The New Phrenology

Robert Stern

Phrenology was Franz Gall’s (1758–1828) and later Johann Spurzheim’s (1776–1832) attempt to correlate behavior with skull shape. Gall published his work between 1810 and 1819 in four volumes, titled The Anatomy and Physiology of the Nervous System in General and of the Brain in Particular, with Observations on the Possibility of Discovering the Number of Intellectual and Moral Dispositions of Men and Animals Through the Configurations of Their Heads. Gall’s original study found twenty-seven different cranial areas overlying specific and separate brain “organs.” By the end of the phrenology vogue in the early twentieth century, the number of “bumps” had risen to forty-two.

Phrenology proved very popular in mid-nineteenth-century America. Both Edgar Allan Poe and Walt Whitman were fervent believers. Poe provided phrenological descriptions of characters in his writings, and Whitman even published the results of his phrenology exams five times. Sarah Josepha Hale, author of Mary Had a Little Lamb and the editor of Godey’s Ladies Book (the most popular women’s magazine in America during the mid-nineteenth century), claimed that phrenology was “second only to Christianity as a force for the elevation and improvement of the status of women” (Hothersall 1995). Celebrities such as Clara Barton and Joseph Smith had their heads measured and assessed by phrenologists. President James Garfield was a believer, having had his head examined several times.

Much like today’s school counselors, phrenologists acted as life coaches, advising clients on their education and marriages, but based on head bumps rather than SAT scores. The appeal of phrenology in America lay in its emphasis on the detection of moral characteristics via physical characteristics and the belief that intellectual and physical attractiveness could be improved through the exercise of moral and religious living (Lintern 2012). Phrenologists developed and used precise measurements and rating scales. Electricity was eventually introduced into the cities, and phrenologists kept up with the times, investing in the Lavery Electric Phrenometer, patented in 1905 to measure skull bumps “electrically and with scientific precision.” (See also Geoffrey Dean, “Phrenology and the Grand Delusion of Experience,” Skeptical Inquirer, November/December 2012.)

Today we are living amid a modern phrenology. But instead of assessing moral characteristics, the new phrenology asserts that by demonstrating smaller brain regions—specifically the hippocampus and the amygdala—in patients with psychiatric disorders, it can provide scientific evidence of child abuse. Questions on the rating scales and self-report questionnaires have changed, and MRIs and fMRIs have replaced the scalp calipers and the Lavery Electric Phrenometer, but today’s phrenology is little changed from that of over a century ago—it’s still reading bumps and correlating them to morality. Only this time the bumps are on the inside of the skull.

The Right Tool for the Job?

MRIs and fMRIs (f is for functional) are astonishing machines. They’re big and noisy, and they allow physicians and scientists to look inside a body without cutting into it or using radioactive X-rays. The MRI consists primarily of a giant magnet powerful enough to align the water molecules in a body. By pulsing on and off and allowing the molecules to alternately fall back into random motion and subsequently realign, computer images can be constructed. Painlessly, and without fear of subsequent radioactive damage to DNA, fMRIs go one step further. They estimate organ function by measuring the blood oxygen being used by cells. Simply put, the MRI measures structure and the fMRI measures function. Both can be enormously helpful in detecting brain anomalies. But as with any tool, it depends upon whether or not it is used appropriately. Attempting to prove a history of child abuse with either tool is GIGO: garbage in, garbage out. Neither machine is capable of imaging patient history, and neither can prove or disprove that child abuse occurred. They can only image size and oxygen use.

Hippocampal Structure and Function

The Hippocampus (Latin for “seahorse”) is a structure that looks less like a seahorse than a tadpole in a yoga position with its tail flipped over its head. There are two hippocampi. Each arcs around the middle of the brain in the left and right hemispheres as part of a series of structures known as the limbic system. The limbic system is involved in the creation and regulation of emotions, learning, and memory, and the hippocampus acts primarily, though not exclusively, to process the formation of memories. If the hippocampi are severed, new memories of events are unable to be formed. Older memories, which reside globally in the brain, remain intact.

Chronic stress can affect hippocampal volume. The hippocampus shrinks as its neural connections to other parts of the brain atrophy, and the cells themselves die (Higgins and George 2007, 72). However, by the 1960s it was discovered that the hippocampal cells could regenerate, not only in mice and rats but in other mammals and in humans as well. Several decades and thousands of rodents later, that discovery has held up in replicated studies: the hippocampus regenerates cells in an attempt to return to normal functioning, i.e., the hippocampus responds to its environment. 

Those Who Do Not Study Pseudoscience Are Destined to Repeat It

Ever since Freud published The Aetiology of Hysteria in 1896, child (sexual) abuse has been assumed to be the etiology for psychiatric disorders. Freud disavowed his theory in 1897, admitting to his friend Wilhelm Fleiss that he exaggerated not only the number of patients he treated for hysteria but also their improvement—none of them improved as a result of his repressed child abuse theory or its practice (Freud 1897). 

  The term hysteria is no longer used; it’s considered sexist, although men have been diagnosed with hysteria since the nineteenth century and observed with hysterical behaviors long before then. Instead, hysteria has been subdivided into separate diagnoses based on specific patient behavioral presentations: borderline personality disorder for the erratic, impulsive types; histrionic personality disorder for overly emotional, attention-seeking types; somatization and conversion disorders for those displaying medically unexplainable dysfunctions of the body; dissociative disorders for those displaying medically unexplainable disorders of memory and mind. There are others, and they’re all hysteria.

These days, patients who have been diagnosed with any of the post-hysteria diagnoses are often simultaneously diagnosed with post traumatic stress disorder (PTSD). The presumption is that they must have been the victims of child abuse. This is ironic, as it’s the very theory that Freud himself discarded over a century ago.

Trauma, PTSD, the Hippocampus, and Some Wrenches

In studies, adults diagnosed with any of the post-hysteria diagnoses who have reported child abuse have undergone MRI imaging, and they have been found to have smaller hippocampal volumes than matched healthy controls. Proof positive? Hardly. Children with documented histories of abuse whose brains have been scanned tend to show no differences or even larger hippocampal volumes than their age- and sex-matched controls (Woon and Hedges 2008; De Bellis et al. 2001). 

Throwing another wrench into the abuse/hippocampal volume theory, Mark Gilbertson compared the hippocampal volumes of male Vietnam War vets diagnosed with PTSD to their twin brothers, as well as to a set of healthy controls. The researchers found that the vets exhibited smaller hippocampi than the controls; however, so did their non-combat twins (Gilbertson et al. 2002). In other words, rather than a result of trauma, smaller hippocampal volume may in fact be a factor in the causes of trauma. These findings have been replicated (van Rooij et al. 2015). 

Relying on brain images to diagnose specific mental disorders guarantees problems. First, is the imprecise nature of mental health diagnoses—diagnoses are based on behavioral symptoms. Any diagnosis can exhibit wildly different behaviors; there are no objective physical or laboratory signs. At least, not yet. Schizophrenics can appear depressed but so can unhappy or anxious people. Demented people can appear normal at first but be incapable of stating what year it is, how old they are, where they live, or who they live with when asked. Depressed patients can appear to be demented, psychotic, or intellectually disabled. Delusional people will often come across as perfectly normal until the right “button” is pushed and a string of paranoid rants spews forth.

Second, smaller hippocampal volumes have been found in persons with other psychiatric disorders: schizophrenia, Alzheimer’s dementia, major depression, PTSD, borderline personality, antisocial personality, panic disorder, ADHD, autism, and obsessive-compulsive disorder, to name only a few. In addition, hippocampal volume losses are also common in nonpsychiatric disorders: epilepsy, alcoholism, traumatic brain injury, Down syndrome, post-cardiac arrest, Parkinson’s disease, congenital adrenal hyperplasia, Turner’s syndrome … the list goes on.

Third, and most unfortunately for the child abuse/smaller adult hippocampi proponents, normal adult hippocampal volumes vary from 1,700 cubic mm to 5,680 cubic mm depending on age, sex, genetics, differences in boundary definitions, thickness of the MRI’s “sliced” images, etc. (Honeycutt and Smith 1995). Scans of the adults reporting child abuse fall within the range of normal, as do those of the depressed, borderline, PTSD, and even dissociative identity disorder (DID) patients. You’d expect the DID patients to have numerous outgrowths of their hippocampi to manage the memories of all those alter personalities. None has been found.

The Amygdala—Size and Symptoms

Along with smaller hippocampi, smaller amygdalae have often been implicated in the child abuse/adult mental illness studies. The amygdala, an almond-shaped structure sitting at the head of the hippocampus, is involved with fear responses secondary to sensory and cortical input. If the amygdala is damaged, fight or flight responses are hindered or suppressed. Several of the trauma studies found that the borderline patients reporting child abuse often have smaller amygdala volumes, and these studies are cited as more proof that child abuse leaves measurable damage later in life. However, when imaged with fMRIs the amygdalae of borderline patients were shown to be more—not less—active than the controls (Soloff et al. 2017). 

A Rose by Any Other Name

Beliefs rarely die; they evolve. Since the memory wars of the 1980s and 1990s, multiple personality disorder has been renamed DID (dissociative identity disorder) and Freudian repression has become Janetian dissociation, but the belief that child abuse leaves permanent damage has not dissipated. Today, the battle has moved from the television and the courts to professional journals. The once common term satanic ritual abuse has been softened and is now called ritual or just organized abuse; however, the purported long-term damages of child abuse have expanded. Child abuse studies now include not only the post-hysteria diagnoses (especially DID), but also schizophrenia, bipolar disorder, ADHD, antisocial personality disorder, eating disorders, and, of course, depression and anxiety. There have also been attempts to change the name of borderline personality disorder to complex-PTSD to emphasize the purported child abuse etiology of the disorder.1 Though proponents insist that those are separate diagnoses, the symptoms are the same. The extensive overlapping of all the disorders constitutes a diagnostic hash in which one diagnosis cannot be adequately differentiated from another, and two or more disorders are often diagnosed simultaneously. The terms may have changed, but we’re still looking at hysteria.

Size Isn’t Everything

Can function be inferred from cranial size? Neanderthals led lives that were nasty, brutish, and short—whether due to starvation or the whims of the weather or diseases or savage saber-toothed animals or even those upstart Homo sapiens. You might expect smaller hippocampal and amygdala and cortical volumes due to all that abuse. Unfortunately, no known Neanderthal brains have survived; it would be a fascinating study. Their skulls, however, have survived, and those have been measured. Neanderthals had larger brain volumes than we currently do. More than 200 cubic centimeters larger (Us ~ 1,350 cu. cc. Them ~ 1,600 cu. cc).

Figure 1: Massive ventricular enlargement, in a patient with normal social functioning

(A) CT; (B, C) T1- weighted MRI, with gadolinium contrast; (D) T2- weighted MRI. LV=lateral ventricle. III=third ventricle. IV=fourth ventricle. Arrow=Magendie’s foramen. The posterior fossa cyst is outlined in (D).

Reprinted from The Lancet, Vol. 370, L. Feuillet et al., Brain of a white collar worker, page 262, copyright 2007, with permission of Elsevier.

At the opposite end of the size argument is the disturbing scan of a French “white-collar worker” (Figure 1) whose only complaint to his doctor was “mild left leg weakness.” Ninety percent of his brain was squashed flat due to hydrocephalus, an abnormality in which cerebral fluid cannot drain out of the brain ventricles and pressure builds up within the skull. Yet the examining doctor found the man to be functioning normally (Feuillet et al. 2007).

Technology Won’t Rescue Faulty Thinking

Gall, Spurzheim, and others had theorized that behaviors such as philoprogenitiveness and metaphysical spirit were generated in specific regions of the brain and that those regions could be assessed via the skull by means of careful measurements. Their theory was wrong. Behavioral attributes and morals aren’t generated in single, specific regions of the brain. They are complex activities that require extensive global activity: initiation of thought, memory of the necessary functions, assessment of the sensory inputs, memory of how to produce the motor requirements necessary to complete the acts, judgments regarding the acts, and dozens of other activities. Thinking and behaving is complicated business.

This is not to dismiss modern research on brain structures and functions—indeed, much important work is being produced. Child abuse and neglect may yet show definitive structural and functional differences shown in properly designed studies. However, most of the current studies that assess child abuse, psychiatric diagnoses, and hippocampal volumes have not been designed to determine if there are causal correlations; they were instead designed to show that there are causal correlations.

Despite the picture painted by the new phrenologists, there is more to a person’s life than abuse. Children grow up; get vaccinated (we hope); get into fights with friends, siblings, and parents; learn to roller skate and fall down; do well or poorly in school and sports; go on dates; go to college; get jobs; lose jobs; get married; have children; get divorced; etc. These are just a very few snapshots in an average life. Yet we know nothing about brain changes brought about by any of these acts. And until we can differentiate any and all of these acts from child abuse in the scans (and currently we cannot), the new phrenology is little better than the old phrenology. 



  1. Genetic studies of borderline twins and their relatives point to heredity as the more important etiology, though numerous environmental factors are involved as well (Distal et al. 2008).


  • De Bellis, M.D., J. Hall, et al. 2001. A pilot longitudinal study of hippocampal volumes in pediatric maltreatment-related posttraumatic stress disorder. Biological Psychiatry 50(4)(August 15): 305–309.
  • Distal, M.A., T.J. Trull, et al. 2008. Heritability of borderline personality disorder features is similar across three countries. Psychological Medicine 38(9)(September): 1219–29.
  • Feuillet, L., H. Dufour, et al. 2007. Brain of a white-collar worker. The Lancet 370(9583)(July 21): 262.
  • Freud, S. 1897. Letter to Wilhelm Fleiss Sept. 21, 1897. In Masson 1985.
  • Gilbertson, M.W., M.E. Shenton, et al. 2002. Smaller hippocampal volume predicts pathologic vulnerability to psychological trauma. Nature Neuroscience 5(11): 1242–1247.
  • Higgins, E., and M. George. 2007. The Neuroscience of Clinical Psychiatry. Philadelphia, PA: Wolters Kluwer/Lippincott Williams & Wilkins, Inc.
  • Honeycutt, N.A., and C.D. Smith. 1995. Hippocampal volume measurements using magnetic resonance imaging in normal young adults. Journal of Neuroimaging April 5(2): 95–100.
  • Hothersall, D. 1995. History of Psychology, 3rd edition. New York: McGraw-Hill, Inc.
  • Lintern, M. 2012. Phrenology in Victorian America. The Garfield Observer: The Blog of James A. Garfield Historical Site (August 31). Available online at
  • Masson, J.M. (ed.). 1985. The Complete Letters of Sigmund Freud to Wilhelm Fliess, 1887–1904. Cambridge: Harvard University Press. Available online at
  • Soloff, P.H., K. Abraham, et al. 2017. Hyper-modulation of brain networks by the amygdala among women with borderline personality disorder: Network signatures of affective interference during cognitive processing. Journal of Psychiatric Research 88(May): 56–63.
  • Van Rooij, S.J.H., M. Kennis, et al. 2015. Smaller hippocampal volume as a vulnerability factor for the persistence of post-traumatic stress disorder. Psychological Medicine. Cambridge University Press. pp1–10. doi:10.1017/S0033291715000707.
  • Woon, F.L., and D.W. Hedges. 2008. Hippocampal and amygdala volumes in children and adults with childhood maltreatment-related posttraumatic stress disorder: A meta-analysis.  Hippocampus 18(8):729–36.