A tale of two cultures: How L. Luca Cavalli-Sforza bridged the gap between science and the humanities Authors: Linda Stone, Paul F. Lurquin Published: 2024 ABSTRACT This article retraces the career of geneticist L. Luca Cavalli-Sforza, from his days as a student researcher to his tenure as a Stanford University professor, and beyond. We show how Cavalli-Sforza’s untiring curiosity, enthusiasm, and global knowledge led him to make incisive contributions to topics as diverse as bacterial genetics and human evolution, both biological and cultural. In an academic world where hyperspecialization is the norm, Cavalli-Sforza stood out as a scientist unafraid to promulgate and apply multidisciplinary approaches to complex issues of human origins and culture. Non credo nell’ esistenza di una vera barriera fra umanisti e scienziati… Non credo nell’ esistenza di una vera barriera fra umanisti e scienziati… (I do not believe in the existence of a true barrier between the humanities and the sciences…) (I do not believe in the existence of a true barrier between the humanities and the sciences…) L. Luca Cavalli-Sforza in L’Evoluzione della Cultura (2004) (1, p. 13) L. Luca Cavalli-Sforza in L’Evoluzione della Cultura (2004) (1, p. 13) This account of Cavalli-Sforza’s career and numerous accomplishments is based on a variety of sources. First, we published Cavalli-Sforza’s biography in 2005 (2). This book was written following extensive interviews with him and collaborators in California, Oregon, and Washington State. Second, we read a large number of his scientific articles and books, including books not translated into English but available in Italian and French versions. Two important examples are “L’Evoluzione della Cultura” (1, in Italian) and Cavalli-Sforza’s autobiography “La Gẻnẻtique des Populations: Histoire d’une Dẻcouverte” (3, in French, translated from the Italian). Finally, we also found inspiration in Anthony Edward’s 2021 biographical memoir of Cavalli-Sforza (4). Throughout both our careers in academia, we have never met a scientist as fearless, innovative, and imaginative as L.L. Cavalli-Sforza (hereafter, simply Cavalli). He is also the only scholar we have ever met who was able convincingly to bridge the widely perceived gap separating the sciences and the humanities. Cavalli was a polymath. He largely ignored disciplinary boundaries, eagerly seeking out and collaborating with specialists from many different fields. Cavalli’s career was marked by a number of sea changes, from medicine to bacterial genetics, to a combination of demography, human genetics, and matrimonial habits, to human population genetics, to linguistics and archaeology, and to the study of the transmission of culture. He saw nearly all of these endeavors through a strong evolutionary lens. It was also Cavalli’s habit to conduct several research projects in parallel, making a strictly chronological recounting of his scientific work a real challenge. Nevertheless, it is possible to roughly demarcate time periods during which a given project was predominant. These periods are identified in the text. Cavalli’s involvement in the foundation of the quantitative study of cultural evolution could not have been predicted easily from his initial scientific interests. Early in his career, Cavalli contributed groundbreaking results to the fields of bacteria genetics (1948 to 1959) and, later (but already starting modestly in 1952), to issues of gene flow, drift, and phylogeny in human populations. He continued to work on the latter topics for the rest of his life. His concentration on cultural evolution, however, developed over a number of years, becoming a major focus (among several others) after meeting Marcus Feldman at Stanford University in 1971. But first, let us see how his scientific life got started. 1940 to 1948: First Encounters with Genetics and Statistics Cavalli’s scientific career began in the field of medicine. He studied medicine for 6 y in Italy, first in Turin, then in Pavia, receiving his medical degree in 1944, at the age of 22. Even in these earliest years of his life and scientific studies, we see a few trends that were later echoed throughout his long career. One trend was continual shifts in his professional focus. Another was his expanding knowledge of mathematics and interest in quantification. A third was international travel and, with it, exposure to new scientific ideas. All of these trends are relevant to his eventual development of a theory of cultural evolution with Marcus Feldman in 1973. Cavalli was unimpressed with medical school. As he told us, attendance at lecture was optional and hospital practice was not mandatory. Of course, this medical education was taking place during the chaotic conditions of WW II. Nevertheless, there was one positive aspect of this medical school laxity and disorganization: It allowed students like Cavalli to simultaneously pursue other interests. For Cavalli, this led to a collaboration with a fellow third-year medical student, Giovanni Magni, on a study of the virulence of Bacillus anthracis and Pneumococcus in mice. Two years later, in 1942, Cavalli and Magni were encouraged by their medical school authorities to travel to Berlin, Germany, to work on these bacteria under the guidance of geneticist N.W. Timofeev-Ressovsky, a Soviet defector. In passing, it is remarkable that even in wartime Italy, medical students were able to devote significant amounts of time to research while learning the practice of medicine. This is quasi-impossible to imagine today, anywhere in the world. The collaboration with Magni and his work in Berlin led to Cavalli’s first of several publications on the virulence of these particular bacteria (5, originally in German and later translated into English in 1947). And it is in this work and this publication that we see Cavalli’s first emphasis on the importance of quantification. From this time on, practically all of Cavalli’s publications insist on quantification in biological research, in particular with respect to frequencies of observed phenomena and their occurrence in time. This insistence continued into Cavalli’s later studies of human culture. Cavalli’s first serious exposure to genetics also took place during his years as a medical student. This was thanks to one of his professors, Adriano Buzzati-Traverso, who became Cavalli’s mentor (and later, uncle-in-law). Cavalli confessed that prior to this he deeply disliked Mendel’s laws as presented in his general biology courses (3, p. 28), probably due in part to his meager familiarity with mathematical thinking (3, p. 58). This was of course to change drastically, notably a little later when he decided to get private math lessons from a faculty of the Milan Polytechnic. Buzzati-Traverso had been a student of genetics at Iowa State University before WW II. Following his studies in the United States, he brought back to Italy textbooks on genetics and statistics and showed them to Cavalli, who then began to consider that genetic research might be what he really wanted to do. Meanwhile, as noted above, Cavalli became an MD in 1944. Now, with his freshly minted medical degree, Cavalli rapidly discovered that medicine was not his calling. He actually found the practice of medicine depressing (2, p. 32) and considered himself useless in hospital situations (3, p. 52). He soon resigned from the hospital where he had been working. But then, of course, he needed a job. As we saw, Cavalli had published several articles on bacterial pathogens during his medical studies. But that was not enough to secure a university position, which he desired, with Italian academia in full turmoil after Italy’s WW II defeat. Cavalli eventually landed a job at the Istituto Sieroterapico of Milan, a pharmaceutical lab that needed a researcher for its immunology section. There, his colleagues having learned that he had studied statistical methods (thanks to Buzzati and his American books), Cavalli found himself in high demand to explain said methods. This is when he discovered his true passion for numerical calculations and statistics (3, p. 58). These were of course baby steps, but it seems that Cavalli had found his first statistical footing, starting roughly in 1945. In addition, while at the Milan institute, he and his colleague N. Visconti di Modrone had studied the quantitative effects of UV radiation and nitrogen mustard on Escherichia coli, looking for resistant mutants. One of these mutants, Hfr C, would change Cavalli’s professional life. 1940 to 1950 Working with a Master Statistician Cavalli did not find his position at the pharmaceutical lab particularly stimulating. Adding to that the absence of interest in basic biological research in postwar Italy, Cavalli yearned to spend time in a country where this type of research was taken seriously. This country was to be England, and the laboratory that of Kenneth Mather at the John Innes Institute, then located near Wimbledon. Cavalli had won an Italian fellowship to spend a few months there. Mather was interested in the study of polygenic traits, a concept he had originated himself. Thus, Cavalli decided to study polygenes in Drosophila. Rather predictably, given the timing and the novelty of the concept of polygenes, these studies did not result in any publishable data, but Cavalli learned even more statistics while at the John Innes Institute. From there, Cavalli went to Stockholm to attend the 8th International Congress of Genetics to present his results on the effect of UV and nitrogen mustard on E. coli. There, Cavalli by chance bumped into Ronald Fisher, father of modern statistics and eminent geneticist, and introduced himself. To Cavalli’s enormous astonishment, Fisher offered him a job on the spot, at Cambridge University, to work on bacterial genetics. Cavalli surmised that Mather, one of Fisher’s former students, may have mentioned his name to Fisher. Cavalli accepted the job right away, no questions asked (3, p. 69). Fisher also happened to be interested in chromosome recombination and gene mapping. Here, the number of crossing-over events is of the essence, hence the attractiveness of bacteria such as E. coli that can be grown in large numbers. Thus, Cavalli set up a bacterial genetics lab in a small, former tea room and there started offering a bacterial genetics course, which he thinks may have been the first in the world. Interestingly, Cavalli confessed struggling to comprehend Fisher’s books on statistics (3, p. 69) which he thought user-unfriendly! Fisher was well aware of Joshua Lederberg and Edward Tatum’s 1946 discovery of conjugation and recombination in E. coli (working in Madison, Wisconsin) (6). Fisher then asked Cavalli to try to reproduce these observations, which Cavalli did. For this, Cavalli requested strains from Lederberg and said he wanted to stay in touch regarding future results. Now, the frequency of conjugation events in Lederberg’s strains was painfully low. Cavalli then treated one of these strains with the mutagen nitrogen mustard and was able to isolate a resistant mutant that could effect recombination at a frequency 10,000 times higher than that of the original strain. He called this strain Hfr (later known as Hfr C) for high frequency of recombination. This strain, and others like it, proved to be instrumental in both the understanding of sex in bacteria (7, 8) and the establishment of the operon model through work performed mostly by François Jacob, Jacques Monod, and Elie Wollman (9, 10). Sadly, Cavalli’s stay in England lasted only about 2 y. In spite of his success with his Hfr strain, Fisher was unable to find further funding for Cavalli, who also found the lack of interest of English biochemists and bacteriologists in his work quite dispiriting. In 1950, he decided to return to Italy, where he was able to resume his old position at the Istituto Sieroterapico of Milan. The isolation and initial characterization of Cavalli’s Hfr strain were never published in English. Rather, these observations were merely described in a cryptic bulletin (11) put out in Italian by the Milanese pharmaceutical company that rehired Cavalli in 1950. 1950 to 1958: Collaborating with the Lederbergs: Solving the Riddle of Bacterial Sex Using the F+/F-/Hfr System Cavalli’s collaboration with Joshua and Esther Lederberg took place for years without face-to-face contact, entirely by air mail. These exchanges, slow as they were, went very smoothly, as reported to us by Joshua Lederberg (2, p. 48). The experiments done to elucidate the nature of E. coli K12 mating were simple enough, involving only water baths to grow and coincubate the relevant strains, sterile glass pipettes and dilution test tubes, and petri plates containing selective growth medium. The thinking behind the interpretation of the results, however, was complex and even characterized as “rabbinical” by James Watson (2, p. 42). The Lederbergs and Cavalli (and later John Jinks) hypothesized that their E. coli cells came in two categories: a donor designated F+ (at the time sometimes called “male”) and a gene recipient designated F− (“female), where F stood for “fertility.” A mating of F+ with F− cells resulted in 100% of the F− recipient being converted into F+. Occasionally, chromosomal genes from the F+ were transferred to some of the F− cells, resulting in the production of recombinants. Cavalli’s Hfr strain acted as an enhanced chromosomal gene donor but with a twist: contrary to ordinary F+ x F− crosses, where the resulting progeny were all F+, Hfr x F− crosses yielded recombinants of which very few were F+ and a wide majority were F−. The interpretation of these results was the following: a) F+ cells contain a factor F that renders them capable to transfer this factor to factor-less F− cells. Sometimes chromosomal genes are transferred along with the F factor, b) In an Hfr strain the F factor becomes integrated at a specific spot in the bacterial chromosome and transfers chromosomal genes at high frequency, c) The locus of integration of the F factor usually breaks during conjugation, explaining why the F factor is rarely transmitted to F− cells. And this is indeed how conjugation works: The F factor, now known to be a 100 kbp episome, that is, a plasmid able to become integrated within the E. coli chromosome, initiates the transfer of the linearized donor chromosome into the recipient cell. The structure allowing this transfer (perhaps a specialized pilus) can break, therefore preventing the transfer of F itself, which then enters the recipient last if no breakage occurs. Once the donor chromosome enters an F− cell, it recombines with the host chromosome. Cavalli and Jinks demonstrated in 1956 that the site of integration of F into the chromosome of their Hfr strain is near the genes responsible for lactose metabolism (the Lac region) (8). Sometimes, integrated F can leave the bacterial chromosome and, if the excision does not take place at the exact ends of F, the latter can carry with it the adjacent Lac region, becoming what is called an F’Lac plasmid. This plasmid (and other F’s) became instrumental in the experiments that led to the formulation of this great paradigm of molecular genetics, the operon model (9, 10). More or less in parallel, William Hayes, an Irishman working at the University of Manchester, England, after receiving strains and advice from Cavalli, started his own conjugation experiments. His results duplicated those of Cavalli and the Lederbergs with one important difference: In the Hfr strains that he isolated independently (HfrH), the point of insertion of F was different, and the donor chromosome entered recipient F− cells in the opposite direction suggesting, to Jacob and Wollman that the E. coli chromosome is in fact circular (9). All these observations and interpretations were at the time generally received with a solid dose of skepticism. One English researcher went so far as to say “How can you believe this bacterial sex thing, knowing it was discovered by a Jew, an Italian, and an Irishman!” (2, p. 47). Presumably, this was meant to be a joke. But then, Cavalli and Lederberg themselves had a hard time believing the results of Jacob and Wollman that showed that gene transfer between Hfr strains and an F− strain occurred rather slowly and gradually, allowing the accurate timing, and hence mapping, of the genes transferred by the donor to the recipient. Nevertheless, Cavalli and Lederberg, this time working together in Madison, Wisconsin, in 1958, were able to replicate the two Frenchmen’s results. There was no controversy. The F system is a joy to teach to mesmerized undergraduates because its logical development (no doubt streamlined by the passage of time) is so elegant and so rich with applications. Today, many conjugative plasmids able to mobilize themselves, and other plasmids are known. However, only F integrates and mobilizes E. coli K12 chromosomal genes. Interestingly, another plasmid, called pTi, residing in the plant pathogen Agrobacterium tumefaciens, can transfer some of its own genes to plant cells and integrate them into the host chromosomes. This is an odd case of sexual gene transmission of sorts between a prokaryote and a eukaryote. This property is used to introduce foreign genes into plants, including plants for human consumption (12). The final collaboration between Cavalli and Joshua Lederberg addressed the nature of mutations in bacteria: Are bacterial mutations truly spontaneous or are they somehow induced by conditions under which bacteria are grown in the laboratory? Performing what they called a “sib-selection” experiment, they demonstrated that mutants do indeed appear spontaneously (13). This was the end of Cavalli’s contributions to bacterial genetics, with the F system his crowning achievement, in collaboration with the Lederbergs. Cavalli himself felt that this work was his best contribution to the field (4). But by this time, Cavalli was already beginning to think about applying some mathematical models he had learned at Cambridge to the genetics of human populations. This thinking resulted in Cavalli’s first foray into human population genetics: genetic drift based on research on consanguineous marriages, first in the Parma valley, and subsequently in all of Italy, and next, the building of the first human phylogenetic tree. 1954 to 2004: Genetic Drift in Italy Following his fruitful collaborations with American (the Lederbergs) and British (John Jinks) scientists, Cavalli found himself again isolated at the Istituto Sieroterapico in Milan, where he saw no prospect to continue a career in bacterial genetics. Also, he felt the need to find a research topic where international competition from well-equipped laboratories would be less of a problem. If ever scientific competition were to take place, he wanted to arrive first, not second (3, p. 134). Meanwhile, between 1951 and 1961, Cavalli had been given the opportunity to teach statistics and genetics at the universities of Parma and Pavia, without actually holding a professorship. In 1961 he became a professor at Parma, moving to Pavia in 1963. Once he secured his first professorship, Cavalli started embarking on multiple research projects. In fact, running several projects in parallel became a hallmark of his career. The first such project was with a Parma university student, Antonio Moroni, who was put in charge of estimating genetic variation in the Parma valley. This was clearly a major change in Cavalli’s professional trajectory. This change, however, was to allow him to apply his statistical and mathematical skills to human genetics. Also, explaining further his shift from bacterial genetics to human genetics, Cavalli once told us that “humans are more charismatic than E. coli!” (2, p. 51). Moroni, a catholic priest, had informed Cavalli that the Catholic Church had detailed records of dispensations awarded for consanguineous marriages in the Parma valley dating back to 1563 (date of the Council of Trento). These archives thus contained precious demographic information about past generations. Cavalli and his associate Moroni measured the variation in three blood types—the ABO, MN, and Rhesus (Rh) polymorphic genes—in the Parma valley as a function of village altitude. Variation was higher in the highest-altitude villages, (which also had highest isolation plus lowest population density and highest degree of consanguinity), suggesting to Cavalli that drift rather than natural selection was an important factor in the evolution of this population. To test this hypothesis, Cavalli decided to use a computer to run drift simulations. The University of Pavia had purchased its first computer in 1963, which Cavalli and colleague Gianna Zei used to simulate a population of 5,000 individuals from the 20 Parma valley villages situated at the highest altitude, using the three polymorphic Mendelian genes ABO, MN, and Rh, whose frequencies they had previously determined. The simulations yielded results in agreement with the observed data. Thus, the importance of drift (and also migration) in human population genetics started occupying a prominent place in Cavalli’s thinking at the time. Interestingly, population geneticist Motoo Kimura visited Pavia in Fall 1963, interacted with Cavalli, and lectured on mathematical and population genetics. Kimura’s revolutionary neutral theory of evolution (14), where drift plays an important role, may have germinated in his mind at that time, as per Cavalli’s recollection of events. The Parma valley results were presented at several scientific meetings at the time but were only formally published in book form in 2004 (15). This book further includes a massive amount of demographic and genetic data (gathered over many years) covering other parts of Italy. In this work one notes in particular the use of surnames (which are patronymic in Italy) to estimate the impact of drift and migration on human gene frequencies. This reasoning was based on the idea that surnames rarely mutate (e.g., through clerical errors or out-of-wedlock births) and are largely selectively neutral (barring times and places where racism and xenophobia are prevalent). Cavalli, Moroni, and Zei tabulated over 500,000 consanguineous marriages as recorded in the Catholic Church archives (to determine variation in time) and over 18 million telephone directory entries (to determine variation in space) in order to test their hypothesis that natural selection—contrary to drift—only plays a small role in the genetic structure of the population. These demographic results turned out to be consistent with actual gene frequencies at least in the Sardinian population. Today, applying this type of investigation to all of Italy remains a work in progress. As work on gene frequencies in the Parma valley was progressing, Cavalli decided to extend his research to the evolution in space and time of human populations at the level of the whole planet, a much more ambitious project. For this, he hired Anthony Edwards, a young mathematical geneticist. 1962 to 1965: First Human Evolutionary Tree Anthony Edwards, who had been a student of R. A. Fisher and then a postdoc at Cambridge, was invited by Cavalli to spend time in Pavia in 1962. Cavalli’s idea was to build an evolutionary tree using frequency data for five polymorphisms, the blood groups—ABO, MN, Rh, Duffy, and Diego—already available for five continents. Each continent was represented by three geographically isolated populations to minimize the effect of migration. Then, the genetic distance between each pair of populations was calculated (using FST or equivalent) and the length of the branches estimated by the method of least squares (4). And so was born the very first human evolutionary tree comprising 15 human populations. The tree showed that the most divergence was found between Africans and Australian aborigenes. Being unable to determine the position of the root, Cavalli and Edwards arbitrarily put it in the middle (3, p. 193). This result was presented at the 11th International Congress of Genetics in 1963 and published in 1964 (16). At the same time, however, Edwards thought of projecting the genetic data summarized in the tree onto a map of the world, thereby generating a first map of suggested human migrations that could have led to the observed genetic variation (17). The refinement and expansion of these first results were to occupy Cavalli and collaborators for the next four decades. 1968 to 1969: Cavalli Further Hones His Mathematical Skills with Walter Bodmer Cavalli first met Walter Bodmer in 1962 when they team-taught a population genetics course at Stanford University. Both Bodmer (then a young PhD from Cambridge University) and Cavalli had been invited by Joshua Lederberg, then Chair of the Genetics department in the Medical School. This collaboration continued for a full academic year when Cavalli spent a sabbatical at Stanford. Ultimately, they published two massive tomes on mathematical population genetics, in 1971 and 1976, respectively (18, 19). By writing these books, they collated an enormous amount of information regarding cutting-edge mathematical approaches to genetics. This was also done to each other’s mutual benefit. Indeed, in 2003 Cavalli told us that “I learned math from Bodmer and he learned human genetics from me” (2, p. 81). With this new, improved mathematical baggage, Cavalli was ready to tackle his next project, the history and geography of human genes. From this time on, Cavalli’s base of operations became Stanford University, where he became a professor of genetics in 1971, following a long-standing job offer by Joshua Lederberg, by now a Nobel Laureate. 1971 to 2008: The Neolithic Transition, the History and Geography of Human Genes, and the Y Chromosome Plant geneticists and agronomists had long suspected that one of the places where several crops were domesticated was in the Middle East, and more specifically, in the area known as the Fertile Crescent, located in the upper region of the Tigris and Euphrates rivers. This domestication took place during the Neolithic era, roughly 12,000 B.P., as determined by radiocarbon dating and archaeological findings. The cultivation of these crops then spread from the Middle East into Europe over a period of several thousand years. The Neolithic transition, as it is called by some, was accompanied by technical advances, such as the invention of sickles, grinding stones, and pottery. Cavalli’s interest in prehistory had originated in 1960 from visits he had made to Sardinia where he saw megalithic towers built about 3,800 B.P. (the nuraghi) and similar structures he saw in Apulia (the specchie). The populations responsible for building these structures both practiced agriculture. He then wondered how agriculture could have spread from Mesopotamia to Western Europe. At that time, the prevalent theory was that this spread was a case of cultural diffusion, where the principle and techniques of agriculture were transmitted culturally (that is, verbally or by example) in a northwestward fashion, but not by people migrating and taking these techniques with them. The latter would be a case of demic diffusion. In 1966 and subsequent years, Cavalli had witnessed a case of demic diffusion of agriculture in Africa, in which Bantu agriculturists occupied space originally occupied by hunter–gatherer “Pygmy” (meaning short statured) people. He then surmised that perhaps something similar may have happened in Europe during the Neolithic transition. This was a revolutionary idea at the time. In 1970, Cavalli attended a conference in Romania devoted to mathematics in archaeology and history. There, he met an American archaeologist, Albert Ammerman, who had done his PhD work in England using a computer to analyze assemblages of stone tools. Cavalli invited Ammerman to Pavia where the latter initially established a map of radiocarbon-dated cereals (indicating the presence of agriculture) found in archaeological sites. The oldest sites (8,500 B.P. or older) were found in the Middle East while the most recent ones (6,000 B.P. or less) were found in Western and Northern Europe. Therefore, the rate of diffusion of agriculture was about 1.1 km per year. Further, their map summarizing dates of the first arrival of wheat in Europe clearly showed a gradient extending from the Southeast to the Northwest. Applying Fisher’s wave of advance theory, Ammerman and Cavalli showed that a moving population starting at a density of 0.1 inhabitant/square km at the origin, having a growth rate of 3.9% per year, and reaching saturation of the land at a density of 5 inhabitants/square km, would be able to expand from Northern Syria to the British Isles, taking about 4,000 y to cover about 4,000 km. These numbers were compatible with radiocarbon dating of cereals and were also compatible with the rate of expansion observed in some modern African populations using primitive agricultural techniques. This modeling, however, did not prove that demic diffusion necessarily took place and, even if so, to what extent relative to cultural diffusion. In 1978, Cavalli and coworkers, Alberto Piazza and Paolo Menozzi, performed a principal component analysis of the distribution of 95 “classical” genes (no DNA markers involved) that clearly showed a genetic gradient in Europe. This gradient was a good match for the gradient representing the spread of agriculture in Europe. Their interpretation of the genetic gradient was that Neolithic farmers expanded out of the Middle East into Europe, according to the demic diffusion model (20). Cavalli and associates did not dismiss cultural diffusion, however. They believed that the spread of agriculture was the result of both phenomena, without knowing yet their relative importance. The demic diffusion model was initially received with a great deal of skepticism by archaeologists and some geneticists but these criticisms were then addressed by Cavalli’s group (21). Over time, Cavalli’s model became much more generally accepted, confirmed by others, and refined by Y-chromosome DNA and ancient DNA analysis. Today, it is thought that demic diffusion accounts for about 60% of the agricultural spread rate with 40% accounted for by cultural diffusion (22). Cavalli’s demic diffusion hypothesis also stirred up another debate, this time involving the fields of archaeology and linguistics. The controversy was over the geographic origin and subsequent spread of the Proto-Indo-European (PIE) language family. One theory, proposed by archaeologist Marija Gimbutas (23) held that the PIE homeland was in the steppe region north of the Black Sea, dating back 5,000 to 5,500 B.P., spreading first to Europe and later to Iran and South Asia. A competing theory, developed by archaeologist Colin Renfrew (24) held that the PIE homeland was in the area of Anatolia, Turkey dating back much earlier, to 10,000 B.P. Renfrew held that these PIE speakers were agriculturalists who spread their language along with agriculture as they expanded west toward Europe. Cavalli’s demic diffusion data strongly supported Renfrew’s model over that of Gimbutas. Later studies (25, 26) have confirmed both Cavalli’s demic diffusion model and Renfrew’s Anatolian homeland of PIE. On a larger, global level, Cavalli, Menozzi, and Piazza then extended principal component analysis to a study of 120 different genes (represented by classical markers) from 6,633 different locations corresponding to 1,915 different ethnic groups. Their results were published in the massive The History and Geography of Human Genes that came out in 1994 (17). In this book, Cavalli and coworkers present a grand synthesis of genetic, paleoanthropological, linguistic (see the next section), mathematical population genetics, and statistical data. By then, Allan Wilson’s group at the University of California, Berkeley, had published evidence—based on mitochondrial DNA polymorphisms—that the common ancestor of all human maternally transmitted mitochondrial DNA lived in Africa some 200,000 B.P. (27). Cavalli’s results based on principal component analysis were entirely compatible with an African origin, followed by worldwide migrations. An African origin of modern humans was also in line with contemporaneous archaeological findings. Cavalli summarized the results of his team in two entertaining books destined for the general public (28, 29). But the time had come to move on from classical (protein) markers to DNA. Cavalli’s laboratory started investigating DNA polymorphisms thanks to junior associates proficient in the new DNA technology. They began by investigating 100 single nucleotide polymorphisms in nuclear DNA as well as chromosomal microsatellite polymorphisms and determined that these diverged most in Africa, also consistent with an “Out-of-Africa” origin for humans. However, some critics argued that chromosome recombination unnecessarily complicated the interpretation of results. Thus, in 1994, Cavalli and coworkers started studying the polymorphisms of the nonrecombining portion of the Y chromosome in order to trace the evolution and migrations of modern humans. Contrary to mitochondrial DNA, which is transmitted from the mother to her children of both sexes, the Y chromosome is exclusively transmitted from the father to his sons. Such research would decide whether the male lineage of modern humans also appeared in Africa. Cavalli assumed that whatever polymorphisms they would discover in Y-chromosome DNA would be selectively neutral, and their persistence due to drift alone. Peter Underhill, who became lead investigator in Cavalli’s lab, was put in charge of the Y-chromosome project. At that time, a single mutation in the nonrecombining region was known. After a whole year of repetitive sequencing experiments, Cavalli’s team discovered a second mutation. Clearly, a faster technique was needed to speed up the process. Purely by chance, Underhill then met Peter Oefner when the latter gave a seminar at Stanford. There, Oefner explained his new technique to detect DNA polymorphisms: denaturing high-performance liquid chromatography. This technique detected polymorphisms in minutes rather than weeks (at best) and could be multiplexed; it allowed the discovery of hundreds of Y DNA polymorphisms within months after its first application. Analyzing 205 new polymorphisms in DNA collected from 1,962 men, Cavalli and coworkers found that these polymorphisms could be distributed among ten patrilines (also called haplogroups), with an origin in Africa’s Great Rift Valley perhaps as early as 140,000 B.P. (30). This “Out-of-Africa” model of human origins was further buttressed by a sophisticated statistical study of the correlation between genetic and geographic distance which favored a serial founder effect indeed located in Africa (31). Then followed the great expansions, first inside Africa (about 100,000 B.P.), and later, starting about 50,000 B.P., all the way to Australia, China, Japan, and New Guinea. Meanwhile, about 45,000 to 30,000 B.P., a subpopulation dispersed from the Middle East to Europe and Central Asia. Polynesia and the New World were occupied last. All these numbers and migration paths have been somewhat refined over the years since the first publication of these data, but the original findings remain largely valid and were widely publicized (28, 29). 1988 to 2008: The Coevolution of Genes and Languages There is a long history to the idea that there is a correlation between human ethnic and linguistic diversification. Indeed, Charles Darwin in his book “On the Origin of Species” (32) proposed this basic idea, writing that “…a genealogical arrangement of the races of Man would afford the best classification of the various languages…” Of course, it was only much later that it became possible to test this idea. In 1988, both Robert Sokal, a statistician at the State University of New York-Stony Brook, and Cavalli’s group at Stanford and Pavia published correlations between genetic distances and linguistic and geographical distances (33, 34). Cavalli’s group favored an approach involving principal component maps, phylogenetic trees, and archaeological data encompassing the whole world, at first considering only 120 gene frequencies grouped into 42 populations. These numbers were to be much surpassed in a subsequent study (see below). Sokal used a different statistical methodology, involving 97 gene frequencies and 10 cranial variables, but not including archaeological data, in 3,466 samples of populations in Europe only. The use of two different methodologies by these authors led to a lively debate regarding which approach was the most accurate. In the end, an agreement concerning population movements into Europe from the Middle East during the Neolithic (not initially the main goal of these studies) was reached, with the work of Cavalli’s group receiving confirmation (2, p. 72). Cavalli’s exploration of the evolution of languages rests in part on the work of Stanford linguist Joseph Greenberg and his student Merritt Ruhlen. Contrary to a majority of linguists, these two envisioned that language “superfamilies” had roots in time much deeper than was then accepted, perhaps as deep as 100,000 y. Cavalli became a supporter of this idea and set out with collaborators Alberto Piazza, Paolo Menozzi, and Joanna Mountain to determine whether a massive human phylogenetic tree could be matched with a large number of linguistic families also represented as an evolutionary tree. And indeed, they found such a tree, incorporating about 1,800 populations and close to 80,000 gene frequencies (35). The correlations between the two trees immediately suggested that genes and languages coevolved. In other words, as groups of people migrated, they took their languages with them. Then, as subgroups split up, they differentiated genetically and linguistically, especially if contact between the subgroups stopped due to geographical distance. As expected for groundbreaking research, these results were criticized. For some, the Greenberg/Ruhlen linguistic classification was controversial and unrealistic. For others, the concept of population as used by the geneticists was too ill-defined to be of use. However, attitudes among linguists have changed: Fourteen years after the publication of the genetic and linguistic coevolution model, Cavalli delivered an invited address to the International Conference on the Origin and Evolution of Languages in Paris. In it, Cavalli was able to affirm that “the expansion of humans to the whole inhabitable Earth favored a differentiation of genes and languages and the two evolutions went on in parallel” (2, p. 134). 1991 to 2008: The Human Genome Diversity Project In our interactions with him, Cavalli was always upbeat and positive about his work, present and past. The Human Genome Diversity Project (HGDP), however, was a bittersweet experience for him. The aim of the HGDP—at first a committee of scientists interested in human gene polymorphisms—was to collect DNA samples globally in order to determine the genetic relationships between human populations (including indigenous populations) and provide information on disease that could have an underlying genetic origin. Cavalli became de facto leader of the project and the main target of its virulent critics. Meanwhile, blood samples were taken in a large variety of locations, their leukocytes immortalized, and samples gathered at the Centre d’Etude du Polymorphisme Humain in Paris. Unfortunately, the HGDP was accused of fostering racism by categorizing people into specific biological classes. In fact, Cavalli was an outspoken opponent of racism (1, 3, 36–38). In addition, as Cavalli frequently pointed out to us, if one is to compare populations genetically, a necessary step is to categorize them into units with some sort of label. Next, the HGDP was accused of “biocolonialism” and “biopiracy” and the participating geneticists referred to as “gene hunters” or “vampires.” These critics viewed the HGDP as an organization whose goal was to mine indigenous populations for their genes in order to develop genocidal biological weapons or to patent these genes for the lucrative development of medicines. Finally, opponents saw HGDP members as an assembly of Western, hegemonic individuals deciding what the origins of non-European peoples really were, regardless of their own traditions. Cavalli took these—often acerbic—criticisms very seriously and discussed them repeatedly with his opponents, largely to no avail. Today, the HGDP no longer exists in its initial form, but work on DNA sequencing of the acquired samples continues. Cavalli found the negative reactions to the HGDP disheartening but found some solace in his associate Peter Underhill’s quotation: “It is easy to recognize a pioneer: his back is full of arrows” (2, p. 176). One positive outcome of the HGDP was that it drew attention to a thorny problem in global genetic studies of human populations: How can scientists acquire a representative sample of the human species (39)? Should they use geographical, linguistic, ethnic, or other categories? And how, in any case, can these categories be precisely defined? Various approaches have been suggested but none of them so far meeting with wide agreement in the scientific community. 1971 to 2018: Cultural Evolution Cavalli’s interest in cultural evolution was from the beginning embedded within his larger concern with human evolution and prehistory. By the early 1970s, Cavalli had already dealt with some human cultural traits when he combined genetic data with that from archaeology and linguistics to propose prehistoric human migratory paths and to hypothesize about the spread of agriculture. This work led to his growing belief that the study of human cultural expressions, combined with genetic data, could help confirm or reject proposals about human prehistory. From this base, Cavalli gradually came to believe that culture itself evolves. By his own account (3), the spark that lit the fuse of cultural evolution in his mind was his encounters with African hunter–gatherers. Between 1966 and 1985, Cavalli mounted 10 expeditions to study Pygmy populations in central Africa. He worked mostly with the Aka of the Central African Republic and Cameroon but also observed groups such as the Mbuti in Zaire (now the Democratic Republic of the Congo) and the Baka in the Republic of the Congo. Cavalli undertook these expeditions to study genetic differences between Pygmy populations and other African groups, but he soon became intrigued with the hunting-gathering way of life. It was also through his work with these African groups that he met anthropologists such as Colin Turnbull and Barry Hewlett. He later collaborated with Hewlett in studies of the Aka. In many respects, hunter–gatherers share similar features even though globally they occupy quite different environments. They exhibit informal leadership; they live in small, nomadic groups; they have intimate knowledge of their physical, floral, and faunal environments; and they show egalitarian social relationships, including a relative gender equality. Like many observers before him, Cavalli was impressed with the peaceful, gentle way of life of these hunter–gatherers. Globally, hunter–gatherer culture now faces extinction as outside groups continue to encroach upon these peoples’ land and extract natural resources from their habitats. Many hunter–gatherers are becoming disadvantaged minority groups within nation-states. As anthropologist John Bodley put it, they are “victims of progress” (40). From his very first expedition, Cavalli was struck by how completely different the hunter–gatherer way of life is from that in contemporary industrialized cultures, such as his own. He knew that all of humanity’s ancestors had lived in this hunter–gatherer mode for over 200,000 y and that over the last 10,000 y, radical cultural changes had taken place. It struck him that human culture, like human biology, also evolves, and he became curious to explore how, when, where, and through what processes human culture moved from hunter–gatherer adaptations to contemporary ways of life around the globe. Given his background in genetics, it is not surprising that his first idea about cultural evolution focused on comparisons between cultural and genetic diversification and transformation. Although not the first to ponder this comparison, he, along with Marcus Feldman, took it to a new level and in a unique direction. A turning point in Cavalli’s career and a time of advancement in his thoughts on cultural evolution came when he accepted a permanent position at Stanford University. He was already quite familiar with Stanford, having had a sabbatical year there and having taught summer courses there before that. In addition, as noted earlier, two of his important colleagues, Joshua Lederberg and Walter Bodmer, were at Stanford. Then 1 d, soon after his arrival at Stanford, Cavalli gave a seminar there for a group of mathematical biologists. Here he presented some of his new ideas about analogies between biological and cultural evolution, offering some tentative mathematical models. In the audience was Marcus Feldman, a mathematical biologist from Australia and a new assistant professor at Stanford. At the end of the talk, when the room had cleared, he and Cavalli further discussed mathematical issues and solved a few mathematical problems on the spot. With this encounter, Cavalli and Feldman began a long-term professional collaboration and friendship. Working closely together over the next years, they developed their theory of cultural evolution, publishing their major paper in 1973. It is noteworthy that this article is quite critical of the simplistic idea that IQ is majorly determined by genes (41). Later they expanded their ideas in the book, Cultural Transmission and Evolution: A Quantitative Approach in 1981 (42). This theory built upon specific parallels between biological and cultural evolution. Thus, while biological evolution operates through the effects of natural selection on genetic mutations, in cultural evolution, cultural mutations (innovations, new ideas) undergo a process of cultural selection (acceptance or rejection) through choices made by individuals. The core feature of this theory is transmission. Cavalli and Feldman distinguished particular modes of cultural transmission, noting that some of them (especially when compared to biological evolution) could lead to rapid and radical change, while other modes were conservative and tended to preserve current cultural ideas, knowledge, and ways of behavior. For example, what they called “vertical transmission” (parent or elder to child) of cultural knowledge and ideas would tend to be conservative while “horizontal transmission” (cultural ideas passed among persons of the same generation) can spread rapidly through a population. Subtypes of these and other modes of transmission, and their effects, were also developed and later, Feldman and associates incorporated into their work the concepts of cultural drift, niche construction (43), and ideas about gene–culture interaction. As for how individuals actually come to accept or reject innovations (or even how these innovations come about in the first place), Cavalli and Feldman admitted that this is not precisely known and may have to await a better understanding of neurological mechanisms in the human brain. By the 1970s, it was well understood that culture is humanity’s primary means of adaptation to the environment. Like Cavalli and Feldman, other scholars had tried to understand how this works and how and in what sense culture evolves. In the field of anthropology in the mid-twentieth century, writers like Leslie White (44) and Julian Steward (45) proposed mechanisms of cultural evolution and in many cases tried to use quantitative methods. Their work was followed in later decades by other anthropologists such as Marvin Harris (46). Many of these evolutionary schemes in anthropology were materialist and neo-Marxist, proposing that a prime mover of cultural evolution was technology, modes of production, or a “techno-environmental” base of culture that, as it changes, will affect other features of culture such as social organization, political structure, even religion. One of the earlier writers, Leslie White, even proposed a parallel between biological and cultural evolution, which was that in both cases there is a development from simple to complex with increasing specialization of parts (44). He proposed that the control of energy prompted cultural evolution and could serve as a quantitative measurement of evolutionary change. What is unique about the Cavalli–Feldman theory is that it looks not at technological or other core, basic features of culture, but at what lies behind all cultural expressions, namely human ideas, and how these are transmitted or not over the generations. Their theory focuses directly on the individual and sees the individual actor as the agent of cultural change, much as individual organisms play the primary role in genetic change in populations. As a result, their proposal of cultural evolution is the closest analogue to biogenetic evolution. Following the original publication of these ideas, Cavalli, Feldman, and others undertook studies to evaluate this theory of cultural evolution. Many of these supported the relative strength and power of vertical transmission. One example was Cavalli and Feldman’s quantitative study of Stanford University students, as reported in their 1981 book. This study showed that although the students sampled differed from their parents in a number of behaviors and attitudes, they showed strong similarity with them in terms of religious beliefs and political orientations. The study also showed more differences among students and their friends along these lines than between students and their parents. These results may reflect that parents vertically transmit religious and political values early on to children, at least among certain populations in the United States. Another statistical study in 1981 (47), while not referring to cultural transmission theory, showed how childhood experiences in certain types of family structure in France (ranging in degrees of patriarchal authoritarianism) may influence political orientations later in life, such as a favoring of monarchy, socialism, or capitalism. In terms of cultural transmission, this study supports the notion that vertical transmission within different types of French families may determine what type of broader sociopolitical system is supported by individuals. Finally, Cavalli and Barry Hewlett’s study of cultural learning among the Aka hunter–gatherers of central Africa (48) showed that the core body of knowledge necessary for survival among these foragers is vertically transmitted to children at early ages. Also, this vertical transmission continues beyond infancy and childhood. This transmission may account for why Aka culture is conservative and resists change. Much later, Cavalli and Hewlett, along with C.R. Gugliemino and C. Viganotti (49) effectively applied cultural transmission theory to account for the distribution of various cultural traits among several African societies. Although largely ignored within cultural anthropology, the Cavalli–Feldman theory of cultural evolution considerably advanced the new and now growing field of evolutionary anthropology (50). It also influenced evolutionary areas such as coevolution (51) and dual inheritance theory (52) and can potentially benefit other fields such as human ecology and demography (53). The continuing contributions of this theory lie, first, in its focus on cultural transmission through individual actors; second, the use of quantitative methods; and third, its multidisciplinary approach. Cavalli continued his cultural and scientific endeavors after permanently returning to Italy, in 2008. In 2009 there appeared an encyclopedia in 12 volumes entitled La Cultura Italiana (54), which Cavalli edited. This ambitious project covers all of Italian culture, from the origin of its population, to Italian language, to the cities and dwellings, to science and technology, to music, and to art. In 2012. at age 90, Cavalli curated an exhibition entitled “Homo sapiens—The great history of human diversity” an interactive and multimedia exhibit organized in Rome, at the Palazzo delle Esposizioni. This Renaissance man remained intellectually involved and influential until the end. He passed away on August 31, 2018, in Belluno, Italy. Epilogue In his autobiography (3), Cavalli makes clear his belief that brain research constitutes the new frontier of the life sciences. In this volume, he also maps out how Mendel’s ideas on transmission genetics gradually evolved into the understanding of molecular genetics, that is, how genes work. We are of the opinion—also based on his autobiography plus our conversations—that Cavalli envisioned a similar path for the Cavalli–Feldman models of cultural transmission. Thus, we think he hoped that these models would serve as catalysts for the understanding of the neurophysiology of the formulation of ideas, that is, innovations that are pivotal cultural elements in these models.