Constitutional Theories of Crime

Lombroso's atavistic criminal was defined by discrete anomalies: a sloping forehead, an asymmetric skull, unusually long arms, read as throwbacks to an earlier evolutionary stage. By the 1910s that model had drawn sustained empirical criticism, most forcefully from Charles Goring's statistical comparison of English convicts and non-convicts, which found no distinct physical criminal type.

Constitutional theorists did not abandon biology in response. They relocated it, from isolated stigmata to the whole physical and chemical constitution of the person, understood as a fixed inheritance that shaped temperament and, through temperament, conduct.

The premise had three working parts, each pursued by a different research programme. The first held that gross body build (height, musculature, fat distribution) correlated with temperament closely enough to predict behavioural tendencies, including delinquent ones.

The second held that criminality itself was a heritable trait, transmissible down family lines much like eye colour, and provable by tracing pedigrees. The third held that the endocrine glands regulated impulse and emotional stability directly, so that a malfunctioning thyroid or adrenal gland could produce, in effect, a biochemical predisposition to crime.

What unified the three was a shared assumption that criminality had a stable internal cause, physical rather than volitional, discoverable by measurement rather than by casework. Much of the associated field infrastructure, anthropometric measurement, family-history interviewing, endocrinological testing, had been developed for other purposes (public health, psychiatry, eugenics) and was simply redirected at the question of crime.

That borrowing explains both the initial credibility of the constitutional school and its later collapse: its methods carried the assumptions of the fields they came from, including eugenic assumptions about heredity that did not survive the rise of Mendelian genetics.

Positivist criminology more broadly, of which constitutional theory is one branch, insisted that criminal conduct could be studied with the same causal methods as any other natural phenomenon, treating the offender rather than the offence as the unit of analysis.

Constitutional theorists pushed that commitment further than most of their positivist peers: later biosocial researchers would treat physique or hormone level as one probabilistic input among many, but the constitutional school of the 1920s and 1930s often wrote as though the physical trait alone settled the outcome.

That overreach, more than the underlying interest in biology itself, is what the field eventually corrected. Read against that backdrop, the constitutional school is best understood not as a fringe curiosity but as mainstream criminology for roughly two decades, taught in universities and cited in courtroom-adjacent psychiatric evaluation before its evidentiary base gave way.

The German psychiatrist Ernst Kretschmer set out the first influential constitutional typology in Korperbau und Charakter (Physique and Character), published in 1921. Working from clinical observation of psychiatric patients, Kretschmer proposed four physical builds: the pyknic type, rounded and thickset with a tendency toward middle-age stoutness; the leptosomic (or asthenic) type, narrow and slight with a flat chest; the athletic type, broad-shouldered and muscular; and a residual dysplastic type for builds that fit none of the other three cleanly.

Kretschmer's real interest was psychiatric, not criminological. He linked the pyknic build to a cyclothymic temperament, sociable and prone to swings between elation and depression, and associated it statistically with manic-depressive illness in his patient samples. He linked the leptosomic and athletic builds to a schizothymic temperament, withdrawn, sensitive, and prone in its extreme form to schizophrenia. The dysplastic type carried no consistent temperament association in his scheme.

Later criminologists, reading Kretschmer's typology for its own purposes, mapped these temperaments onto offending patterns: the cyclothymic pyknic type toward impulsive, opportunistic offences such as fraud or petty theft committed in a mood of excitement, and the schizothymic leptosomic or athletic type toward premeditated violence or offences carried out with emotional detachment.

Kretschmer himself never made this extension systematically, and the criminological uptake rested on clinical impression rather than a matched offender-versus-non-offender comparison of the kind Goring had already shown was necessary. The typology's lasting contribution was methodological: it treated physique and temperament as a package to be measured together, a template Sheldon adopted directly two decades later.

Kretschmer's sample was drawn almost entirely from psychiatric inpatients rather than convicted offenders, a design choice that limited how far his temperament findings could travel into criminology even before anyone tried the extension. His statistical method was also comparatively crude by later standards: he sorted patients into his four types by clinical eye rather than standardised anthropometric measurement, so the correlations he reported between build and diagnosis rested on the same rater's judgement for both variables.

Kretschmer revised the typology across several editions of his book through the 1920s, refining the boundary cases he had originally lumped into the dysplastic category, but he never added the matched non-patient control group that would have let him test whether his types distinguished psychiatric patients from the general population at all.

Kretschmer's terminology also entered general psychiatric vocabulary well beyond criminology, and versions of the pyknic-leptosomic-athletic distinction persisted in European clinical writing into the 1950s even after the specific criminological extension had fallen out of use.

The American psychologist William Sheldon replaced Kretschmer's four discrete types with a continuous rating system he called somatotyping, set out in Varieties of Human Physique (1940). Sheldon scored each individual on three components, endomorphy (soft, rounded, fat-storing build), mesomorphy (hard, muscular, bone-and-tissue build), and ectomorphy (lean, fragile, large-surface-area build), each on a seven-point scale, so that a person's somatotype was a three-number profile rather than a single label.

As with Kretschmer, Sheldon paired each physical component with a temperament: viscerotonia (relaxed, sociable, comfort-seeking) with endomorphy, somatotonia (assertive, energetic, risk-taking) with mesomorphy, and cerebrotonia (restrained, inhibited, inward-looking) with ectomorphy. In Varieties of Delinquent Youth (1949), Sheldon somatotyped 200 young men referred to a Boston rehabilitation home and reported that they skewed heavily toward mesomorphy compared with a general male population, concluding that the muscular, high-somatotonia build predisposed toward the impulsive, physically assertive conduct that produces delinquency.

Sheldon's 1949 study drew heavy methodological criticism: it used no matched non-delinquent control group scored by the same rater, and Sheldon's own moral judgements about his subjects coloured his physique ratings, a criticism that echoes the same rater-bias problem that undermined Lombroso's craniometry. The claim needed an independent test with a proper control group, which is exactly what Sheldon and Eleanor Glueck supplied next.

Sheldon defended the somatotype rating as an objective anthropometric measurement, taken from standardised nude photographs assessed against a numerical scale, and he argued this made his method more rigorous than Kretschmer's clinical typing by eye. Critics countered that photographic rating still left substantial room for the rater's prior expectations to shape the score, particularly because Sheldon rated physique and assigned delinquency status himself rather than using blinded, independent raters.

The temperament components, viscerotonia, somatotonia, and cerebrotonia, were themselves derived from personality inventories Sheldon designed around his own typology, so the whole system risked confirming its own assumptions at every stage of measurement. The 1949 sample itself was narrow in another way: it drew only from young men already referred to one Boston-area rehabilitation programme, so it said nothing about how mesomorphy related to delinquency in the general population, only within a population already selected for behavioural problems.

That narrowness is exactly what a matched control group is designed to correct for, which is why the discipline treated Sheldon's finding as suggestive rather than settled until an independent, properly controlled sample could be tested. The distinction between a suggestive finding and a settled one is a recurring theme across this whole chapter, and it is the same distinction the Gluecks' work in the next section was designed to resolve.

Harvard researchers Sheldon Glueck and Eleanor Glueck had already assembled one of the best-controlled delinquency samples of the era for their study Unraveling Juvenile Delinquency (1950): 500 persistently delinquent boys matched, boy for boy, with 500 non-delinquent boys on age, ethnicity, IQ, and neighbourhood. In Physique and Delinquency (1956), the Gluecks applied Sheldon's somatotyping method to that same matched sample, giving the mesomorphy claim the controlled test it had lacked.

The result partly confirmed Sheldon: mesomorphic build was indeed overrepresented among the delinquent boys relative to the matched non-delinquent group, roughly 60 percent of the delinquents rated predominantly mesomorphic against about 30 percent of the control group. That correlation was real and had survived a genuine control comparison, unlike Sheldon's uncontrolled 1949 sample.

What the correlation did not show was single-cause determinism. The Gluecks' own multi-factor data, drawn from the same Unraveling Juvenile Delinquency project, showed that family cohesion, parental supervision, and neighbourhood disadvantage predicted delinquency at least as strongly as body build, and that mesomorphic boys raised in stable, well-supervised homes were not disproportionately delinquent.

Later commentators revisiting the Gluecks' data have argued the mesomorphy-delinquency association most plausibly operated indirectly: a muscular build may make a boy more likely to be recruited into a delinquent peer group, or more likely to be treated as physically threatening by teachers and police, rather than mesomorphy driving offending through temperament alone. The clean biological determinism Sheldon had proposed did not survive its own best test.

The controlled result mattered for practice as much as for theory. Juvenile courts and reform schools in the United States had shown real interest in Sheldon's uncontrolled claim during the 1940s, on the hope that physique could serve as an early screening flag for at-risk boys.

The Gluecks' finding that family cohesion and supervision predicted delinquency independently of build, and often more strongly, undercut any case for using body type as a screening tool on its own, and no juvenile justice system adopted somatotyping as a formal assessment method after Physique and Delinquency appeared.

That episode is a useful case study in how a single well-controlled replication can settle a question a decade of looser studies could not: it did not falsify Sheldon's correlation, but it did falsify the policy conclusion that correlation seemed to license. Framed this way, the mesomorphy finding survives today mainly as a textbook example of how a correlation can be real, replicable, and still not carry the causal weight a theory wants to place on it.

A separate constitutional argument tried to show that criminality itself passed down family lines like a Mendelian trait. Richard Dugdale's The Jukes: A Study in Crime, Pauperism, Disease and Heredity (1877) traced roughly 700 known descendants of a New York frontier family through parish, almshouse, and prison records and reported disproportionate rates of crime, pauperism, and disease across generations. Dugdale himself was cautious, weighing poor environment alongside heredity as a cause, but the study was popularised as proof that criminal tendency bred true.

Henry Goddard's The Kallikak Family: A Study in the Heredity of Feeble-Mindedness (1912) made the hereditary-taint argument far more starkly. Goddard, director of research at the Vineland Training School in New Jersey, traced two branches descending from a single Revolutionary War-era man: one line, founded with a woman Goddard identified as 'feebleminded', which he claimed produced generations of paupers, alcoholics, and criminals, and a second, legitimate line founded with a woman of respectable stock, which Goddard reported as almost entirely free of those outcomes. He read the contrast as evidence that feeblemindedness, and the criminality Goddard treated as one of its symptoms, was a single recessive Mendelian trait.

Both studies collapsed under later scrutiny for the same underlying reason: they never separated inherited trait from inherited disadvantage. Children raised by impoverished, poorly educated, and socially stigmatised parents inherit the poverty and stigma along with any genes, and a pedigree chart cannot distinguish the two pathways.

Goddard's evidence held up even worse than Dugdale's: historian J. David Smith's 1985 re-examination of Goddard's field records found that many of the 'feebleminded' branch's case notes were vague, second-hand, or plainly shaped by the field workers' assumptions about the families they visited, and some of the archival photographs used to illustrate 'degenerate' features had reportedly been retouched.

Once Gregor Mendel's genetics were properly absorbed into human heredity research in the 1920s and 1930s, the pedigree studies' claim that a complex social outcome like criminality behaved as a single-gene trait became untenable on genetic grounds as well as methodological ones.

The hereditary-taint argument did not stay confined to family case studies. It fed directly into the American eugenics movement of the early twentieth century, which used exactly this kind of pedigree evidence to justify compulsory sterilisation laws, beginning with Indiana's 1907 statute and eventually enacted in over thirty US states. The United States Supreme Court upheld the constitutionality of such a law in Buck v.

Bell (1927), with Justice Oliver Wendell Holmes writing that three generations of the affected family had already proven the point. That legal and legislative uptake is why the pedigree studies matter beyond their own methodology: the same evidentiary weaknesses that undermined the Jukes and Kallikak claims had already been used to justify policy affecting thousands of people before geneticists caught up with the science.

The third constitutional claim looked inside the body's chemistry. Max Schlapp, a neuropathologist, and Edward Smith set out the endocrine argument in The New Criminology (1928), proposing that malfunctioning ductless glands, the thyroid, adrenal, pituitary, and gonads, disrupted the hormonal balance that ordinarily kept emotion and impulse under control. An overactive thyroid, on their account, could produce the irritability and impulsiveness behind assault, while glandular imbalance more broadly could explain crimes that seemed to lack any rational motive.

Endocrine theory had genuine institutional traction in the late 1920s and 1930s, when several American prison systems briefly experimented with glandular examinations of inmates, but it faded quickly once basic endocrinology matured: hormonal variation turned out to be far more common in the general population, and far less tightly linked to specific behaviour, than Schlapp and Smith's clinical case histories had suggested.

No controlled comparison ever reproduced their claimed gland-to-crime pathway, and by the 1940s the theory had dropped out of mainstream criminology, though it left a durable legacy in reminding later biosocial researchers that hormones (testosterone and cortisol among them) remain worth studying as one input among many, not a standalone cause.

The field's explicit reaction to body-type, pedigree, and gland theories alike came from the Boston psychiatrist William Healy and the psychologist Augusta Bronner, whose casework at the Judge Baker Foundation led them to reject any single-cause account of delinquency.

In New Light on Delinquency and Its Treatment (1936), building on Healy's earlier The Individual Delinquent (1915), they argued that each delinquent case had to be investigated individually across biological, psychological, family, and social dimensions at once, since the same offence could arise from entirely different combinations of causes in different children.

In Britain, Cyril Burt reached a parallel conclusion independently in The Young Delinquent (1925), tabulating some 170 contributing factors across nine broad categories, from poor discipline to defective intelligence to broken homes, none of them sufficient alone.

Multiple-factor theory, as this position came to be called, is better described as a research method than a causal theory: it prescribes how to investigate delinquency (case by case, factor by factor) without committing to any single mechanism, and it became the direct bridge to the integrated biosocial and developmental theories that dominate the field today.

The move to multiple-factor theory anticipated much of what modern integrated criminology now takes for granted. When Terrie Moffitt proposed her influential developmental taxonomy in 1993, distinguishing adolescence-limited from life-course-persistent offending, she drew on exactly the kind of interacting biological, family, and social variables that Healy, Bronner, and Burt had insisted on investigating case by case sixty years earlier, now organised into a testable developmental model rather than a loose case-method.

Twin and adoption studies in modern behavioural genetics likewise treat heredity as one weighted input estimated statistically across a population, not a single pedigree read as proof, which is the corrective the Jukes and Kallikak studies never received in their own era. Read this way, multiple-factor theory was not a retreat from biology but the discipline learning to hold biological, psychological, and social causes in the same frame at once.