1 The Status of What Is Life? in Molecular Biology

In February 1943, right in the depths of the Second World War, far away in neutral Ireland Erwin Schrödinger (Figure 1) delivered a series of public lectures at Trinity College, Dublin under the grandiose title ‘What Is Life? (The Physical Aspect of the Living Cell)’. By this time, Schrödinger was already widely acclaimed as one of the most distinguished physicists of modern times.

Figure 1 Portrait of Erwin Schrödinger (1887–1961)

(Credit: Bettmann/Getty Images)

Born in Vienna in 1887 into a cultured, middle-class family, Schrödinger’s dream had been to study physics under the great Ludwig Boltzmann, one of the founders of statistical mechanics. However, when he entered the University of Vienna in 1906, he was distraught to learn that Boltzmann had taken his own life just a few weeks earlier.Footnote ¹ Instead, he was taught theoretical physics by Boltzmann’s successor, Fritz Hasenöhrl, and experimental physics by Franz Exner, both of whom would exert a lasting influence on Schrödinger.

After obtaining his doctorate and serving on the Italian front during the First World War, Schrödinger held a series of short-term appointments at Jena (as assistant to Max Wien), Stuttgart, and Breslau, before taking up a professorship in theoretical physics in Zurich in 1921. Five years later, at the relatively late age of thirty-eight, Schrödinger’s groundbreaking discovery of wave mechanics made him one of the leading physicists of the day. In 1927, he accepted Max Planck’s chair in Berlin upon the latter’s retirement (and after Arnold Sommerfeld declined it), joining an extraordinary cohort of physicists that included Albert Einstein, Walther Nernst, and Max von Laue.

When the Nazis rose to power in 1933, Schrödinger renounced his chair—probably the most prestigious in the world of physics at the time—in protest. His resignation was an act of principle, as Schrödinger himself was not Jewish.Footnote ² He accepted a temporary position as fellow of Magdalen College, Oxford, and soon after arriving in England, he received the news that he was to share the Nobel Prize in Physics with Paul Dirac.Footnote ³

Schrödinger’s time in Oxford (vividly recounted in Clary Reference Clary2022; see also Hoch and Yoxen Reference Hoch and Yoxen1987) was altogether not a happy one, so when the opportunity arose to return to his beloved homeland in 1936, he moved back to Austria to take up a chair in Graz.Footnote ⁴ Schrödinger later described this decision as “an extremely foolish thing to do” (Schrödinger Reference Schrödinger1992: 182), as less than two years after his return the Anschluss left him at the mercy of the Nazis, who had not forgotten his disgruntled departure from Germany in 1933. He wrote a public letter, published in a local newspaper, expressing his deep shame at not having realized sooner the true destiny of his homeland, in the hope that it would suffice to placate the authorities.Footnote ⁵ The tactic failed, and he was summarily dismissed from his post.

Schrödinger hastily fled to Rome—leaving most of his belongings behind, including his Nobel medal—and from there made his way eventually to Dublin, where he had accepted the invitation by the Taoiseach Éamon de Valera (himself a former mathematician) to direct the School of Theoretical Physics at the newly established Dublin Institute for Advanced Studies (DIAS). He would remain in Ireland for seventeen years, returning to his beloved Vienna for good in 1956 (shortly after the end of the Soviet occupation of Austria), where he spent his final years. He died in 1961 aged 73.Footnote ⁶

DIAS was set up to provide for Schrödinger what the Princeton Institute for Advanced Studies provided for Einstein (as well as for other émigrés): a scientific haven where he could work in peace completely unencumbered by external events. Like Einstein, Schrödinger spent a great deal of his later years unsuccessfully trying to derive a unified field theory that would encompass gravitation and electromagnetism (see Halpern Reference Halpern2015), but he also found time to pursue many other projects. One of the statutory obligations of DIAS was the delivery of annual public lectures, and Schrödinger was more than happy to take on this responsibility on a number of occasions. For the 1943 edition, he decided to discuss the relationship between physics and biology, choosing the intriguing title ‘What Is Life?’.

The lectures were held on three consecutive Fridays in February at the Physics Theatre of Trinity College, Dublin (Figure 2a). In attendance was de Valera, along with cabinet ministers, diplomats, socialites, and artists. The number of people who tried to crowd into the lecture hall was so great that it was necessary to repeat each lecture the following Monday for those who were turned away (Moore Reference Moore1989: 395). Shortly after the lectures, Anny, Schrödinger’s wife, remarked in a letter to Born that “[p]eople came one hour before the beginning […] equiped [sic] with newspapers and books and sweets—just like for an Opera Première”.Footnote ⁷ In total, Schrödinger estimated “an audience of about four hundred which did not substantially dwindle” between lectures (Schrödinger Reference Schrödinger1944: 1). The public excitement evoked by the lectures was reported in the Irish Times and even received international coverage by Time magazine.

Figure 2 (a) DIAS notice for the ‘What Is Life?’ lectures at Trinity College, Dublin (reproduced with permission of the DIAS Archives) and (b) first American edition of What Is Life?

As soon as he completed the lectures, Schrödinger began preparing them for publication, making arrangements with the Irish publisher Cahill & Co. He appended a brief epilogue titled ‘On Determinism and Free Will’, where he referred to the mystical Hindu philosophy of Vedanta in a way that implied a rejection of Christianity. In conservative Catholic Ireland, this was perceived as offensive and unacceptable, and Schrödinger was warned that his book would not be published unless he removed the epilogue. Schrödinger refused, and the publisher decided to cancel the book’s publication, despite already having produced the final proofs (Gribbin Reference Gribbin2013: 237–238).

Undeterred, Schrödinger sent the manuscript to his friend Frederick G. Donnan, a physical chemist at University College London, who recommended Cambridge University Press as a suitable publisher and negotiated its publication with them on Schrödinger’s behalf.Footnote ⁸ The terms of publication were settled in January 1944 and What Is Life? (Figure 2b) was published in December of that year.

It was an instant success. The book’s captivating title, its short length, and Schrödinger’s eloquent and accessible style—not to mention his world-class reputation—ensured a wide readership. Marketed at 6s a copy in Britain and $1.75 in the United States (where it was published by MacMillan), the book sold briskly despite wartime austerity. After the war ended, it became even more popular. Within three years, the publisher had already arranged for the book to be translated into Italian, German, Swedish, Spanish, French, and Hungarian.Footnote ⁹ By 1948, at least sixty-five reviews of What Is Life? had been published in scientific journals and the popular press (Yoxen Reference Yoxen1979: 45).

Some of the most prominent biologists who reviewed the book included geneticists Hermann Joseph Muller, J. B. S. Haldane, and Cyril Darlington, immunologist Peter Medawar, and physicist-turned-biologist Max Delbrück. The book also prompted responses from many others. Nature alone published three such responses in the space of two years (Brabazon 1945; Manton Reference Manton1945; Butler Reference Butler1946). Among physicists, cosmologist George Gamow was deeply impressed by it. He organized a conference in 1946 sponsored by the US National Academy of Sciences in Washington, DC titled ‘The Physics of Living Matter’ featuring a number of outstanding physicists, including Niels Bohr, John von Neumann, Edward Teller, and Leo Szilard, as well as Delbrück, who kicked off the proceedings by stating that What Is Life? had been the stimulus that had brought them together (Moore Reference Moore1989: 403). Later, while working on the hydrogen bomb at Los Alamos, Gamow discussed the book with his younger colleagues and two of them authored a paper stimulated by those discussions (i.e., Reitz and Longmire Reference Reitz and Longmire1950). Gamow even used Schrödinger as the basis for one of his fictional characters in his popular science book Mr. Tompkins Learns the Facts of Life, describing him as:

Gamow playfully puts his finger here on what is often considered to be the book’s greatest achievement, which is that it supplied an intellectual impetus for the migration into biology of many young and disaffected physicists after the war (see, e.g., Fleming Reference Fleming1968). By singling out the gene as the material carrier of life and showing how it could be fruitfully investigated by appealing to physical principles, What Is Life? is credited with helping precipitate the discoveries that resulted in the molecular revolution in biology.

Horace Freeland Judson, in his interview-based history of molecular biology, has observed that “[e]verybody read Schrödinger” (Judson Reference Judson1979: 244). This is not quite the overstatement that it might seem. It is remarkable to find so many of those who are now regarded as ‘founders’ of molecular biology speaking—often in very personal terms—about how inspiring Schrödinger’s book was. Box 1 compiles some of the most striking pronouncements to this effect. What one ought to make of such declarations is, of course, another matter, which we shall have the opportunity to examine in Section 5. Taken at face value, however, they paint a clear picture of the book’s almost mythical status in molecular biology.

As Box 1 shows, all three recipients of the Nobel Prize awarded for the momentous discovery of the double helical structure of deoxyribonucleic acid (DNA)—James Watson, Francis Crick, and Maurice Wilkins— independently acknowledged Schrödinger’s decisive influence. Their model of the double helix drew on the base-parity rules discovered by Erwin Chargaff, who was also inspired by the book. In 1953, shortly after the publication of the two famous Nature papers describing the DNA structure and its hypothesized role in the replication of genetic information (Watson and Crick Reference Watson and Crick1953a, Reference Watson and Crick1953b), Crick wrote a letter to Schrödinger to let him know of the importance of What Is Life? in their discovery (Figure 3). Watson, for his part, has repeatedly credited Schrödinger with setting him on the path to the double helix. As early as 1966, he remarked that “from the moment I read Schrödinger’s What Is Life? I became polarized toward finding out the secret of the gene” (Watson Reference Watson, Cairns, Stent and Watson1966: 239). Watson has also mentioned that the first book he planned to write was “an elegant little successor to Schrodinger’s What Is Life? that I would call This Is Life” (Watson Reference Watson2001: 120), which over time developed into his landmark textbook The Molecular Biology of the Gene (Watson Reference Watson1965).

Figure 3 Letter from Francis Crick to Erwin Schrödinger, 12 August 1953

(reproduced with permission of the DIAS Archives)

The book also influenced Seymour Benzer and Gunther Stent, two members of Delbrück’s legendary ‘phage group’ at Caltech, whose work was instrumental in the establishment of molecular biology.Footnote ¹⁰ Salvador Luria—another member of the phage group—recalled finding Schrödinger’s book “exciting” (Stent Reference Stent1968: 395). François Jacob, who shared the Nobel Prize with André Lwoff and Jacques Monod for their foundational work on gene regulation, was inspired by reading What Is Life? as a young man (Jacob Reference Jacob1988: 198) and later wrote about the book’s impact on the rise of molecular biology in his well-known book on the history of biology, The Logic of Life (Reference Jacob1973). Lwoff and Monod were influenced by Schrödinger as well. Lwoff ’s Biological Order (Reference 88Lwoff1962) includes an extended discussion of What Is Life? in its final chapter, and Monod’s Chance and Necessity (Reference Monod1972) is even more indebted to Schrödinger’s biological ideas, as we will see in Section 5. Other Nobel laureates who have recognized the book’s influence include Sydney Brenner (known for his pioneering work on the genetic code and developmental genetics) and Joshua Lederberg (who discovered genetic recombination in bacteria). The latter has claimed that Schrödinger’s book persuaded him early on in his career “that biology was not simply a form of ‘stamp collecting’ but had a real intellectual content” (quoted in Bernstein Reference Bernstein2016: 163).Footnote ¹¹

The impact of What Is Life? has not been restricted to molecular biology. It has also inspired those exploring the biological dimensions of self-organization and non-equilibrium thermodynamics (as we will see in Section 4), as well as those interested in the biological application of information theory and cybernetics (as we will see in Section 5). In addition, it has attracted the attention of theoreticians of the calibre of Freeman Dyson (Reference Dyson1985), Stuart Kauffman (Reference Kauffman, Murphy and O’Neill1995), Robert Rosen (Reference Rosen, Buckley and Peat1996), Terrence Deacon (Reference Deacon2011), and Paul Davies (Reference Davies2019).Footnote ¹² Schrödinger’s book has been discussed in even wider contexts. For instance, the infamous Trofim Lysenko referred to it as a conspicuous example of ‘bourgeois pseudo-science’ in his political crusade against Mendelian genetics in the Soviet Union.Footnote ¹³

Today, What Is Life? is almost always referred to as a ‘classic’, and it is not unusual to see it described as one of the most important and influential scientific books of the twentieth century. Given its considerable fame, it is rather ironic that the actual content of the book is seldom discussed beyond three soundbites, which are repeated ad nauseam in commentaries of the book. Specifically, we tend to be told that What Is Life? puts forward three key theses:

1. 1. The hereditary substance is an aperiodic crystal that contains within it a code-script for development.

2. 2. Organisms comply with the second law of thermodynamics by feeding on negative entropy.

3. 3. The study of living matter is likely to eventually result in the discovery of new laws of physics.

Now, these statements are not wrong—Schrödinger does indeed make these three striking claims—but they do not, on their own, capture what Schrödinger is trying to accomplish in the book. In the last analysis, What Is Life? is not a book about genes, nor is it about entropy. Moreover, all three slogans, despite seeming straightforward enough, are frequently misunderstood, as we will see later. It is therefore a rather pressing matter to look back at the argument that Schrödinger advanced in his book, as well as to the reasons that prompted him to do so, as serious misconceptions about what the book is actually about remain to this day.

The recent celebrations commemorating the 75th anniversary of the ‘What Is Life?’ lectures in 1943, and of the book’s publication a year later, provided a fitting occasion to revisit Schrödinger’s biological ideas and to reflect on their contemporary relevance. In September 2018, Trinity College, Dublin hosted a most impressive international conference titled ‘Schrödinger at 75—The Future of Biology’ with a stellar roster of eminent scientists, including half a dozen Nobel laureates, and with the distinguished philosopher Daniel Dennett giving the keynote address.Footnote ¹⁴ Ostensibly an exploration of the book’s legacy, the event was in fact an examination of cutting-edge research in a variety of scientific disciplines. The speakers, while dutifully acknowledging the greatness of What Is Life? in their introductory remarks, for the most part avoided engaging with the book in any meaningful way. It is hard not to view this as a missed opportunity.

The same is unfortunately true for most of the articles that were published to mark the occasion. Both Nature (i.e., Ball Reference Ball2018) and Science (i.e., Sigmund Reference Sigmund2019) featured special commemorative essays, which, while providing serviceable overviews of the book, its author, and its influence, did not try to offer new insights.Footnote ¹⁵ The situation is not much better when we look back at the 50th anniversary celebrations. Trinity College, Dublin also organized a commemorative conference on that occasion, which brought together a cadre of scientific celebrities, including Roger Penrose, John Maynard Smith, Stephen Jay Gould, Lewis Wolpert, Leslie Orgel, and the aforementioned Kauffman. The proceedings were published by Cambridge University Press with the title What Is Life? The Next Fifty Years (Murphy and O’Neill Reference Murphy and O’Neill1995). Sadly, when one inspects the individual contributions, one is disappointed to discover that, aside from Kauffman’s, none really examine the contents of the book in any depth.

What we find, then, is that despite being “widely regarded as a philosophical cornerstone of modern molecular biology” (Newman Reference Newman1988: 360), efforts to carefully engage with the book’s argument are surprisingly rare. Most discussions of What Is Life? fall under two categories. Those in the first group typically summarize the book by invoking the three soundbites mentioned earlier and then rehearse the usual familiar tropes about the book’s influence (e.g., Symonds Reference Symonds1986; Witkowski Reference Witkowski1986; Sarkar Reference Sarkar1991; Dronamraju Reference Dronamraju1999; Sarkar Reference Sarkar, Harman and Dietrich2013; Moberg Reference Moberg2020). Those in the second group mischievously try to avoid examining What Is Life? altogether by addressing the question ‘What Is Life?’ instead, which they construe capaciously enough to allow them to discuss any topic they wish (e.g., Elitzur Reference Elitzur1995; Margulis and Sagan Reference Margulis and Sagan1995; Fuller Reference Fuller2021). There are some notable exceptions, of course. Three particularly insightful commentators are Kauffman (Reference Kauffman, Murphy and O’Neill1995, Reference Kauffman2000), Lenny Moss (Reference Moss2003), and Jean-Jacques Kupiec (Reference Kupiec2009, Reference Kupiec2010).

The comprehensive analysis of What Is Life? that I present in this Element builds on the insights afforded by these three authors. It also draws on a range of archival materials, and on recent (e.g., Sloan and Fogel Reference Sloan and Fogel2011; Sloan Reference Sloan2012; Loison Reference Loison2015) as well as old and unpublished (e.g., Yoxen Reference Yoxen1977) historiographical studies. Having introduced Schrödinger, explained the context of his Dublin lectures, described the production of the book, and reviewed its influence on leading molecular biologists, let me now outline the four goals that I have set myself for the rest of this Element.

The first is to set the record straight on what Schrödinger is actually arguing for in What Is Life? (Section 2). The second is to retrospectively evaluate Schrödinger’s main ideas in relation to current science—clearing various persistent misunderstandings about how we should interpret their historical significance along the way—in order to assess how well they have stood the test of time (Sections 3 and 4). The third is to propose a new way of thinking about the impact that the book has had on the development of molecular biology (Section 5). And the fourth is to provide a novel account of what drove Schrödinger to write What Is Life? in the first place (Section 6).

To anticipate in a bit more detail what lies ahead, I will be arguing that Schrödinger’s emphasis on the rigidity and specificity of the hereditary material (which stemmed from his attempt to explain biological order from physical principles) shaped how molecular biologists came to think about the structure and behaviour of macromolecules, resulting in a mechanical, deterministic, and genocentric view of the cell (and of development) that was instrumental in defining the agenda of molecular biology during the second half of the twentieth century. We will see, however, that cracks have begun to appear, as the shortcomings of this once-dominant view are becoming increasingly apparent. Regarding the genesis of What Is Life?, my contention shall be that Schrödinger turned to biology because he hoped that he would find in the molecular structure of living matter the means to salvage the mechanical and deterministic worldview of classical physics that he felt had become undermined by the orthodox Copenhagen interpretation of quantum mechanics. But more about that later.

My overarching aim is to help philosophers, historians, and biologists understand how we should read and think about this perennially popular book today, 80 years after its initial publication, and also to show that in the present context, it really does matter (even if one is utterly indifferent to history) what Schrödinger had to say about ‘the physical aspect of the living cell’ back in 1944. By systematically reconsidering the book’s origins, argument, impact, and legacy, this Element aims to shed light on how molecular biology got to be where it is, and where it is likely to go next.

2 Reconstructing the Argument in What Is Life?

What Is Life? is a fairly short book—it is roughly as long as this Element. The first edition ran to 91 small pages. It is also accessible, with only five references to the technical literature and less than ten equations from start to finish. The prose is lively, with vivid imagery that often verges on the poetic (e.g., “if you are given a single radioactive atom, its probable lifetime is much less certain than that of a healthy sparrow” (Schrödinger Reference Schrödinger1944: 78)).

Despite all of this, the book does not make for easy reading. Stent (Reference Stent, Cairns, Stent and Watson1966: 3) was right to point out that “though it seems to be clearly written, the clarity turns out to be deceptive”, adding that “most readers must have had the uneasy feeling from time to time that perhaps they had not really understood what Schrödinger was trying to get across”. Schrödinger himself acknowledges this on the very first page, stating that his Dublin audience “was warned at the outset that the subject-matter was a difficult one and that the lectures could not be termed popular” (Schrödinger Reference Schrödinger1944: 1).

My impression, however, is that what can make the book difficult to follow is not the subject matter per se, but the circuitous route that Schrödinger chooses to make his point, which reflects his own learning of the subject.Footnote ¹⁶ Accordingly, the book slides back and forth between speculations about the nature of the genetic material based on considerations of statistical mechanics and quantum theory on the one hand, and textbook-like expositions of Mendelian genetics coupled with reports of radiation-induced mutation experiments conducted on the fruit fly Drosophila melanogaster on the other. Schrödinger’s central claims are not presented in one go but are instead articulated piecemeal. There are also numerous physical, mathematical, and biological digressions—some important (such as the thermodynamic discussion of how the organism avoids succumbing to entropic decay), others considerably less so. In this section, I will reconstruct Schrödinger’s line of reasoning as straightforwardly as possible and highlight the far-reaching implications that he draws from it.

Let us start with the title. Although the book is titled What Is Life?—an audacious choice that Sent (Reference Stent, Cairns, Stent and Watson1966: 3) described as “a piece of colossal nerve”—Schrödinger does not address that question. The one that actually concerns him is less extravagant and more concrete, namely: what is the nature of biological order? The best way to approach the book, I think, is to regard it as Schrödinger’s concerted attempt to answer this question.

As a physicist, Schrödinger begins by considering the kind of order described by physics. Chapter 1 is titled ‘The Classical Physicist’s Approach to the Subject’. Statistical mechanics teaches us that individual atoms are incapable of exhibiting orderly behaviour on their own because they are continuously subject to the disruptive stochastic effects of thermal agitation at any temperature above absolute zero. This is why most physical laws are statistical. Lawful regularities only emerge upon consideration of immense numbers of microscopic particles, which collectively display macroscopic patterns of order. Schrödinger refers to this as the order-from-disorder principle, and he takes it to be as fundamental in physics and chemistry as “the fact that organisms are composed of cells is in biology” (Schrödinger Reference Schrödinger1944: 9). He gives several examples to illustrate it.

One of them, depicted in Figure 4a, concerns what happens when you fill a glass vessel with fog—a visible aerosol of minute water droplets. Over time, the fog gradually sinks to the bottom with a well-defined velocity, determined by the viscosity of the air and the magnitude and specific gravity of the droplets. Still, if you observe one of the droplets under a microscope you find that it does not sink steadily with constant velocity but instead performs highly irregular movements, known as ‘Brownian motion’, as a consequence of thermal agitation. So, although the behaviour of any given droplet is random and disorderly as it sinks, the overall behaviour of the fog is regular and orderly. In general, the larger the number of participating particles in a physical process, the more precisely we can predict its behaviour. This is commonly referred to in statistics as the ‘law of large numbers’.

Figure 4 Schrödinger’s illustrations of the order-from-disorder principle in What Is Life?. (a) The vessel on the left shows the regular sinking of fog over time. The downward arrow on the right delineates the irregular trajectory of an individual water droplet. The lawful behaviour of the fog reflects a statistical average of the combined behaviour of all its constituent droplets; (b) The concentration of the permanganate molecules in the vessel of water gradually decreases from left to right as the molecules diffuse until they are equally distributed. This orderly process is constituted by numerous ‘random walks’ performed by the permanganate molecules; (c) The oxygen molecules in the quartz tube, taken collectively, predictably align themselves with the magnetic field

(adapted from Schrödinger Reference Schrödinger1944)

A related example, shown in Figure 4b, is the familiar process of diffusion. If you add a small amount of a coloured substance—say, potassium permanganate—on one side of a vessel filled with water, over time it will spread evenly until it is uniformly distributed throughout the vessel. Notice that this is not due to any force driving the molecules away from the crowded region toward the less crowded one. It merely reflects the statistically predictable result of all permanganate molecules randomly being knocked about by the surrounding water molecules.

A third example, represented in Figure 4c, is the phenomenon of paramagnetism. A quartz tube filled with oxygen gas that is exposed to a magnetic field will result in the oxygen molecules orienting themselves parallel to the field, like a compass needle. But the molecules do not all turn in parallel at once. They keep changing their orientation incessantly due to thermal agitation. It is only because there is such a huge number of them that they appear on average to be aligned with the field.

Now, a “naïve physicist,” Schrödinger writes, might be forgiven for supposing it to be obvious that the astonishing orderliness of an organism must likewise be based on the lawful macroscopic patterns of behaviour displayed by enormous ensembles of interacting molecules. But here comes the twist: “this expectation,” he continues, “far from being trivial, is wrong” (ibid.: 18).

Schrödinger’s reasoning is as follows. The order of an organism is “essentially determined” by its genes (ibid.: 20), and we know from experimental evidence that a gene molecule is not much larger than a million atoms. This number, he observes, “is much too small (from the [law of large numbers] point of view) to entail an orderly and lawful behaviour according to statistical physics” (ibid.: 30). Because genes are so tiny, they should not be able to reliably code for heritable traits, given that they are firmly in the grip of thermal agitation. And yet we know for a fact that genes are remarkably stable, “with a durability or permanence”, he writes, “that borders upon the miraculous” (ibid.: 49).

Schrödinger illustrates this baffling predicament with the striking example of the ‘Habsburg lip’ (Figure 5), a genetic trait afflicting the Habsburg rulers that resulted in a protruding lower jaw, and which was faithfully preserved in this famous European family for many hundreds of years despite having a molecular basis and therefore being permanently subject to the relentless turbulence of thermal agitation.Footnote ¹⁷

Figure 5 Portraits of members of the Habsburg dynasty across four centuries displaying the distinctive ‘Habsburg Lip’. (a) Charles V, Holy Roman Emperor (1500–1558); (b) Rudolf II, Holy Roman Emperor (1552–1612); (c) Ferdinand II, Holy Roman Emperor (1578–1637); (d) Philip IV of Spain (1605–1665); (e) Leopold I, Holy Roman Emperor (1640–1705); (f) Charles II of Spain (1661–1700); (g) Archduke Charles, Duke of Teschen (1771–1847); (h) Archduke Albrecht, Duke of Teschen (1817–1895)

The puzzle that Schrödinger sets out to solve in the book is this: how do we reconcile the small size of genes with their extraordinary stability in the face of constant stochastic perturbations?

Given the inability of classical physics to account for the permanence of the gene, Schrödinger turns to the new quantum mechanics, in particular to the 1926–1927 Heitler–London theory of the chemical bond.Footnote ¹⁸ The interatomic forces postulated by this theory explain the crystalline rigidity exhibited by solid matter. Schrödinger hypothesizes that the stability of the gene is afforded by the fact that its constituent atoms are also “bound together by those ‘solidifying’ Heitler–London forces” (ibid.: 60). The gene is a large molecule that is stable, he argues, because it “presents the same solidity of structure as a crystal” (ibid.).Footnote ¹⁹ However, unlike the ordinary crystals known to physicists, which display regular and periodic configurations, Schrödinger postulates that the structure of the gene must be nonrepetitive—or ‘aperiodic’—in the sense that every atom, or group of atoms, plays an individual role that is not equivalent to (or interchangeable with) that of many of the others. He vividly exemplifies the contrast between periodic and aperiodic crystals by comparing it to the difference “between an ordinary wallpaper in which the same pattern is repeated again and again in regular periodicity and a master of embroidery, say a Raphael tapestry, which shows no dull repetition, but an elaborate, coherent, meaningful design traced by the great master” (ibid.: 3).

The reason the genetic material must be aperiodic, Schrödinger argues, is that only by exhibiting a nonrepetitive molecular structure could it possibly specify in such minute physical dimensions the detailed set of instructions required for the unfolding of ontogenic development. Schrödinger refers to this as the ‘hereditary code-script’, and he justifies its existence in the following terms:

Schrödinger illustrates this enormous number of possible specifications with the example of the Morse code, where only two signs, dot (·) and dash (—), are enough to codify the entire alphabet.Footnote ²⁰ “[I]t is no longer inconceivable”, Schrödinger concludes, “that the miniature code should precisely correspond with a highly complicated and specified plan of development and should somehow contain the means to put it into operation” (ibid.: 62).Footnote ²¹

As Schrödinger sees it, then, “the mechanism of heredity is closely related to, nay, founded on, the very basis of quantum theory” (ibid.: 47). But quantum mechanics does not just come to the rescue in What Is Life? to safeguard the durability of the gene; Schrödinger also appeals to it to explain the relatively rare occurrence of genetic mutations. By likening the gene to a stable state of a quantum-mechanical system, he suggests that we can think of mutation as the spontaneous and discrete transition of the system to an alternative ‘isomeric’ configuration (i.e., one with the same atoms stably rearranged) through the surmounting of a steep energy threshold. A gene can thus be expected to occasionally ‘jump’ from one allele form to another, thereby increasing the heritable variation in a population in a way that can be potentially acted upon by natural selection.

Let us now turn to the last two chapters—in many ways the climax of the book—where Schrödinger articulates the main lesson he wishes to draw from his detailed examination of genetics. Taking stock, he asserts that the “admirable regularity and orderliness” that characterizes the processes of life, which is “unrivalled by anything we meet with in inanimate matter”, turns out—against the physicist’s expectations—to be deterministically “controlled by a supremely well-ordered group of atoms” (ibid.: 77) specifying the hereditary code-script. These orderly biological processes, he writes, are “guided by a ‘mechanism’ entirely different from the ‘probability mechanism’ of physics” (ibid.: 79). Life is subject to what Schrödinger calls the order-from-order principle, which he explicitly contrasts with the order-from-disorder principle described at the start of the book (recall Figure 4).

In this crucial respect, organisms are analogous to machines, whose orderly and predictable operation is not a consequence of statistical regularities either but of the fact that they are constituted by solid-state structures large enough to remain unaffected by the stochastic perturbations of thermal agitation. To the extent that a piece of clockwork behaves mechanically, it is as if it operated at absolute zero. In the same way, the behaviour of organisms resembles “that purely mechanical (as contrasted with thermodynamical) conduct to which all systems tend, as the temperature approaches the absolute zero and the molecular disorder is removed” (ibid.: 69–70). Schrödinger thus arrives at his final “conclusion that the clue to the understanding of life is that it is based on a pure mechanism, a ‘clock work’” (ibid.: 82) that “also hinges upon a solid—the aperiodic crystal forming the hereditary substance, largely withdrawn from the disorder of heat motion” (ibid.: 85).

Schrödinger anticipates that as further progress is made in the elucidation of the order-from-order hereditary mechanism—progress that he predicts will come not from further physical theorizing but from experimental work in “biochemistry under the guidance of physiology and genetics” (ibid.: 68)—we should expect to find new laws of physics operating in the organism: “living matter, while not eluding the ‘laws of physics’ as established up to date, is likely to involve ‘other laws of physics’ hitherto unknown” (ibid.: 68–69). This famous pronouncement, which Schrödinger notes “was my only motive for writing this book” (ibid.: 68), has frequently been misunderstood as implying a veiled defence of anti-reductionism in biology. However, a close reading of the text dispels this interpretation.

Schrödinger compares the present inability of the known laws of physics to fully account for the behaviour of an organism to the inability of an engineer who, knowing how a steam engine works, encounters an electric generator, such as a dynamo, for the first time. Both of these mechanical contraptions contain many of the same materials, but unless the engineer has studied electrical phenomena, he will be unable to understand the workings of the dynamo. Even so, Schrödinger adds, “[h]e will not suspect that the dynamo is driven by a ghost because it is set spinning by the turn of a switch, without furnace and steam” (ibid.: 76). Instead of concluding that the laws of physics break down when applied to the dynamo, the engineer will simply realize that he is being confronted with the behaviour of matter under a new set of conditions that he has not yet analyzed. Like the dynamo, the organism is just a new kind of physicochemical machine, governed by non-statistical, yet-to-be-discovered, order-from-order laws—laws, that is, that can explain how the microscopic order in the hereditary substance is amplified to produce the macroscopic order exhibited by the organism.Footnote ²²

Schrödinger makes an important digression from his main line of argument in the penultimate chapter to consider the bearing of the second law of thermodynamics on the organism, “forgetting at the moment all that is known about chromosomes, inheritance, and so on” (ibid.: 70). This well-known law negates the possibility of a perfectly efficient transformation of heat into work. It describes the irreversible tendency for the disorder in a system, measured in terms of entropy, to increase until the system reaches thermodynamic equilibrium. In a remarkably elegant discussion—undoubtedly one of the most memorable in the whole book—Schrödinger shows that there is no contradiction between nature’s inexorable trend to become increasingly disordered (as mandated by the second law) and life’s prodigious ability to preserve and propagate order (by means of reproduction). By eating, drinking, breathing, and so on, the organism imports the energy it needs to maintain its ordered state and thereby compensate for “all the entropy it cannot help producing while alive” (ibid.: 72). In other words, the organism eludes (at least for a time!) the inert state of thermodynamic equilibrium that we call ‘death’ by drawing into itself free energy from its surroundings—which Schrödinger refers to as ‘negative entropy’. However, it is only able to do so at the expense of increasing the entropy (in the form of heat and other metabolic waste products) of the environment.

Schrödinger does not go on to explain the “organism’s astonishing gift of concentrating a ‘stream of order’ on itself and […] of ‘drinking orderliness’ from a suitable environment”, though he does note that this physiological capacity to maintain order “seems to be connected to the presence of the ‘aperiodic solids’, the chromosome molecules” (ibid.: 77).

What Is Life? ends with a five-page epilogue, ‘On Determinism and Free Will’, that had not been part of the original lectures. Schrödinger first notes that the physical processes taking place in the brain when one is thinking are “if not strictly deterministic at any rate statistico-deterministic” (ibid.: 87). Addressing the physicists in particular, he emphasizes that, “in my opinion, and contrary to the opinion upheld in some quarters, quantum indeterminacy plays no biologically relevant role in them” (ibid.). As I will show in Section 6, this is an extremely revealing statement, even though its significance has escaped the notice of almost all commentators of the book.

Finally, Schrödinger tries to reconcile two seemingly contradictory statements: “(i) My body functions as a pure mechanism according to the Laws of Nature” and “(ii) Yet I know, by incontrovertible direct experience, that I am directing its motions” (ibid.). He does so by retreating from science altogether and adopting the Vedanta metaphysical belief that individual consciousness is an illusion. We are all aspects of one single, omnipresent, eternal being—‘Brahman’, which is somewhat akin to God in the Judeo-Christian tradition. Hence Schrödinger’s conclusion: “I am God Almighty”, which, he is the first to admit, “sounds both blasphemous and lunatic” (ibid.: 88).

Most readers of What Is Life? were utterly baffled by the epilogue, to the extent that many chose to ignore it altogether. In his review of the book, Muller described it as a piece of “straight old-fashioned mysticism”, adding that it is “startling to find it in a serious work by an otherwise so responsible scientist” (Muller Reference Muller1946: 92). Haldane, for his part, wryly quipped that “[a] mechanist must either give a mechanistic account of mind, or turn a somersault. In his epilogue, Schrödinger does the latter with very great elegance” (Haldane Reference Haldane1945: 376).

3 A Critique of the Order-from-Order Principle

We begin now our appraisal of Schrödinger’s argument in What Is Life? as it strikes the modern reader, starting with an examination of the order-from-order principle, which Schrödinger stresses is “the real clue to the understanding of life” (Schrödinger Reference Schrödinger1944: 82). This principle accounts for the transmission of biological order and locates it in the code-script that Schrödinger thinks is contained in the genes. A first question that the reader of What Is Life? might have is how original Schrödinger’s ideas really are, so this seems as good a place as any to begin our analysis.

The most easily recognizable influence on What Is Life? is a 1935 German-language paper titled ‘On the Nature of Gene Mutation and Gene Structure’ co-authored by Drosophila geneticist Nikolai Timoféeff-Ressovsky, radiation physicist Karl Zimmer, and the aforementioned Delbrück (Timoféeff-Ressovsky et al. Reference Timoféeff-Ressovsky, Zimmer and Delbrück1935, TZD hereafter), which Schrödinger alludes to several times in the book.Footnote ²³ TZD is divided into four sections, one written by each co-author and a conclusion jointly written by the three. Schrödinger draws on the detailed review of gene mutation research that Timoféeff-Ressovsky provides in his section, and especially on the quantum-mechanical model of the gene and of mutation that Delbrück offers in his. Indeed, one of the chapters in What Is Life? is titled ‘Delbrück’s Model Discussed and Tested’, where Schrödinger even declares that “[i]f the Delbrück picture should fail, we would have to give up further attempts” (Schrödinger Reference Schrödinger1944: 57). The notion that the gene is a well-defined material entity of a molecular order of magnitude, that it exhibits remarkable chemical stability, and that mutation can be construed in terms of stable rearrangements in its atomic structure, are all claims in What Is Life? that Schrödinger explicitly takes from TZD.

This has led some prominent commentators, such as embryologist and geneticist Conrad Hal Waddington, to go as far as to describe What Is Life? as “a re-writing of the classical paper which we used to refer to as TZD” (Waddington Reference Waddington1969: 321). Similarly, the famed structural chemist Max Perutz claimed that “the chief merit of What Is Life? is its popularization of [TZD] that would otherwise have remained unknown outside the circles of geneticists and radiation biologists” (Perutz Reference Perutz1987a: 558).

Still, if one pays closer attention to what Schrödinger argues than to whom he cites, it becomes clear that he owes a greater intellectual debt to Muller than to Delbrück in his thinking about genes and in his postulation of an order-from-order principle. Muller had begun working in Morgan’s legendary ‘fly room’ at Columbia, and he was only the second geneticist to be awarded the Nobel Prize, after Morgan himself. By the early 1920s, Muller was passionately arguing—against most of his geneticist colleagues—that genes, far from being purely hypothetical units conveniently postulated to account for inheritance patterns observed in crossbreeding experiments so as to make them amenable to mathematical treatment (see, e.g., Johannsen Reference Johannsen1923), are in fact real “ultramicroscopic particles […] [that] play a fundamental role in determining the nature of all cell substances, cell structures, and cell activities” (Muller Reference Muller1922: 32). While most early geneticists, including Morgan, displayed a keen awareness of the complex relationship existing between genes and their developmental expression, Muller defended a far more reductionistic and deterministic view that afforded genes ontological and causal priority in the explanation of biological order.

Muller’s views are perhaps most forcefully expressed in his manifesto ‘The Gene as the Basis of Life’ (Reference Muller1929), where he argues, inter alia, that genes are: (i) well-defined physical units composed of chain-like arrangements of elementary parts; (ii) capable of ‘autocatalysis’ in a process analogous to crystallization; and (iii) the primary agents responsible for the organism’s morphological and physiological features. Schrödinger follows Muller in all these respects. He takes for granted that genes are “the most vital parts of an organism” (Schrödinger Reference Schrödinger1944: 2) and, in effect, “the material carrier[s] of life” (ibid.: 3). Also like Muller, Schrödinger suggests that genes are composed of long sequences of repeatable elements, and he uses the crystal analogy to explain their properties.

Muller even anticipated the need for physicists to get involved in the study of the gene. In a 1936 address to the Soviet Academy of Sciences titled ‘Physics in the Attack on the Fundamental Problems of Genetics’, Muller remarked that “genes have properties which are most unique from the standpoint of physics”, and whose elucidation “may throw light not only on the most fundamental questions of biology, but even on fundamental questions of physics as well” (Muller Reference Muller1937: 210). Muller closed his address by calling on physicists to join him in examining these “ultimate particles of life itself” (ibid.), proclaiming that “[t]he geneticist himself is helpless to analyze [them] further. Here the physicist […] must step in. Who will volunteer to do so?” (ibid.: 214). Eight years later, Schrödinger took on precisely this challenge in What Is Life?, though at no point does he discuss or even cite Muller’s work.Footnote ²⁴ I shall have much more to say about Delbrück and TZD, as well as about Muller, in Section 6.

The suggestive notion of an order-from-order principle guiding the operation of the organism, then, is not Schrödinger’s own invention, but rather reflects the bold genocentric views of several early-twentieth-century geneticists, especially Muller. The belief that there is a rigid, unchanging, static structure inside the body that determines its form and function precedes even the advent of genetics. It is implicit in the writings of several late-nineteenth-century biologists, most notably August Weissman (Reference Weismann1893), who famously demarcated what he called the ‘germ-plasm’ from the rest of the body (or ‘soma’), and who considered the former not just the bearer of inheritance but the central directing agency in control of development.

What Schrödinger contributes to the order-from-order principle—aside from making it explicit and deriving its necessity from purely physical considerations—is the groundbreaking idea of a hereditary code-script. This marks the introduction of the term ‘code’ into genetics, which, of course, proved to be an extremely fruitful metaphor. The extraordinary thing about Schrödinger’s proposal is that it precedes the widespread adoption of information-talk in biology. Indeed, it arrives not only a decade before the discovery of the double helix, but even before DNA was conclusively identified as the hereditary substance (the crucial experimental demonstration by Oswald Avery, Colin MacLeod, and Maclyn McCarty was published the same year as What Is Life?).Footnote ²⁵

One of the most common claims made about the historical significance of What Is Life? is that Schrödinger’s code-script was the forerunner of the genetic code—the quasi-universal set of rules by which nucleic acid base sequences are ‘translated’ into the amino acid sequences that make up proteins. The race to ‘crack’ the genetic code (recounted in Cobb Reference Cobb2015; see also Kay Reference Kay2000) was one of the most dramatic events in molecular biology during the period immediately following the discovery of the double helix. (The code was not completely deciphered until 1966.)

Despite the frequency with which this attribution is made (e.g., Symonds Reference Symonds1986; Pauling Reference 90Pauling and Kilmister1987; Moore Reference Moore1989; Sarkar Reference Sarkar1991; Kauffman Reference Kauffman, Murphy and O’Neill1995; Hendrickson Reference Hendrickson, Gumbrecht, Harrison, Hendrickson and Laughlin2011; Gribbin Reference Gribbin2013; Sarkar Reference Sarkar, Harman and Dietrich2013; Sigmund Reference Sigmund2019), a careful reading of What Is Life? suggests that it is based on a misunderstanding of what Schrödinger means by ‘code-script’. This is due to the unfortunate ambiguity inherent in the word ‘code’. Often, this word is used to mean a cipher, that is, a system for exchanging one set of characters with another. The Morse code is an example, as it provides a standardized method for replacing the letters in the alphabet with dots and dashes. The genetic code is also a cipher, as it refers to the correspondence rules that specify how triplets of nucleotide bases (or ‘codons’) can be translated to amino acids.

However, the word ‘code’ can be used in a way that does not involve the translation of a message. It can refer instead to any system or collection of rules or instructions. Think of the highway code, or the Napoleonic code, or a code of conduct. This appears to be closer to what Schrödinger has in mind when he speaks of a hereditary code-script: a complex set of instructions that stipulates the nature and timing of the processes by which cells function and multicellular organisms develop. Notice that while the genetic code concerns only the relation between nucleic acids and proteins, the scope of Schrödinger’s code-script extends far beyond that, “involving all the future development of the organism” (Schrödinger Reference Schrödinger1944: 61).

The reason this has not been generally acknowledged is that Schrödinger himself, as we saw in Section 2, resorts to the Morse code—a cipher—when describing his code-script. It is important to remember, though, that he does so only to illustrate its combinatorial richness. Schrödinger’s point is that “an almost unlimited number” (ibid.) of distinct developmental determinations are possible if the relatively short association of atoms composing the code-script are arranged in different permutations (just as in the Morse code, only two signs in combinations of four or less are more than sufficient to specify all 26 letters of the alphabet). At no point when discussing his code-script does Schrödinger refer to (or imply the existence of ) specific one-to-one correspondence rules between two sets of sequences, or two kinds of substances, which is what a cipher—like the genetic code—essentially involves.Footnote ²⁶

This important interpretive point has been made recently by several commentators (Kay Reference Kay2000: 61–62, Kogge Reference Kogge2012: 627–631, and especially Walsby and Hodge Reference Walsby and Hodge2017). I agree with their analysis, but I think it does not go far enough. Not only is Schrödinger’s code-script a rule-code for development rather than a cypher; I believe that we should regard it as the direct precursor to a biological concept that has proven to be almost as influential as the genetic code, namely the genetic program. Proposed 17 years after the publication of What Is Life? by Jacob and Monod (Reference Jacob and Monod1961)—and simultaneously by evolutionary biologist Ernst Mayr (Reference Mayr1961)—the genetic program “has come to be widely regarded as a fundamental explanatory concept for biological development” (Keller Reference Keller2000: 74). “It equates the genetic material of the egg with the magnetic tape of a computer” (Jacob Reference Jacob1973: 9), and it is considered to play “a decisive role in laying down the structure of an organism, its development, its functions, and its activities” (Mayr Reference Mayr1997: 123). According to this notion, the fertilized egg is assumed to contain a program (akin to a computer program) that directs and controls the developmental process by executing a predetermined set of operations according to algorithmic instructions encoded in its genome.

While the genetic program enjoyed a great deal of popularity during the latter half of the past century, in recent decades it has become increasingly apparent that it fails to provide an adequate understanding of development. Much has been written about the conceptual inconsistencies and empirical inaccuracies of the genetic program (see, e.g., Newman Reference Newman1988; Nijhout Reference Nijhout1990; Lewontin Reference Lewontin2000; Longo and Tendero Reference Longo and Tendero2007; Walsh Reference Walsh2020). I myself have argued in earlier work (i.e., Nicholson Reference Nicholson2014) that the genetic program serves to legitimize three deeply problematic theses concerning the role of genes in development, namely:

1. 1. Neo-preformationism: genes fully specify the outcome of development.

2. 2. Developmental computability: complete knowledge of how genes interact during development should enable us to ‘compute’ the embryo.

3. 3. Genetic animism: genes initiate, direct, and control development.

In what follows, I want to emphasize the remarkable extent to which Schrödinger’s code-script prefigures the concept of the genetic program by showing how the descriptions that Schrödinger offers of it in What Is Life? already imply an unequivocal commitment to all of these theses.

Let us start with neo-preformationism. Preformationism refers to the ancient belief that the structure of the adult organism is already present in miniature form as a ‘homunculus’ encased in either the egg or the sperm, and that development simply consists in the mechanical enlargement of that structure. Today’s neo-preformationism is more subtle: “the organism is not pre-formed in the head of the sperm, but the head of the sperm and the egg nucleus do carry an immensely complex, species-specific, regulatory program for the stepwise process of embryonic development” (Davidson Reference Davidson2009: R217). What development amounts to, in the words of Medawar (Reference Medawar1965: 1329), is “an unfolding of pre-existing capabilities, an acting-out of genetically encoded instructions”. The genome is thus assumed to contain all the information required to specify the organism. Neo-preformationism appears to have provided an important impetus for the Human Genome Project. One of its leading advocates, Nobel laureate Walter Gilbert, memorably declared that the “[t]hree billion bases [of a human’s DNA] sequence can be put on a single compact disk (CD), and [in the future] one will be able to pull a CD out of one’s pocket and say, ‘Here is a human being; it’s me!’” (Gilbert Reference Gilbert, Kevles and Hood1992: 96)

Now, if we look at the passage in What Is Life? where the term ‘code-script’ is first introduced, we find that Schrödinger does so in order to defend neo-preformationism. He begins by discussing “the four-dimensional pattern” of the organism, by which he means “the whole of its ontogenetic development from the fertilized egg cell to the stage of maturity” (Schrödinger Reference Schrödinger1944: 20). He observes that “this whole four-dimensional pattern is known to be determined by the structure of that one cell, the fertilized egg”, specifically by the chromosomes in its nucleus. “It is these chromosomes”, Schrödinger continues, “that contain in some kind of code-script the entire pattern of the individual’s future development and of its functioning in the mature state” (ibid.). He expresses the same idea later, when he asks himself “how this tiny speck of material, the nucleus of the fertilized egg, could contain an elaborate code-script involving all the future development of the organism” (ibid.: 61).

Let us move on to developmental computability, which articulates the deterministic expectation that an embryo can in principle be computed from the complete data set of a fertilized egg. The question is this: “given a total description of the fertilized egg—the total DNA sequence and the location of all proteins and RNA—could one predict how the embryo will develop?” (Wolpert Reference Wolpert1994: 270).Footnote ²⁷ If the fertilized egg contains a program, then development must involve the predictable execution of an algorithmic succession of predetermined steps. The prospect of ‘computing the embryo’ is probably most closely associated with Wolpert, who began exploring it in the 1970s (e.g., Wolpert and Lewis Reference Wolpert, Lewis and Thorbeke1975). His hope was that by deciphering the commands coded in the program, it might be possible to uncover the precise sequence of developmental steps without becoming hopelessly entangled in all the messy molecular and cellular details of the process. Brenner pursued a similar research program during the same period (see de Chadarevian Reference de Chadarevian1998).Footnote ²⁸

Looking back at What Is Life?, it is striking to find Schrödinger endorsing developmental computability so emphatically—and dramatically—in one of the most famous passages in the whole book, where he invokes the unforgettable image of the Laplacian demon to describe the deterministic character of his code-script:

Finally, let us consider genetic animism. This is the belief that DNA exerts executive control over the organism’s operations. It reflects the tendency to ascribe causal powers to genes, and to regard them not just as the material carriers of heredity, but as the primary agents of life. It is manifested when developmental geneticists speak of ‘master control genes’ (e.g., Gehring Reference 85Gehring1988) and it is a recurring theme in popular science books, such as Richard Dawkins’ bestseller The Selfish Gene (Reference Dawkins1976). Genetic animism stems from what Evelyn Fox Keller has called the ‘discourse of gene action’, which characterized the thinking of some early American geneticists like Muller, for whom the cell is “only a by-product, originally, of the action of the gene material” (Muller Reference Muller1929: 918). Keller herself puts it best: “Part physicist’s atom and part Platonic soul, [the gene] was assumed capable simultaneously of animating the organism and of directing (as well as enacting) its construction” (Keller Reference Keller2000: 47).

Schrödinger speaks of genes in exactly this way in What Is Life?. He states that they “play a dominating role in the very orderly and lawful events within a living organism. They have control of the observable large-scale features which the organism acquires in the course of its development, [and] they determine important characteristics of its functioning” (Schrödinger Reference Schrödinger1944: 19, my emphasis). He also attributes agency to the code-script, noting that “the code-script must itself be the operative factor bringing about the development” (ibid.: 62). Indeed, he recognizes (in another famous passage) that

Sometimes, Schrödinger expresses more than one of the three theses in a single sentence, such as when he notes that “the miniature code should precisely correspond with a highly complicated and specified plan of development [neo-preformationism] and should somehow contain the means to put it into operation [genetic animism]” (ibid.: 62). It is difficult not to be reminded of such passages when reading Jacob and Monod’s introduction of the genetic program concept 17 years later, when they conclude—clearly echoing Schrödinger—that “the genome contains not only a series of blue-prints [neo-preformationism], but a co-ordinated program of protein synthesis and the means of controlling its execution [genetic animism]” (Jacob and Monod Reference Jacob and Monod1961: 354).

Today, these three theses are perceived as seriously misguided by a growing number of biologists—not to mention philosophers of biology (see, e.g., Oyama et al. Reference Oyama, Griffiths and Gray2001). There are good reasons for this. The information required to specify an organism does not come preformed in the DNA. It emerges progressively through the interaction of DNA with other cellular components, as well as with the environment. Development is not the gradual unfolding of the organism from a prespecified genetic plan. It is a highly dynamic and heterogeneous process of construction involving the confluence of numerous interacting causal factors, only some of which have their basis in the DNA.

Similarly, developmental computability is an illusion. Gene expression—like any molecular process—is subject to stochastic fluctuations. Developmental ‘instructions’ reflect probabilistic tendencies, not deterministic mandates. The goal to ‘compute the embryo’ is irreconcilable with the fundamentally stochastic character of the molecular and cellular processes that underlie development. We will return to this point in Section 4.

Lastly, genetic animism attributes a causal agency to genes that they simply do not possess on their own. It is difficult to see how genes could possibly be responsible for initiating, directing, and controlling development, given that DNA is not an inherently active molecule, but rather requires activation from without. By itself, DNA is inert, relatively unstructured, and non-functional. To be functional, it needs to be embedded in an already organized, living cell. Despite what is often claimed, genes are not ‘self-replicating’. Nor do they ‘make’ proteins, or ‘switch on’ molecular pathways. All these processes require the orchestrated involvement of many other molecules inside the cell, especially proteins. In other words, it is only in the presence of the highly structured cellular environment that any talk of ‘gene action’ even makes sense.

Crucially, it is not just the function of genes that is a product of the cellular context, but also their stability. This is the real key to solving the puzzle that lies at the heart of What Is Life?. Schrödinger was undoubtedly perceptive in reasoning that the material substratum of the code-script would need to be bound together by the same covalent bonds that make crystals so stable. The DNA double helix is indeed stabilized by long chains of phosphodiester covalent bonds linking the external sugar–phosphate strands of the helix, as well as by the hydrogen bonds connecting the complementary pairs of nucleotide bases inside the helix (three in the case of cytosine and guanine, two in the case of adenine and thymine). But the stunningly reliable transmission of order during reproduction that we observe in the perpetuation of genetic traits like the Habsburg lip (recall Figure 5) is not primarily a consequence of the crystal-like stability of the DNA molecule.

The reason the replication, transcription, and translation of genetic information is so astonishingly accurate is that there is an entire army of proteins in the cell whose job it is to proofread and correct any errors accrued during these processes.Footnote ²⁹ It is not that genes behave as if they were at absolute zero; genetic processes are as susceptible to thermal agitation as anything else! It is just that the cell devotes a considerable amount of its energetic resources to minimizing the disruptive effects of stochasticity. Genes do not impose stability on the cell; it is rather the other way round. To put it another way, the stability that so deeply impressed Schrödinger is not an intrinsic property of the structure of the genetic material, but a continuous accomplishment of the cell operating as a whole.

Summing up, what should we make of the order-from-order principle in light of current knowledge? In one sense, Schrödinger was right to suppose that there is a “stream of order” (Schrödinger Reference Schrödinger1944: 77) transmitted from parent to offspring, and that it is intimately connected to the hereditary substance. It is also the case that this substance has an aperiodic atomic structure—exactly as Schrödinger predicted—that allows it to store inordinate amounts of data, which helps account for the amazing diversity of organisms. There is even a code-script of sorts after all, as the genome does specify the amino acid constitution of the proteins that carry out most functions in the cell.

On the other hand, Schrödinger was wrong to assume that the code-script determines all of the organism’s observable characteristics (i.e., its phenotype). This mistake was compounded by the molecular biologists who derived the idea of the genetic program from his code-script, as they took genes to provide not just instructions to synthesize proteins, but a blueprint to make an organism. The phenotype cannot be read from the information stored in the genes. Genetic information acts only as a resource for development, not as a step-by-step guide. It only acquires meaning in the context of the cell and of the external environment. Schrödinger was also wrong to endow genes with legislative and executive powers. DNA does nothing alone. It is the whole cell that acts. Even the stability of genes is an achievement of a coordinated set of subcellular processes—which is what a molecular biologist today would answer if asked about the puzzle that troubled Schrödinger in What Is Life?.

Finally, we now know that the order transmitted in reproduction is not exclusively encoded in the DNA. Although the amino acid sequence of proteins derives from DNA, protein function is primarily determined by cellular context (Kupiec Reference Kupiec2010; Nicholson Reference Nicholson2019). Moreover, many subcellular compartments and structures, such as membrane-bound organelles, as well as the plasma membrane itself, serve as templates for their own replication—they are the only source of ‘information’ for their own regeneration. Proteins do get incorporated into them, but they are always already formed and thus constitute distinct and irreducible repositories of cellular order.Footnote ³⁰ It is a mistake to think of the information encoded in the DNA as the only thing that gets transmitted from parent to offspring. Membranes and organelles, basal bodies and microtubule organizing centres, are all inherited as well. These channels of epigenetic inheritance have been the subject of much recent research (see Jablonka and Lamb Reference Jablonka and Lamb2014; Bondiuransky and Day Reference 83Bonduriansky and Day2018). Rudolf Virchow was right when he memorably declared in the mid nineteenth century that it takes a cell to make a cell. None of the cell’s constituents, including DNA, is the sole basis of the order it exhibits. As Moss (Reference Moss2003: 91) has aptly put it, “Schrödinger’s order-from-order descriptor well characterizes the cell as a whole—but only as a whole”.

4 A Critique of the Order-from-Disorder Principle

The other key principle that Schrödinger discusses in What Is Life?—indeed, the one he starts with—is the order-from-disorder principle. It is to this notion that we now turn. As we saw in Section 2, this principle describes how predictable patterns of orderly behaviour emerge when a large number of entities is considered collectively, even if the composing entities, taken individually, display no order whatsoever. Schrödinger discusses and even graphically illustrates several examples of order-from-disorder phenomena in the first chapter of What Is Life?, three of which are reproduced in Figure 4. This principle is the foundation upon which the edifice of statistical mechanics is built. Let us briefly review the origins of this field, as this will help us understand the significance of the principle.

The rise of statistical thinking in the nineteenth century led to a profound transformation in our understanding of nature (see Porter Reference Porter1986). Chance, which had previously been regarded as a sign of our ignorance of causal relations, became ‘tamed’—to use Ian Hacking’s (Reference Hacking1990) suggestive phrase—when it was realized that robust statistical regularities could be identified in large data sets. Even in the absence of causal knowledge, it became possible to make reliable predictions using the language of probability. Initially, the adoption of the statistical method in physics was a matter of practical convenience. When James Clerk Maxwell developed the kinetic theory of gases—which explains the macroscopic properties of gases (e.g., volume, pressure, temperature) in terms of the positions and velocities of their constituent microscopic particles—the application of statistics allowed billions of gas particles to be treated as a single ensemble, thereby avoiding the utterly impractical task of having to calculate the trajectory of each individual particle. It was later realized, however, that even if it were possible to track the trajectory of every particle, this would not tell us anything that we could not already surmise from a statistical treatment of the ensemble as a whole. In the statistical mechanics of Boltzmann and J. Willard Gibbs, it is the behaviour of the collective, rather than that of any given individual, that is of interest. The fact that statistics focuses on populations at the expense of ignoring individuals came to be regarded as a strength rather than a limitation. For these physicists, statistics was no longer a convenient methodological tool; it had become an indispensable explanatory resource.

A sublime illustration of this is Boltzmann’s statistical explanation of the second law of thermodynamics. Confronted with the problem of reconciling the irreversible trend for the entropy in a system to increase (a phenomenon later described by Arthur Eddington as the ‘arrow of time’) with the reversible nature of the mechanical interactions of the microscopic particles in that system, Boltzmann showed that this inevitable entropic increase reflects only a statistical tendency for the system to move towards more probable, and more disordered, microscopic arrangements. To put it another way, the observed directionality of macroscopic processes that the second law prescribes arises from the expected statistical behaviour of the ensemble of microscopic particles over time. The law therefore expresses a probability, not an absolute certainty—though, given the astronomical number of particles typically involved (recall the law of large numbers discussed in Section 2), it is a probability that can safely be regarded as a certainty for all practical purposes.

By the early twentieth century, it had become clear that the lawful regularities observed for many physical and chemical phenomena were similarly statistical—all following the same order-from-disorder principle discussed in the first chapter of What Is Life?.Footnote ³¹ Towards the end of the book, Schrödinger poetically describes this principle as “our beautiful statistical theory of which we [physicists and chemists] were so justly proud because it allowed us to look behind the curtain, to watch the magnificent order of exact physical law coming forth from atomic and molecular disorder” (Schrödinger Reference Schrödinger1944: 80).

Nevertheless, we should not forget that one of the main theses in What Is Life? is that the order-from-disorder principle, despite its overwhelming importance in physics and chemistry, is irrelevant to the explanation of biological order. As Schrödinger asserts at the start of the book, “it is in relation to the statistical point of view that the structure of the vital parts of living organisms differs so entirely from that of any piece of matter that we physicists and chemists have ever handled physically in our laboratories or mentally at our writing desks” (ibid.: 2–3). In this section, I will examine this contention. We will see that a compelling case can be made, contra Schrödinger, that the order-from-disorder principle is as central to explanations in biology as it is to explanations in physics.

The simplest way to make this case is to point to Charles Darwin’s theory of evolution, which notably deals with populations, as opposed to individuals. No individual ever evolves; it is rather the population, considered over many generations, that is subject to evolutionary change. The activities of individual organisms may be disorderly or unpredictable, but as long as there is heritable variation in the population that leads to differences in reproductive success among its members, then evolution by natural selection (an order-producing process if there ever was one!) will inevitably follow (Lewontin Reference Lewontin1970).Footnote ³² Modern evolutionary biology is statistical through and through. In fact, the triumvirate credited with supplying the mathematical foundation of evolutionary theory, namely Francis Galton, Karl Pearson, and R. A. Fisher, contributed as much to the establishment of mathematical statistics as to the theory of evolution. Indeed, they viewed the two as inseparable (Pence Reference Pence2022).Footnote ³³ In evolutionary biology, just as in statistical mechanics, collective patterns of order can be reliably predicted despite—or rather, because of—the unpredictable nature of individual events. As Fisher paradoxically put it, “[t]he effects of chance are the most accurately calculable, and therefore the least doubtful, of all the factors of an evolutionary situation” (quoted in Porter Reference Porter1986: 319).Footnote ³⁴

Population genetics and evolutionary biology are not outliers. Order-from-disorder phenomena are studied in other fields that deal with the dynamics of biological populations, such as ecology, where the Lotka-Volterra predator-prey equations are sometimes compared to the ideal gas law in physics (e.g., O’Dwyer Reference O’Dwyer2020), or ethology, where collective animal behaviours—like the trajectories of flocking birds—are routinely investigated using models drawn from statistical mechanics (e.g., Bialek et al. Reference Bialek, Cavagna, Giardina, Mora, Edmondo, Viale and Walczak2012). Thus, it is simply not true that the order-from-disorder principle plays no role in the living world.

But perhaps we are being uncharitable to Schrödinger. After all, What Is Life? is concerned with the order displayed by an individual cell, or organism, not by a population. Can Schrödinger’s claim be maintained if it is understood in this more restricted way? On the face of it, if we consider a single cell, Schrödinger cannot be faulted for downplaying the order-from-disorder principle. We now know that most functionally important macromolecules in the cell exist in such minuscule quantities that the law of large numbers does not apply to them.Footnote ³⁵ As a result, the internal architecture of the cell must impose constraints on where specific individual molecules find themselves at any given time in order to maximize the probability that they will interact in a way that ensures that all cellular functions are carried out appropriately and in a timely fashion (see Nicholson Reference Nicholson2019).

The most dramatic example is the DNA molecule in a bacterial cell, of which there is usually only one copy, and yet the cell’s survival depends on the information in this one molecule being appropriately replicated, transcribed, and translated in a timely fashion by an array of other molecules. Note, though, that replication, transcription, and translation are just as probabilistic as any other sequence of chemical reactions. Indeed, the stochastic character of gene expression accounts for why there is so much variation between genetically identical cells—the reason being that the molecular processes that underlie them, owing to thermal agitation, never occur in exactly the same way or take precisely the same time.Footnote ³⁶

The cell, then, appears to make do without relying on order-from-disorder regularities. What about the multicellular organism? Recall that Schrödinger introduces his hereditary code-script idea not merely, or even primarily, to account for the order exhibited by the cell, but to explain the formidably complex multicellular process of development. We already saw in Section 3 that the genetic program model, which I suggested is the direct successor to Schrödinger’s code-script, faces numerous problems. Do order-from-disorder approaches to development fare any better?

Despite the preponderance of preformationist, order-from-order reasoning in the history of embryology, the process of development is actually quite amenable to a characterization in terms of statistical mechanics (see, e.g., García-Ojalvo and Martínez Arias Reference García-Ojalvo and Martínez Arias2012). The spatiotemporal organization of the developing embryo requires highly coordinated sequences of cellular differentiation events. These events result from ‘decisions’ made by individual cells about their fate, which are in turn prompted by the expression of certain genes. Although these cell fate decisions are intrinsically stochastic and are not reproducible on an individual basis, they nevertheless result in highly regular—almost deterministic—spatiotemporal patterns at the level of the whole ensemble of cells (the embryo). In other words, order at the macroscopic scale (i.e., the reliable generation of tissues and organs) predictably arises out of disorder at the microscopic scale (i.e., the noisy molecular process of gene expression). It is possible in principle, then, to understand development as an order-from-disorder phenomenon.

What makes statistical mechanics so appealing as a theory is that it provides a systematic way of connecting the macroscopic properties of a system to its microscopic constituents. Interestingly, this has become one of the chief goals of developmental biology: to establish links between events at the molecular (or microscopic) scale and those at the multicellular (or macroscopic) one. And researchers are increasingly looking to the conceptual resources of statistical mechanics to aid them in this task. For instance, Boltzmann’s aforementioned confrontation with the entropic arrow of time is clearly being evoked when researchers emphasize the need “to explain the ‘arrow of time’ in gene expression dynamics” (Teschendorff and Feiberg Reference Teschendorff and Feinberg2021: 461). This refers to the challenge of accounting for the macroscopic directionality of gene expression patterns as cells progressively differentiate to form distinct tissues given the microscopic reversibility of the activation and repression of the individual genes involved. Another example is the suggestion that pluripotency is a statistical property of stem cell populations that typically arises out of highly variable molecular microstates (see MacArthur and Lemischka Reference MacArthur and Lemischka2013).

A useful way of comparing order-from-order and order-from-disorder accounts of development is to consider their respective stances on determinism. Schrödinger’s code-script, and the genetic program that arose from it, takes the developmental process to be determined all the way down to the molecular level. The obvious determinacy of development at the macroscopic level (i.e., the fact that the journey from fertilized egg to adult organism reliably follows a well-defined series of morphological stages that results in an essentially predictable outcome) is seen as a direct consequence of the deterministic action of genes, which are assumed to contain a detailed blueprint for development, or at least a complex set of algorithmic instructions for its programmatic execution.

In contrast, order-from-disorder models of development do not infer determinism at the microscopic scale on the basis of the observed determinacy at the macroscopic scale. This point was perceptively made half a century ago by the organicist developmental theorist Paul Weiss (Reference Weiss1973), who illustrated it in a memorable diagram, redrawn in Figure 6. Weiss pointed out that the preservation of topographic relations during two successive stages of development allows us to experimentally identify regions of an early embryo as being the predictably earmarked forerunners for the formation of specific organs (e.g., heart, brain, liver, kidneys). Nevertheless, such clear-cut correlations between stages do not hold at the cellular level (and even less so at the molecular level). That is, if we follow the differentiation trajectories of particular cells in these prospective organ areas from the earlier to the later period, we find them taking far more fortuitous routes, differing individually from case to case. Thus, although the coarse-grained development of organ rudiments can be mapped out in a precisely determinable fashion, the fine-grained molecular and cellular processes that underlie it, when considered individually, remain indeterminate and unpredictable. Macroscopic order, in this model, far from being the mechanical amplification of an underlying microscopic order, arises in the system despite the indeterminate and disorderly activities of the individual microscopic elements. Weiss himself pithily described the resulting situation as “determinacy in the gross despite demonstrable indeterminacy in the small” (ibid.: 21–22).Footnote ³⁷

Figure 6 Schematic depiction of the persistence of the topographic pattern of whole embryonic regions despite a lack of regularity in the positions of their constituent cells. The left side shows the macroscopic view of the regions of an embryo between developmental stages S₁ and S₂. At t₂, every embryonic region (A₂, B₂, C₂, D₂, E₂) can be traced back to a corresponding region (A₁, B₁, C₁, D₁, E₁) at t₁. Yet if we take a microscopic view of the same process, shown on the right side, we find that even two very closely related embryos (top and bottom) exhibit a great deal of variation in the precise trajectories undertaken by their cells (depicted as symbols) in each embryonic region

(based on Weiss Reference Weiss1973)

The case for understanding development as an order-from-disorder phenomenon can be strengthened further by broadening our conception of the order-from-disorder principle itself. So far, we have taken Schrödinger’s own formulation of the principle at face value, as we have assumed that linear statistical effects are the only way of getting order out of disorder. But we now know that this is not so. Eight years after What Is Life? appeared, the mathematician Alan Turing published a landmark paper titled ‘The Chemical Basis of Morphogenesis’ (Reference Turing1952) that suggested an alternative way for order to arise from disorder. Turing showed that a system of hypothetical chemical substances (which he called ‘morphogens’) reacting and diffusing together can spontaneously produce periodic patterns—peaks and valleys of concentration—in the form of spots, stripes, or spirals not unlike those found in the living world. He did this by providing a pair of differential equations and demonstrating mathematically that by adjusting the relevant parameters (such as the reaction rate and diffusion coefficient of each morphogen) the system could be made to produce a wide array of spatial patterns.

Despite its novelty (or perhaps because of it), Turing’s paper was all but ignored for twenty years, save for a few exceptions (e.g., Waddington Reference Waddington1962; Maynard Smith Reference Maynard Smith1968). However, in 1972 mathematical biologists Hans Meinhardt and Alfred Gierer revived Turing’s ‘reaction–diffusion model’ and showed that a system of two interacting morphogens—specifically, a slow-diffusing activator coupled with a fast-diffusing inhibitor—could in principle account for the basic properties of biological patterning processes (Gierer and Meinhardt Reference Gierer and Meinhardt1972). Since then, numerous kinds of stable biological patterns have been successfully recreated in computer simulations using variations of this model, including the spiral arrangement of leaves on a plant stem (a phenomenon called ‘phyllotaxis’), the distribution of feather buds, and the pigmentation patterns of mollusc shells (Kondo and Miura Reference Kondo and Miura2010; Ball Reference Ball2015). Perhaps the most impressive reaction–diffusion computer models are those that faithfully replicate the characteristic markings on the fur coat of feline predators and the elaborate stripe patterns of tropical fish, examples of which are shown in Figure 7a and 7b.Footnote ³⁸

Figure 7 Examples of kinds of order-from-disorder processes not considered by Schrödinger in What Is Life?. (a) Reaction–diffusion simulation of the ‘rosette’ markings of a jaguar (adapted from Ball Reference Ball2015); (b) Reaction–diffusion simulation of the radial stripe patterns of the map pufferfish Arothron mappa (adapted from Kondo and Miura Reference Kondo and Miura2010); (c) Patterns consisting of travelling waves of chemical concentrations produced by a Belousov–Zhabotinsky reaction in a petri dish

(Credit: David Maitland)

Still, the striking resemblance of these computer simulations to the real patterns displayed by organisms does not by itself establish that Turing-style reaction–diffusion systems are causally responsible for them. In fact, the experimental identification of diffusing activator and inhibitor morphogens has proven elusive. Turing’s theory suffered a serious blow early on when it was discovered that the stripe-like patterns in the early Drosophila embryo are not produced by an elegant reaction–diffusion system (as had been conjectured by some of the theory’s earliest advocates) but instead results from the sequential activation of an unwieldy cascade of genes according to a gradient of mRNA molecules laid down in the egg prior to fertilization (Akam Reference Akam1989). This important finding, which earned its discoverers the Nobel Prize in 1995, suggested that at least some aspects of Drosophila development are best approached from a conventional order-from-order perspective.Footnote ³⁹ In turn, this made most developmental biologists skeptical of the idea that reaction–diffusion systems are genuinely involved in biological pattern formation. As a result, Turing’s theory in biology has remained, until very recently, nothing more than a tantalizing hypothesis—a possible explanation awaiting empirical confirmation (cf. Harrison Reference Harrison1987; Keller Reference Keller2002: chap. 3; Maini Reference Maini2004).

The situation has changed dramatically in the last fifteen years, as real-life Turing-style morphogens have now been identified for a considerable number of developmental phenomena.Footnote ⁴⁰ To mention just one example, the regularly spaced ridges in the mammalian mouth palate (which are called ‘rugae’) appear to be the product of a reaction–diffusion system involving the proteins fibroblast growth factor and Sonic hedgehog, which operate respectively as activator and inhibitor. Nevertheless, it is becoming apparent that Turing’s original model is probably too simple to realistically capture the actual dynamics of patterning processes during morphogenesis (a possibility that Turing already contemplated in his Reference Turing1952 paper). Although the existence of reaction–diffusion systems is no longer in doubt, it is likely that most developmental patterns are produced not by simple activator–inhibitor pairs, but by more complex systems involving three or more factors—perhaps in combination with the activation of gene regulatory networks.Footnote ⁴¹

Be that as it may, the key lesson for us here is that Turing’s model provides precisely what Schrödinger thought was inconceivable: a means of producing order during development in the absence of a pre-existing genetic template (i.e., a code-script). As Ball has aptly observed, “Turing seems to have identified one of nature’s general mechanisms for generating order from macroscopic uniformity and microscopic disorder” (Ball Reference Ball2015: 9). Notice that this order-from-disorder mechanism is totally different from the one that Schrödinger speaks so fondly about in What Is Life?, as its physical foundation is not late-nineteenth-century statistical mechanics but mid-twentieth-century non-equilibrium thermodynamics. Indeed, as it turns out, ‘Turing patterns’ are but one example of a much larger set of order-from-disorder phenomena that can be brought together under the banner of self-organization.Footnote ⁴²

Around the time that Turing was developing his theory, the Soviet chemist Boris Belousov used a mixture of reagents to create oscillating patterns in vitro, which were further explored by Anatoly Zhabotinsky in the 1960s. The so-called ‘Belousov–Zhabotinsky reaction’ (Figure 7c), became one of the first well-understood instances of chemical self-organization. The connections between these travelling waves and Turing’s stationary patterns began to emerge shortly thereafter in the pioneering work of Ilya Prigogine and his group at the Free University of Brussels, who named them ‘dissipative structures’ because they are sustained by the dissipation of energy in a non-equilibrium process. Dissipative structures are everywhere. Familiar examples include candle flames, whirlpools, and tornadoes. Prigogine’s greatest achievement (for which he was awarded the Nobel Prize in 1977) was to show how self-organization arises in nature—that is, to explain how the macroscopic patterns of order displayed by dissipative structures arise spontaneously from non-linear molecular interactions and become dynamically stabilized in far-from-equilibrium conditions through an ongoing flux of energy and matter (see Nicolis and Prigogine Reference Nicolis and Prigogine1977; Prigogine and Stengers Reference Prigogine and Stengers1984).

The crucial biological significance of these developments is that organisms themselves are also dissipative structures—albeit of a vastly more complex kind than any of the aforementioned examples. Thermodynamically speaking, a cell is a membrane-bound open system (open, that is, to the flow of both energy and matter) that maintains itself in an irreversible, low-entropic ‘steady state’ far from equilibrium.Footnote ⁴³ Not only is the cell as a whole a dissipative structure, but research in the last couple of decades has confirmed that many subcellular assemblies (including the mitotic spindle, the Golgi apparatus, and the nucleolus) are highly dynamic steady-state systems capable of spontaneously self-organizing into morphologically and functionally distinct configurations through intrinsically stochastic interactions (Kirschner et al. Reference Kirschner, Gerhart and Mitchison2000; Karsenti Reference Karsenti2008; see also Nicholson Reference Nicholson2019).Footnote ⁴⁴

Empirical findings of this kind serve as powerful reminders that the ‘information’ that specifies the cell’s spatiotemporal organization is not wholly encoded in the genome. It is highly misleading to speak of a genetic blueprint for the cellular architecture when we know that so much of it is generated by self-organization in the absence of an external template or a global plan. Of course, this is not to say that genes are not important; it is only to assert that they do not causally bring about the spatiotemporal organization of the cell, as is implied by the idea of a code-script or a program. Instead, gene products are released into a cellular milieu that is always already structured, and they exert their influence under the physical constraints of the existing four-dimensional order—much of which is shaped and reconstituted by ongoing self-organizing processes.

It is time to take stock of what our analysis in this section has revealed. The order-from-disorder principle, as Schrödinger understands it, is as indispensable to biology as it is to physics. Statistical mechanics is indeed founded on it, but so is modern evolutionary biology. Recent research suggests that it is likely to become indispensable in developmental biology as well, and not just because of the problems and limitations of competing order-from-order models. In addition, it is clear in hindsight that Schrödinger’s construal of the order-from-disorder principle in terms of linear statistical regularities is overly restrictive, as there are other ways in which order can emerge from disorder in nature, most notably by self-organizing processes. Today, we know that such processes are crucial not just for pattern-formation during morphogenesis (as Turing predicted), but more fundamentally for the generation and maintenance of the cell’s internal architecture.

We can conclude, then, that life is a manifestation of order-from-order and order-from-disorder processes. The former are necessary for its propagation, but the latter play a key role in its conservation. Both are essential for its perpetuation over longer (evolutionary) time scales.

In Section 5, we will turn our attention to the historical reception of What Is Life? and challenge the assessment that professional historians have given of the book’s role in the development of molecular biology. Before doing so, however, it is necessary to dispel some persistent misconceptions about the legacy of Schrödinger’s book in discussions of self-organization and non-equilibrium thermodynamics, as these are topics that we have addressed in this section.

One of the most surprising things about the way What Is Life? is discussed today is that it is often described as a foundational document in biological thermodynamics—despite the fact that, as we saw in Section 2, Schrödinger only discusses thermodynamics as an aside in the penultimate chapter to point out that metabolism is what keeps the organism from succumbing to entropic decay. Although Schrödinger never uses terms such as ‘self-organization’, ‘steady state’, ‘open system’, or ‘far-from-equilibrium conditions’, it is possible to read him, with the benefit of hindsight, as having anticipated certain features of dissipative structures (as Prigogine would later call them). If this was all that was claimed, there would hardly be a need to even mention it. But many contemporary advocates of non-equilibrium thermodynamics go much further, and in the process, they do serious violence to the facts.

Their first mistake is to credit Schrödinger with identifying and resolving the conflict between life’s propensity towards order and the inexorable increase in disorder mandated by the second law of thermodynamics. Although Schrödinger explains away the apparent contradiction lucidly and concisely in What Is Life?, he was certainly not the first to do so. The tension between life and the second law was identified almost a century earlier by the founders of thermodynamics, particularly Hermann Helmholtz and Lord Kelvin. Thereafter, various late-nineteenth-century authors tried to elucidate the peculiar kinds of stability exhibited by organisms. Herbert Spencer spoke of a ‘moving equilibrium’, Gustav Fechner of a ‘tendency toward approximate stability’, Emil Du-Bois Reymond of a ‘dynamic balance’, and so on. Already by the 1920s and 1930s, biochemists were describing the thermodynamics of living systems in recognizably modern terms.Footnote ⁴⁵ Donnan, who as I indicated in Section 1 was one of Schrödinger’s most important influences, wrote about the topic in 1928 in terms that clearly foreshadow the famous remarks we find in What Is Life?:

The second mistake that biological thermodynamicists make is to anachronistically project back the modern meaning of order-from-disorder as self-organization onto Schrödinger’s Boltzmannian conception of order-from-disorder, thereby occluding the vital role played by statistical mechanics in What Is Life?. Again, although Schrödinger does discuss how the organism takes in energy and matter from its surroundings as a means of evading entropic decay, the possibility of organization spontaneously emerging in far-from-equilibrium conditions—that is, of order-from-disorder in the modern sense of non-equilibrium thermodynamics—does not even occur to him. At any rate, it is a notion that runs directly counter to what Schrödinger is arguing for in his book.Footnote ⁴⁶

But the misinterpretations of What Is Life? go deeper still. Recent commentators have claimed that Schrödinger argues that the order-from-disorder principle is just as central to the explanation of life as the order-from-order principle (e.g., Hendrickson Reference Hendrickson, Gumbrecht, Harrison, Hendrickson and Laughlin2011; Sigmund Reference Sigmund2019), or that he answers the question ‘What is life?’ twice, first using his order-from-order principle and then using his order-from-disorder principle (e.g., Walsh Reference Walsh2015), or—perhaps most bizarrely of all—that he prophetically envisioned two research programs for future biology: one based on his order-from-order principle (which resulted in molecular biology) and another based on his order-from-disorder principle (which resulted in non-equilibrium thermodynamics) (e.g., Murphy and O’Neill Reference Murphy and O’Neill1995; Hendrickson Reference Hendrickson, Gumbrecht, Harrison, Hendrickson and Laughlin2011).

One can find all of the above misconceptions in the writings of the ecological thermodynamicist Eric Schneider. Schneider brands the tension between life and the second law ‘The Schrödinger Paradox’, falsely claiming that “Schrödinger was the first to emphasize the need to grapple with life from a thermodynamic perspective” (Schneider and Sagan Reference Schneider and Sagan2005: 16).Footnote ⁴⁷ He declares anachronistically that Schrödinger “proposed that to study living systems from a nonequilibrium perspective would reconcile biological self-organization and thermodynamics” (Schneider and Kay Reference Schneider and Kay1994: 26), and even that Schrödinger “expected that such a study would yield new principles of physics” (ibid.)—thereby misinterpreting the oft-quoted remark about ‘new laws of physics’ we discussed in Section 2. He also mistakenly takes Schrödinger to be arguing that “life was comprised of two fundamental processes; one ‘order from order’ and the other ‘order from disorder’” (ibid.: 25) and that these principles “outlined two future sciences: the molecular biology that has proved to be such a force in the world, and the thermodynamics of biology that has yet to prove its mettle” (Schneider and Sagan Reference Schneider and Sagan2005: 7–8).

It seems that Schneider himself is responsible for fabricating and spreading these myths about What Is Life?. Their first appearance in the literature is in a brief letter that Schneider wrote to Nature in 1987 (i.e., Schneider Reference Schneider1987). One does not find them in any of the earlier works devoted to biological thermodynamics (e.g., Morotwitz Reference Morowitz1970; Dyson Reference Dyson1985; Wicken Reference Wicken1987). When What Is Life? is mentioned in these earlier treatments of the subject, it is generally in a critical vein to point out that Schrödinger overestimated the importance of his order-from-order principle. Before Schneider, nobody appears to have conflated Schrödinger’s Boltzmannian order-from-disorder principle with the Prigoginian notion of self-organization, as commentators today so frequently do. In fact, Prigogine went out of his way to avoid this precise confusion, declaring that he “introduced the term ‘dissipative structures’ to contrast such structures from the equilibrium structures […] based on Boltzmann’s order[-from-disorder] principle” (Nicolis and Prigogine Reference Nicolis and Prigogine1977: 4) and that he prefers to “call this order ‘order through fluctuations’ to contrast it with the Boltzmann order[-from-disorder] principle” (ibid.: 5).Footnote ⁴⁸

So why have Schneider and others gone to so much trouble to artificially inflate the importance of What Is Life? for the development of biological thermodynamics? The answer, I think, is obvious. Schrödinger is, after all, one of the most revered physicists of the twentieth century—a veritable “deity of science”, in Schneider’s own words (Schneider and Sagan Reference Schneider and Sagan2005: 12). To be able to claim him as the founder of your discipline is to bestow a considerable amount of respectability on what you are doing. Moreover, if the spectacular successes of molecular biology can be plausibly interpreted as the process of working out the implications of Schrödinger’s order-from-order principle, then, provided that you grant yourself the license to liberally reconstruct his order-from-disorder principle so that it refers to self-organization (as opposed to the regularities described by statistical mechanics), you can triumphantly proclaim without much fear of embarassment that “[p]erhaps fifty years from now What Is Life? will be seen as prophetic for its treatment of the thermodynamics of living systems rather than for the prediction of the structure of the gene” (Murphy and O’Neill Reference Murphy and O’Neill1995: 3). This is exactly what Schneider means when he “advocate[s] flipping over Schrödinger’s record and listening to its other side” (Schneider and Sagan Reference Schneider and Sagan2005: 24).

But enough about thermodynamics. Let us return now, without further delay, to the more famous first side of ‘Schrödinger’s record’.

5 Rethinking Schrödinger’s Impact on Molecular Biology

Despite the problematic appropriation of What Is Life? as a foundational document in biological thermodynamics, most readers might reasonably assume—given the wealth of evidence presented in Section 1—that the book’s reputation as a cornerstone of molecular biology remains firmly secured. In fact, the opposite is the case. Surprisingly, perhaps, a consensus of sorts has emerged among historians of science that Schrödinger’s book actually had little discernible impact on the rise of molecular biology, and that its importance for the field was recognized only retrospectively by its foremost practitioners. Historians have drawn attention to the fact that the bombastic declarations of influence collected in Box 1 were all made after the establishment of molecular biology (Wilkins was the first to do so in his Nobel Lecture of 1962, and others followed shortly thereafter), and that such flamboyant pronouncements were really nothing more than thinly veiled self-serving attempts to confer an aura of epistemic prestige on the new biology of the post-war period by conveniently associating it with the intellectual authority of one of the greatest physicists of the twentieth century.Footnote ⁴⁹

Pnina Abir-Am was one of the first to ‘deconstruct’ these autobiographical attributions of influence by leading molecular biologists as “efforts to gain legitimacy as the proponents of a new scientific ultra-discipline” (Abir-Am Reference Abir-Am1985: 104) and as attempts to “define their own history while ‘proving’ their affiliation to a hero-scientist, thereof projected as an ancestor of molecular biology” (ibid.: 113). Abir-Am identifies three reasons for the persistent appeals to Schrödinger’s What Is Life?. First, as a founder of quantum mechanics—a discipline that had radically transformed physics some decades earlier—Schrödinger was perceived as a respectable revolutionary, which is precisely the image that molecular biologists sought to convey of their own field: revolutionary (to help attract newcomers) yet respectable (to gain institutional acceptance as an autonomous new area of research and teaching). Second, Schrödinger’s intrepid theorizing from physical principles and his apparent lack of concern with chemical details exemplified the new approach to biology that molecular biologists hoped to promote. And third, since Schrödinger was no longer alive, it was “safe to invoke him as a real ancestor, i.e., part of a cult of the notable dead, [as] he could not interfere or reply that he neither anticipated molecular biology nor took an interest in it” (ibid.: 104).

Others have made similar claims. Lily Kay (Reference Kay2000: 59) discusses how the pronouncements in Box 1 “serve to buttress this ‘founding father’ narrative”, thereby “reinforcing the canonization process”. Leah Ceccarelli (Reference Ceccarelli2001: 65) writes that the molecular biologists making those pronouncements were jumping on the “ex post facto bandwagon” by “draw[ing] on the authority of a text that would rationalize their professional choices to others” (ibid.: 66). And Keller (Reference Keller1990: 404) remarks that “Schrödinger’s legacy [in molecular biology] depended so little on the utility of any of his particular biological arguments, and so much on disciplinary politics”.

In effect, what these historiographical analyses suggest is that What Is Life? had already lost its topicality as a ‘scientific object’ by the 1950s, and “by the 1960s the molecular biologists were looking back with nostalgia to a ‘historical object’” (Witkowski Reference Witkowski1986: 267). It is this “later decontextualization and reinvention of What Is Life? [that] enhanced its prescience and durability” (Kay Reference Kay2000: 61), allowing it, only in retrospect, to “become the stuff of prophecy” (ibid.: 62)—a science classic that molecular biologists “mostly read for reassurance” (Rosen Reference Rosen, Buckley and Peat1996: 168).

We are told that autobiographical declarations of influence so long after the fact cannot be trusted, as “it is virtually impossible to believe most of what one is told about things that happened 30 to 40 years ago” (Symonds Reference Symonds1986: 224). And even if What Is Life? did inspire some to take up the molecular study of life, “the impact of Schrödinger’s views on biological matters has been, in the strictly scientific sense, negligible” (Dowdle Reference Dowdle1989: 104). His book cannot “be said to have provided any actual suggestions for further research that proved to be useful” (Keller Reference Keller1990: 403).

The negative tenor of these modern assessments has been fuelled to a considerable degree by the surprisingly disparaging remarks of the two Nobel laureates who were asked to reappraise What Is Life? on the occasion of Schrödinger’s centennial (Kilmister Reference Kilmister1987), namely Linus Pauling and Perutz. Pauling, who briefly worked under Schrödinger in 1927, first makes the point that given the extent to which molecular biology is indebted to principles of modern chemistry that are themselves founded on quantum mechanics (such as the Heitler–London theory of the covalent bond discussed in Section 2), we are justified in asserting that “Schrödinger, by formulating his wave equation, is basically responsible for modern biology” (Pauling Reference 90Pauling and Kilmister1987: 228). However, his evaluation is drastically different when it comes to What Is Life?. “When I first read this book, over 40 years ago”, he writes, “I was disappointed. It was, and still is, my opinion that Schrödinger made no contribution to our understanding of life” (ibid.: 229). Pauling’s criticisms focus on the book’s brief treatment of thermodynamics, which he considers “vague and superficial to an extent that should not be tolerated even in a popular lecture” (ibid.).Footnote ⁵⁰ He particularly chastises Schrödinger for speaking confusingly of ‘negative entropy’ instead of employing the more theoretically appropriate concept of free energy, and he contemptuously declares that by coining this nonsensical notion, What Is Life? “made a negative contribution” to biology (ibid.).Footnote ⁵¹

Perutz, for his part, admonishes Schrödinger for, among other things, his outdated presentation of genetics, his failure to recognize key theoretical insights known at the time (such as George Beadle and Edward Tatum’s ‘one gene–one enzyme hypothesis’, published in 1941), and his ignorance of chemistry, which helps resolve the apparent contradiction between the stability of the gene and the statistical laws of physics. Perutz’s (Reference Perutz and Kilmister1987b: 243) caustic conclusion is that “a close study of his book and of the related literature has shown me that what was true in his book was not original, and most of what was original was known not to be true even when the book was written”. Perutz also lambasted What Is Life? in a commentary in Nature (i.e., Perutz Reference Perutz1987a), as well as in a brief note in the April 5 1987 issue of The Scientist, which he scornfully titled ‘What Is Life? Fiction, Not Science’.Footnote ⁵²

I suspect that the virulence of these remarks, which contemporary historians have gleefully quoted when giving their damning verdicts on the book’s scientific value, goes some way towards explaining the puzzle with which I started this Element—the question of why, despite the book’s enduring fame and classic status, there has been so little appetite among recent scholars to engage seriously with Schrödinger’s argument, beyond paying lip service to the three soundbites listed in Section 1.

Now, the historians have been absolutely right to draw attention to the strategic ways in which prominent molecular biologists in the 1960s started deploying Schrödinger’s reputation to validate their own interests and endeavours. But it is important not to overstate the case to the point where the impression conveyed is that it took twenty years for molecular biologists to start paying attention to What Is Life?. There is actually plenty of evidence to the contrary, some of which was already adduced in Section 1, including the recently unearthed letter that Crick sent to Schrödinger just weeks after the publication of the double helix model of DNA in 1953 (recall Figure 3). Yet even the more moderately revisionist reappraisals that do acknowledge that Schrödinger’s book was widely read when it came out but maintain that its chief importance does not lie on what it said, or even on how it was said, but rather on who said it, are still problematic. In this section, I will argue—against the prevailing historiographical consensus—that the ideas Schrödinger advanced in What Is Life? (and to a lesser extent, also the way he chose to present them) did in fact exert a considerable influence in shaping and consolidating the research agenda of molecular biology during the second half of the twentieth century.

As we saw in Sections 2 and 3, Schrödinger’s central thesis is that the source of all cellular order (and, by implication, all organismic order) resides in the code-script embedded in the fixed, solid-state structure of the genetic material, which protects it from the randomizing effects of thermal agitation. Now, Schrödinger does not tell us how the order in the aperiodic crystal is used or relayed so that the cell can perform its operations in an orderly way. He anticipates that the “detailed information about the function of the genetical mechanism” will emerge not from physics but from further experimental studies in “biochemistry under the guidance of physiology and genetics” (Schrödinger Reference Schrödinger1944: 68). What he does tell us is that this future research will disclose, as Moss (Reference Moss2003: 60) carefully words it, “new higher-level laws or principles that explain the ability of living systems to parlay high levels of order between the chemically stable but metabolically inert aperiodic crystal and the growing and metabolizing, but entropically vulnerable, apparatus of the cell and organism”.

One of the principal claims I want to propose in what follows is that Schrodinger’s vision of the cell as a microscopic machine operating deterministically according to non-statistical, mechanical principles—and thereby impervious to the disruptive effects of stochasticity—provided a tacit conceptual framework for molecular biology within which empirical results could be interpreted, as well as a direction towards which further research could be oriented. It is in this context, I believe, that we should understand Schrödinger’s famous quip about “[n]ew laws to be expected in the organism” (Schrödinger Reference Schrödinger1944: 76). What he had in mind was laws (or principles, or mechanisms) that would account for the genocentric, order-from-order logic that he predicted would soon be discovered operating at the heart of the cell.

I begin by emphasizing what I take to be a crucial implication of Schrödinger’s argument, namely that in the absence of statistical regularities, the order encoded in the structure of the aperiodic crystal must somehow be reliably transmitted to other cellular components, especially to the proteins, so that these can individually express it through their respective functions in a way that similarly eludes or overcomes the raging Brownian storm of the molecular realm. This Schrödingerian understanding of cellular order as the summative result of the order of its separate macromolecules is, I would argue, part of what makes molecular biology so different from the older biochemical tradition that preceded it, which had assumed the cell to be a homogeneous solution or colloidal suspension—a sort of ‘bag of chemicals’—governed by the statistical, order-from-disorder laws of chemical kinetics. Richard Lewontin expounds this point in a remarkably crisp passage that deserves to be quoted in full:

This composite (as opposed to systemic) conception of cellular order, with its emphasis on the specific structure of individual molecules, which interact with one another mechanically and predictably, is exactly what one finds in the writings of the pioneers of molecular biology. Jacob and Monod express it clearly in their influential 1961 paper in which the genetic program concept is introduced when they describe the genome as “a mosaic of independent molecular blue-prints for the building of individual cellular constituents” (Jacob and Monod Reference Jacob and Monod1961: 354). It is also apparent in the way proteins are characterized as miniature machine tools. It is the reason why Perutz calls haemoglobin a “molecular lung” and “an organ on a molecular scale” (quoted in Judson Reference Judson1979: 213), and why Medawar asserts that “[t]here is no dividing line between structures in the molecular and in the anatomical sense” (ibid.). As Judson remarks, for molecular biologists, “genetical events, like biochemical ones, now really felt as though they were rightly explained only when they could be conceived mechanically, in terms of the pieces and links and angles and local electrical charges that molecules are made of ” (ibid.: 215–216).Footnote ⁵³

In noting molecular biology’s exaltation of molecular structure, it is worth saying something about one of the key ideas in What Is Life?, namely the concept of an aperiodic crystal. A number of commentators have been baffled by Schrödinger’s unorthodox use of the word ‘crystal’ to describe the genetic material. Waddington (Reference Waddington1969: 321) described it as “an exceedingly paradoxical phrase”. Some have wondered why he did not resort to a more suitable chemical term, such as ‘macromolecule’ or ‘polymer’. Both Crick (in Judson Reference Judson1979: 245) and Perutz (Reference Perutz and Kilmister1987b: 241) take this odd terminological choice as evidence that Schrödinger had little understanding of chemistry. However, a very different interpretation suggests itself when we recall why Schrödinger invokes the idea of a crystal in the first place, which is that he is trying to solve the paradox of the permanence of the gene by appealing only to physical principles, especially the Heitler–London theory derived from quantum mechanics, which accounts for the structural stability of molecules and of solid matter.

As I indicated in Section 2, it is by virtue of the ‘solidifying’ interatomic forces described by this theory that Schrödinger concludes that the hereditary substance must be a large molecule exhibiting the solid-state rigidity of a crystal. Notice that Schrödinger arrives at this conclusion completely independently of any chemical considerations about genes. For this reason, Robert Olby (Reference Olby1974: 242) has suggested—plausibly, in my view—that “Schrödinger deliberately avoided using chemical evidence”. By accounting for the stability and mutability of genes in quantum-mechanical terms, “Schrödinger made the facts of genetics meaningful to the physicist”, so that “a physicist reading this book could get excited about genetics” (ibid.: 245). If Olby is right, then the way Schrödinger chose to make his case really did make a difference to the physicists who read What Is Life? upon its publication.Footnote ⁵⁴

We come now to what is widely regarded as the most important innovation of the molecular revolution in biology: the enthusiastic, wholesale adoption of the notion of information. Information language entered the biological discourse in the post-war period not once but twice (cf. Keller Reference Keller1995; Sarkar Reference Sarkar and Sarkar1996; Kay Reference Kay2000), and Schrödinger’s influence is evident in both importations, as I will show in a moment. My main contention here is that Schrödinger’s explicit principle of order-from-order evolved into a tacit principle of organization-from-information that decidedly shaped the agenda of molecular biology. To see how this happened, let us consider the two introductions of information-talk into biology in turn.

The origin of information theory is usually traced to the mathematical theory of communication developed by Claude Shannon in a 1948 paper, which was republished in book form the following year with an introductory essay by Warren Weaver (Shannon and Weaver Reference Shannon and Weaver1949).Footnote ⁵⁵ Shannon was concerned with the engineering problem of increasing the accuracy of the transmission of a message between sender and receiver. His concept of information is a measure of the uncertainty involved in this communication process; its numerical value is determined by a probabilistic function that formally resembles Boltzmann’s entropy formula in thermodynamics. This prompted Shannon (on the advice of von Neumann) to repurpose the term ‘entropy’ for use in this new context.Footnote ⁵⁶

The year 1948 also saw the publication of Norbert Wiener’s Cybernetics: Control and Communication in the Animal and the Machine, which despite its mathematical content became an international bestseller and helped make the new transdisciplinary field of cybernetics a worldwide cultural phenomenon in the 1950s. Wiener’s formal treatment of information, like Shannon’s, also made use of the concept of entropy, but he defined the former as the negative of the latter, citing Schrödinger’s proposal of ‘negative entropy’ in What Is Life? when making this identification (Wiener Reference Wiener1948: 18–19). Shortly thereafter, Léon Brioullin (Reference Brioullin1949, Reference Brioullin1956) developed Wiener’s notion and rebranded it as ‘negentropy’—also crediting Schrödinger with the original insight.

The attempt to apply information theory to biology was an unmitigated failure. It is probably best exemplified by the Quixotic efforts of Henry Quastler, who employed Shannon’s formalisms in the hope of transforming biology into an information science (see Kay Reference Kay2000: 115–127). The proceedings of a symposium he organized in 1952, titled Essays on the Use of Information Theory in Biology (Quastler Reference Quastler1953), are replete with astonishing calculations of the ‘information content’ of various biological entities, from genes and proteins to cells and organisms. A human being, according to one estimate, was deemed to contain 5 × 10²⁵ bits of information!

Wiener, for his part, corresponded with Haldane, who conveyed his enthusiasm about the mathematical arguments in Cybernetics—though that was probably because he was one of the very few biologists who could actually understand them. One of Haldane’s collaborators at University College London, Hans Kalmus, did publish a paper on the potential of Wiener’s framework for genetics (i.e., Kalmus Reference Kalmus1950), but it was completely ignored and promptly forgotten, being cited only once by another author until historians rediscovered it in the 1990s. Finally, Brillouin’s theoretical notion of negentropy was only seriously taken up by a few biologists, most notably by his compatriot Lwoff, who devoted considerable attention to it—as well as to Schrödinger’s idea of negative entropy—in his 1960 Compton Lectures at MIT, later published as Biological Order (Lwoff Reference 88Lwoff1962, see especially chapter 6).Footnote ⁵⁷

On the whole, these technical accounts of information proved too abstruse and too unwieldly to be of any real use in biological research. An important limitation is their exclusive focus on the syntactic aspects of a message; they say nothing about its semantic content (i.e., its meaning), which is what biologists are most likely to care about. Shannon himself was well aware of this, and he often cautioned others against the improper application of information theory outside the realms of communication and engineering (see, e.g., Shannon Reference Shannon1956). Wiener, by contrast, promoted the idea of information as part of his cybernetic worldview with almost messianic zeal. Information, for Wiener, was not just a new measure, but a new kind of basic ingredient of the universe, existing in its own distinct domain. “Information is information,” he proclaimed, “not matter or energy. No materialism which does not admit this can survive at the present day” (Wiener Reference Wiener1948: 155). His collaborations with physiologists such as Arturo Rosenblueth and Walter Cannon also meant that Wiener, unlike Shannon, not only recognized, but went out of his way to stress the underlying similarities between organisms and servomechanisms, effectively eliminating any ontological barriers still separating the biological from the mechanical (recall that the subtitle of Cybernetics is Control and Communication in the Animal and the Machine). In tirelessly promoting his cybernetic cause, Wiener targeted the biologists even more so than the physicists—he spoke not of order (as Schrödinger had done) but of organization, which is the all-important biological keyword that he strategically chose to identify with information.

When the concept of information finally infiltrated the biological discourse, it did so alongside a whole battery of related (and, at the time, fashionable) cybernetic terms—for example, ‘code’, ‘message’, ‘feedback’, ‘control’, ‘program’—that were applied loosely and metaphorically. It is in this non-technical, non-mathematical sense that ‘information’ re-entered biology in 1953 when Watson and Crick famously suggested in their second Nature paper on the double helix model of DNA that “the precise sequence of the bases is the code which carries the genetical information” (Watson and Crick Reference Watson and Crick1953b: 965). Information theory might have been a biological dead end, but information language quickly proved to be tremendously productive in guiding experimental research into the structure and function of cellular macromolecules. Perhaps most significantly, information-talk helped make intelligible the vital connection between nucleic acids and proteins (see, e.g., Gamow Reference Gamow1955). As Crick memorably argued—echoing Wiener—in his seminal 1958 paper ‘On Protein Synthesis’, the crucial relation between DNA, RNA, and protein is not the flow of matter, or the flow energy, but the flow of information, clarifying that “[i]nformation means here the precise determination of sequence, either of bases in the nucleic acid or of amino acid residues in the protein” (Crick Reference Crick1958: 153).

Moreover, in adding the proviso that information always flows from nucleic acid to protein but never the reverse—a contention he lightheartedly dubbed the ‘Central Dogma’ due to the absence of the required evidence to confirm it—Crick established the genocentric, order-from-order logic of molecular biology that Schrödinger had envisaged. The genetic material has a privileged causal (and therefore explanatory) role because it stores, replicates, and transmits the hereditary information. It also directs and controls the synthesis of proteins and is thus indirectly responsible for all cellular operations. The Central Dogma came to be described in textbooks with the catchy slogan that every student of molecular biology learns: ‘DNA makes RNA makes protein’.

There is something unmistakably Schrödingerian about the conflation in this slogan of the specification of a message with the command to carry it out; it is hard not to be reminded of Schrödinger’s description of the hereditary code-script. Keller has also noticed this: “If the genetic code is a message”, she writes, “it is a very particular kind of message: it is an order. Cast in the imperative, it says, ‘Make an enzyme!’ As Schrödinger so aptly observed, this is no ordinary code: it is ‘law-code and executive power in one’” (Keller Reference Keller1995: 95). The conflation nevertheless proved to be extremely fertile, leading eventually, as I argued in Section 3, to Jacob and Monod’s formulation of the genetic program in 1961, with its exaltation of DNA as the ‘master molecule’ in charge of the cell and of development—just as Schrödinger’s code-script is.

I now wish to argue further that molecular biology’s understanding of cellular order assumes two distinct kinds of specificity—informational (or genetic) and structural (or spatial)—that represent elaborations (in the case of the first) and vindications (in the case of the second) of ideas we already find in What Is Life?.

Informational specificity can be inferred directly from Crick’s (Reference Crick1958: 153) definition of information quoted earlier. It is the postulate that the semantic content of a genetic message is specified by the exact arrangement (“precise determination”) of the linear order (“sequence”) in which its elements (“bases” or “amino acid residues”) follow each other in aperiodic succession. As we saw in Section 2, Schrödinger tacitly relied on this very same postulate to justify the existence of his code-script. A non-repetitive yet “well-ordered association of atoms”, he reasoned, “appears to be the only conceivable material structure that offers a variety of possible (‘isomeric’) arrangements, sufficiently large to embody a complicated system of ‘determinations’” that could “precisely correspond with a highly […] specified plan of development” (Schrödinger Reference Schrödinger1944: 61–62). Although Schrödinger did not suggest that the code-script needs to be deciphered (in the way that it was later found that the information encoded in the DNA must be ‘transcribed’ and ‘translated’), he did clearly indicate that its semantic content (the “plan of development”) is determined by what molecular biologists after Crick Reference Crick1958 came to understand as the postulate of informational specificity.

Structural specificity is an equally important postulate in molecular biology. The flow of information from nucleic acid to protein described by the Central Dogma only takes us from nucleotide sequence to the corresponding amino acid sequence—this is what Crick referred to as the ‘Sequence Hypothesis’ in his influential Reference Crick1958 paper. However, most cellular functions are performed by proteins folded in space, which often interact and associate with other proteins to form complexes and networks. This is where structural specificity comes in. It is based on the old chemical principle of ‘stereospecificity’, or ‘stereocomplementarity’, which dates all the way back to Emil Fischer’s lock-and-key model of the enzyme–substrate relation of 1894. It was later extended to immunology by Paul Ehrlich to understand the antibody–antigen relation. But it was Pauling who developed a general theory of stereocomplementarity in the 1940s that he argued applied to all biological processes at the molecular level, and which became foundational for molecular biology (Mertens Reference Mertens2019). It was used by Watson and Crick to understand the complementary relation between nucleotide bases in the DNA double helix, and more broadly to explain the process of molecular recognition that underlies the functions of most proteins, as well as the way proteins come together to form larger assemblies in the cell. What proteins do when they interact with other molecules (including other proteins) is determined by the uniqueness of their shape, or ‘conformation’, which is itself determined by the precise sequence of their amino acid ‘building blocks’. This led to a general theoretical picture of how the virtual, one-dimensional order encoded in the genes (information) results in the material, four-dimensional order manifested by the cell (organization). Schrödinger’s order-from-order thus became molecular biology’s organization-from-information.

Although the postulate of structural specificity developed completely independently of What Is Life?, there is an important sense in which Schrödinger’s argument helped consolidate it as a fundamental pillar of molecular biology (cf. Kupiec Reference Kupiec2009, Reference Kupiec2010). It relates to the emphasis that stereocomplementarity places on stability and rigidity as necessary attributes of macromolecules. The timely execution of cellular operations relies on the ability of proteins to discriminate each other’s conformations and physically interlock by means of non-covalent linkage sites to form transient stereospecific complexes. In turn, this requires them to adopt perfectly complementary geometric shapes that fit exactly into one another, so as to exclude all other potential binding partners. It is therefore necessary for their structure to be rigid and stable—like a solid—as only then can they achieve the extreme specificity that is required for their function.Footnote ⁵⁸

This is precisely the view of macromolecular interactions that, as I indicated earlier, follows directly from the argument in What Is Life?. Although Schrödinger appeals to physical rather than to chemical considerations, the conclusion he arrives at is essentially the same: the functionally important components of the cell (i.e., the nucleic acids and proteins) must be stable, rigid structures that behave like solids, or crystals, which interact with one another specifically and deterministically (rather than generically and statistically) like cogs in a microscopic machine, thereby escaping the stochastic perturbations that are typical of the molecular realm.

This mechanical, deterministic, and genocentric conception of life that became dominant with the rise of molecular biology—and which was ardently promoted in popular books such as Biological Order (Lwoff Reference 88Lwoff1962), Of Molecules and Men (Crick Reference Crick1966), Chance and Necessity (Monod Reference Monod1972), and The Logic of Life (Jacob Reference Jacob1973)—was not strictly speaking new. It has a distinctively Cartesian flavour, as many commentators have observed.Footnote ⁵⁹ Nevertheless, it constituted a radical departure from the biological views that had prevailed in the immediately preceding decades. Earlier biochemical and physiological research (e.g., Needham Reference Needham1936; Schoenheimer Reference Schoenheimer1942; see also Nicholson Reference Nicholson, Nicholson and Dupré2018) had tended to emphasize the plastic and dynamic character of biological systems, reflected in the continuous metabolic turnover of their material constitution, as well as in the highly heterogeneous nature of their underlying processes, assumed to be subject to stochasticity and only statistically predictable.

Although I certainly would not want to claim that Schrödinger is fully, or even primarily, responsible for prompting this ontological shift in mid-twentieth-century biology—the story, as with everything in history, is obviously more messy and complicated—I do want to suggest that What Is Life? played a hitherto unappreciated role in shaping and consolidating molecular biology’s theoretical understanding of cellular order. As this section has shown, there was no twenty-year lag between the book’s publication and its reception (as some historians would have us believe), even if the book did acquire a new political significance in the mid 1960s when it began to be used as a propagandistic tool to boost the epistemic prestige of molecular biology. I actually agree with Kay when she writes that “Schrödinger’s four-dimensional Raphaelesque pattern [contained in the code-script] was digitalized and collapsed into a one-dimensional Boolean message inscribed on a magnetic tape” (Kay Reference Kay2000: 66). Where I disagree with her is that I take this reformulation of Schrödinger’s ideas in the language of information and cybernetics to be evidence of their continued inf luence, not of their initial neglect and subsequent ‘reinvention’.

While I have made the case for Schrödinger’s impact on molecular biology in general terms, one could also make it by attending to the intellectual trajectories of individual scientists (so that we do not have to rely only on their retrospective autobiographical declarations). In a recent study, Laurent Loison has done just that in relation to Monod, showing that his biological thinking was profoundly affected by his reading of What Is Life? (Loison Reference Loison2015). Monod’s case is particularly interesting because, unlike other pioneers of molecular biology who originally came from physics, he was trained as a biologist from the outset. This means that he was taught to think about life in terms of the very principles he would later repudiate as a champion of molecular biology. Though initially concerned with providing statistical, non-deterministic explanations of biological regularities (such as the exponential growth of bacterial cultures), Monod later came to adopt the exact opposite conception, interpreting cellular order as a product of the rigid, clockwork-like precision of its macromolecular components. And one of the key factors that led him to change his mind was Schrödinger’s argument in What Is Life?.

Evidence for this can be found in Monod’s unpublished manuscripts and lecture notes. For example, in the notes for his 1958 Dunham Lectures at Harvard, Monod directly credits Schrödinger for having emphasized the very features—stability and rigidity—that came to characterize molecular biology’s understanding of macromolecules (Figure 8). Elsewhere in his notes, Monod comments on the “very high precision” of “the protein-synthesizing process”, and again mentions Schrödinger when observing that the ribosome “appears to work mechanically, like a clock or a precision machine tool, rather than statistically (like what?) (Schrödinger)” (quoted in Loison Reference Loison2015: 396).

Figure 8 Excerpt from Monod’s unpublished lecture notes for his 1958 Dunham Lectures at Harvard in which Schrödinger is mentioned

(reproduced with permission of the Pasteur Institute Archives)

This crucial distinction between mechanical and statistical forms of order, which lies at the core of Schrödinger’s argument in What Is Life?, is one that Monod repeatedly invokes to illustrate the difference between the ‘old’ and the ‘new’ biology. Consider the following remark from Enzymatic Cybernetics, a book-length French-language manuscript completed in 1960 that Monod never got round to publishing in his lifetime (it was finally published in 2021):

Furthermore, Monod makes it clear that Schrödinger’s ideas were absolutely critical to this transition. This is how he dramatically expresses the point in a public lecture delivered in Geneva in 1965:

Monod’s philosophy of molecular biology was presented in mature form in his widely read Chance and Necessity (Reference Monod1972), whose revealing subtitle is An Essay on the Natural Philosophy of Modern Biology. In many ways, this book represents the fully fledged realization of Schrödinger’s biological vision, even though What Is Life? is not actually cited, presumably because by then its message had been fully assimilated and had become too familiar to require mentioning.

Loison (Reference Loison2015) has interestingly suggested that we can think of Monod’s theoretical agenda in molecular biology between 1950 and 1970 as an attempt to identify the new order-from-order laws or principles that Schrödinger expected would be eventually discovered. These would ostensibly include many of the tenets we have discussed in this section, such as informational and structural specificity, Crick’s Central Dogma, and his and Jacob’s concept of the genetic program. If this is the right way to think about this history, then Schrödinger’s order-from-order research program can only be regarded a resounding success.

Be that as it may, the subsequent development of molecular biology in the late twentieth century only served to further consolidate Schrödinger’s mechanical, deterministic, and genocentric view of the cell. Monod (Reference Monod1972: 98) had already claimed that “[w]ith the globular protein we already have, at the molecular level, a veritable machine”. Thereafter, the ‘molecular machine’ concept came to be used to characterize most functionally specialized macromolecular complexes in the cell (Alberts Reference Alberts1998; Frank Reference Frank2011; see also Nicholson Reference Nicholson2019). Similarly, the postulate of structural specificity was invoked to justify the appeal to wiring diagrams and design charts in schematic representations of metabolic, regulatory, and signalling pathways, which are still today frequently depicted as fixed, solid-state circuits deliberately made to resemble the circuit boards of electronic engineering. A typical example is reproduced in Figure 9.

Figure 9 Wiring diagram of metabolic pathways in the bacterium Baumannia cicadellinicola depicted as a solid-state circuit board

(adapted from Wu et al. Reference 94Wu, Daugherty, Van Aken, Pai, Watkins, Khouri, Tallon, Zaborsky, Dunbar, Tran, Moran and Eisen2006)

Now, as I anticipated at the end of Section 1, one of the reasons why it really does matter that Schrödinger is (at least partly) responsible for molecular biology’s view of the cell is that much of this view is increasingly at odds with experimental findings. Since the turn of the twenty-first century, a growing number of biologists have been systematically calling into question many of its ontological assumptions (see, e.g., Rose Reference Rose1998; Kirschner et al. Reference Kirschner, Gerhart and Mitchison2000; Lewontin Reference Lewontin2000; Oyama et al. Reference Oyama, Griffiths and Gray2001; Moss Reference Moss2003; Woese Reference Woese2004; Karsenti Reference Karsenti2008; Kupiec Reference Kupiec2010; Heams Reference Heams2014; Noble Reference Noble2016; Ball Reference Ball2023).

We already discussed in Section 3 some of the major problems with the genetic program, and with genocentric, order-from-order explanations of cellular activities more generally. And in Section 4 we saw how statistical (as opposed to mechanical) models are increasingly being used to understand the molecular basis of development, as well as how non-classical instances of order-from-disorder not foreseen by Schrödinger (such as reaction–diffusion systems and other self-organizing processes) are now acknowledged to play an indispensable role in the generation and maintenance of order at the cellular and multicellular levels. In much the same way, recent research has shown many of the ideas we have discussed in this section to be seriously problematic. However, as I have already examined these problems at length in previous work (i.e., Nicholson Reference Nicholson2019, Reference Nicholson, Holm and Serban2020), I shall resist the temptation of doing so again here. Instead, let me just mention some of the most pertinent and surprising findings, and the interested reader can consult the aforementioned publications for the full analysis.

Stability and rigidity, it turns out, are not necessary features of biological macromolecules. The structure of proteins in their native state is soft and fluid, not hard and rigid. Far from displaying a single conformation, most proteins stochastically alternate between different ordered states. Many proteins (known as ‘intrinsically disordered proteins’) do not have an ordered conformation at all but instead roam the cell as unfolded polypeptide chains capable of binding to a wide range of substrates. Promiscuity, not specificity, is the rule for most proteins (including enzymes): what a protein does in the cell is as much a product of its cellular context as its amino acid sequence. The same polypeptide chain can partake in a variety of functions depending on when and where it is expressed, and on what partners it associates with (a phenomenon sometimes referred to as ‘moonlighting’).

Protein-protein interactions are not stereospecific interlockings but probabilistic collision events—they are contingent and opportunistic, not the solid-state manifestation of a pre-existing genetic blueprint. This is why circuit-like depictions like Figure 9 can be so misleading. There is no wiring physically connecting the proteins as there is in a real electronic circuit; the ‘wiring’ in such diagrams rather reflects the questionable reification of ethereal ‘information flows’. Moreover, the depicted network of proteins represents only one of the potentially innumerable ways in which those same proteins might interact at different times and under different conditions. Larger macromolecular assemblies (e.g., ribosomes) also lack the structural rigidity and operational precision of what might be expected of true ‘molecular machines’, and they move stochastically rather than mechanically when performing their functions.Footnote ⁶¹

Overall, there is something decidedly odd about the habit of molecular biologists to downplay stochasticity, if not to disregard it altogether, in their representations and explanations. After all, stochasticity is a well-understood physical phenomenon known as far back as the nineteenth century to be inescapable at the molecular scale at any temperature above absolute zero. If I may be allowed to end this section on a more speculative note, perhaps this is, in the final analysis, the real legacy of What Is Life?. It is not inconceivable that Schrödinger’s theoretical view of the cell as a microscopic clockwork operating deterministically according to non-statistical, mechanical principles granted molecular biologists the license to dismiss the effects of stochasticity (despite being inevitable from a strictly physical point of view), freeing them up to focus on meticulously characterizing the structure of individual macromolecules—drawing attention to their crystal-like rigidity (ostensibly confirmed by the use of available methods like X-ray crystallography) and emphasizing their functional specificity (assumed to be genetically determined)—all while largely ignoring the turbulent, destabilizing influences of their physical milieu. One can easily imagine how this attitude might have encouraged the appeal to conceptual models borrowed from the macroscopic domain—where stochastic fluctuations are negligible and can be safely ignored—such as the highly suggestive engineering metaphors spawned by cybernetics and, later, computer science that, as we have seen, are so distinctive of molecular biology.

6 Understanding Schrödinger’s Motives in What Is Life?

In the preceding sections, we have taken Schrödinger to task for having promoted an inappropriately genocentric view of living systems, which he assumed to operate deterministically in a way that wrongly ignored the effects of stochasticity on their underlying molecular processes. Now, the reader might feel a bit uneasy about indictments issued with the benefit of hindsight, as it would be unfair to chastise Schrödinger for making claims whose problems only became apparent many years (or even decades) later. This would be as absurd as blaming him for failing to predict the future course of science. One can certainly think of criticisms of What Is Life? that would be unjust for this reason—for instance, if we scolded Schrödinger for believing that the hereditary substance is protein rather than DNA, or for not having recognized that self-organizing processes are a crucial form of order-from-disorder in living systems.

One of my aims in this section is to demonstrate that the criticisms I have presented of Schrödinger’s conception of cellular order do not fall under this category. The reason is that there is nothing necessary or inevitable about this conception—either today or in 1944. Schrödinger’s biological views, far from mirroring the conventional wisdom of the historical period when they were formulated, were not even orthodox in their own time. Instead, they reflect deliberate choices that Schrödinger made when he was working out the argument that he would later present in What Is Life?. To see why Schrödinger chose to argue the way he did in his book, we first need to identify his motives for turning to biological issues in the first place.

This is not a trivial question. Why would a 56-year-old theoretical physicist with no formal training in biology decide to take on the thorny subject of ‘the physical aspect of the living cell’ and go on to publish an entire book about it? A number of hypotheses have been proposed to explain this. Some have suggested that Schrödinger’s interest in biology can be traced back to his childhood. His father, Rudolf Schrödinger, had a lifelong amateur interest in botany—he wrote several papers on the phylogeny of plants, and at the time of his death he was vice-president of the Zoological–Botanical Society of Vienna (Moore Reference Moore1989: 116)—so it is possible that conversations at home instilled in him a love for biology at an early age.Footnote ⁶² In his Autobiographical Sketches, written at the end of his life, Schrödinger recalled how as a schoolboy he “virtually devoured The Origin of Species” and “soon became an ardent follower of Darwinism” (Schrödinger Reference Schrödinger1992: 174–175).

Later, as a student at the University of Vienna, Schrödinger became fascinated by Richard Semon’s book The Mneme as a Conservative Principle (Reference Semon1904)—a highly speculative Lamarckian treatise that sought to explain the phenomena of heredity, development, instinct, and consciousness using the concept of a ‘mneme’, a sort of cellular repository of organismic memory that Semon claimed governed all biological processes. According to Schrödinger’s biographer Walter Moore, Semon’s book “had a major influence on the development of his philosophical ideas” (Moore Reference Moore1989: 46). In 1925, just before his monumental discovery of wave mechanics, he wrote a long essay expounding his philosophical views—published decades later as My View of the World (Schrödinger Reference Schrödinger1964)—where he revisited the themes of Semon’s book, and which reveals his continued interest in the intriguing idea of a genetic ‘memory’ preserved through the generations and immune to the ravages of time. Keller (Reference Keller1995), Kay (Reference Kay2000), and Morange (Reference Morange2020) have all argued that this longstanding preoccupation is what drove Schrödinger to write What Is Life?.

Others have sought the book’s origins in Schrödinger’s general intellectual disposition. Neville Symonds (Reference Symonds1986), who studied with him in Dublin in the late 1940s, points out that Schrödinger had exceptionally wide interests. He was a modern-day ‘renaissance man’ with a desire for unified knowledge and who enjoyed making brief intellectual excursions into areas other than his own. Indeed, we should not forget that, besides biology, Schrödinger published booklets (also based on public lectures) on philosophy of science (Schrödinger Reference Schrödinger1951), the Ancient Greeks (Schrödinger Reference Schrödinger1954), and the mind-body problem (Schrödinger Reference Schrödinger1958). A version of this hypothesis was previously advanced by Olby (Reference Olby1971) and it has been put forward again more recently by Gould (Reference Gould, Murphy and O’Neill1995).

While acknowledging that all of the above factors might well have played a role, I suspect that the primary cause for the genesis of the book lies elsewhere. Specifically, what I want to suggest—and here I follow Edward Yoxen (Reference Yoxen1977, Reference Yoxen1979) and Phillip Sloan (Reference Sloan2012)—is that it is impossible to make sense of Schrödinger’s engagement with biology without carefully considering his position in the heated debates that took place during the interwar period regarding the interpretation and extension of quantum mechanics. As I will show in what follows, only by attending to this context can we understand what Schrödinger is really up to in What Is Life?.

The key point to bear in mind when placing Schrödinger in relation to the physics community of his time is that he was fundamentally at odds with most of his contemporaries (with the notable exception of Einstein) regarding the way to interpret the formalizations of quantum mechanics—a science that Schrödinger himself had helped create with his wave mechanics in 1926. While his own wave equation is deterministic, the orthodox interpretation of quantum mechanics articulated by Bohr and Werner Heisenberg, known as the Copenhagen interpretation, is not. Schrödinger (like Einstein) could never bring himself to accept this interpretation, and he grew ever more frustrated with its proponents, who invariably reinterpreted his wave mechanics in probabilistic and instrumentalist (i.e., antirealist) terms, much to his dismay.Footnote ⁶³

Feeling increasingly out of step with the dominant paradigm, Schrödinger published a three-part paper in the German-language journal Die Naturwissenschaften sardonically titled ‘The Present Situation in Quantum Mechanics’ (Schrödinger Reference Schrödinger1935a) where he devised his famous cat thought experiment, which he intended as a reductio ad absurdum of the Copenhagen interpretation. Schrödinger imagined a cat inside a sealed box along with a minuscule amount of radioactive material. If this material happens to decay, then a device releases a hammer, which smashes a vial of poison, which kills the cat. If the material does not decay, the cat lives. As radioactive decay is a quantum-mechanical phenomenon, the entire box must be regarded as a giant quantum system. What this means according to the Copenhagen interpretation is that until a measurement is made (i.e., the box is opened and the cat is observed), the cat remains in a blurry indeterminate state of superposition—both dead and alive—which is, of course, absurd. Ironically, though, instead of undermining physicists’ confidence in the validity of the Copenhagen interpretation (as Schrödinger had hoped), his cat only served to strengthen it, eventually becoming the definitive symbol of the transcendent weirdness that has come to be generally associated with quantum mechanics.Footnote ⁶⁴

Schrödinger’s distaste for the Copenhagen interpretation has a lot to do with his scientific background, particularly with the enormous influence that Boltzmann exerted on his intellectual development. In his inaugural address to the Prussian Academy of Sciences in 1929, Schrödinger declared that Boltzmann’s “line of thought may be called my first love in science. No other has ever thus enraptured me or will ever do so again” (Schrödinger Reference Schrödinger1935b: 13). Much later, in his Autobiographical Sketches, he asserted that “no perception in physics has ever seemed more important to me than that of Boltzmann—despite Planck and Einstein” (Schrödinger Reference Schrödinger1992: 168).

Besides a lifelong fascination with statistical mechanics (see, e.g., Schrödinger Reference Schrödinger1946), what Schrödinger inherited from Boltzmann is the philosophical conviction (which Boltzmann propounded in explicit opposition to the phenomenalist attitude urged by Ernst Mach, the other major influence in Vienna at the time) that physical theory should provide realistic, internally consistent, and causally determinate descriptions (or visualizable ‘pictures’) of the microscopic entities responsible for macroscopic phenomena. This familiar expectation—generally associated with the notion of ‘classical physics’ (a term that Boltzmann himself coined in 1899)—became progressively eroded in the twentieth century, especially with the rise to prominence of Bohr’s views, which encouraged younger physicists to develop mathematical models without worrying about their physical (i.e., spatiotemporal) visualizability. As it became apparent that it would not be possible to provide observer-independent, non-contradictory descriptions of quantum phenomena, Schrödinger became increasingly distraught and alienated from most of his physicist colleagues, who simply accepted the profoundly counterintuitive, observer-dependent, and intrinsically indeterministic character of the Copenhagen interpretation.

For Schrödinger, this attitude towards the atomic world—defined by Bohr’s ‘principle of complementarity’ (which holds that entities can exhibit mutually exclusive properties that cannot be measured simultaneously), as well as ideas of ‘acausality’ and spatiotemporally discontinuous ‘jumps’—was bad enough as a theoretical perspective in physics.Footnote ⁶⁵ But what incensed him infinitely more was the philosophical use of the Copenhagen interpretation in tackling fundamental problems in domains beyond physics; something that Bohr and Jordan both did repeatedly, in different ways, in the contexts of biology and psychology.

In a number of public lectures and papers (collected in Bohr Reference Bohr1999), Bohr suggested that in these two sciences one encounters an epistemological situation that is analogous to the one found in atomic physics. Just as light, under different experimental setups, can be described either in terms of waves or of particles but not both simultaneously, life can be described either mechanistically (by means of physicochemical analysis) or teleologically (by means of direct observation) but not both simultaneously—as the former involves killing the organism, thereby destroying its purposive character. The two descriptions stand in a complementary relation, according to Bohr, as they are mutually exclusive yet jointly necessary. Complementarity also obtains in the psychological consideration of mental states, as it is not possible to observe our own thoughts without affecting them in the process. We can either analyze our emotions or experience them, but not both simultaneously. Bohr argued that the age-old dispute between determinism and free will could be solved by viewing it from the perspective afforded by his principle of complementarity.

In a way, Jordan went even further than Bohr. Not content with merely drawing analogies, Jordan argued that quantum mechanics (in its orthodox Copenhagen interpretation) was itself responsible for biological and psychological phenomena. In a speculative paper published in Die Naturwissenschaften titled ‘Quantum Mechanics and the Foundational Problems of Biology and Psychology’ (Jordan Reference Jordan1932), and in many other publications thereafter, Jordan articulated his ‘amplifier theory’ of the organism, which postulates that the acausal behaviour of individual atoms can be amplified through the structure of organic molecules and larger subsystems so as to ‘direct’ or ‘steer’ their macroscopic behaviour, thereby transferring acausality to whole organisms. According to Jordan, this amplified acausality of organisms—not explainable in classical or mechanical terms—is ultimately responsible for our inner experience of free will, which is a manifestation of quantum acausality magnified from the atomic level to neurological processes in the brain.Footnote ⁶⁶

It is these ideas that I believe provide the right context to understand Schrödinger’s incursion into biology, and where we should locate the inception of the argument that would eventually appear in What Is Life?. My contention, which I anticipated at the end of Section 1, is that Schrödinger turned to biology because he hoped that he would find in the molecular structure of living matter the means to salvage the mechanical and deterministic worldview of classical physics that he felt had become undermined by the Copenhagen interpretation of quantum mechanics. His chief motivation was to prevent proponents of this interpretation from using biological phenomena as a stepping stone to ground the claim that quantum indeterminacy provides a physical foundation for free will (which Schrödinger thought could not be justified scientifically). He did this by defending the idea that life is at its core a strictly deterministic phenomenon, and that it is deterministic owing to the prodigious stability of its genetic material, which is safeguarded by the Heitler–London theory derived from quantum mechanics. So, quantum mechanics comes to the rescue for Schrödinger, just as it does for Jordan, but what it offers him is the exact opposite: not indeterminacy and freedom, but determinacy and clockwork precision.Footnote ⁶⁷ By strategically severing the biological link connecting the physical to the neurological (or psychological), free will would need to be accounted for by other, preferably non-scientific, means.

When we look more closely at the circumstances in which Schrödinger developed the argument for What Is Life?, we find in them compelling evidence in support of this hypothesis. Let me mention three specific examples (each, incidentally, obtained from a different archival source).

In February 1933, exactly ten years before his ‘What Is Life?’ lectures in Dublin, Schrödinger gave a talk at the Prussian Academy of Sciences in Berlin titled ‘Why Are Atoms So Small?’ where he outlined what he later called in What Is Life? “the naïve physicist’s approach to the subject” (Schrödinger Reference Schrödinger1944: 4). Indeed, the book includes in its opening chapter a subsection with the same title as his Berlin talk from a decade earlier (i.e., ibid.: 4–6). Interestingly, this talk appears to have been partly conceived in response to the Reference Jordan1932 paper by Jordan that we discussed above. In September 1932, Schrödinger had written to his former colleague Karl Przibram at the Vienna Institute for Radium Research (and brother of Hans Przibram, director of the Vienna Institute for Experimental Biology) about the effects of Brownian motion on microorganisms. When Jordan’s paper came out in November, he wrote a second letter in which he referred to it critically. Although Schrödinger’s letters to Przibram appear to have been lost, Przibram’s second reply—writing also on behalf of his brother Hans—is very revealing:

Schrödinger never published a reply to Jordan in Die Naturwissenschaften; all that appeared in print was a brief summary of his Berlin talk (i.e., Schrödinger Reference Schrödinger1933). Nevertheless, what survives of his exchange with Przibram indicates a clear connection between Jordan’s extravagant attempt to explain free will in terms of amplified quantum indeterminacy in organisms and Schrödinger’s desire to tackle questions at the interface of physics and biology—more than a decade before he wrote What Is Life?.

A second line of evidence can be found in Schrödinger’s correspondence with Donnan in the late 1930s and early 1940s, which provides an insightful, behind-the-scenes look into how the ideas in What Is Life? gradually took form. In 1935, Donnan sent Schrödinger an old paper of his (i.e., Donnan Reference Donnan1918) that prefigures one of the central theses of What Is Life?, namely that “[t]he essentially ‘biological’ aspect of the science of living things is that it is fundamentally concerned, not with statistical averages [like physics and chemistry], but with sequences of events pertaining to particular individual units” (ibid.: 285; my translation). In 1943, Donnan sent him another of his papers (i.e., Donnan Reference Donnan1928) that, as I already indicated in Section 4, anticipated Schrödinger’s famous discussion in What Is Life? of how organisms comply with the second law of thermodynamics. Donnan’s Reference Donnan1928 paper also included the remark that “[i]t is difficult to resist the comparison of the developing embryo with the building of a house” and “to the plans of an invisible architect” (ibid.: 1561), which is almost identical to Schrödinger’s memorable description of the code-script as “architect’s plan and builder’s craft—in one” (Schrödinger Reference Schrödinger1944: 21).Footnote ⁶⁹

Schrödinger’s exchanges with Donnan were partly concerned with the problem of how to account for the autonomy [Eigengesetzlichkeit] of living matter in a non-vitalistic way that would not violate the laws of physics and chemistry. The somewhat paradoxical solution that Schrödinger eventually arrived at was to tie the distinctive orderliness of life to the failure of statistical physics to adequately explain it. Although statistical physics fails, Schrödinger explains to Donnan that it does so instructively, as it allows us

In privately discussing these matters with Donnan, Schrödinger made telling comments that provide additional support for the hypothesis outlined earlier. For instance, when referring to a paper where Donnan had mathematically described the historical character of organisms with reference to Bohr’s principle of complementarity (i.e., Donnan Reference Donnan1936), Schrödinger lamented the “disaster” that “Bohr’s sham-philosophy has produced” when applied to other sciences—an application that he irreverently dubbed “Kopenhagen twaddle”. He asserted that “physicists do wrong in using their claim of representing the ‘fundamental science’ to try and impose some petty fundamental methodological principle of their science to others”.Footnote ⁷¹ In fact, prompted by Donnan’s Reference Donnan1936 paper, Schrödinger published a note in Nature titled ‘Indeterminism and Free Will’ where he criticized the suggestion made by some physicists “that the apparent indeterminacy […] of living matter might be connected with the theoretical indeterminacy of modern physics”, adding—revealingly—that their motivation for doing so “is evidently the hope (whether outspoken or concealed) of extracting from the new physical dogma a model of free-will”. Schrödinger argued that this hope is illusory, and that “free-will actions do not call for a special ‘indeterminist’ explanation any more than other events” (Schrödinger Reference Schrödinger1936: 13).

Schrödinger’s correspondence with Darlington offers further evidence for my hypothesis. When What Is Life? was nearing publication, Schrödinger wrote to Darlington to ask permission to use some of his illustrations. Darlington replied, granting Schrödinger the permission he requested, and expressing his delight that a physicist of his calibre had taken up problems in genetics, despite admitting that he often quarrelled with physicists over the ‘question of indeterminacy’. Schrödinger’s response is enormously revealing, not just because it shows, once again, his underlying concern with free will, but because it lays his cards on the table in a way that clearly discloses what he was trying to do in What Is Life?. He begins by assuring Darlington that his quarrel over indeterminacy is not with him, and then writes the following by way of explanation:

Schrödinger’s motives were clear—at least to Darlington. In one of the most perspicacious analyses of What Is Life?, Darlington explains that the main takeaway of Schrödinger’s book is that the extreme stability of the chromosome molecules “enables the organism […] to escape to a very large extent the quantum indeterminacy of inorganic matter. We can no longer skip merrily (as some did a short while ago) from quantum mechanics to free will. The organism now has a say in the matter” (Darlington and Mather Reference Darlington and Mather1950: 170, my emphasis).

Only in the light of the above considerations, I think, does it become possible to understand why Schrödinger insisted on adding his quirky epilogue ‘On Determinism and Free Will’ to What Is Life?, despite not having been part of the original lectures, and not being obviously connected to their content. They also explain why, with the proofs already in hand, Schrödinger chose not to go ahead with the book’s publication when the initial Irish publisher made it conditional on him removing the epilogue, as we noted back in Section 1.Footnote ⁷³ Although to the biological reader the epilogue is by far the least relevant, least convincing, and most muddled section of the entire book (and this is as true today as it was in 1944), if the argument presented here is correct, then it is also perhaps the most significant for understanding why the book was written. And not because of what the epilogue actually claims, but because of Schrödinger’s dogged determination to include it as his “own, necessarily subjective” five-page philosophical coda to his 86-page examination of ‘the physical aspect of the living cell’ (Schrödinger Reference Schrödinger1944: 87).Footnote ⁷⁴

Nevertheless, because Schrödinger made no effort to make his intentions clear (neither Jordan nor Bohr are cited or even mentioned in What Is Life?), the point of his epilogue was lost on most readers, who found it perplexing, slightly embarrassing, and not worthy of comment—unless, of course, it was to mock it, and neither Haldane (Reference Haldane1945) nor Muller (Reference Muller1946) missed the opportunity to do so in their respective reviews of the book, as we noted at the end of Section 2. An interesting exception here is Delbrück, who in his own review of What Is Life? quoted Schrödinger’s revealing declaration in the epilogue that “contrary to the opinion upheld in some quarters, quantum indeterminacy plays no biologically relevant role” in the cell (Schrödinger Reference Schrödinger1944: 87), and then perceptively inferred that “[t]he opinions here referred to are presumably those of Bohr and those of Jordan” (Delbrück Reference 84Delbrück1945: 371).Footnote ⁷⁵

We can now return to the point with which we started, when I suggested that the conception of cellular order that we find in What Is Life? is not one that a non-biologist would have reasonably inferred upon becoming acquainted with the pertinent technical literature. It is rather a conception that reflects the particular philosophical agenda that drove Schrödinger to develop an interest in the topic—an agenda that we have uncovered in the preceding pages. One consequence of this is that, in his hope of finding in living matter a way of salvaging the deterministic worldview of classical physics, and in striving to keep it from becoming further eroded by the shocking indeterminacy discovered at the heart of quantum mechanics, Schrödinger only read biological literature that validated his existing deterministic predilections. The uncompromisingly genocentric view of the cell and of development that Schrödinger defended in What Is Life? is not one that he could have extracted from the writings of most biochemists and embryologists of the time (should he had taken the time to read them)—it would not even had been the obvious message to derive from the work of most geneticists of that period. But it was the view best suited for him to make the case he wanted to make.

There is plenty of evidence for this. For example, the Schrödinger Papers at the Austrian Central Library for Physics contain a notebook labelled ‘Warum’ (Figure 10), which Schrödinger used to prepare his Berlin talk of 1933 (‘Warum’ is the first word of the talk’s original German title: ‘Warum sind die Atome so klein?’). Of particular interest is a small card enclosed within it (also reproduced in Figure 10) where Schrödinger wrote down the papers that he read in preparation, presumably after his correspondence with the Przibram brothers in late 1932. These are: Alexander and Bridges Reference Alexander and Bridges1929, Boycott Reference Boycott1929, Hetler and Bronfenbrenner Reference Hetler and Bronfenbrenner1929, and Muller Reference Muller1929. Out of the four, Muller’s paper is by far the longest and most substantive, and the ‘Warum’ notebook shows that Schrödinger studied it carefully (see also Sloan Reference Sloan2012 for a complementary analysis of Muller’s influence on Schrödinger).

Figure 10 Front cover of the notebook Schrödinger used to prepare his Berlin talk of 1933 with an enclosed card listing some of the biological literature that he read in preparation

(reproduced with permission of the Austrian Central Library for Physics)

The reason why this matters is that Muller’s views on the gene, which we examined back in Section 3, stand—particularly as presented in the Reference Muller1929 manifesto that Schrödinger studied—as some of the most fervently reductionistic and deterministic expressed by any geneticist in the first third, if not the first half, of the twentieth century. They certainly did not reflect the views of most geneticists (let alone most embryologists and biochemists), at least not at that time. To learn that Schrödinger was so taken with them in the early 1930s, precisely when his biological ideas were starting to take shape, helps explain why, although Muller’s work is not even cited in What Is Life?, it is really Muller’s views—much more than Delbrück’s (which are discussed at great length)—that Schrödinger channelled in the book when he proposed his genocentric, order-from-order conception of the cell, as I already suggested in Section 3.

This becomes even more apparent when we compare Delbrück’s views as presented in TZD (the key claims of which we reviewed in Section 3) with the interpretation that Schrödinger gave of them in What Is Life?. As Sloan has acutely observed (Sloan Reference Sloan2012; see also his co-authored introduction to TZD in Sloan and Fogel Reference Sloan and Fogel2011), Schrödinger misrepresented the main conclusions of TZD in order to make them consistent with his own genocentric and deterministic commitments.Footnote ⁷⁶

The Mullerian view of the gene that Schrödinger defends so enthusiastically in What Is Life? is actually described in TZD, specifically in the penultimate paragraph of the paper’s closing theoretical summary, which was most likely written by Delbrück:

Crucially, however, in the paper’s final paragraph, this Mullerian view is emphatically rejected by the authors:

The failure to recognize this serious discordance between TZD and What Is Life? has meant that Delbrück’s research program in biology (which stemmed from the desire to confirm Bohr’s biological views) has often been erroneously equated with Schrödinger’s (which stemmed from the desire to refute those very same views). This has caused widespread confusion. For example, Schrödinger’s oft-quoted remark in What Is Life? that he expected ‘new laws of physics’ to be discovered in biology has tended to get conflated with Delbrück’s Bohr-inspired search for a paradox that would reveal the limits of ‘classical’ descriptions of biological phenomena and require some form of complementarity to fully explain them (see Delbrück Reference Delbrück1949). The first to make this unfortunate conflation appears to have been Stent (Reference Stent, Cairns, Stent and Watson1966: 4), and many others have made the same mistake since (e.g., Carlson Reference Carlson1971: 152; Jacob Reference Jacob1973: 259; Perutz Reference Perutz and Kilmister1987b: 242; Davies Reference Davies2019: 6; Sigmund Reference Sigmund2019: 43).Footnote ⁷⁷

Returning to Schrödinger’s genocentrism, the key point is that it was, at least to some extent, a choice. Schrödinger did not inevitably arrive at it after surveying the biological literature of the time. What the historical evidence indicates instead is that he first adopted it and then used it to decide what biological papers to read and how to interpret them. Yoxen has reported that Paul Ewald, the physicist who drew Schrödinger’s attention to TZD in 1942, told him that “Schrödinger appeared not to be interested in a thorough survey of the literature in biology, but rather wanted to know of a few articles that would corroborate his point of view” (Yoxen Reference Yoxen1979: 35), and also that Schrödinger’s angry retort to Ewald’s additional reading suggestions was: “‘What do you think, I can’t read the entire literature!’” (quoted in Yoxen Reference Yoxen1977: 147).

What this implies is that, as I argued at the start of this section, there is nothing improper or unfair about taking Schrödinger to task for his unapologetically genocentric view of the cell. And the same could be said for his brazen disregard of the effects of stochasticity at the molecular level. As we noted in Section 5, the impact of random perturbations on microscopic processes has been a well-documented phenomenon—not just in physics but also in biology—for a very long time.Footnote ⁷⁸ Indeed, it is not difficult to find extended discussions of it in the early twentieth-century literature (e.g., Thompson Reference Thompson1917). Thus, it would have been far from obvious in 1944 that one could realistically describe ‘the physical aspect of the living cell’ (not to mention the far more complex process of development) without acknowledging the disruptive effects of statistical fluctuations. In fact, Delbrück criticized Schrödinger’s argument for precisely this reason in his 1945 review of What Is Life?:

Many commentators have complained that much of the biology that Schrödinger discussed in his book was already out of date even before he wrote it. But according to what I have argued in this section, this is not terribly surprising, and not even all that problematic. What Is Life? was never intended to advance biology; to judge it according to that standard is to fundamentally misunderstand why it was written. Schrödinger was a physicist, not a biologist. The biological argument he developed in the book, despite the impact it later happened to exert, was for Schrödinger simply a tool—a means to an end. Once he had made his point (even if it was lost on most of his readers!) he turned his attention to other matters. As Symonds (who studied with Schrödinger) puts it, “[s]omewhere along the line the problems tackled in What Is Life? confronted him, were thought about, the lectures were given and the book written, and then the episode was forgotten as he moved on to think about something else” (Symonds Reference Symonds1986: 226).Footnote ⁷⁹

This explains why, despite the encouragement (both publicly and in private correspondence) of leading geneticists such as Haldane and Darlington, Schrödinger never wrote about genetics again; even after describing it in What Is Life? as “easily the most interesting [science] of our days” (Schrödinger Reference Schrödinger1944: 41).Footnote ⁸⁰ It also explains why he did not take the opportunity to properly update his book for its second edition of 1948—or at any point thereafter—in light of later biological discoveries, which he could have easily interpreted as decisive vindications of his argument (as I suggested in Section 5). And it is probably also why he did not bother responding to Crick’s letter following the momentous discovery of the double helix in 1953 (recall Figure 3).Footnote ⁸¹

7 Conclusions: What Is Life? 80 Years On

What Is Life? has become “part of the folklore of biology” (Symonds Reference Symonds1986: 221). The book keeps “providing nourishment for historians, sociologists and philosophers of science who have commented on it, […] or on the comments on the comments on it” (Perutz Reference Perutz1987a: 555). Its popularity among scientists also shows no signs of abating. Morange (Reference Morange2020: 74) has recently remarked that “[m]odern molecular biologists feel quite at home studying the pages of Schrödinger’s book” because “[t]hey share Schrödinger’s determinist vision of the gene”. 80 years on, the book continues to be widely read. It also remains widely misunderstood. We have seen that virtually all of the book’s memorable expressions—‘aperiodic crystal’, ‘hereditary code-script’, ‘negative entropy’, ‘new laws of physics’—have lent themselves to a fascinating array of interpretations, some more misleading than others. It appears that “[r]eaders of What Is Life?”, as Olby (Reference Olby1974: 246) pithily observed, have “found in it what they were looking for”. This should remind us to take autobiographical declarations of the book’s influence (such as those collected in Box 1) with a grain of salt, though we should be equally wary of overly dismissive reassessments (like those of Perutz and Pauling, and of the historians that have followed in their footsteps) for the same reason.

Schrödinger’s book offered a physical framework to think about two central features of life, heredity and metabolism, though he was evidently far more interested in the former than in the latter. Schrödinger’s bias was embraced by the molecular biology pioneers who sought in What Is Life? a legitimation of their own concern with the molecular structure of the genetic material. It is interesting that many of these pioneers—most notably Delbrück and his group—used bacteriophage viruses as their experimental system, as these are purely parasitic, crystallizable entities in which the metabolic function has been lost and only the hereditary function survives. Dyson (Reference Dyson1985: 5) has perceptively noted that in his book, “Schrödinger’s view of what constitutes a living organism resembles a bacteriophage more than it resembles a bacterium or a fruit-fly”, which in hindsight proved perfectly appropriate.Footnote ⁸²

The title of Schrödinger’s book is too good, and too obvious, to have been original. Schrödinger was not the first to write a book titled What Is Life? (e.g., Windle Reference Windle1908; Gaskell Reference Gaskell1928), nor, for that matter, has he been the last (e.g., Margulis and Sagan Reference Margulis and Sagan1995; Regis Reference Regis2009; Pross Reference Pross2012; Nurse Reference Nurse2021).Footnote ⁸³ Even so, ‘What is life?’ has come to be almost universally regarded as ‘Schrödinger’s problem’ (Olby Reference Olby1971) or as ‘Schrödinger’s question’ (Rosen Reference Rosen, Buckley and Peat1996). This is amusing and ironic; not only because Schrödinger does not even try to answer ‘his’ question, but because one of the salient consequences of his book’s success—and of the molecular revolution that ensued—was the banishment of that previously central question from the biological discourse (Shostak Reference Shostak1998; Morange Reference Morange2008). As Jacob (Reference Jacob1973: 299) famously declared, “[b]iologists no longer study life today”. With the rise of molecular biology, the old concern with the organization of living systems was replaced with a new fixation on the structure and function of genes, which came to be viewed as the master controllers of life—just as Muller had envisioned in his Reference Muller1929 manifesto.Footnote ⁸⁴

Without wanting to downplay the book’s role in rallying young and disaffected physicists to the cause of elucidating the nature of the hereditary substance, I have suggested in this Element that we should not lose sight of the actual argument it put forward. What Is Life? should be remembered not simply because of who wrote it (despite the author’s preeminent scientific reputation) but also because of what it claimed. As we have seen, the mechanical, deterministic, and genocentric view of the cell articulated in its pages was instrumental in shaping and consolidating the agenda of molecular biology during the second half of the twentieth century. Molecular biologists’ conception of cellular order as organization-from-information is recognizably Schrödingerian, and so is their understanding of development as the execution of a pre-existing genetic program. Principles such as informational and structural specificity, the Central Dogma, and even the pervasive metaphorical appeals to molecular machines and solid-state electronic circuits can all be construed as theoretical extensions of the argument that Schrödinger laid out in What Is Life?—even if some emerged independently of it.

A key point to remember about Schrödinger’s argument is that it can be held empirically to account. It might well have turned out to be true that cellular and organismic order is conserved and transmitted by means of a self-executing, preformationist code-script embedded in the solid-state structure of an aperiodic crystal that is immune to the ravages of thermal agitation. As it happens, however, there is now overwhelming empirical evidence that this is not in fact the case. In this respect, to think carefully about the problems with Schrödinger’s argument—as we have done so in this Element—is to reflect on the limitations of conventional molecular biological explanations of macromolecular, cellular, and developmental phenomena.

Despite the book’s popularity and influence, Schrödinger’s original reasons for writing it have been completely forgotten—or so I have argued. The fact that his dispute with Jordan and Bohr regarding the extension of quantum mechanics to biology and psychology did not go anywhere shows, in a sense, how triumphant Schrödinger has been. Leaving aside the perennial problem of free will, which continues to be discussed today with as much vigour and panache as it was in the 1930s and 1940s (and, like most fundamental philosophical problems, is unlikely to ever be conclusively resolved), the empirical question of the bearing of quantum indeterminacy on cellular processes was decisively settled, and in Schrödinger’s favour. Jordan’s quantum biology faded into oblivion in the immediate post-war period, and molecular biologists have shown no appetite to reconsider the relevance of quantum effects ever since.Footnote ⁸⁵

I hope to have convincingly shown in this Element that philosophers, historians, and biologists all stand to benefit immeasurably from revisiting What Is Life? today. And not because what the book says is true, but because reading it helps us understand how we ended up with our current image of the cell, and how this image is likely to evolve in the decades to come. Recalling Judson’s (Reference Judson1979: 244) remark, perhaps it is time, once again, that “[e]verybody read Schrödinger”.