Dr Herbert earned her medical degree at the Columbia University College of Physicians and Surgeons. She holds a doctoral degree from the University of California, Santa Cruz, where she studied evolution and development of learning processes in biology and culture in the History of Consciousness program and then did postdoctoral work in the philosophy and history of science. She trained in pediatrics at Cornell University Medical Center and in neurology and child neurology at MGH. For her neuroimaging research and its implications, she received the first Cure Autism Now Innovator Award. She is co-chair of the Environmental Health Advisory Board of the Autism Society of America (ASA) and directs ASA's Treatment Guided Research Initiative (TGRI). Some of her papers are available on her website, www.marthaherbert.com.

Alternative Therapies (AT): You have a very broad educational background, including your doctoral work on the evolution and development of learning processes in biology and culture. What brought you to study and then practice medicine? More specifically, pediatric neurology and your focus on autism?

Martha Herbert, MD: My PhD is actually from a program called the History of Consciousness at the University of California at Santa Cruz. My dissertation was on evolution and development of learning processes using the texts of some social theorists, including Piaget, Habermas, and Marx. But using these texts as pivot points, it was able to be very broad. When I did my defense, my committee told me that while we are in a "history of consciousness" program where most people ignore the name, I was actually taking the name of the program seriously in what I was studying. I had a remarkable dissertation advisor, John Halverson, who's now passed away, who was able to encompass this approach and led me through a great range of readings. The History of Consciousness pro-gram at that time was a precious interdisciplinary haven, located on the magnificient and vision-inspiring campus of UC Santa Cruz, where this kind of breadth was fostered.

I had a particular interest in emotion, partly from some therapeutic processes that I had been trained in that involved catharsis. This got me fascinated with the biology of laughing and crying and yawning and fear expressions and trembling, which led me to pursue literature on these things, thinking that I wanted to do a dissertation on the evolution of emotions, but I didn't find much that was very useful to the questions I had developed from my observations and experience. I mean, there was Darwin, but there really wasn't enough of a systematic body of literature that could have allowed me to do something like that taking a dynamical approach.

While I was doing that work, I took an independent study with a neuro-linguist on the neural anatomy of emotion, and I be-came completely entranced with brain anatomy. It was so intricate and detailed. And it was a welcome relief from some of the more speculative work that I'd been reading in anthropology, philosophy, and psychology. I had started college years before as a biology major, and so I always had kind of a first love of biology, and getting into the brain anatomy while I was in graduate school provided me with a focus that felt like going home and, at the same time, linked to these broader questions.

I had been thinking about medical school, partly because of the science and partly because of the chance to legitimize the clinical side of this therapeutic work that I had done. I resisted this thought because I was getting a PhD and to start over and get an MD was a bit daunting. So I worked with some evolutionary biologists and comparative psychologists studying the evolution of emotion. But the bottom line was that they weren't asking the questions that I wanted to ask. So I bit the bullet and finished off my pre-med degree and got into various medical schools, and I chose Columbia.

When it came time to apply for residency, I wrote about a project I'd done as an elective about the relationship of brain and neuropsychological evaluations to brain localization. When I wrote about that interest in my application for the neurology pro-gram, my application was picked out of a pile at Mass General because the hospital had a lab that was doing exactly that.

When I got there, I still really wanted to study the biology of emotion. But when I finally had enough time late in my training to attend some professional societies that focused on emotion, I found that their concepts of emotion were very static: there's fear and there's anger and there are these behaviors and these facial expressions and these brain regions associated with each distinct emotion. That was really remote from what my own thinking had been, which was that emotion is a dynamical process that in-volves integrating information that's disparate and maybe full of painful cognitive dissonance. The emotional processing allows you to bring different pieces together into some kind of a new synthesis. There was nothing even remotely like that in emotion research. And there was no way that I was going to go into a social-sciences emotional research field; it was completely arid, bor-ing to me.

Meanwhile, I was handed a big pile of brain images of children with autism. I got 93 MRI (magnetic resonance imaging) scans, and I started analyzing the data. That's how I got into autism. They just basically handed me this giant pile of already col-lected data and said, "Do something with it." So I rolled up my sleeves and did imaging analysis and then data analysis. It took several years. As I got into the data analysis, I realized that this was something very, very different than what everyone had been telling me it was. Because the analyses that we'd initially set out to do were complete failures.

But alongside the lack of significance of the correlations between specific brain regions and specific behaviors there were other things that were crying out to be noticed-in particular, the striking and widely distributed increases in white matter volumes in both autism and specific language impairments, as well as the alterations in brain asymmetry in both conditions that were dis-tributed very widely beyond regions "specific" for language and that were distributed in almost the same way in both ostensibly distinct conditions. These pervasive changes blew the old modular model and also blew the model that these conditions are clearly distinguishable, and I had to fight with my senior colleagues to get them to take these other findings seriously. I really came into my own integrity as I said to myself, "I'm not going to try and look for what I'm supposed to be looking for. I'm going to call what I see and do whatever it takes to say what I think about it." That's when I began to open up to the incredible challenges that autism poses to just about everything. That's how I got into autism.

AT: Were you aware of any other researchers who were translating emotional components to the physical and biological compo-nents of brain?

Dr Herbert: There may have been, but partly it was where I was looking at the time-as an academic graduate student I was look-ing for "scientific" and "rigorous" literature with pretty conservative criteria for meeting those bars. When I looked in the neuro-logical literature for the neurobiology of laughing and crying, I read about things like gelastic seizures (ie, uncontrollable laughter disconnected from emotion) but not emotion as information processing. I didn't find what I was looking for either in the psychiat-ric or cognitive neuroscience literature I read. The closest I came was some neuroanatomical work by Sanides on the evolution of cortical and limbic connectivity that reversed commonly held assumptions about the order of evolution of primary and associa-tional processing and commentaries upon this work, and I still hope to get back to pursuing and writing up what I found in this literature. From time to time I read about things like neurobiology of ritual trance or cathartic therapies where the questions were closer to those my earlier training and experience had posed, but there was a big gap between, say, Wilhelm Reich or Art Janov or bio-energetics and what you can rigorously support or even explain to anyone in an academic situation. There was a yawning chasm. And I didn't see bridges at that time. Then I got into neuro-anatomy, but I got into it in, again, a static framework. We were studying brain volumes as indicators of fixed brain deficits. So while it had its fascinations, it wasn't an immediate bridge. Dama-sio's work on emotion bridged somewhat, but by the time it appeared I was already well down this other anatomical path.

AT: Has your affiliation with Harvard been a boon to your work as a researcher?

Dr Herbert: I got into the residency program at Harvard, and I never left. I went to Mass General, which is a premier teaching hos-pital at Harvard Medical School, to do my neurology training, and I've been there ever since. Mass General has a marvelous aris-tocratic tradition of graciously persistent inquiry and discourse, and I was privileged to learn from some of the senior people in the department who had so deeply shaped the field. I was hired by Harvard to be a trainee, and then I stayed. I stayed in the imaging analysis lab, and now I'm doing all kinds of other things, too. Certainly having this affiliation increases my profile. As for the envi-ronment there, it's tolerant in many ways, which I cherish, and there are remarkable people at Harvard, MIT, and other places to talk with, but it's not nurturing. So you can think broadly, but as a researcher you have to kind of make your own way.

AT: Is the TRANSCEND program the first of its kind in approaching autism and these other neuro-developmental disorders in an interdisciplinary manner?

Dr Herbert: Yes, probably-at least in the way we're doing it. In the brain domain in autism and neurobehavioral conditions, you have a predominance of the gene-brain-behavior model that says, "Oh, it must be genetic, and therefore the gene directly affects the brain, and the brain causes the behaviors." Then there's another part of brain investigation done by neuropathologists, who look at brain tissue in people who have passed away, where they have the resolution to look more at the level of cellular biology and chemistry. There's a little bit of chemistry in brain imaging if you do spectroscopy, otherwise not. And then there are people who do pathophysiology, such as biochemistry or immunology, which is more often than not systemic rather than brain-based.

In much neuroanatomical work in autism, interpretation of data has focused on identifying clues that could suggest develop-mental processes that might have gone awry, that would in turn point to potential genetic mechanisms. I spent a long time trying to do this myself, and I have written a series of literature reviews and read most of what has been there to read. But my work over the years has led me to question the strongly held assumption that autism is a neuro-developmental disorder that is wired in before you're born, a "static encephalopathy."

The "static encephalopathy," hard-wired assumption is certainly entrancing. It sort of makes sense because after all, autism does start early, and it sure seems like a life sentence. Even so, I began to realize that there are alternatives to the ways some of the findings being used to support this idea are being interpreted. For example, some brain studies looked at tissue in people who died and saw cellular changes that looked like they probably happened before the individual was born. But this was an interpretation of the ar-rangement of cells. One example is tightly packed cells in the limbic system; another example is changes in the brain stem. These brain tissue changes were found in less than a dozen brains each. So on the basis of a small number of brains, global inferences were made that this must have all happened in the third to fourth week of gestation or the 30th week of gestation. This interpretation be-came a "fact" that actively blocked funding of postnatal processes in autism-I have watched this blocking occur in grant review processes.