The Origins of the Clonal Selection Theory of Immunity

A Case Study for Evaluation in Science


(From FASEB. Journal February 1995, vol 9, 164-166 with copyright permission from the Editor, Sandra Jacobson)

The role of Ehrlich
The role of Jerne
The roles of Talmage and Burnet

End note (circa 2000)
End note (Sept 2010)


The advent of the clonal selection theory in the 1950s transformed immunology. The implications of the theory were explored extensively by the Australian F. Macfarlane Burnet, who received a Nobel prize in 1960 (1, 2). However, the prize was awarded for "the discovery of immunological tolerance," and not for the clonal selection theory (3). Nevertheless, the theory is widely recognized as "Burnet’s clonal selection theory" (4-6).

    In a monumental review of the development of ideas in immunology in the first half of the 20th century, Silverstein in 1989 carefully traced the origins of the clonal selection theory and the roles of Paul Ehrlich, Niels Jerne, David Talmage, and Burnet (6). Recently, the immunologists Cruse and Lewis (7) pointed out that "Talmage never received the recognition he deserved for his seminal contribution" and they attempted to "rectify this oversight." Their article is in two parts. The first part gives a historical outline similar to Silverstein’s; the second is a transcript of a conversation with Talmage.

      It appears to me that Cruse and Lewis have succeeded admirably in drawing attention to Talmage’s role. In so doing, however, they may have paid insufficient attention to various political and human factors, they may have failed to take into account acts of omission as well as acts of commission, and they may have been less than fair to Paul Ehrlich and his collaborators. It is apparent that the dialectical approach, by which supporters of different scientists deliver one-sided appraisals of their champion’s role ("unabashed advocacy"; 8), is not the best way for properly attributing credit in science. There is an objective record, which should be open to reasoned and dispassionate analysis.

    I should make clear that I am a laboratory researcher who has had a long-standing interest both in theoretical immunology (9) and in the objective evaluation of scientific performance (10). I am not a professional historian of science. I have not examined exhaustively the writings of Ehrlich, but I am sufficiently familiar with his writings and those of the other protagonists to be able to offer a perspective different from that of Cruse and Lewis. The attribution of credit for the clonal selection theory demonstrates dramatically the propensity for error of scientific evaluation, and reinforces the view that we should redesign our evaluation systems taking this error-proneness into account.

The role of Ehrlich

In simple form, the clonal selection theory of immunity states that an antigen selects from among a variety of lymphocytes those with receptors capable of reacting with part of the antigen. As a result of this interaction, specific lymphocytes are activated to proliferate (clonally expand). They may then secrete molecules of antibody that can combine with the antigen. If the antigen is part of the surface of a virus or bacterium, then the antibody labels that organism as foreign ("not-self"). The organism is then ingested by phagocytic cells and degraded.

blood cell smear

Some cells of the immune system as they appear in blood. Lymphocytes at bottom. Phagocytic cells at top.

     Ehrlich suggested that antigens interact with receptors borne by cells (11, 12) and this results in the secretion of excess receptors (antibody). He did not suggest that individual cells would have homogeneous receptors of one specificity (the key postulate of the clonal selection theory), nor did he suggest that responding in this way to antigen might be a dedicated function of a particular cell or tissue (lymphoid tissue). In this respect, he differed from his contemporary Metchnikoff who ascribed the production of antibodies to macrophages.

    Ehrlich did suggest that erythrocytes would not have this function and that the function might be a specialized characteristic of "haemopoietic tissue." He also implied that some tissues might respond better to certain antigens than others. Thus, his collaborator Von Dungern wrote (12), "the most varied cells, according to the kind of side-chains [receptors] they possess and the affinities thereby brought about, are probably able to produce immune body [antibody]" (my italics and parentheses).

      This implies that Ehrlich recognized the possibility of some differentiation among antibody forming cells, but this idea was not extrapolated to the extreme of one receptor and antibody specificity per cell.

     The key feature of Ehrlich’s "selective" model was that there was a preexisting repertoire of specificities for a variety of antigens. The antigens then would act to select from among the specificities. It is unlikely that Ehrlich had any appreciation of the number of receptors that a cell might bear, so there was no reason for him to think that a single cell might not carry receptors of a wide variety of specificity's. Ehrlich did, however, appreciate that some antibody (natural antibody) might be released without the necessity of previous interaction of the cell with antigen. Furthermore, he appreciated that this free antibody might serve to buffer receptor-bearing cells against interaction with antigen. He understood that antibody molecules would have a distinct structure, and that parts of the molecule that react with complement might differ from parts reacting with specific antigen. He also recognized that antibodies themselves are potential antigens and that distinct anti-antibodies might be raised against different parts of the antibody molecule. He also introduced the idea of a mechanism of self/not-self discrimination:

"which prevents the production within the organism of amboceptors [antibodies] directed against its own tissues. In this horror autoxicus, we are dealing with a well-adapted regulatory contrivance" (12).

Ehrlich’s main problem was, in the words of Cruse and Lewis (7), that:

"It was inconceivable that there would be preformed receptors for antigens that the animal body would never see."

Or in Ehrlich’s words (11):

"It would not be reasonable to suppose that there were present in the organism many hundreds of atomic groups destined to unite with toxines, when the latter appeared, but in function really playing no part in the processes of normal life, and only arbitrarily brought into relationship with them by the will of the investigator. It would indeed be highly superfluous, for example, for all our native animals to possess in their tissues atomic groups deliberately adapted to unite with abrin, ricin, and crotin, substances coming from the far distant tropics."

     It was then well known that antibodies could be raised against these substances. Thus, in this respect he had discredited his own hypothesis before the work of Landsteiner described by Cruse and Lewis (7).

    Theories once conceived have to be marketed. This requires that the points where they fit in with current knowledge be emphasized and ad hoc assumptions be avoided. To have postulated that receptors or tissues played a dedicated role in antibody formation would have been a big step at the end of the 19th century. The idea that cells would need to take up nutrients was established, and whether he believed it or not, it was easiest for Ehrlich to postulate that receptors for nutrients also served the antibody role. Indeed, we now know he was partially correct. After reaction with antibody, bacteria are digested by phagocytic cells and their breakdown products are then recycled for use by the host organism. In terms of Ehrlich’s overall concept, the source of the receptors is a minor quibble.

The role of Jerne

Jerne in 1955 (13) presented a theory that Talmage (14) acknowledged 2 years later as "a basically similar concept" to Ehrlich’s. However, because of Ehrlich’s position regarding whether antibodies are dedicated molecules or have some other function, in the interview with Cruse and Lewis, Talmage reverses this position: "The ideas of Ehrlich and Jerne are fundamentally different" (7). Yet, Jerne’s "natural selection theory" simply restated the Ehrlich viewpoint that there is a preexisting repertoire from which antigen selects. Actually, Ehrlich went further than Jerne, introducing the idea of cells being activated as a result of receptor-ligand (antigen) interaction. I suspect that if Ehrlich’s theory were rewritten today using modern terms and then submitted together with Jerne’s paper to a naive observer for dating, it would almost certainly be dated after Jerne and before Talmage and Burnet

     Jerne proposed that antigen-antibody complexes are taken into cells where the antibody is then replicated. One commentator has found it "inconceivable" that Jerne "could propose a self-replicating protein, given what was know about DNA structure and genetics in 1955" and is "surprised" that the transmitting editor (Delbruck) agreed to submit the paper to the Proceedings of the National Academy of Sciences USA (15).

     There is much we do not know and may never know. Jerne did not cite Ehrlich’s work in 1955 (13). Perhaps he was unaware of it (16, 17). Some immunologists, however, were fully aware of it and contributed to the 1954 centennial celebration of Ehrlich’s birth (18). We do not know how many journal editors had rejected papers like Jerne’s on the grounds that the papers merely reiterated Ehrlich. Nor do we know how many potential Jerne-like authors failed to write up their ideas because they recognized that they were derivative of Ehrlich. We also do not know whether what Cruse and Lewis describe as "the abandonment of Ehrlich’s selective theory" in the 1930s and 1940s was complete, or whether many immunologists still took the Ehrlich viewpoint, but felt it was politically correct to remain silent. Acts of omission get less attention from historians of science than do acts of commission. Although we should try to use positive evidence (documentation) to reconstruct the past, we should also continue to explore the probabilities of what was not documented.

The roles of Talmage and Burnet   

Talmage, in a review early in 1957 (14), wrote:

" . . . it is tempting to consider that one of the multiplying units in the antibody response is the cell itself. According to this hypothesis, only those cells are selected for multiplication whose synthesized product has affinity for the antigen injected. This would have the disadvantage of requiring a different species of cell for each species of protein produced, but would not increase the total amount of configurational information required on the hereditary process."

Photograph of David Talmage, circa 1970, kindly supplied by Dr. Talmage
He discussed supporting evidence from:
  • 1) The kinetics of the antibody response.

  • 2) Immunological memory.

  • 3) The fact that myeloma tumours often result in what Putnam and Udin had described (19) as the: "massive production of one globulin randomly selected from the family of normal globulins."

Thus, disparate observations in a variety of areas were suddenly unified. In writing the review, Silverstein said Talmage had "first hinted at" the clonal selection theory (6). Burnet had already dismissed the review as a "brief suggestion" (20). Although one can quibble about the definition of the clonal selection theory, it seems to me that Talmage’s 1957 contribution, although he did not use the phrase "clonal selection," was much more than a hint or brief suggestion.

burnet01.gif (815289 bytes)

      It is probable that Burnet had independently conceived clonal selection before receiving an advanced copy of Talmage’s 1957 manuscript (14). But as far as I am aware, there was no documentation of any kind. Wallace, in the middle 19th century, was the first formally to publish the theory of evolution. However, Darwin had conceived the idea some years before, and this was documented.

    Certainly, Burnet had thought deeply about immunological problems for many years, was a leading theoretician as well as leading experimentalist, and was a man of high integrity (4, 5). In the interview with Cruse and Lewis, Talmage gives the view that Burnet "truthfully had developed the idea before he received my paper" (7).

     Burnet may well have believed this. There is a twilight zone when ideas wander around the mind, crystallizing and then redissolving in the face of doubt and uncertainty. Gradually, the uncertainties fade and some form of documentation begins. Talmage’s paper may have removed some of the uncertainties. It seems unlikely that if Burnet had written anything before the arrival of Talmage’s work, he would not have kept it. On the other hand, Talmage must have gone through the crystallization process many months, if not years, before. Talmage’s final manuscript (14) is quite long, has 117 references including one to Ehrlich, and had probably gone through many drafts.

    Burnet’s short paper (1), dated 21st October 1957, is described as a "preliminary account" and cites Talmage’s paper. It seems probable that the entire paper was drafted after the receipt of the Talmage manuscript. In 1956 Burnet had published a book maintaining the position that an antigen directs the formation of (rather than selects) specific antibody (21). This viewpoint had dominated immunology for much of the century (6). It may be that, even with the Talmage manuscript in hand Burnet was still too wedded to the directive approach to cast it aside easily. These doubts are confessed in his autobiography (3) in which Burnet says that he published in the "obscure" Australian Journal of Science as:

"an attempt to eat one’s cake and have it." "If, perish the thought, there was something very wrong about the clonal selection concept, the publication in an out-of-the-way Australian journal would mean that very few people in America or England would see it."

Submission to an Australian journal also guaranteed rapid publication; submission to America or Europe could have resulted in postal and reviewing delays.

      It appears that Burnet was not alone in toying with clonal ideas. Recently, Cohn has reported (15) that in 1953 Pollock had wondered:

"Why don’t immunologists consider the possibility that antigen selects for cells?"

In 1955 Cohn and Lennox:

"had discussed the idea that one cell made one antibody, and had come up with two approaches to measuring the antibody secreted by single cells."

Failure to document in their case seems to have been motivated by the desire to do the two experiments before others thought of doing them. For this reason Cohn:

"played ideas like a winning poker game, close to the chest."

Later he was to become more open with ideas (15).

    It is of interest that, although mentioning Ehrlich and his work, in neither his 1957 account (1) nor his 1959 book (2) does Burnet actually cite Ehrlich in his reference list. In his autobiography (3) Burnet devotes a chapter to clonal selection. He states that:

"I regard the development of the clonal selection theory of immunity as my most important scientific achievement."

Later he reiterates that:

"I believe my most important contribution to science was the concept of clonal selection as applied to immunological theory."

The qualifications I have italicized imply that Burnet recognized that his role was developmental. Yet, in his autobiography he does not mention Talmage, and he mentions Ehrlich only in the context of a carcinoma line he had introduced. A subsequent biography by Sexton likewise does not mention Ehrlich or Talmage (5). Both Burnet (3) and Sexton (5) mention Jerne, to whom is attributed a major role in drawing Burnet’s attention to Ehrlich’s work (but not to Ehrlich). In a 1966 autobiographical account (16), Jerne mentions neither Ehrlich nor Talmage.



The proper evaluation of the roles of the major protagonists in the conception and development of the clonal selection theory is an ongoing task for professional historians of science (17). With his background in virology and immunology, Burnet was ideally prepared for the clonal selection approach. It is possible that the verdict of history will be that he spread himself too widely in his scientific interests. In 1957 he was much engaged in the cancer problem (5). Talmage seems to have been more focused. Yet those very qualities that denied Burnet priority in conceiving the clonal selection theory appear to have uniquely qualified him for the subsequent task of exploring its implications. Here Burnet’s genius came into its own.

     Does it really matter who originated the clonal selection concept in immunology? Doesn’t the question pander to the worst features of competitiveness that may be impeding communication and collaboration between scientists? Not if the evaluation of an individual’s contribution toward a major research advance is seen as part of an ongoing process of evaluation that lies at the heart of the research endeavour. Every day thousands of decisions are made by peers. As authors, peers decide whose papers will be cited and whose will not. As reviewers, they decide whose papers will be published and whose will not, and whose research projects will be funded and whose will not. It would seem of vital importance for the healthy development of science that peers get these decisions right. Getting it right with regard to the few, highly visible issues should serve as a guide to getting it right with regard to the multiplicity of less visible issues. The error-proneness of evaluation in science demonstrated here has important implications for the design of evaluation systems, as discussed elsewhere (10, 22-24).



Evaluation of high achievement forms part of an evaluation continuum operating throughout the scientific enterprise. Correct evaluation of individuals and their work means that the best work is published and the most able individuals obtain research grants and awards. It is important that evaluations be carried out fairly and objectively. This paper considers evaluations of the roles of Ehrlich, Jerne, Talmage, and Burnet in the conception and development of the clonal selection theory, which revolutionized our understanding of body defences against foreign organisms. These evaluations show varying degrees of bias; in particular, the major role of Ehrlich tends to be overlooked. Contradicting the conventional wisdom, the objective record shows that Talmage, not Burnet, first conceived clonal selection. The error-proneness of evaluation in science suggests that our evaluation processes should be redesigned to take this into account.

Note added in proof of the published paper: Further information on Jerne's role is provided by an article entitled "Darwinian overtones: Niels K. Jerne and the origin of the selection theory of antibody formation" by Thomas Soderqvist (1994, J Hist Biol 27, 481-529). [Here reference #17]


1. Burnet, F. M. (1957) A modification of Jerne’s theory of antibody production using the concept of clonal selection. Australian Journal of Science 20, 67-69.

2. Burnet, F. M. (1959) The Clonal Selection Theory of Acquired Immunity. Cambridge University Press, Cambridge.

3. Burnet, F. M. (1968) Changing Patterns, an Atypical Autobiography. W. Heinemann, Melbourne, Australia.

4. Ada, G. L. (1989) The conception and birth of Burnet’s clonal selection theory. In Immunology 1930-1980: Essays on the History of Immunology. Edited by Mazumdar, P. H., pp. 34-44, Wall and Thomson, Toronto.

5. Sexton, C. (1991) The Seeds of Time. The Life of Sir Macfarlane Burnet. Oxford University Press, Melbourne, Australia.

6. Silverstein, A. M. (1989) A History of Immunology. Academic Press, San Diego.

7. Cruse, J. M. and Lewis, R. E. (1994) David W. Talmage and the advent of the cell selection theory of antibody synthesis. Journal of Immunology 153, 919-924.

8. Forsdyke, D. R. (1995) Jerne and positive selection. Immunology Today 16, 105.

9. Forsdyke, D. R. (1968) The liquid scintillation counter as an analogy for the distinction between self and not-self in immunological systems. The Lancet 1, 281-283.

10. Forsdyke, D. R. (1993) On giraffes and peer review. FASEB Journal 7, 619-621.

11. Ehrlich, P. (1900) On immunity with special reference to cell life. Proceedings of the Royal Society of London 66, 424-448.

12. Ehrlich, P. (1906) Collected Studies on Immunity. J. Wiley & Sons, London.

13. Jerne, N. K. (1955) The natural selection theory of antibody formation. Proceedings of the National Academy of Science, USA. 41, 849-857.

14. Talmage, D. W. (1957) Allergy and immunology. Annual Reviews of Medicine 8, 239-256.

15. Cohn, M. (1994) The wisdom of hindsight. Annual Reviews of Immunology 12, 1-62

16. Jerne, N. K. (1966) The natural selection theory of antibody formation: ten years later. In Phage and the Origins of Molecular Biology. Edited by Cairns, J., Stent, G. S. and Watson, J. D., pp. 301-312, Cold Spring Harbor Laboratory Press, New York.

17. Soderqvist, T. (1994) Darwinian overtones: Niels K. Jerne and the origin of the selection theory of antibody formation. Journal of the History of Biology 27, 481- 529.[This was a "Note added to proof" in the formally published version: see also: Science As Autobiography: The Troubled Life of Niels Jerne by Thomas Soderqvist, (2003) Yale Univ. Press.]

18. Witebsky, E. (1954) Ehrlich’s side-chain theory in the light of present immunology. Annals of the New York Academy of Science 59, 168-181.

19. Putnam, F. W. and Udin, B. (1953) Proteins in multiple myeloma. A physiochemical study of serum proteins. Journal of Biological Chemistry 202, 727-743.

20. Burnet, F. M. (1967) The impact of ideas on immunology. Cold Spring Harbor Symposium on Quantitative Biology 32, 1-8.

21. Burnet, F. M. (1956) Enzyme, Antigen, and Virus: A Study of Macromolecular Pattern in Action. Cambridge University Press, Cambridge.

22. Forsdyke, D. R. (1989) A systems analyst asks about AIDS research funding. The Lancet 2, 1382-1384.

23. Forsdyke, D. R. (1993) Bicameral grant review: how a systems analyst with AIDS would reform research funding. Accountability in Research 2, 237-241.

24. Forsdyke, D. R. (1994) A theoretical basis for accepting undergraduate academic record as a predictor of success in a research career. Implications for the validity of peer review. Accountability in Research 3, 269-274.

End note (circa 2000)

   This paper was published in February 1995. Later in that year David Talmage was awarded the Sandoz (now Novartis) prize for basic immunology at the Ninth International Congress of Immunology in San Francisco.

   However, there is life after Immunology! In 1998 David Talmage and Richard Sanderson published a paper in Physics Essays (11, 53-59) entitled: "On the Unification of Gravity and Inertia".

Photograph of David Talmage in 1988, kindly supplied by Dr. Talmage


End note (Sept 2010)

     Thomas Soderqvist's paper was published in the autumn of 1994 and came to my attention when this paper was at the proof stage, so it could not be added to the published list of references (added here as ref. 17). In 2003 a revised and abridged translation of Soderqvist's 1998 Danish biography of Jerne was published and I added it to the list (ref. 17). However, I did not get round to reading the book until September 2010. Some facts emerged that cast new light both on the issues raised in my above 1995 article, and in my work on positive selection of lymphocyte repertoires discussed elsewhere in these webpages Click Here.

    We learn that Jerne's scientific background was in serology - studying with much mathematical elegence the titre of antibodies in immune sera for purposes of standardization. He worked in the Standardization Department of the Danish State Serum Institute (from which he was called upon to advise WHO on international standards). Here, he was able to prepare his doctoral thesis under ideal circumstances. As he himself stated:

I just picked up the phone in Serum Institute and said, 'I want twenty sterilized plates with agar and twenty mice and a rabbit and so forth, this afternoon at two o'clock,' and everything was there.

    This is a situation many research students would envy. In the course of his experiments Jerne serendipidously discovered a serum activity that could have been an antibody that had pre-existed prior to the organism's encounter with antigen - already dubbed "natural antibody" in the literature. This led to the idea that an incoming antigen might encounter a wide spectrum of pre-existing antibodies. Jerne had high expertise in the phenomenon of antibody avidity, so knew that a large number of antibodies of variable avidity would suffice to recognize the potentially infinite spectrum of foreign antigens (Ehrlich's dilemma). Among the natural antibodies one might bind an antigen better than others. The antigen would select that antibody and take it to a cell where, somehow, more of the same kind of antibody would be synthesized. The natural antibody population would, somehow, have been randomly generated, not co-opted from some other function (e.g. nutrition) as had been suggested by Ehrlich. Furthermore, Jerne speculated that the randomization might be on-going and might occur in the thymus.

      Apart from translations of Ehrlich's above mentioned papers (refs. 11 & 12), in 1957 The Collected Works of Paul Ehrlich became available. Ehrlich wrote (1897):

To attribute what could be called inventive activity to the body or to it cells, enabling them to produce new groups of atoms as required, would involve a return to the concepts current in the days of natural philosophy. Our knowledge of cell function and especially of synthetic processes would lead us rather to assume that in the formation of antibodies, we are dealing with the enhancement of a normal cell function, and not with the creation at need of new groups of atoms. Physiological analogues of the group of the specifically combining antibodies must exist beforehand in the organism or in its cells.

Being fluent in German, Jerne could have read the Ehrlich originals prior to 1955. In his 1994 paper Soderqvist makes clear that Jerne did indeed know of Ehrlich's side chain theory. In the biography it is stated that in an early draft of his 1955 natural selection theory paper Jerne had actually discussed Ehrlich's side chain theory, but had omitted this discussion from the final version. Jerne conceded that he might have been unconsciously influenced by his prior reading of Ehrlich's papers but, as far as he was aware, the key idea came out of the blue one day when walking near Knippel Bridge.

   Soderqvist suggested that Jerne "wanted to be unique, that he opted for originality rather than displaying his connectedness with tradition." The idea was completely new to Jerne and greatly excited him. Jerne wanted to claim it as his own. But he did not take it any further. It is possible that others who had come across natural antibodies in their work had thought of, but dismissed, the natural selection idea because, as Cohn pointed out (above), protein directly begetting protein (more antibody) was inconsistent with genetics and molecular biology. If a cell could somehow "read" a protein and turn that information into nucleic acid, it might work, because nucleic acid can beget nucleic acid, and that nucleic acid can then beget more of the protein. But no mechanism for directly "reading" a protein into nucleic acid was known then, or has since emerged (i.e. this so-called central dogma of Crick has stood the test of time).

   The fact that Jerne did know of, and had studied, Ehrlich's work is surprising, because it now seems such a short step to go from the selection of antibodies to the selection of cells. Jerne did not take that step. But, as indicated in my above article, others were already toying with the idea. It was in the air. Indeed, Soderqvist tells us that a few months after publication of the natural selection theory paper a young colleague, Jørgen Spärck, had suggested that cells were the selection unit, but Jerne "was evasive and did not accept it."

   I am here reminded of William Bateson and chromosomes (see the Cock-Forsdyke biography, 2008). Bateson and Saunders had discovered that certain Mendelian characters were inherited with each other more frequently than chance would allow (i.e. the characters were in what we came to call a linkage group). And at a meeting in New York in 1902 Bateson had noted that the factors responsible for the Mendelian characters (now referred to as genes) were likely strung along chromosomes. Yet he did not take the further step to suggest that characters were linked because they belonged to the same chromosome. One explanation for this is that, with minimal funding, he was totally occupied in combating a well-funded establishment ("the biometricians") that was ferociously attacking Mendel's work. First things first. Linkage was then a minor detail.

  In Jerne's case, other scientists were not a distraction, there were plentiful funds, and clonal selection was not a minor detail. Soderqvist portrays a man of high talent who, in his early years, wandered aimlessly. Lost in philosophical speculation he also had, to put it mildly, a distracting private life. However, it seems that he was well placed to jolt Burnet out of his Lamarckist line of thought to embrace clonal selection. For this, for his seeing a way through Ehrlich's dilemma, and for pointing out the need for some randomization process by which different antibody specificities would be generated, we remain in Jerne's debt. For more on this Click Here.

Next for: Genes and Antibodies, 1959 (Click Here)

Go to: Theoretical Immunology Index. (Click Here)

Go to: Videolectures page (Click Here)

Go to: Homepage (Click Here)

This page was established circa 1998 and last edited on 17 Jan 2018 by Donald Forsdyke