The holist tradition in twentieth century genetics. Wilhelm Johannsen's genotype concept

Abstract

The terms ‘genotype’, ‘phenotype’ and ‘gene’ originally had a different meaning from that in the Modern Synthesis. These terms were coined in the first decade of the twentieth century by the Danish plant physiologist Wilhelm Johannsen. His bean selection experiment and his theoretical analysis of the difference between genotype and phenotype were important inputs to the formation of genetics as a well-defined special discipline. This paper shows how Johannsen's holistic genotype theory provided a platform for criticism of narrowly genocentric versions of the chromosome theory of heredity that came to dominate genetics in the middle decades of the twentieth century. Johannsen came to recognize the epoch-making importance of the work done by the Drosophila group, but he continued to insist on the incompleteness of the chromosome theory. Genes of the kind that they mapped on the chromosomes could only give a partial explanation of biological heredity and evolution.

Nils Roll-Hansen is professor emeritus of history and philosophy of science, Department of Philosophy, Classics, History of Art and Ideas, University of Oslo. Studied plant physiology in Oslo and California. Later philosophy and history of science in Oslo and abroad. The Lysenko Effect: The Politics of Science (2005) on how science policy undermined the autonomy of science in the Soviet Union. Eugenics and the Welfare State (1995, 2005), co-edited with Gunnar Broberg, on eugenics and sterilization policy in Scandinavia. Papers on Pasteur and spontaneous generation, the origins of classical Mendelian genetics, plant breeding, environmental science, eugenics, etc. Major present interests include history of genetics and the political importance of the differences between basic and applied science.

The concept of the gene has been a topic of discussion throughout the history of genetics. Growing knowledge about the material basis of heredity has continuously led to new and more sophisticated definitions (see, e.g. Gerstein et al. 2007; Müller-Wille & Rheinberger, 2012). Particulate concepts, the gene as a large molecule or part of such a molecule, dominated until well after the discovery of DNA structure and its function in protein production. However, with the rapid development of molecular genetics and embryology during the last half century the idea of a particulate ‘gene’ has become more and more elusive. This development has undermined traditional ideas about the nature of biological heredity and evolutionary change. There is a widespread feeling among biologists that the Modern Synthesis (neo-Darwinism) of the mid-twentieth century led biology into a ‘reductionist’ paradigm that needs to be broken for new progress to be made. [This ‘revolutionary’ situation is well summed up by Noble 2013. ENCODE (2013) represents the insistence on DNA molecules as carriers of heredity. Doolittle (2013) criticizes the ENCODE view. Ball (2013) gives science journalists the impression of a radically open and uncertain situation. Müller-Wille & Rheinberger (2012) present an historian of science's interpretation.] The introduction to the present collection of papers argues the need for an integration of evolutionary studies with general physiology, and it points in particular to the need for a better ‘understanding of the genotype-phenotype relation’ (Noble et al. 2014).

The present feeling that a broad physiological perspective is lacking stands in sharp contrast to the situation at the birth of genetics around 1900. The new discipline was very much a result of combined efforts in a broad range of biological specialties. Traditional methods of descriptive natural history and the new experimental approaches of the late nineteenth century were both essential. Systematics, evolution, physiology, cytology, embryology and practical breeding were in close contact at the birth of genetics. Hugo de Vries and Wilhelm Johannsen were plant physiologists by training. William Bateson was an evolutionist, Thomas Hunt Morgan an embryologist. Johannsen was well acquainted with recent cytological discoveries and discussions as well as with the ‘rediscovery’ of Mendel's laws in 1900 (Roll-Hansen 2009, 2014).

So, how did the later feeling of a divorce between studies of evolution on the one hand, and physiology and biochemistry on the other come about? What role did a narrow genocentric view play in the development of genetics itself, and in its historiography? How should the present popularity of ‘epigenetics’ be understood? Organisms are like their parents in kind and still differ because of each individual's history from conception on. Preformationism and epigenesis alike face the challenge of providing a set of mechanisms that can explain both the similarity and the differences. Recognizing a role for both aspects was central to the founding of genetics a century ago. With the vast new knowledge about molecular biological structures and processes that has accumulated in recent decades a new round of critical discussion and reflection seems due.

In this paper I will show how the terms ‘genotype’, ‘phenotype’ and ‘gene’ originally had a different meaning from that in the Modern Synthesis. These terms were coined in the first decade of the twentieth century by the Danish plant physiologist Wilhelm Johannsen. In a paradigm experiment of selecting beans he demonstrated a striking stability of biological type (Johannsen, 1903), which he later called genotype (Johannsen, 1909). This discovery of ‘hard heredity’ became the basis of genetics as a special line of experimental investigations, leading to the particulate gene concept of classical chromosome theory, and eventually to the discovery of DNA as a key to chemical explanations of heredity. However, in Johannsen's holist interpretation the genotype and not the gene was the basic concept. Genes were derivative ‘parts’ of the genotype. They were theoretical entities that could be handled experimentally by observing statistical differences in phenotypes of organisms belonging to different genotypes. The genes had a material basis but could only be identified as ‘parts’ of the genotype of an organism interacting with its environment. As the idea of the genotype as consisting of particulate genes orderly located along the chromosomes developed in the 1910s and 1920s, Johannsen developed a subtle and insistent criticism that has not lost its relevance. Moreover, he was not alone in this criticism. Nevertheless, a historiography of genetics formed when the Modern Synthesis was at its high point still makes it hard to see this clearly.

Johannsen's genotype theory

The only formal academic exam of Wilhelm Johannsen born 1857, dead 1927 (1857–1927) was as a pharmacist from the University of Copenhagen. However, he studied with leading plant physiologists in Germany and received advanced scientific training at the Carlsberg laboratory under the chemist Johan Kjeldahl and the microbiologist Emil Chr. Hansen. Kjeldahl is known for his method to analyse organic nitrogen. Hansen is famous for his pure line breeding of yeast, a method that revolutionized beer brewing by stabilizing the production. Among Wilhelm Johannsen's discoveries was a method for breaking winter dormancy in plants by treatment with ether, widely used in the gardening business (Roll-Hansen 2009).

Through chemical analyses of barley grain, in the interest of brewing, Johannsen came to plant breeding. In 1892, he was appointed lecturer in botanical subjects at the Royal College of Agriculture (Landbohøiskolen), and in 1905 called to the University of Copenhagen as professor of plant physiology. Inspired by Claude Bernard's famous Lecons sur les phénomènes de la vie communs aux animaux et vegetaux (Lessons on the phenomena of life common to animals and plants), from early on, he had been interested in general questions of biology. Johannsen's informal education and broad experience with practical as well as theoretical scientific questions is typical of many innovative natural scientists of this period (Roll-Hansen 2009).

Early on, Johannsen had an interest in the history of biology from Aristotle on. There was a broad interdisciplinary scientific milieu in Copenhagen in the late nineteenth and early twentieth century. Lively discussions in the Academy of Sciences included humanists as well as natural scientists.

The distinction between ‘phenotype’ and ‘genotype’ as well as the term ‘gene’ was first used in Johannsen's somewhat formidably Germanic 500 page treatise Elemente der exakten Erblichkeitslehre (Introduction to the elements of an exact theory of heredity) published in 1909. Fortunately, there also exists a short and pointed account in English, ‘The Genotype Conception of Heredity’ (Johannsen 1911).

Johannsen introduced the word ‘gene’ (in German ‘Gen’) to designate the ‘something’ that is found in the gametes and participates in determining specific properties in a developing organism. He presented the word ‘gene’ as a suitable short version of ‘pangene’, which had been used by de Vries. His theory of ‘intracellular pangenesis’ was fundamentally different from Darwin's theory of ‘pangenesis’. Darwin's hereditary particles (‘gemmules’) moved between cells. De Vries's hereditary particles (‘pangenes’) were restricted to staying within one cell as did Johannsen's ‘genes’. Johannsen's ‘genes’ represented established facts, but beyond that he did not want to make any claims about their nature, whether they were material particles, dynamic states or something else. In a much discussed passage he explained:

Das Wort Gen ist völlig frei von jeder Hypothese; es drückt nur die sichergestellte Tatsache aus, dass jedenfalls viele Eigenschaften des Organismus durch in den Gameten vorkommende besondere, trennbare und so mit selbstständige ‘Zustände’, ‘Grundlagen’, ‘Anlagen’ – kurz, was wir eben Gene nennen wollen – bedingt sind. (Johannsen, 1909: 124) [English translation (by N.R.): “The word gene is fully free from any hypothesis: It only expresses the securely ascertained fact that at least many properties of the organism are conditional on individual, separable and thus independent ‘states’, ‘basis’, ‘dispositions’ found in the gametes – briefly, just what we want to call genes”.]

The word “genotype” was introduced in the following chapter discussing Johannsen's famous bean selection experiment.

The experiment had been designed to test Galton's ancestral law of heredity. Galton like other orthodox Darwinians assumed that in normal unitary populations heredity varied continuously, both in time and in amount. This had been contested for instance by Hugo de Vries and William Bateson who claimed that stepwise and infrequent hereditary change (e.g., “mutation”) was the main source of evolutionary change. Galton's law said that in each generation parents will have offspring that differ from the average of the population by a fraction of the parents own difference. For instance, Galton showed that in humans tall parents get children that are tall, but less tall than the parents. According to this law selection for a certain character, like tallness, would gradually change the hereditary character of a population. This could provide an explanation for the evolution of species by means of natural selection.

The ingenuity of Johannsen's experiment was to consciously make use of a plant which was to a high degree self-fertilizing. In normal populations of such plants the individuals would be highly homozygous due to inbreeding. (Johannsen used a batch of Princess beans which he bought on the market. Galton's favorite experimental plant was sweet pea, also a self-fertilizer.) From a number of individual seed Johannsen grew new sub-populations. He called them “pure lines,” which he defined as descendants from one single parent. Johannsen found that within each of these sub-populations selection had no effect on the average property (e.g., weight of the beans). Thus Galton's ancestral law did not hold. However, if all the pure lines were added together to make one big population, the law gave a correct description of the result of selection. Johannsen concluded that this was due to a selection between stable genotypes. In other words, the continuous variation in heredity that the law of ancestral heredity had been assumed to demonstrate was an illusion.

Johannsen nevertheless saw Galton as his main predecessor and source of inspiration. In my view Ernst Mayr (1982), Frederick Churchill (1974) and William Provine (1971) have not correctly understood the difference between Johannsen and the biometricians. Their interpretation of Johannsen was too much under the spell of the Modern Synthesis (Roll-Hansen 2009).

Galton's use of experiment and statistics was a new start in the study of heredity (Sloan 2000). Johannsen's 1903 monograph was dedicated to ‘the highly deserving creator of the exact theory of heredity Francis Galton F.R.S. in gratitude’ (Johannsen 1903a). However, when Johannsen sent this paper to Karl Pearson and Frank Raphael Weldon, the contemporary wardens of Galton's biometric heritage, the response was dismissive. Johannsen could not be taken seriously because his knowledge of statistics was deficient. Among the biometricians in England only Udny Yule responded with appreciation: Johannsen's results were highly interesting, his statistics were primitive but adequate for the purpose. Nevertheless, Yule thought that in the longer run continuous change in heredity could still be significant and thus important for the evolution of species. On his visit to England in 1904, Johannsen also met with William Bateson, the ardent Mendelian and arch opponent of the biometricians. However, Bateson was not enthusiastic. He probably saw in Johannsen too much of a biometrician.

Johannsen's international reputation grew rapidly in the following years. Not least in the USA, breeders and geneticists took keen interest in his genotype idea. A special symposium on ‘The Study of Pure Lines of Genotypes’ was organized at the annual meeting of the American Society of Naturalists in December 1910 with Johannsen as keynote speaker. Because of illness he was prevented from attending, but his paper ‘The Genotype Conception of Heredity’ was very positively received. It was published in the March 1911 issue of The American Naturalist. The following winter Johannsen visited the USA giving lectures around the East coast and in the Middle West. This tour Johannsen himself felt as the high point of his scientific career.

During the first decade of the twentieth century Wilson was intensely involved in debates leading to radical modification of the idea of preformation (Baxter, 1976). The old idea was that the germ contained some kind of model or blueprint of the features that an organism would develop. It can be characterized as ‘morphological preformation’. This thinking is reflected in the thinking of early geneticists. For instance, Bateson and his school experimented with ‘unit characters’ of inheritance suggesting a direct link between factors and characters. The practice of naming the factors (genes) by their associated characters also had some lingering affinity to morphological preformation. However, it soon became clear that the relation between hereditary factors and characters was complex. It appeared that one character could depend on several factors and that one factor could affect several characters. A more systemic and holistic view developed, which can be characterized as ‘physiological’ or ‘chemical’ preformation (Roll-Hansen, 1978a). By 1910, there was little doubt among geneticists that in principle the relation between hereditary factors and characters was many-to-many. Johannsen underlined this point in his 1910 lecture. It is misleading to talk about a gene ‘for’ a specific character (Johannsen, 1911: pp. 147–148).

Johannsen's 1911 paper gives a succinct and high profiled presentation of his genotype theory. It represents ‘the modern view of heredity’ as opposed to the old ‘transmission’ view. Since Hippocrates, the dominant conception had been ‘transmission of the parent's (or ancestor's) personal qualities’, i.e. the phenotype. However, Mendelian hybridization and pure line experiments had shown that this was fundamentally mistaken. What an offspring receives from its parents is the genotype inherent in their gametes. Moreover, each generation produces gametes independently of its personal qualities. This reminds of Galton's ‘stirp’ and Weismann's ‘Keimplasma’, but according to Johannsen they had not freed their thinking from morphological preformation. Weismann even developed an elaborate speculative theory about embryological development based on a kind of morphological preformation (Johannsen, 1911: pp. 129–131).

Johannsen and the chromosome theory

From early on there was tension between Johannsen's theory of the genotype and the chromosome theory that conceived heredity as carried by a set of particles located on the chromosomes. In 1910, he was quite agnostic about the relevance of recent cytological discoveries about the behaviour of chromosomes during cell divisions. He suggested that such phenomena might be ‘regarded rather as consequences or manifestations of the divisions, repartitions and segregations of genotypical constituents (and all other things in the cell) than as their causes’. He saw no good reason to localize the ‘factors of heredity’ in the nucleus. The organism as a whole is ‘penetrated and stamped by its genotype constitution’. Johannsen came close to some present versions of developmental systems theory when he stated: ‘All living parts of the individual are potentially equivalent’ (Johannsen, 1911: pp. 153–154). One reason for this negative attitude to chromosome theorizing was his deep scepticism concerning Weismann's speculations about embryological development. As a botanist Johannsen knew well that even highly differentiated plant cell could regenerate a whole organism. There is no well-defined ‘germ-line’ in plants.

A few years later, in the second edition of his textbook Elemente, Johannsen was more positive and discussed the chromosome theory and its gene concept at some length. He still worried, however, that Morgan was falling into the morphological misconceptions of Weismann. It would be ‘mistaken, or at least premature, to conceive of the genotypic elements as spatially discrete and separate – like the parts of chromosomes are’. A long footnote discussed Morgan's use of Janssens's ‘Chiasmatypie’ model. Johannsen appreciated Morgan's craving (‘Drang’) for ‘deeper insight’, as long as the ideas were recognized as a speculative hypotheses (‘ein Stück morphologischer Dialektik’). The overall judgement was: ‘The cytological foundation of the whole speculation is hardly secured’ (Johannsen, 1913: p. 605).

By the third edition of Elemente, in 1926, Johannsen was full of praise for the achievements of Morgan and his Drosophila group of researchers. Their mapping and analysis of genes and chromosomes was the most outstanding contribution to genetics since its founding in the first decade of the twentieth century. In this edition of the textbook, Morgan received as much attention as Darwin, Mendel, Galton, Bateson and de Vries. Nevertheless, Johannsen also had clear opinion about the limitations of the chromosome theory. It did not give a full explanation of heredity and had contributed little if anything to the understanding of evolution (Johannsen, 1923a,b).

In a short paper of 1923, one of very few that he wrote in English, Johannsen summarized his radical criticism. His target was the various concepts of hereditary units, whether of particulate nature as in Darwin or Weismann or ‘unit characters’ without a particulate basis as Bateson's ‘allelomorphs’. Johannsen insisted that a genotype is much more than a sum of hereditary factors as studied by Mendelian genetics. The existence of genotypes, which are manifested in individual organisms, is the basic and empirically well supported principle of the genotype theory. Genes, on the other hand, are more hypothetical entities derived from differences in phenotypic characters in crossing experiments.

Johannsen untiringly stressed that a phenotypic character is the response of the whole genotype, and not only of one or more genes. From the ‘physiological or chemico-biological standpoint’ we must see the characters or parts of a developed organism as ‘Reactions’ of the genotype as a whole, he insisted. In this sense ‘there are no unit-characters at all!’ (Johannsen's emphasis, N.R.) He admitted to having been himself at first ‘somewhat possessed with the antiquated morphological spirit in Galton's, Weismann's and Mendel's viewpoint’ (Johannsen, 1923a: p. 136).

In Johannsen's view, the existence of the genotype was well documented as a fact, while the genes had a status of theoretical entities inferred from statistically measured phenotypic differences between populations belonging to different genotypes. On this interpretation, Johannsen's concepts of genotype and gene refer to existing entities. They are not simply instruments for systematic treatment of observations, and making predictions. The instrumentalist interpretation that is common in recent history and philosophy of genetics does not take into account the basic role of Johannsen's intuitive realism (e.g. Schwartz, 2000; Waters, 2007).

This realist understanding of genotype and gene is the framework for Johannsen's insistence on the limited explanatory power of contemporary genetics, exemplified by its most advanced theory, the chromosome theory. Existing genetic science was

very far from the ideal of enthusiastic Mendelians, viz. the possibility of dissolving genotypes into relatively small units, be they called genes, allelomorphs, factors or something else. Personally I believe in a great central ‘something’ as yet not divisible into separate factors.

Moreover, he continued with the much quoted humorous remark that:

The pomace-flies in Morgan's splendid experiments continue to be pomace-flies even if they lose all ‘good’ genes necessary for a normal fly-life, or if they be possessed with all the ‘bad’ genes, detrimental to the welfare of this little friend of the geneticists. (Johannsen, 1923a: p. 137.)

But, as already noted, Johannsen had a very high regard of the pioneering work of Morgan and his co-workers in the Drosophila group. Johannsen's point was simply that this gene–chromosome model tended to overlook some fundamental problems. Even scientists were not immune to careless old-fashioned ‘morphological’ language, he warned. Johannsen was an observant critic of the misleading effects of a facile and superficial use of language. Unfortunately his little 1914 book on ‘false analogies’ (Falske Analogier), elaborating on Bacon's idols, only exists in Danish (Johannsen, 1914).

Johannsen and Darwinism

Darwin's theory of evolution by natural selection was under heavy criticism in the late nineteenth and early twentieth century. Erik Nordenskiöld in The History of Biology simply declared that ‘Darwin's theory of the origin of species was long ago abandoned’ (Nordenskiöld, 1928: p. 486). That natural selection did not operate as Darwin thought ‘must certainly be taken as proved’, the question is ‘whether it exists at all’ (Nordenskiöld, 1928: p. 616). More recently the biology historian, Peter Bowler, characterized the period as The Eclipse of Darwinism (Bowler, 1983: p. 5). He borrowed the phrase from Julian Huxley who had used it in the introduction to Evolution: The Modern Synthesis (Huxley, 1942: pp. 22–28). Of course the issue was, not descent, not evolution as such. Neither did most scientific critics of Darwinism doubt an essential role for natural selection. The controversy was primarily over the nature of the hereditary variation that made evolution at all possible. In other words: What were the consequences of the new genetic discoveries for the theory of evolution?

The evolution of species was the overarching topic when Johannsen sketched his programme for an experimental science of heredity in 1896. This new science would serve as an independent and necessary basis for future more adequate explanations of evolution: ‘Evolution needs a science of heredity, but not vice versa’ (Johannsen, 1896: p. 12). The traditional Linnean species were broad classes that often included many subspecies and varieties. Like many contemporary biologists Johannsen had an interest in the origin of the most limited classes, often called ‘elementary species’. This was a question accessible to experimental methods. It was also a field where evolution overlapped with plant and animal breeding.

In 1903, more or less simultaneously with the bean selection paper, Johannsen published a popular article in Danish on ‘Darwinism, as seen from the point of view of heredity theory’ (Johannsen, 1903b). He pointed out that in contemporary discourse ‘Darwinism’ was a label for quite diverse views. Even Darwin himself had, in the course of the many editions of the Origin, gradually moved ‘away from natural selection as the only or main factor in the origin of species and approached more and to the Lamarckian view, that the conditions of life quite directly can form organisms and thus even without selection produce new forms, new species’ (Johannsen, 1903b: p. 527). Johannsen identified four main hypotheses or ‘theories’ about the source of hereditary change that were ‘at the moment on the agenda’:

1. The theory of ‘re-imprinting’ by environmental influence (‘Omprægningste orien’).

2. Mutation (‘Mutationsteorien’)

3. Successive selection (‘Den suksessive Selektions Teori’)

4. Hybridization (‘Bastard-Teorien’)

Johannsen emphasized that these were not exclusive alternatives (Johannsen, 1903b: p. 528). They could be combined in various ways as Darwin, for instance, had done.

The theory of ‘successive selection’ was ‘Darwinism in the narrowest sense’. It took as the starting point, on which selection could work, ‘the familiar common variability’. In Johannsen's view, the successive selection theory was based on the idea that continuous variation, often called ‘individual’ or ‘fluctuating’, was in considerable measure inherited. This was just what he tested in his bean selection experiment, with a negative outcome. For Johannsen ‘Darwinism in the narrowest sense’ included Alfred Russel Wallace and August Weismann as well as the biometricians (Johannsen, 1903b: pp. 532–533). They had all missed the significance of the difference between underlying biological type and the appearance of individual organisms – which he later called ‘genotype’ and ‘phenotype’.

Johannsen often expressed his criticism of ‘Darwinism in the narrow sense’ in strong words. In a 1915 survey of the experimental basis of evolutionary theories, for instance, he concluded that Darwin's ‘theoretical presuppositions with respect to inheritance were in principle incorrect’ and thus ‘the Darwinian theory of selection finds absolutely no support in genetics’ (Johannsen, 1915: p. 659). But even if genetics had hardly made any positive contributions to the big question about ‘the origin of species’, it had at least cleared the ground of some false ideas that were still popular and found even in some obscure corners of biological science (Johannsen, 1923b: pp. 102–103).

To sum up: Johannsen's criticism of Darwinism objected to the idea that selection as such, whether natural or artificial, was the main cause of hereditary change. In his view, selection could only choose between differences in heredity that already existed. His bean selection experiment became paradigmatic, in the sense of Thomas Kuhn. It demonstrated how hereditary variation and variation due to environment could be clearly separated. This experimental separation paved the way for genetics as a science about ‘hard’ heredity (Mayr 1982), i.e. genotypes and genes that were generally stable through the generations and only changed occasionally and in steps. Hybridization and mutation of individual genes were the two main sources of hereditary variation, for selection to work on, according to Johannsen's genotype theory.

Thus, the two basic ideas of neo-Darwinism, mutation of individual genes and their recombination in sexual reproduction were standard elements in the discourse on biological heredity and evolution by 1910. The neo-Darwinian so-called Modern Synthesis of the 1930s and 1940s developed a more precise formal mathematical treatment rather than a new theoretical understanding. On the contrary, it tended to narrow down the evolutionary perspective of genetics by losing sight of the limitations of the chromosome theory that Johannsen, among others, pointed to.

This tendency to a narrow view of genetics is well illustrated in the writings of Ernst Mayr (Roll-Hansen, 2009). He was an active contributor to the Modern Synthesis and major influence on the history of genetics as it came of age in the 1970s. In his championship of the Modern Synthesis, Mayr fundamentally misinterpreted Johannsen. He had not really understood the distinction between genotype and phenotype, claimed Mayr. Johannsen's confusion was ‘particularly obvious in his treatment of continuous variation’. And such confusion unfortunately ‘affected the interpretation of virtually all genetic and evolutionary phenomena from the Darwin period to the 1940s’ (Mayr, 1973: p. 130). Johannsen mistakenly saw the genotype as a statistical feature of the population and not a property of each individual organism, (Mayr, 1982: pp. 782–783). August Weismann was in Mayr's view the farsighted pioneer of neo-Darwinism: ‘We now know that Weismann's basic idea – a complete separation of the germ plasm from its expression in the phenotype of the body – was absolutely correct’ (Mayr, 1985: p. 315). Mayr appreciated the fundamental importance of the distinction between phenotype and genotype, but his misreading of Johannsen is a clear symptom of how strong a reductionist gene-centred view of evolution was into the last decades of the twentieth century.

Concluding remarks

Johannsen's criticism of chromosome theory and neo-Darwinism was by no means unique.

In the 1920s and 1930s there was a lively debate on these issues. Views similar to Johannsen thrived, especially in the German cultural sphere. This alternative culture of genetics and individual development (embryology) took a back seat in the middle decades of the century, and it has long been neglected both in the historical mind of working geneticists and in the professional account of historians. A good description has been given by Jonathan Harwood (1993), Styles of Scientific Thought. The German Genetics Community 1900–1933.

Some of the geneticists that fled from Nazi Germany to the USA in the 1930s felt they were outsiders. Richard Goldschmidt was respected as a maverick with wild ideas about cytology and macro-mutations as driving forces of evolution, ‘the hopeful monster’, and became an influential although marginal figure in American genetics. Victor Jollos's research on ‘lasting modifications’ (‘Dauermodifikationen’) had meagre conditions. He was not happy in the new World and died already in 1941, 54 years of age. The interest in soft heredity with neo-Lamarckian affinities also had indigenous American representatives, such as T.M. Sonneborn. C.H. Waddington was a prominent British geneticist who found contemporary neo-Darwinism inadequate for evolutionary and embryological problems. In the first decades after World War II the combination of chromosome theory and neo-Darwinian Modern Synthesis became even more dominant.

The plant physiologist Wilhelm Johannsen clearly saw his own efforts as a contribution to what can be called the ‘integration of evolutionary biology with physiological science’ (Noble et al. 2014: p. 1). Johannsen's genotype theory represented a holistic approach, emphasizing the level of the whole organism. It rejected the chromosome theory that claimed to have an answer to the basic evolutionary problem of hereditary variation. The Modern Synthesis built on a particulate gene concept fundamentally different from that of Johannsen. It largely neglected the limitations he pointed to.

Johannsen insisted that his genotype theory implied that the ‘morphological’ view of heredity had been definitively disproved. The biological heredity of an organism contained no blueprint or model of its parts or properties. To this extent, he rejected preformation. On the other hand, however, the genotype represented a measure of preformation that seems unavoidable to any theory of heredity. In this balancing, or perhaps better, integration of preformation and epigenesis he resembles the cytologist Edmund B. Wilson, a central pioneer of the chromosome theory (Roll-Hansen, 1978a). Wilson was the mentor for Morgan and his Drosophila group, and he invited Johannsen to give a seminar series at Columbia University, the home of Drosophila genetics, during his 1911–1912 visit to the USA.

Johannsen's favourite physico-chemical metaphor for the genotype was the structure of chemical molecules. This suggested that genetic stability and change had a basis in chemical structure and reactions. Bateson was another important critic of excessive trust in the chromosome theory. But his speculation on the material basis of heredity was in terms of physics rather than chemistry. Bateson used the analogy with stable physical processes, e.g. standing waves on a string (Coleman, 1970).

In my interpretation, Johannsen's saw the genotype as the empirically established fundamental fact and the genes as derived entities. He thus appears to place the basis of genetic stability on a level higher than that of DNA and other molecules, and seems to be in tune with an important role for ‘higher order principles’ (Noble et al. 2014: p. 7). But how could one make sense of modern molecular genetics in Johannsen's scheme? As opposed to the phenotype which can be ‘directly described, measured, weighed, chemically analysed, etc. … the genotype is something we reach by inference, though we can maintain there is reality in it’ (Johannsen, 1916: p. 131). Does this imply that the whole molecular machinery belongs to the phenotype? Where is the genotype then to be found? Throughout his life, Johannsen referred to Aristotle when explaining biological heredity in general and popular contexts. It is not easy to see how Aristotle's conception of organisms as constituted by matter and form, can help in interpreting the genotype theory. But Johannsen's fondness for Aristotle (Roll-Hansen, 1978b) does at least remind us that biology, and even that most modern branch genetics, is built on ideas that are not easily ‘reduced’ to the physical and chemical theories that seem to rule modern molecular biology.

Additional information

Competing interests

None declared.