Retroactivity in Science: Latour, Žižek, Kuhn

1 Latour on Partial Existence

If asked, I would not hesitate to name Latour as one of the most significant philosophers of the past half-century.^[8] This does not prevent me from strongly disliking one of his specific arguments, found in an article from the millennial year of 2000: “On the Partial Existence of Existing and Non-Existing Objects.”^[9] Here, Latour recalls an episode from 1976 in which the mummy of the Egyptian Pharaoh Ramses II was flown to Paris, where a team of French experts determined that Ramses had died of tuberculosis, a disease unknown – though apparently active – in the Pharaoh’s own time (d. 1213 BCE). Spurred by his memory of this event, Latour poses and answers what looks at first like a rhetorical question:

Let us accept the diagnosis of “our brave scientists” at face value and take it as a proved fact that Ramses died of tuberculosis. How could he have died of a bacillus discovered in 1882 and of a disease whose etiology, in its modern form, dates only from 1819 in Laënnec’s ward? Is it not anachronistic? The attribution of tuberculosis and Koch’s bacillus to Ramses II should strike us as an anachronism of the same caliber as if we had diagnosed his death as having been caused by a Marxist upheaval, or a machine gun, or a Wall Street crash.^[10]

Scientific realists never viewed Latour favorably during his lifetime; for them, the preceding paragraph could only serve as further proof of his “irrationality.” True enough, no one would contest the claim that Marxist upheavals, machine guns, and Wall Street crashes all appeared in the world long after the lifetime of Ramses, so that none of these could be listed as plausible causes of death for the Pharaoh. Tuberculosis would strike most observers as a different case, however. After all, there is no obvious contradiction in saying that the disease existed in ancient times but was not discovered until the nineteenth century. Yet Latour does see a contradiction in claiming that a thing can have effects prior to its discovery. The latter part of his article offers the additional example of Louis Pasteur and microbes, a topic he had already covered a year earlier in Pandora’s Hope.^[11] Here too, Latour finds it impossible to speak about what microbes were doing “all along” prior to their discovery by humans, which required the mobilization of a vast network of human and non-human witnesses.

Latour is still better known in the social sciences than in philosophy proper. At the time of his death in 2022, he had received such major international awards as the Holberg Prize in Sweden and the Kyoto Prize in Japan. Idolized by two generations of students in anthropology, sociology, and science studies, he is usually the first person who comes to mind whenever there is talk of Actor-Network Theory.^[12] Yet he is still not the subject of an article in the authoritative Stanford Encyclopedia of Philosophy, and only in the Netherlands can one regularly find his books shelved in the “Philosophy” section rather than scattered piecemeal under multiple headings. Analytic philosophers tend to dismiss Latour as a puffy relativist and social constructionist who insults science by reducing it to human power struggles; continental thinkers are baffled by his analyses of such apparently pedestrian topics as how the price of apricots is determined in Paris, and by his general lack of interest in Kantian transcendental philosophy.^[13]

Nonetheless, with We Have Never Been Modern in 1991, Latour wrote one of the philosophical classics of the late twentieth century.^[14] In this book, he defines modernism as an artificial separation between a “pure” realm of nature that moves with the rigid necessity of dead clockwork and an equally “pure” sphere of human culture that merely imposes a grid of arbitrary values onto an otherwise cold universe of dead grey matter. Latour redefines both of these pseudo-categories in terms of the unified concept of “actors.” Anything is real as long as it has an effect on something else, leading to the wonderfully broad result that everything from atoms of nickel and trilobite fossils to Sherlock Holmes and Daffy Duck must be analyzed in exactly the same terms. How does each of these actors affect other actors? Nothing else matters. It is a daring ontology, though one guaranteed to annoy those with intense pre-existing commitments to other ways of arranging the universe. This includes not only hardcore scientistic naturalists, but also committed continental “readers” of “texts” who have no idea how or even why philosophy might come to terms with bona fide inanimate objects.

Even those who are sympathetic to Latour’s approach might raise at least two fundamental objections: (1) By contending that actors should be analyzed into nothing more than the effects they have on other things, Latour risks sliding into the ultra-relational ontology already defended by Alfred North Whitehead, one of his intellectual heroes.^[15] Latour writes as follows in his article on Ramses: “As Whitehead shows in his cosmology, realism and relativism [i.e., relationism- g.h.] are synonyms.”^[16] (2) Latour responds to the modernist divide between pure nature and pure culture not only by pointing to the existence of numerous hard-to-classify “hybrid” entities such as the ozone hole and genetically modified corn, but goes further by implying that all entities are hybrids. This means that whenever we talk about anything, we must imagine a human chaperone on the scene, co-constituting its reality. Thus it is with tuberculosis, which cannot be said to pre-exist the collective scientific apparatus that first identified it. Stated more generally, Latour’s conception of science forbids us to announce purported facts about entities outside the mind, since his standpoint entails that any scientific proposition is merely shorthand for all the various human and non-human actors that had to be assembled to support a given claim.^[17]

Science might think of itself as pointing to entities that exist outside the act of pointing at them, but Latour denies this is possible.^[18] It is true that we speak of “X-rays” or “electrons” as if these were autonomous entities existing in a vacuum devoid of relations, but that is merely because we have “black-boxed” the stories of their discovery for the purposes of mental economy. Every scientific fact is an achievement to which numerous human and non-human entities contribute, and every fact forever carries the baggage of its entire historical past. And given that science (as far as we know) is only conducted by humans, there is always a human being present – at least tacitly – in any statement about reality. Thus, we find that despite Latour’s intense opposition to Kant, he resembles his German predecessor in holding that any statement about reality is really a statement about human thought’s relation to reality. Although Latour is much too interested in non-human things to be viewed as any sort of idealist, he is very much a philosopher of science rather than a philosopher of nature. And while we can imagine nature without any humans present, it is obviously impossible to imagine the same for science. In this respect, Latour’s dual critique of the thought/world modernist dyad is not quite as even-handed as he tends to claim. Both thought and world are reconceived in terms of “action,” but action means relation, and relation for Latour means the demise of anything existing in itself, apart from all human witnesses.

From this brief sketch, it is easy to see why Latour is skeptical toward the theory that Ramses II died of tuberculosis. A standard scientific realist thinks of the word “tuberculosis” as caused by a bacillus existing in the world outside the mind, whether we like it or not. For Latour, it refers instead to a long chain of historical actors and laboratory experiments that together institute the purported existence of this disease. It would be easy for realists to accuse him of a pettifogging relativism on this point, and many have already done so. But in all fairness, it should be remembered that tuberculosis is not the sort of thing that we can simply open our eyes and see. The reason its existence took millennia to detect is that a complex series of mediators was needed to establish its reality. We should also recall that science does not provide direct intuitive visions of reality in its own right, as mathematics at least claims to do, since we know full well that the sciences are fallible and in constant change. It may turn out a few decades from now that what we know as “tuberculosis” will have been a botched and primitive oversimplification of the disease that now seems to have killed both ancient figures like Ramses II and such nineteenth-century luminaries as Frédéric Chopin and Robert Louis Stevenson. If so, then the insistence of French scientists in 1976 that they knew tuberculosis to be the true cause of Ramses’s death could eventually look as quaint as the former belief in “phlogiston” and “caloric” does to chemists today.

What I mean to say is that we should not dismiss Latour’s position on Ramses too flippantly; it involves a serious philosophical claim. This is not to say that we ought to accept it, given the rather extreme fashion in which he forbids a statement from pointing to anything beyond its own extensive conditions of production.^[19] In terms of analytic philosophy, we could say that Latour treats statements as all meaning, no reference. Whereas Kripke treats names as “rigid designators” that continue referring to the same person even if everything we thought we knew about them turns out to be false, Latour’s position amounts to the claim that names are shorthand descriptions for a vast series of antecedent actions. This is especially true of scientific objects, which gradually come into focus through a sequence of “trials of strength.” As Latour puts it in Science in Action, speaking of a famous discovery by Marie and Pierre Curie:

[“Polonium”] defeated uranium and thorium at the sulphurated hydrogen game; it defeated antimony and arsenic at the ammonium sulphur game; and then it forced lead and copper to throw in the [towel], only bismuth went all the way to the semi-final, but it too got beaten down during the final game of heat and cold! At the beginning of its definition the thing is a score list for a series of trials.^[20]

Returning to the case of tuberculosis, Latour calls it a form of Whig history to claim that tuberculosis was there all along before we discovered it.^[21] When we extend tuberculosis into the distant past, this comes at a heavy price, given how much modern equipment is needed to identify the supposedly autonomous existence of the disease prior to its discovery. This leads Latour to make the intriguing point that technology provides a better model than science for any state-of-the-art ontology, since everyone admits that technological objects “never escape the conditions of their production.”^[22] It is true that machine guns cannot be transported back into Ancient Egypt, but equally true that machine guns cannot exist independently in the present, given all the “know-how, bullets, oil, repairmen, and logistics necessary to activate it.”^[23] Actors exist only through their relations, and given that the past is always understood from the standpoint of the present, there is no independent past characterized by a specific way things were back then.

Latour takes all this to be the necessary result of recent achievements in the history and sociology of science, which he proudly states has “raised enough problems, monsters, and puzzles … to keep philosophers, metaphysicians, and social theorists busy for decades.”^[24] Just as no one will ask “where was Louis Pasteur before 1822?,” the year of his birth, so too there is little point in claiming that tuberculosis existed prior to its own recent birthdate in medical history. This does not mean, of course, that Pasteur would have agreed with Latour on this point. As Latour openly admits, although Pasteur and Félix Archimède Pouchet were on opposite sides in the debate over spontaneous generation, they were joined in ascribing a “boasting and empty privilege [… to] nonhumans over human events.”^[25] Rather than overcoming the modernist thought/world divide like Latour himself, Pasteur and Pouchet simply asserted the “world” pole of modernity, despite their opposed beliefs in either the falsity (Pasteur) or truth (Pouchet) of germs springing into existence ex nihilo. But in order to demonstrate the existence of “contamination” in non-sealed flasks, Pasteur had to mobilize a veritable army of actors, which he conveniently erased from his account of germs existing in the world outside the mind.

Latour effectively endorses – as Žižek also does – the old German Idealist argument that we cannot think a thing outside thought without turning it into a thought. The main difference is that Latour is less interested in “thought” than in material laboratory practice, but the essential resemblance remains. Nothing really exists apart from whatever factors lead us to acknowledge its existence. For it is not only humans who have histories.^[26] We cannot think of things in the world as “substances” that possess “a little bit more” than their histories.^[27] The only thing that makes Pasteur “right” is that we now live in a world of institutions (not substances) organized according to his discoveries.^[28] Today, most of us refrigerate our milk and accept vaccinations from doctors: the best possible proof that Pasteur won the argument with Pouchet. For Latour this means that the words “constructivism” and “realism” are synonyms.^[29] What we usually call realism, he holds, is nothing more than a political epistemology designed to place scientists in a position of power over sociologists and historians.^[30] Here we catch sight of the permanent dispute that divides Latour from his natural enemies, the scientific realists.

2 Žižek’s Grain of Scientific Realism

The colorful Slovenian philosopher Slavoj Žižek belongs to the same generation as Latour, being just two years younger than the late lamented Frenchman. Their mutual dealings were minimal during Latour’s lifetime, and to this day, few scholars even refer to both authors in the same publication. Žižek grew up in communist Yugoslavia, and his intellectual formation owes much to the philosophy of German Idealism and the psychoanalysis of Jacques Lacan, along with the rather unorthodox Marxism imbibed during his youth. Latour was a product of Gaullist and post-Gaullist France, and his initial influences were Christian theologians and Anglo-American social scientists. Even so, with the theme of retroactivity, we have a point of significant overlap between them.

Žižek’s fondness for retroactive structures has a different impetus from Latour’s own, arising as it does from his passion for both psychoanalysis and German philosophy in the late eighteenth and early nineteenth centuries. In a first sense, Žižek is a vehement convert to Kant’s Copernican Revolution in philosophy, which counts for Latour as a modernist disaster that artificially separates the “thought” and “world” poles of reality. But for Žižek (who resembles Quentin Meillassoux on this point) it is only the Kantian recourse to transcendental philosophy that prevents us from sliding into old-fashioned dogmatic realism.^[31] As Žižek tells his interviewer Glyn Daly: “Kant was the first philosopher … Pre-Kantian philosophy cannot think [the] transcendental aspect.”^[32] One of the features of standard realism that Žižek finds disturbing is that in his view, it is ultimately responsible for “ideology.” A typical Marxist account of ideology would tell us that whereas the material conditions of production are the true explanation for all that happens in human history, ideology misascribes many events to such superstructural byproducts as culture, religion, or the “bourgeois” philosophy found in such figures as Kant and G.W.F. Hegel. Despite his own sympathies for Marx, and his notable tendency to defend the role of economics against present-day cultural Marxists, Žižek does not think that economics is what saves us from ideology. Instead, in a way more in keeping with his German Idealist side than his Marxist one, Žižek treats philosophical realism as the core of ideology.

The best remedy for this, he holds, is an insight into retroactive projection as constitutive of the real. For instance, in his 2012 book Less Than Nothing, Žižek repeats his view of two decades earlier that anti-Semitism can be blamed on a realist interpretation of Jewishness. Just as he had claimed in The Sublime Object of Ideology, a true anti-Semite would not merely say that Jews conspire and swindle, but rather that they conspire and swindle because they are Jews.^[33] In other words, anti-Semitism could never arise from an event-oriented philosophy such as Actor-Network Theory, but only from a theory of substance that treats Jewishness as a hidden autonomous core, one that is capable of generating specific devious acts. Summarizing the anti-Semitic standpoint in Less Than Nothing more than twenty years later, Žižek once again blames it on the essentialist formula that “there is some mysterious ingredient in Jews, an essence of being-Jewish, which causes them to be degenerate […]”^[34] This realist je ne sais quoi is treated as the beating heart of all ideology, in the pejorative Marxist sense of the term.

It is interesting that, for Žižek, the only escape from the ideological trap is found in a structure through which the subject retroactively posits its own antecedent conditions. When the young Karl Marx accepts Aristotle’s realist premise that things are the way they are whether we know them or not, Žižek responds decisively and critically.^[35] Human freedom, in the Slovenian’s conception, does not mean being free of the universal system of causes and effects, but entails instead the freedom to choose which past causes will henceforth determine us.^[36] In this sense, Žižek upholds “a suspension of the Principle of Sufficient Reason” just as strong as Meillassoux’s. But whereas Meillassoux emphasizes the necessity of all contingency, Žižek (like his friend Alain Badiou) is more interested in the opposite phenomenon: the way we contingently choose which necessities will bind us.^[37] He is quick to add that our free choice of reasons for our actions is “not only epistemological but also ontological,” thereby slamming the door on any form of traditional realism like the sort that tempted the young Marx.^[38] Žižek expresses fondness for T.S. Eliot’s view that one cannot simply read the works of the tradition, since one inevitably transforms the tradition with the production of any new work. This is yet another example of how there is no pre-existent past, but only a past that belongs retroactively to a given present moment.^[39] Given this unwillingness to concede the existence of any reality-in-itself apart from how it is thought, Žižek resembles not only Latour, but also Karen Barad, with her allegiance to Niels Bohr’s notion of “complementarity” between thought and world.^[40] Like Barad, the quantum theory-loving Žižek is effectively asking us to “meet the universe halfway.”

If we ask now about Žižek’s position concerning analytic philosophy’s dispute over meaning and reference, he initially resembles Latour (and by extension, Barad) in treating all language as a question of meaning. There would appear to be no room in Žižek’s philosophy for a rigid reference pointing at one and the same thing that endures over time. Indeed, it takes no guesswork to say so, since Žižek tells us directly what he thinks of Kripke’s theory of rigid designation.^[41] In The Sublime Object of Ideology, his critical interpretation of Kripke makes common cause with the anti-essentialist views of his one-time allies Ernesto Laclau and Chantal Mouffe.^[42] More importantly, he takes the delightful step of translating Kripke’s ideas into the terms of Lacanian psychoanalysis. As Žižek puts it, “when we encounter in reality an object which has all the properties of the fantasized object of desire, we are nevertheless somewhat disappointed; we experience a certain ‘this is not it’; it becomes evident that the finally found real object is not the reference of desire even though it possesses all the required properties.”^[43] This, of course, is comparable to how Kripke treats the workings of proper names. Any attempt to define the “meaning” of a name must fail, given that even in cases where all apparent facts about the named person turn out to be false (“Columbus discovered America,” “Einstein invented the atomic bomb”) the name continues to point to this person right here, the same person to whom it pointed all along.

This is what Kripke means with the confusing description of his own theory of reference as a “causal” one; it is causal only the sense that, in order to discover the referent of a name, we need to trace it back to the point when it was first bestowed upon a person, place, or thing. The meaning of the name may have changed greatly in the meantime, but the reference has not. At this stage we might assume that Žižek is set to embrace a Kripkean theory of reference, given his invocation of a desired object irreducible to any of its tangible properties, none of them very compelling to anyone who does not desire it. Yet we must always remember Žižek’s overriding debt to Lacan, and to the absence of any traditional realism from Lacan’s theory of desire (his “Real” is a trauma from outside the symbolic order, and has nothing to do with “reality” in the usual sense).^[44] Žižek amusingly asks: “How could we overlook the libidinal content of these proposition of Kripke? … It is perhaps no accident that Kripke selects as examples objects with an extreme libidinal connotation, objects which already embody desire in common mythology: gold, unicorn […]”^[45] Yet even for a heterodox Lacanian like Žižek, the object of desire is never treated as a real object outside the mind, one that was sitting there all along before we appeared and began to want it. Instead, the object of desire is the Lacanian objet a or “object-cause” of desire, which provides a retroactive explanation for our obsession with a person or object, although it strikes others as just another relatively dull bundle of qualities.^[46]

Even so, Žižek is sometimes unwilling to push things as far in the retroactivist direction as Latour or Barad. His surprising discomfort with hardcore anti-realism about objects is best seen in some remarks about science found in Less Than Nothing. Over the years, Žižek’s excursions into physics have generally shown him to be an ally of the Copenhagen Interpretation of quantum mechanics, with an inclination toward Bohr’s radical “ontological” interpretation rather than Werner Heisenberg’s more moderate “epistemological” one.^[47] Whereas Heisenberg’s uncertainty principle claims that an observer cannot accurately know both the position and momentum of a particle with any exactitude, Bohr’s complementarity goes further, contending that the unobserved particle does not even have a definite position and momentum prior to measurement. As mentioned earlier, Barad sides with Bohr in this dispute, and as a rule I find that Žižek excessively downplays his overlap with Barad’s own position.^[48]

But Žižek’s Bohrian tendencies are dampened in Less Than Nothing as soon as he speaks of the relation between science and nature. The Latour of the Ramses article effectively tells us that we cannot remove tuberculosis from its historical era and transport it back to Ancient Egypt, since there is not one and the same physical nature that endures through the ages in the face of changing human knowledge. The disease landscape of Ramses’s time and our own are different, even if a group of Parisian doctors detect “the same” tuberculosis in both time periods. And while Žižek could safely be expected to enlist in this Latourian view, that is not what happens. Instead, when we least expect it, we find a grain of scientific realism in Žižek’s position:

The epistemological passage from classical physics to the theory of relativity did not mean that this shift in our knowledge was correlated to a shift in nature itself, that in Newton’s time nature itself was Newtonian and that its laws mysteriously changed with the arrival of Einstein—at this level, clearly, it was our knowledge of nature that changed, not nature itself.^[49]

Translated into Kripkeanese, this is equivalent to Žižek saying that once we realize Columbus did not discover America, it is our knowledge of Columbus that changes rather than Columbus himself. Given how different this seems from the Žižek readers know, one might be tempted to write it off as an anomalous passage. But that is impossible, since he goes on to endorse a specific theory about his newly acclaimed nature-in-itself. In a decision less surprising than it might initially seem, he lends his support to the ontology of the analytic philosopher Peter van Inwagen, who essentially holds that nothing exists in the extra-mental world aside from physical simples (and living creatures).^[50] That is to say, in 2012 it appears that Žižek endorses a world of tiny particles, onto which the subject “projects” anything – such as chairs – that might seem more complex. If there is any Kripkean theory of direct reference in Žižek, it turns out that most of the names in this theory refer, ultimately, to something like atoms. This is a far cry from Žižek’s usual form of materialism, which tends to redefine the word in such a way that there is no room for independent matter at all.^[51]

There are two key intellectual influences at work for Žižek that mean little or nothing for Latour: namely, Lacan and Hegel. During my twenty-three years of friendship with Latour, I do not believe he mentioned either psychoanalysis or dialectics even once. This makes a stark contrast with Žižek, whose conception of retroactivity ultimately has Freudian, Lacanian, and Hegelian roots. In his early book The Metastases of Enjoyment (1994), he reminds us of Freud’s case study of the “Wolf Man.”^[52] For what was the cause of the Wolf Man’s severe neurosis? As Žižek summarizes:

the Cause, of course, was the traumatic scene of the parental coitus a tergo [i.e., sexual intercourse from behind-g.h.]– this scene was the non-symbolizable kernel around which all later symbolizations whirled. This Cause, however, not only exerted its efficiency after a certain time-lag, it literally became trauma – that is, Cause – through delay: when the Wolf Man, at the age of two, witnessed the coitus a tergo, nothing traumatic marked this scene; the scene acquired traumatic features only in retrospect, with the later development of the child’s infantile sexual theories, when it became impossible to integrate the scene within the newly emerged horizon of narrativization–historicization–symbolization.^[53]

As usual with Žižek and other members of the Ljubljana School, this structure passes beyond psychoanalysis and is interpreted as being the structure of Hegelian dialectic as well. As Žižek claims on the following page:

It is impossible to miss the Hegelian overtones of this paradox: does not this repetition of the same, this return to the same, which brings about the change of the surface, offer a perfect illustration of Hegel’s thesis on identity as absolute contradiction? Moreover, does not Hegel himself assert that, through the dialectical process, the thing becomes what it is?”^[54]

There are reasons to question this identification of Hegel with Lacan, though Žižek linked them no later than his 1985 Paris doctoral thesis, The Most Sublime Hysteric.^[55] But what interests us here is something else. Namely, how far would Žižek be willing to push this structure beyond the bounds of Hegel? We have seen that in Less Than Nothing, he flirts with a non-retroactive or “Kantian” interpretation of scientific change, in which our thought about nature changes without nature itself shifting. But how might Žižek respond to the case of tuberculosis and Ramses’s mummy? My best guess is that Žižek would treat this more like his example from physics, and in a less Freudian or Hegelian way. We will now see that Kuhn pushes things further than this, though in a different manner from Latour.

3 Kuhn on Discovery and Invention

Although I am enough of a realist to be suspicious of both the Latourian and Žižekian stances on retroactivity, I find myself less bothered by a different sort of argument, this one made by Kuhn. Although his classic book The Structure of Scientific Revolutions is still widely familiar to university students, most readers’ memories of this work are dominated by its central distinction between “normal” and “paradigm-shifting” science.^[56] But the book also cites two examples of retroactivity in the history of science, and they are meant to point to a more general problem.

The first concerns the question of who first discovered oxygen, where the two leading candidates are Joseph Priestley and the tragically guillotined Antoine Lavoisier.^[57] Kuhn’s second example is the discovery of X-rays. Here there is no question that Wilhelm Röntgen deserves the credit, yet there remains the problem of knowing when exactly he discovered them. Kuhn concludes only that X-rays appeared at some point “between November 8 and December 28, 1895.”^[58] It is worth noting that an analogous problem structures Kuhn’s later book on Max Planck and black body radiation, the scene where quantum theory first emerged in 1900–01.^[59] Or did it really emerge in those years? Kuhn’s intriguing conclusion is that quantum theory was first interpreted by Planck as a mathematical device for attacking a recalcitrant problem. Only once he had assimilated the criticisms of Paul Ehrenfest and Albert Einstein – by 1909 – did he come to recognize that quanta must exist as genuine units of physical nature.^[60] In this sense, it is only from the standpoint of 1909 that quantum theory was discovered in 1900–01, which at least sounds close to Latour’s view of tuberculosis.

Yet Kuhn’s theory of retroactivity in science is different in spirit from those of Latour and Žižek. As mentioned, the initial issue for Kuhn involves the difficulty in determining the precise point in time when any scientific discovery occurs. This is less a matter of the tangled claims to priority made by multiple researchers, and more a question of two distinct moments that characterize any breakthrough, and which turn out to have ontological no less than historical weight. Let’s consider more closely the question of who discovered oxygen. Kuhn begins by disqualifying C.W. Scheele in Sweden, despite his being the first to have prepared a pure sample of the gas. The reason is that Scheele’s discovery “was not published until oxygen’s discovery had been repeatedly announced and thus had no effect upon the historical pattern that most concerns us here.”^[61] That leaves us with Priestley in England and Lavoisier in France as the two contenders for the crown.

The problem is that neither of these celebrated scientists has a clear-cut case. Priestley misidentified oxygen in 1774 as nitrous oxide, and again in 1775 as “dephlogisticated” air, though the very existence of phlogiston is now dismissed as a myth. For Lavoisier’s part, he first misidentified oxygen as “the air itself entire,” which any schoolchild now knows to be false, given the variegated mixture of oxygen with other gases found in the earth’s atmosphere. By 1777 Lavoisier had correctly decided that oxygen was something altogether different, though to his dying day, he failed to realize that oxygen can be present even when no combustion occurs. Kuhn’s conclusion is that oxygen was not discovered before Priestley’s first result in 1774, but was surely discovered by 1777 “or shortly thereafter.”^[62]

A lesser historian might have stopped here and offered the muddy conclusion that the history of science, and of everything else, is so hopelessly confused by conflicting human interests that we cannot say anything definite about when a discovery was made. But Kuhn was also a first-rate philosopher, and he proceeded to draw an important philosophical lesson from the dispute over oxygen. The reason for the grey zone of uncertainty between 1774 and 1777, he argues, stems from the duality between the fact “that” something is and the further determination of “what” it is. Clearly, we are in the vicinity of the distinction between meaning and reference: a division of labor in which names point at or refer to something that purportedly exists (“oxygen,” “Einstein”) and definite descriptions mean the properties that belong to these entities. Needless to say, the latter can always be false, as when Lavoisier wrongly held that oxygen is found only in combustion, or when the proverbial person on the street mistakenly identifies Einstein as the inventor of the atomic bomb. There is a solid historical pedigree for this duality between “that” and “what,” one stretching back to the medieval distinction between existence and essence.^[63] The critical point for Kuhn is that his “paradigm shifts” in the history of science cannot happen instantly. Rather, they “involve an extended, though not necessarily long, process of conceptual assimilation.”^[64]

A similar process occurred with Röntgen’s 1895 discovery of X-rays in Würzburg, Germany, though here it happened in reverse order. Priestley and Lavoisier initially believed they had discovered something, though not necessarily something entirely new or unknown, and needed time to work out its properties or “whatness.” Röntgen, by contrast, immediately grasped the whatness of his discovery, noting certain effects in an experiment that should never have occurred. Over the course of an intense seven-week period, he proceeded in the other direction from this whatness to the “thatness” of a previously unknown entity, which even the distinguished scientist Lord Kelvin initially viewed as a hoax.^[65] Kuhn’s vision of puzzle-solving “normal science” conceives of workaday scientific practice as presupposing the existence of a set of basic objects borrowed from the dominant paradigm of one’s era: such as force, mass, and gravity in Newtonian physics. Any paradigm is riddled with unsolved anomalies, and most scientific research involves working out their solutions or further implications. But in the two paradigm shifts just mentioned, a conflict arose between the object and its properties. Either a newly discovered object (oxygen) demands a new survey of its numerous implications, or a newly discovered anomaly (glowing screen effects) leads the scientist backward to the positing of a previously unknown object responsible for observed discrepancies.

For this reason, a paradigm shift necessarily unfolds in a grey zone without any definite point in time where the shift occurs. This is because any object is inherently twofold, consisting both of a unified existence and a vast multitude of properties. Only from the standpoint of 1777 does Priestley’s result of 1774 look like the beginning of a chemical revolution. Only on December 28, 1895, does Röntgen’s Würzburg anomaly of November 8, 1895, emerge as a fundamental discovery in physics. In this way, both the “that” and the “what” of a thing are able to posit the other retroactively as belonging to itself. The same holds, in Kuhn’s account, for the grey zone between Planck’s first quantum papers of 1900–1901 and his full recognition, in 1909, that real quanta must exist in the world as something more than a set of mathematically manipulable properties.^[66]

We will not find this happening in Kuhnian normal science, where the range of established objects is relatively stable; nor will we find it in a scattershot list of effects with no known underlying cause. The latter was the case – Latour has shown – for the pre-Pasteur “hygienists,” arming themselves with vast laundry lists of rules for remaining healthy, though before the discovery of microbes they were unable to make much theoretical headway with these lists.^[67] For Kuhn, a closely related phenomenon to paradigm shifts is that of “crisis” periods in science, characterized by the growing mismatch between an existing paradigm and anomalous observed results. Initially, many scientists will treat this discrepancy as nothing more than a new series of puzzles to be solved: “in such cases normal science continues as usual, aside from a fringe group of would-be innovators who feel a crisis at hand.”^[68] But there are other eras that Kuhn describes as “period[s] of pronounced professional insecurity,” in which there is general awareness that something is wrong, but with no clear idea of how to set the ship back on course.^[69] Consider the latter decades of Ptolemaic astronomy, in which “astronomy’s complexity was increasing far more rapidly than its accuracy,” thus slowly paving the way for the needed introduction of the Copernican system.^[70] In such cases, it often happens that recourse is had to an earlier, premature theory whose time had not yet come: as with the heliocentric astronomy of Aristarchus of Samos in the third century BCE, which can be called premature due to the continuing general success of Ptolemy’s system at the time. A similar thing occurred with Leibniz’s relational view of space and time, which was initially the loser in battle with Newton’s view that space and time were empty and absolute containers, though Einstein’s later breakthroughs made Leibniz’s theory seem the more prescient one.^[71]

One overly easy way to handle the tension between the “that” and the “what” of scientific objects would be to restrict theory change to the side of the “what,” the side of meaning. Someone could argue that science keeps referring to the same real objects while merely evolving in its understanding of the properties of those objects. But as early as The Structure of Scientific Revolutions, there are signs that Kuhn regards this solution as untenable. He openly questions whether there can be any scientific reference without meaning, as in the example of the word “mass” in the systems of Newton and Einstein. Whereas mass for Newton is always conserved, for Einstein it is convertible with energy: e = mc ², as the famous equation goes.^[72] In short, when Newton and Einstein speak of mass they are not really pointing to the same thing at all. The same holds for the meaning of “planet” before and after Copernicus, since in his heliocentric astronomy the moon loses planetary status and becomes a satellite, while the earth becomes a wandering planet and ceases to be the fixed center of the universe.^[73] But if we are not making steady scientific progress toward a referent that remains the same over time, this has the important consequence that we cannot assume human knowledge gradually approaches reality. There is a sense in which Newton and Einstein live in “different worlds,” not just different theories about one and the same world.^[74]

What does this sort of retroactivity mean for the distinction between meaning and reference, especially in contrast with the aforementioned views of Latour and Žižek? Luckily, Kuhn does not leave us entirely in the dark on this point. In 1977 he engaged in public debate with Boyd at the University of Illinois, during which the topic of metaphor was raised. Although Kuhn’s great book had suggested new possible connections between science and literature, he embraced their link in a spirit of caution, carefully listing a dozen or so ways in which the arts and sciences differ.^[75] Nonetheless, Kuhn and Boyd were united in their admiration for Max Black’s “interactive” theory of metaphor, in which the tension between objects and their properties or meaning is made lucidly visible.^[76] For Black, a metaphor creates a new object that cannot be translated without information loss into any particular meaning: we can never say exactly what “man is a wolf” is supposed to say; hence the tension found in good metaphors. An analogous tension between the “that” and the “what” of scientific objects establishes a metaphorical dimension of science that Kuhn is not shy in describing. As he puts it in his discussion with Boyd:

[…] I agree entirely with Boyd’s assertion that the open-endedness or inexplicitness of metaphor has an important (and I think precise) parallel in the process by which scientific terms are introduced and thereafter deployed. However scientists apply terms like “mass,” “electricity,” “heat,” “mixture,” or “compound” to nature, it is not ordinarily by acquiring a list of criteria necessary and sufficient to determine the referents of the corresponding terms.^[77]

For his part, Boyd ties this directly to the philosophical distinction between meaning and reference. And while Kuhn is largely in agreement, he does add an interesting caveat. As he puts it: “Like Boyd, I take this analysis of reference to be a great advance, and I also share the intuition of [Kripke and Putnam] that a similar analysis should hold of natural kinds: … games, birds …, metals …, heat, and electricity.”^[78] Kuhn continues: “[n]onetheless, despite the amount that Putnam and Kripke have written on the subject, it is by no means clear just what is right about their intuition.”^[79]

What troubles Kuhn is an apparent difference between the names of natural kinds on the one hand and of scientific objects on the other. If we say “Boyd” and point at Kuhn’s companion on stage, there is no doubt about the object of reference: a familiar person named Boyd, visible with our own eyes, even if our senses are deluded and no such person exists. Here Kripke’s “rigid designator” is unproblematic. But if we point to the needle of a galvanometer and say “electric charge,” Kuhn contends that this is more closely entangled with the meaning of the term. Namely, the referential phrase “[‘electric charge’] supplies no information at all about the many other sorts of events to which the name ‘electric charge’ also unambiguously refers.”^[80] What Kuhn seems to think is that while a banal sort of reference is possible in the case of natural kinds, there is a more turbulent relation between meaning and reference in the case of scientific objects, one that more closely approximates the character of metaphor. In particular, Kuhn is critical of Boyd’s view that certain analogies in science are exact and can be treated as such. Boyd cites the parallel between the solar system and Bohr’s model of the atom, which treats electrons orbiting the atomic nucleus as being like planets orbiting the sun. Kuhn counters that Bohr never meant this analogy literally, but was always aware of the difference between the two cases. For instance, Bohr’s electrons change orbits in sudden leaps, which obviously does not occur with planets; moreover, electrons can spin in two different directions, something unthinkable for planets. Bohr’s model always remained a model, and even its mathematical expressions make no sense without reference to its character as a model. As Kuhn puts it: “Without [the model’s] aid, one cannot even today write down the Schrödinger equation for a complex atom or molecule, for it is to the model, not directly to nature, that the various terms in the equation refer.”^[81]

One might ask whether this is a good place to compare Kuhn with Latour and Žižek on the topic of metaphor. But there is a sense in which no such comparison is possible. The first point worth noting is that Kuhn discusses metaphor far more often than the other two authors, for whom it plays no central theoretical role. This is no accident: if we accept the implications of Black’s theory that metaphor polarizes between an object-term (“man”) and a properties-term (“wolf”), then it is Kuhn alone of the three who is able to do this theory justice.^[82] When Latour and Žižek discuss retroactivity, they are speaking of a relation between a human mind in the present and a past in the outer world that does not really precede the mind that knows it. For Kuhn, by contrast, retroactivity unfolds primarily on the plane of the object itself, split as it is between a “that” and a “what.” Although Kuhn seems to agree with Latour and Žižek that thought itself constitutes the prior structure of the world, for Kuhn this is not so much due to a struggle between thought and reality, as to one between two distinct sides of the object that thought confronts.

This brings us to a final important difference between Boyd and Kuhn. In one sense, Kuhn declares, “[b]oth of us are unregenerate realists.”^[83] In principle, that is, both authors believe in a world that exists outside the human mind. Yet Kuhn is more suspicious than his colleague about another aspect of traditional realism: the view that not only does nature exist apart from us, but that we are able to “carve it at the joints” with an eventual true theory. Kuhn describes Boyd as holding that “nature has one and only one set of joints to which the evolving terminology of science comes closer and closer with time.”^[84] This means that “it should be possible, if Boyd is right, to display some process of bracketing and zeroing in on nature’s real joints. But my strong impression is that [he] will not succeed.”^[85] Here, Kuhn is wearing his hat as a historian of science; as such, he is sensitive to what we earlier called the “different worlds” brought about by different scientific theories. Not only is there no evidence of a single set of joints in nature – given the intermixing of reference and meaning – but even if there were, the history of science shows no evidence of approaching it. Not only is a final true theory of reality impossible; we cannot even approach it asymptotically or “by degrees,” as conceived in different ways by the strange bedfellows Martin Heidegger and Alvin Plantinga.^[86] Instead, the history of science displays sudden reversals, as well as unexpected revivals of theories that were once left for dead, as with Leibniz’s relational space and time in the wake of Einstein’s General Relativity. Scientific reference is by no means safe from changes in meaning.

4 Conclusion: Meaning, Reference, and Retroactivity

We are now in a position to speak of the differing senses of retroactivity defended by Latour and Žižek on one side and by Kuhn on the other. A hardcore realist philosopher might argue that all three of these authors are idealists. After all, Latour and Žižek uphold the non-realist implications of retroactive projection of later discoveries onto the past, and Kuhn seems to do the same in denying that Newton and Einstein speak of the same world. Indeed, Kuhn might appear to be even more of an idealist than Žižek, since he might not go along with the latter’s claim in Less Than Nothing that nature does not change over time. But Kuhn’s major difference from Latour and Žižek lies elsewhere, and ironically, it is only possible because of his greater willingness than Žižek (in Less Than Nothing) to challenge the existence of a world independent of science. As we have seen, it is the difference between two moments in the scientific object known as the “that” and the “what,” which Kuhn alone raises as a topic.

Recall that for Latour, what we call a “thing” is really just a score list for “trials of strength.” It is not only that we know polonium exclusively through its victories in distinguishing itself from other chemicals, but that its very existence consists in these victories. There is no such thing as “polonium” over and above the ways in which its existence is brought into focus, just as there is no such thing as “tuberculosis bacillus” existing in Ancient Egypt. Whereas David Hume treated objects as nothing but “bundles of qualities,” Latour treats them as nothing but bundles of actions, which means that he truly deserves the title of “empiricist” that he so often sought.^[87] For the Lacan-inspired Žižek as well, we cannot speak of an object of desire as if it were a genuine object apart from the subject’s desiring of it. An object is, in fact, nothing more than a bundle of desires. For Kuhn, however, the scientific object is a distinct moment from whatever qualities we deem it to possess. In this way he makes the same tacit critique of Hume found in Edmund Husserl’s phenomenology, which draws a distinction between the object itself and the various shifting qualities it seems to have at any given moment (“adumbrations”).^[88] This is true despite Husserl’s denial of objects existing in themselves apart from any possible access by consciousness, and equally true of Kuhn despite his doubts as to whether the “object” or “reference” pole can exist apart from the “qualities” or “meaning” pole. But unlike Latour and Žižek, Kuhn at least aspires to be called a realist in the usual sense. How can we reconcile the entanglement of reference and meaning with the purported autonomous existence of reality apart from the mind?

The three philosophers just discussed not only could be accused of defending ontologically idealist positions: all of them have stood accused of it. Žižek’s statement in Less Than Nothing about the independence of the laws of nature from changes in human science is rather atypical of him. But if it represents a lasting change in his standpoint, it could go some way toward satisfying his realist critics. Latour, on the other hand, makes no concessions at all to any independent outside world in his article on Ramses and tuberculosis. Although he did make a clever attempt in Pandora’s Hope to show that his position is somehow more realistic than so-called realism, this was at best a debating trick. For all his dislike of Kantian philosophy, what Latour rejects in the Köngisbergian thinker is his affirmation of the thing-in-itself, not Kant’s contrary tendency to limit human knowledge to situations in which human thought is an ingredient. The latter, of course, is right up Latour’s alley. Yet Latour takes a more anti-realist line than Kant in denying that there is a single objective truth outside the mind, as seen in his years-long assault on the tacit opposition between “multiculturalism” and what he calls “mononaturalism.”^[89] Given the crucial role of science for Latour in co-constructing the very world it studies, he goes so far as to favor a “multinaturalism” of different worlds for different people, something that vanishingly few scientists could accept even in the post-quantum world.

In fact, the disciplinary source of the multinaturalist idea can be found in anthropology rather than physics, as in the increasingly influential work of Philippe Descola and Eduardo Viveiros de Castro: both of them personal friends and close intellectual colleagues of Latour.^[90] We have seen that Latour treats the past as always the result of a retroactive projection from the present, as in his discussions of tuberculosis and the dispute between Pasteur and Pouchet over spontaneous generation. Restated in terms of the analytic philosophy of language, this amounts to the anti-Kripkean claim that there is no reference without meaning. An actor is simply equivalent to everything that he/she/it does right now. There can be no question of an actor on one side and his/her/its actions on the other: as Hume might have put it, an actor is nothing more than a bundle of actions. But since Hume meant little for Latour’s career in any direct sense, we might point instead to Whitehead, a great influence on Latour who actually did owe a great deal to Hume.

As for Kuhn, we have seen that a certain uneasiness on the realism question pervades his work. This is especially clear from his Urbana-Champaign debate with Boyd, in which Kuhn willingly proclaimed his allegiance to realism, even while adding the Latour-like proviso that different scientific paradigms effectively amount to different worlds and not just different views on the same world. But Kuhn’s original contribution to the discussion lies elsewhere, in his distinction between the fact “that” something is and the question of “what” it is. The natural tendency of many readers will be to identify the “that” with the outside world of real existence and the “what” with the phenomenal realm in which the qualities of things become accessible to us. This is how the Polish philosopher Kasimierz Twardowski tried to make sense of his teacher Franz Brentano’s ambiguity as to whether intentional objects inside the mind correspond to anything outside the mind.^[91] For Twardowski, the term “object” played the role of Kuhn’s “that” and “content” did the same work as Kuhn’s “what.” Yet things become more intricate once we see that the “object/that” pair also functions immanently in consciousness, as was accomplished by Husserl when he immanentized the object, due to his denial that anything can exist in itself, inaccessible in principle to the mind.^[92]

What is the evidence that Kuhn is unwilling, in Twardowskian fashion, to link the moment of “that” too tightly with the world outside the mind? As we have seen, the evidence can be found in his vague sense that the Kripke/Putnam theory of names cannot apply equally both to natural kinds and scientific objects. Whereas the natural kind is a well-known “that,” the scientific object is quite often a mysterious one that holds together multiple vaguely understood effects: Kuhn’s example was “electric charge.” The importance of this distinction is as follows. When speaking of familiar natural kinds, such as “rabbit,” “tree,” or “carrot,” Kuhn is effectively pointing to a well-understood object that inhabits the immanent realm of human experience. In such cases, his “that” is something like Husserl’s intentional object or Maurice Merleau-Ponty’s “in-itself-for-us.”^[93] The same does not hold for Kuhn’s discussions of scientific objects, which he likens to the unsettled case of metaphor, where it is difficult to pin down exactly what the object’s identity and qualities may be. Despite his misgivings about any hasty identification of science with literature, Kuhn is clear-eyed about the metaphorical character of the scientific object as opposed to natural kinds.

But what is especially interesting about Kuhn is that he recognizes not only the metaphorical character of the scientific object, but also two separate versions of it. In most cases, we can think of metaphor as the ascription of palpable qualities to an object that somehow eludes our grasp. When reading Hamlet we are invited to link the object “world” with the qualities of a “stage” on which we are merely players; when in the laboratory with Kuhn, we are asked to link the whatness of a moving galvanometer needle with the thatness of the phrase “electric charge.”^[94] The historical example here is Röntgen, who began with the observation of strange effects before determining, over the course of several weeks in 1895, that he was dealing with a previously unknown “that” (the X-ray), for which no place existed in the physics of the time. But eventually, as the X-ray became a better-known object of scientific study, it was domesticated into a natural kind in Kuhn’s sense of the term. The initial gap between the existence or “thatness” of the X-ray and the “whatness” of its obtrusive properties was bridged by human familiarity. It became something like an Empricist bundle of qualities, or like what others have called a “dead metaphor.”^[95]

In Kuhn’s case we also encountered the opposite sort of example, in which rather than encountering mysterious effects (a screen glowing for no obvious reason), an object announces itself directly without it being clear what properties it possesses. This is what happened with oxygen, pursued by Priestley and Lavoisier without immediate success. Both chemists – unlike Röntgen – knew they were face to face with a “that,” but some years were needed to determine the full range of effects it could have. Once this became clearer, the “that” was given the name we all know today: oxygen. Despite the reverse order of the two cases, X-rays and oxygen both fall under the heading of what Kuhn calls scientific objects by contrast with natural kinds. In both cases, the object exists in the world, partly outside our grasp, even though with the passage of time it ceases to be a perturbing novel mystery and transforms into something more like a well-understood natural kind.

It may be worth asking about a different possible tension, one not covered by Kuhn, but there for the taking in his debate with Boyd. Namely, what if there were a tension in play not between a set of palpable qualities and an elusive underlying object, but between a familiar “natural kind” and some newly alarming mysterious properties that had never before been suspected? A pair of examples from Hollywood might be helpful. With Ridley Scott’s Alien, or the monoliths of Stanley Kubrick’s 2001: A Space Odyssey, we struggle to zero in on certain ominous novel objects that were not previous ingredients of our world. To this day, most viewers of the latter film remain confused by the significance of the monoliths, which may even send a chill down our spines when remembered. But how different things are with Alfred Hitchcock’s The Birds, in which familiar natural kinds – birds, our earthly neighbors since the dawn of the human species – take on disturbing new qualities: murderous aggression, unpredictable and intermittent assembly. For what could be more disturbing than when a natural kind cliché such as birds turns into the inverted form of a mysterious scientific object? Marshall McLuhan and Wilfrid Watson spoke in this connection of a shift “from cliché to archteype” in their book of the same title.^[96]

In order to grasp this unfamiliar case of a natural kind gone haywire with unforeseen qualities, we have just taken detours first through Hollywood and then through media theory. But this does not mean that the shift from cliché to archetype is of no significance to science itself. In fact, I suspect that Kuhn’s bread and butter concept, the “paradigm shift,” is best understood as an instance of it. Much scientific research does have to do with the discovery of new objects and the determination of their properties. But there is a sense in which the greatest scientific changes involve something closer to Hitchock’s birds. In Newton’s time, there was nothing unfamiliar about the orbit of the moon, or apples falling from trees. The strangeness of his insight arose, instead, from the unforeseen property shared by both situations: a universal attraction between all physical masses, whether on earth or in the heavens. With Einstein’s special relativity, the cliché object known as a clock suddenly took on the surprising feature of running at different speeds in different frames of reference: a fact in the news once again, in connection with the international need to put clocks on the moon.^[97] In connection with general relativity, public attention is often monopolized by black holes, first deduced by Karl Schwarzschild in the trenches of World War I as an unexpected byproduct of Einstein’s theory.^[98] But the heart of General Relativity’s paradigm shift lies closer to Einstein’s realization that three highly familiar phenomena – gravity, acceleration, and curvilinear motion – were all linked. Like the birds in Hitchock’s iconic film, it is often the most familiar objects that join in strange new combinations, thereby transforming science.