrhythmias to other clinically predictive measures that do

not depend on introspective data from the patient. This predictivity will support the

differential diagnosis of cardiac-based palpitations against psychosomatic etiology, but

modeling how well symptoms correspond to physiological measurements can also serve

to operationalize what I will term “body cognition” and “body knowledge.” Palpitations

are usually defined as unwelcome awareness of cardiac activity (Barsky, 2000), such as

skipping, racing, or thumping heartbeats. Do people with such presenting complaints

have heart rhythm abnormalities requiring medical attention, benign heart rhythm

fluctuations, or normal heartbeats somehow sensed as strange, unpleasant, and abnormal?

Evidence suggests that patients reporting palpitations who have an anxiety disorder are

less likely to have arrhythmias (Abbott, 2005), but the reasons people with normal heart

rhythms report palpitations must be considered a mystery for science, and a challenge

(Barsky, 2000).

Five MD’s respond to my questions about the believability of symptom-report phenomenology


Q-As a doctor, when a patient tells you they are experiencing “palpitations” or “racing heart” or “skipped beats”, what affects whether or how much you believe them? What else would they have to say for you to think they are accurately reporting on their physiological state?

1: “Regarding patients reporting palpitations. I almost always believe them about that. It’s pretty easy to tell when you have palpitations, and when they are significant they scare the shit out of people, so they notice, remember, and report. If I see a patient in atrial fibrillation, and they say they’ve noticed palpitations for 3 months, I will take that as fairly strong evidence that they have had atrial fib with a rapid ventricular rate for approximately 3 months. It’s not a commonly misperceived experience, I don’t think. There’s not many things that masquerade as ‘palpitations’ that aren’t. Maybe gastroesophageal reflux could make people feel that like they are having palpations when in fact they are refluxing, in which case taking a careful medical history would clear this up.
so what effects how much I believe them…
-the patient’s medical history: does the patient have a history of cardiac arrhythmias? Do they know what tachycardia feels like? If yes, then it makes the subjective report very reliable. for example, if a patient is having chest pain and says “this feels just like my last heart attack” then I would take the complaint very seriously.
-the history of present illness: the details of the patient’s chief complaint can be very useful in the interpretation of their “palpitations”. for example, what was the onset? if suddenly the patient noticed (felt) their heart racing, that would be more consistent with an ectopic cardiac foci suddenly misbehaving than if the patient described a gradual increase in awareness of a racing heart (maybe they were getting stressed out, and developed a gradual sinus tachycardia, or maybe they were exercising and developed a physiological tachycardia).
-patient’s mental status: this prolly goes without saying, but is the patient alert and oriented to person, place, time? do they understand why they are in the hospital? do they make sense? can you believe anything they say? are they rational and coherent in they’re descriptions of other physiological processes?”

2 “Even though they are generally synonymous…I take “palpitations” and “racing heart” more seriously because both connote speed and a rhythm problem; whereas, “skipped beats” can mean a normal rate but with an arrhythmia. As for accuracy and belief, I tend to find patterns or repetition of symptoms more believable (which is not accuracy but precision). If it happens once, it is easier to shrug it off.”

3. “I think I generally believe that complaint since there’s no reason to lie about that sensation- not like they’re looking for narcotics. Things that would make me more concerned would be light headedness, loss of consciousness, chest pain, family history of arrhythmia or sudden cardiac death. “

4 “I tend to believe the patient. Whether there is real disease or not it’s a concern and it’s real to them- I rather not assume the patient is just histrionic or borderline. Then I check their vitals, evaluate their risk factors (age, gender, etc.) for the different etiologies. Then based on the possibilities I order studies if necessary (TSH, EKG, Holter, etc.). If work up is negative then reassure, reassure, reassure. “

5 “I would rank my reliance on phrases as: 1) racing heart, 2) palpitations, 3) skipped beats. Mainly because palpitations is such a low-frequency word in general usage. Although I do interpret them all as palpitations per history with equal reliance on patient accuracy. “

What do clinicians come to know about their patient’s heart sensations?


What do clinicians come to know about their patient’s heart sensations? This is not a simple question, as it simultaneously looks at patients as people with bodily experiences, but also as humans understood as systems, as a sort of living machine. What is more intimate than our heart-beating, a familiar yet mysterious sensation we know to be at the very basis of our ongoing experience? Feeling a change in the rhythm or intensity of this fundamental aspect of our embodied existence can be very worrisome. Should clinicians believe patients who complain of cardiac rhythm changes? How accurate are people at detecting medically important heart-beat fluctuations? How should clinicians understand the relationship between symptoms as reported by the patient, and underlying physiological processes? These are complex and multifaceted issues, requiring nimble clinicians who integrate scientific knowledge as well as intuition about what the patient is experiencing bodily. Clinicans develop knowledge of their own bodies through life, and then are required to learn complex anatomical, physiological, and etiological concepts they will use to interpret their patient’s symptom reports. What patients have to say about what is happening in their bodies must be taken seriously, but not necessarily believed. The interrelated problems of how clinicians interpret patient verbal reports, reason about the relation between these reports compared to measurements and scientific models, and then make judgments about the patient’s accuracy in knowing about their own bodies are topics well worth honing in on, and to my knowledge, not throughly explored from a neurophenomenological perspective.

These acts of clinician cognition concerning their patient’s symptoms are framed by an evolving social and professional context. Modern medicine, like the Roman god Janus, stands two-faced, towards healing as an art, but also towards scientific models of disease. In the current era, what is known as “evidence-based medicine” requires an important shift in how clinicians operate, from historically rather unfettered individual judgments in some contexts, to increasingly accepting consensus-developed guidelines formulated from reviews of previous findings. Clinicans who have with great effort developed the ability to intuit diagnoses may have to defend their familiar constructs, criteria, heuristics, and practices if these are not bolstered by peer-reviewed studies, randomized clinical trials, systematic reviews, Bayesian statistical approaches to clinical problem solving, meta-analysis of previous data, and effectiveness metrics. Medical organizations can mandate “best practices” of patient care, “gold standards” of cost-effectiveness for ordering certain tests, references to efficacy criteria that must be satisfied before a program of treatment is established, and more. This ongoing process is transforming medicine, requiring that the traditional art of diagnosis based on years of education and experience be integrated with operationalized definitions, committee-approved metrics, and greater formalization, thus constraining individual opinion and practices in favor of organization-mandated standard operating procedures. Can symptoms based on an individual’s embodied experience be given proper attention in this brave new world of medicine?

I hope that more researchers would address the clinical aspects of neurophenomenology. This is a relatively new and undeveloped area. While William James and Erwin Straus were clinicians, as is Antonio Damasio, other pioneers such as Maurice Merleau-Ponty and Francisco Varela backgrounded medical concerns somewhat (however, if you are unaware of Varela’s haunting work at the end of his life “Intimate Distances -Fragments for a Phenomenology of Organ Transplantation“, it is a must-read.) Shawn Gallagher has made an excellent synthesis of philosophy and clinical studies in “How the Body Shapes the Mind“, a work that bears greater attention from the small community of neurophenomenology researchers.

For my part, I shall focus in on a particular area, palpitations, where changes towards operationalizing and standardizing the definition of “clinically significant” symptoms are occurring, with the aim of modeling the relationship between patient symptom reports and “significant” arrhythmias as revealed on ECG measurements. I will especially focus on how the predictive utility and accuracy of the reports can be operationalized, and attempt to represent for one domain how patients’ verbalization of their phenomenological state can be “mapped” onto measurements of cardiac rhythm abnormalities.

How to operationalize the “body-knowledge” construct so it can be analyzed and measured


I started using the term “body-knowledge” a few years back as a way to label the extent to which people can accurately report symptoms or interior sensations. It is not as of 2010 a popular term. The earliest citation of the term I know of is from a clinical neurology paper by Sirigu, Grafman, Bressler, and Sunderland, (1991): Multiple representations contribute to body knowledge processing: Evidence from a case of autotopagnosia

“Body knowledge” does quite not have the same meaning as “embodied cognition”, “body image”,”body schema”, “interoception”, “visceral perception”, or even “body cognition”, though there is considerable overlap. I typically use the concept body-knowledge to emphasize the verbal reporting of internal states. My epistemology teacher years ago taught me a great idea:

To know something, you have to know that you know it, and to know that you know it, you have to be able to say it.

I wouldn’t defend that as the end-all be-all theory of knowledge, but it works as a heuristic at the least. For now, I use “body-knowledge” to refer to how well people can know and verbally report on what is happening to their physiological states.

To analytically probe this construct, I started looking very deeply at a particular domain: symptom reporting about cardiovascular processes. I have found some useful results from earlier studies that serve as a guide to help approximate how accurate people are when they feel and report palpitations: their heart is racing, they feel irregular beats, heart thumping or pounding, skipped beats, and so forth. Evidently a fair amount of the time people suffering from panic disorder or anxiety “cognize” otherwise benign sensations and report heart problems, and such false positives adds a great deal of expense to the healthcare system.

Symptom report accuracy is a largely unexplored area for the young field of neurophenomenology: how much of what is happening inside our bodies is accessible to our minds? Very little of the existing neurophenomenology literature deals with these issues.

How can the “body-knowledge accuracy” construct be operationalized, analyzed and measured? For the particular domain of palpitations reporting, here are some useful core metrics:

From ‘The Validity of Bodily Symptoms in Medical Outpatients,” (Barsky, 2000) Chapter 19 in The Science of Self Report (Stone, A, ed): -When patients complaining of palpitations undergo 24-hour, ambulatory, electrocardiographic monitoring, 39% to 85% manifest some rhythm disturbance; the vast majority of these arrhythmias are benign, clinically insignificant, and do not merit treatment). Although as many as 75% of these patients with arrhythmias report their presenting symptom during monitoring; in only about 15% of cases do these symptom reports coincide with their arrhythmias.

From Barsky, Ahern, Delamater, Clancy & Bailey (1997): -145 consecutive outpatients referred to an ambulatory electrocardiography (Holter) laboratory for evaluation of palpitations were accrued, along with a comparison sample of 70 nonpatient volunteers who had no cardiac symptoms and no history of cardiac disease. A symptom was considered accurate when it followed within 30 seconds after any demonstrated arrhythmia.

-average positive predictive value (PPV)… is equal to the number of reported symptoms that were preceded by an arrhythmia divided by the total number of symptoms reported (true positives / [true positives + false positives]).

-Ninety-nine palpitation patients (68%) reported at least one palpitation during monitoring. Among those patients who were symptomatic, the mean number of diary symptoms reported in 24 hours was 3.7. The mean PPV for all symptom reports among palpitation patients was 0.399, compared with a mean PPV = .118 for the nonpatient volunteer sample (p = .01).

-the palpitation descriptors most likely to be accompanied by electrocardiographic abnormalities are heart stopping, fluttering, and irregular heartbeat. The least predictive descriptive terms used by the patients were racing and pounding.

-34% of the symptomatic palpitation patients and 11% of the asymptomatic comparison subjects were classified as accurate reporters

experimental physiology, interoception, and body knowledge

interoception, Uncategorized, visceral perception

The work of experimental physiology researchers, such as Gyorgy Adam (1998) and Oliver Cameron (2002), as well as the psychologist James Pennebaker (1982) have opened up our understanding of how our cognition, brain, and peripheral nervous system allow us to know about our internal states. Such research probes the ability of people to sense and perceive processes and events interoceptively or visceroceptively-sensation via the visceral organs, typically the gastrointestinal tract or cardiovascular system. In a series of experiments lasting decades, described in Visceral Perception: Understanding Internal Cognition, Adam and colleagues (1998) painstakingly measured the ability of human subjects to perceive visceral changes. While much of their work is of primary interest to physiologists, for the present purposes it suffices to emphasize these claims and implications:

• The classical conception of the viscera, emphasizing their autonomic (maintaining homeostasis) activity, has not generally taken into account that sensory nerves innervate these organs.

• There are “dim and murky” but experimentally verifiable unconscious and conscious perceptions of these organs’ states.

• Conscious perception of visceral or interoceptive changes will tend to be “contaminated” with skin/surface based “somatic” perception, making “pure” visceral perception very difficult to verify.

• While cognitive neuroscience has fixated on the central nervous system, humans unconsciously and experientially are affected by various peripheral nervous system activities in the insides of their bodies, appropriately titled internal cognition (Adam, 1998).

Adam stresses the provisional, rather than the definitive, state of scientific knowledge of such concepts. Cameron (2002), in Visceral Sensory Neuroscience: Interoception, builds on Adam’s and other research, and attempts to contextualize what is known about internal body perceptions, skin and touch sensations, muscular control of movement, and other sensations. He addresses fundamental questions about how to taxonomize body knowledge, and writes that (pp. 274-275):
Rather than considering interoceptive processes, perhaps defining an overall bodily sense (or more than one-bodily senses) might be more appropriate…Would it not be more appropriate to define (as has been done by others) a bodily sense, including interoception, proprioception, labyrinthine function (i.e., the experience of the body in space), and other afferent information from the body?

The tentative appraisal of Adam and Cameron can only convey a sense of the complex nature of the phenomena at hand. Body knowledge, symptom perception, interoception, visceroception, and internal cognition are overlapping terms which need to be disambiguated. Yet this task is made all the more difficult because of the complexity of the problem of assessing how accurate people are at knowing their internal states.

Models of body knowledge are informed by practical medical and clinical needs. Physicians and clinicians routinely ask patients to report on their bodily sensations, while cognitive scientists, neuroscientists, and experimental psychologists often request subjects to verbally report on their perceptions. How true or “veridical” verbal reports are about objectively-measurable phenomena such as heart rate or blood pressure is a subject of debate. An extensive review by Pennebaker (1982) in The Psychology of Physical Symptoms of numerous results of studies measuring the abilities of normal humans to accurately report on physiological state found some limited evidence for accurate monitoring and reporting, but the bulk of data suggested people are poor at such tasks.

While this view may indeed represent something of a consensus, other authors (Fisher, 1966); (Adam, 1998) emphasize instances of relatively accurate capacity for interoception or perception of internal organ state, accuracy of perception of changes in external stimuli such as light or sound intensity (Stevens, 1975), practical necessity in clinical and medical contexts of asking patients to introspect and report on body state (Heilman and Valenstein, 2006) or numerous other perspectives that variously substantiate the accuracy and/or utility of verbal reports of sensation and perception. New data should be generated to shed light on two important (and overlapping) issues:
-the relative accuracy of verbal reports, which can be provisionally understood as representative fidelity to access to physiological state information.

-the neurobiological mechanism(s) by which people monitor and “get information” about their internal states.

The first issue above typically requires comparison of subject evaluations compared to objective measurements, while the latter usually involves imaging technology, such as EEG, fMRI, and other technologies. A model adequate to explain the results of both data sets will require bridging concepts that serve to link cognitive and biological levels of description. It should not be assumed at the outset whether or not information-processing models are up to that task. It may be that concepts from outside information-processing theory will be required to explain the data.