What are we to make of Galileo Galilei? A scientific hero whose revolutionary ideas were quashed by the institutional authority of the early 17th-century church? A natural philosopher who defended Copernicus’ mathematics and astronomy valiantly but was prone to vanity and arrogance? Or even, as Babette Babich reports that controversial philosopher of science Paul Feyerabend repeatedly asserted of Galileo, a “crook”?
It is important to understand in the first place that to ask this question is not to ask a scientific question – the sciences have absolutely no way of answering a question in this form. True, we could choose to reduce Galileo to his astronomical work and then make an assessment of his heliocentric model based on current data. But this would be grossly unfair to Galileo, for if we do this we’re forced to admit that his model is far from accurate, getting right mainly the placement of the sun at the center of the solar system, as Copernicus had already proposed. Galileo needed Kepler’s insight about elliptical orbits to get close to what we now understand as the cosmology of our solar system – without it, divining between the geocentric and heliocentric models was by no means a slam dunk with the evidence available at that time. Indeed, if we look just after the Galileo affair, we will find the astronomer Giovanni Batista Riccioli in 1651 publishing a list of 126 arguments regarding whether the Earth does in fact move, 49 of them in favor and 77 against.
How then can Galileo be enshrined as a scientific hero of any kind? The question is not a trivial one, and opens the door to extremely important and timely questions about scientific practice that matter even more today than in Galileo’s time. What we cannot legitimately conclude without acting prematurely is that since Galileo supported one fact we accept today as scientifically justified – the Earth moves around the sun – he is automatically a heroic figure. On the contrary, the basis of the heroism being asserted here gains its context from the fact the Galileo opposed institutional authority in his time – which means to truly address such a question today is primarily a historical investigation, and also a philosophical one, since a judgment of heroism is a moral judgment rather than a matter of simple fact.
To answer the question ‘What are we to make of Galileo?’ we must therefore commit to much more than a ‘fact check.’ We must undertake a detailed investigation that is not, in neither form nor content, scientific in nature, for all its deep connections with astronomy. What I wish to do in this discussion, however, is not perform that specific investigation (several books already cover this well) but rather to raise a question about contemporary scientific practice against the backdrop of this ambiguity over whether Galileo is to be seen as a hero or a crook. For the matter of the modes of scientific practice and their tensions with institutional authority are acutely relevant to the crisis of knowledge we face today epitomized by the accusation of ‘fake news.’ And in this regard, we have much more to gain from pondering Galileo than settling the status of a mere astronomical fact.
Three Propositions Concerning Scientific Knowledge
Despite our widespread commitment to scientific discovery, the vast majority of us are quite unprepared for dealing with the complexity of authentic scientific problems. This happens in part because of the faith we possess in the work of the sciences to solve problems. Having witnessed technology utterly transform our planet over the last century we afford to the sciences a tremendous power, one that is not unjustified but which is also highly problematic, in ways that greatly exceed the scope of this particular discussion. Because of our collective faith in scientific research, many of us have come to expect that:
An answer can always be provided by scientific means
A single successful experiment can provide clear answers to our questions
Scientific theories have emerged from such successful experiments
It is no wonder we think like this; we’ve been telling this story since at least the 19th century when an argument between Samuel Taylor Coleridge and William Whewell gave us the term ‘scientist,’ if not perhaps earlier, say, since Boyle’s vacuum pump offered the tantalizing possibility of resolving questions of truth in the laboratory.
Yet all three propositions above are false.
It is this schizophrenic clash between our faith in scientific methods and the unseen yet immense complexities we thus tend to ignore that lies at the heart of the key question we must ask about contemporary scientific research. Once we step beyond merely believing and begin to understand that the work of the sciences is much more fragile than we tend to expect, we may come to recognize that the institutional power that oppressed Galileo is as much a threat to assembling a true picture today as it was in the 17th century.
Not All Questions Can Be Answered Scientifically
This is perhaps the single greatest misunderstanding about the sciences – not every question can be answered by these methods. This is not even one of those points of caution that is superseded by future advances in technique (“in the future, we can answer this, but not now…”). Rather, we must distinguish between questions suitable for answering by scientific methods, questions suitable for answering by other methods, and questions that do not lend themselves to being answered at all.
I foreshadowed this point with the opening question about Galileo – a quintessential example of a problem requiring a historical investigation. The late Mary Midgley was always keen to point to historical methods as an example of questions that can be answered, but in ways that were not in principle scientific. When we want to establish the facts of a prior event, we must make use of all the available evidence, study all the surviving written accounts, and then use deductive reasoning to draw conclusions (often provisionally). Scientific techniques sometimes contribute to this process – if you find a corpse in a bog, carbon dating will get you a time frame, for instance. But these contributions to any given historical puzzle are typically quite minor. What is paramount is a capacity to bring together all the evidence along with our understanding of human life and culture at the relevant place and time. We deduce historical answers through the methods of the detective. That these include scientific evidence, or that other sciences also use deductive reasoning isn’t enough to allow history to be swallowed up by the sciences. On the contrary, these different methods are distinct – and as such, can learn from each other.
As with the historical aspects of the question of Galileo, so with the moral dimensions of the issue – hero versus crook, after all, is more than a simple question of ‘fact checking.’ It requires an understanding of what we mean by heroism, or what justifies the accusation implied in being a crook. Moral or ethical issues belong to the domain of philosophy, but we should not assume from this that philosophers have authority over them – indeed, there is supposed to be no singular source of institutional authority over such matters today, since we are all (quite unlike those living in Galileo’s time) entitled to make our own moral judgments, another point that Midgley was keen to stress.
Much as we hate to admit it, there are also some questions that simply don’t have definitive answers. The very concept of metaphysics is to mark questions beyond (meta) physics i.e. subjects without certain answers. Traditionally, this topic has revolved around theology, but there are also vast landscapes of untestable postulates in ethics, politics, gender, and more besides. That’s not to say mistakes around these issues don’t cause people to erroneously assume that the sciences can muscle in – it happens all the time. It’s rather unsurprising, since it’s easy to confuse the importance of gathering evidence (where experience in a scientific field is usually essential) with the separate process of evaluating it (where non-scientific competences can have just as much bearing).
The reason we value scientific methods for answering some of the tough questions is precisely because where they can be brought to bear, the methods of the sciences can crack some major mysteries wide open. But ‘some’ is the word that gets overlooked in this regard. The destiny of the sciences is not total knowledge of everything but an ever-adapting set of frameworks for understanding the world around us. It is far from clear that we should assume an end point for the scientific adventure – unless, alas, it is human extinction. Rather, a great deal of what we want the scientific community to investigate are questions that relate to what we happen to be doing now, and these will not hold the same salience in the future. The parallax of stars and their apparent sizes is no longer of interest to contemporary astronomers even though it was of vital importance when comparing the differing predictions made by geocentric or heliocentric cosmologies in Galileo’s day. We misunderstand the nature of knowledge production entirely when we imagine a simple kind of ratcheted progress, new discoveries adding to an ever-growing pile of knowledge. On the contrary, the vast majority of all scientific work is destined for immense and eternal obscurity, since it depends for its significance entirely upon the circumstances of its commission.
It is not because the sciences can answer all questions that we esteem their achievements. Rather, it is because when a topic is amenable to scientific study we have a hope of definite answers that are denied to us in most aspects of life. But this yearning for certainty is both a powerful motivating force and an immense liability when it comes to trusting experiments to answer questions for us…
Singular Experiments Reveal Almost Nothing
We’ve all seen those movies where, after a laborious research montage, the scientist finally has a breakthrough and achieves the MacGuffin the heroes desperately need. This is the heroic legend of scientific research epitomized in The Flaming Lips song, Race For The Prize, and it is just as active in our mythology of Galileo as anywhere else. We love to say that Galileo built a telescope, saw that the Earth revolves around the sun, and discovered the truth. But he didn’t do anything of the kind, and the telescope was not even an appropriate instrument to settle that particular argument. Rather, it was Foucault’s pendulum that was to have the pivotal role – and even that it could not have done were it not for the groundwork laid by Ibn al-Shatir, Copernicus, Galileo, and many more besides.
One of the reasons we have adopted this kind of mythic rendering of scientific work is that our way of telling the stories of famous researchers is to repackage their lives to make them into glorious lone heroes for truth, often and especially against a closed-minded dogmatism attributed to religion or government. Since the early 20th century, Galileo has been the poster child for this. Bertolt Brecht’s 1938 play Life of Galileo may have accelerated the adoption of this narrative, although Brecht’s Galileo says much in the service of its author’s philosophy that would have been vile to Galileo himself. Arguably, his fight with the church authorities was closer to the 17th century equivalent of a nerd flame war (and displaying the same degree of ill-judged social awkwardness as that analogy implies) than anything heroic, although the stakes (pun intended) were certainly far higher. Yet whatever the facts of the matter, we love to recognize that Galileo believed what we now believe and thus paint him as a scientific hero, even if his methods were in places rather slapdash. Our desire to glorify scientists is largely about basking in our own smugness anyway: we are right, and we know we are right because we side with science and it is always right.
Notwithstanding the tired rhetoric, singular experiments are nonetheless only decisive in any genuine sense when contrasting explanations for some phenomenon can be tested in a particular apparatus or protocol. Inevitably, therefore, numerous experiments and theories precede the crucial test case in order to provide the scaffolding that made that test worth undertaking. Theories do not come from nowhere. Galileo’s certainty in 1632 that the sun was the center of the solar system came from Copernicus’ diligent mathematics in 1543, which in turn seems to have drawn substantially from Ibn al-Shatir’s pioneering astronomical work in 1375, and so on. But we make Galileo the hero, even though (or especially because) he behaved like an arrogant blowhard. In that sense, he is justifiably called the Father of Modern Science. But it is the Muslim timekeeper from Damascus two and a half centuries earlier who performed the key astronomical observations, and who first produced a theoretical model of the solar system open to empirical verification. In this tale, as with so many in the history of the sciences, the hero is chosen not for the rigor of their methods, but for largely political reasons.
In terms of contemporary research, the exemplar of the illegitimacy of the singular experiment as a carrier of truth is psychology, which after half a century of half-baked theory generation on the back of singular research studies is finally paying the piper for its collective hubris. Instead of working out the complex questions that followed from psychology’s strongest theoretical apparatus, such as Leon Festinger’s cognitive dissonance or Paul Ekman’s basic emotions, a surprising number of 20th century psychologists acted as if there was a gold rush in social experimentation, coming up with hundreds of under-researched, under-challenged, under-developed theories because they placed too much trust in singular experiments. In fact, the measure of the power and efficacy of any scientific field is arguably how well research problems are distributed over a wider community – and psychologists can nail this challenge when they choose to cooperate. But whenever a research gold rush kicks off (as it did in a less problematic way in 19th century paleontology once dinosaur fossils were recognized as significant), everyone is vainly pursuing the glory of putting their personal stamp on a finding. In this regard, the Nobel Prize might not have been beneficial to the sciences.
The power of scientific research always flows from the cybernetic enhancement of individual insight by a wider network of contributors – more experiments, more observations, more critical reflections on how it all fits together. This, in turn, is why it is so disastrous to censor and exclude those scientists with alternate perspectives – the networks become compromised, and the effectiveness of research necessarily falls. If we want to learn a lesson from Galileo, ought it not be that exclusion and censorship are antithetical to scientific investigation? These are the tools of institutional power, represented in the 17th century by the papacy, and today by strange alliances between government agencies and media corporations.
Effective Theories Explain All the Evidence
There is a simple reason that singular experiments are inadequate; namely that it takes more than one source of data to construct a theory worth trusting. But equally, a single theory is also inadequate in the vast majority of real world contexts, because understanding complex phenomena seldom involves circumstances as simplistic as any singular model entails.
Take for instance the hypothesis that a respiratory virus could be stopped in its tracks by asking people to remain in their homes, a prediction sometimes attributed to germ theory. But germ theory only says that certain diseases are spread by infectious agents (germs); it is radically insufficient to predict what happens when nations reorganize their infrastructure to favor something rather akin to universal house arrest. The very proposal itself raises questions that go vastly beyond the scope of germ theory, questions about the specifics of the germs in question (which for a new disease are always ambiguous), questions about where the greatest risks of infection might be found when our evidence is fragmentary and dominated by observational case studies, or questions about which members of society will have to keep moving between houses regardless of the prevailing order. Plumbers and electricians, for instance, cannot work remotely, thus exposing them to greater risks of exposure via homes under a national lockdown than computer programmers or proof readers, whose risks of exposure are not necessarily reduced in such an arrangement either, since if they rarely go out to socialize, ordering them to be housebound has little effect. It simply cannot be known in advance how an unprecedented social arrangement will work in practice – the total collection of theoretical knowledge required to make even the simplest of predictions in this regard is woefully inadequate to the magnitude of the challenge. Sending a small capsule to the moon is child’s play in comparison.
Indeed, this quagmire of situational ambiguities is why those prudent-yet-doomed calls for cost-benefit analyses on national lockdowns could not possibly have been met at the outset even though they are ubiquitous in all other political contexts. We did not make a scientific decision to pursue these unprecedented interventions, we took a bizarre leap of faith. Defending that leap of faith with germ theory is a way of papering over the extent of our ignorance about what we thought we were doing, other than following our vehement intuition that there was something we ought to be doing. The parallel with Galileo is rather too close to ignore, even if his own house arrest was not on health grounds. Once we decide we’re right, the scientific problems are easily brushed aside. We can settle the books later. Once we decide we’re right, we can no more admit that we made a mistake than Galileo, who similarly mistook a singular truth (the Earth revolves around the sun) for a blank check, and so behaved recklessly by writing a book mocking the very pope who had up until that point been defending him from his detractors. Such consequences are much more foreseeable than we sometimes like to admit.
Thus we find the Centers for Disease Control pugnaciously defending the efficacy of community masking on the basis of an impressive range of evidence – lab studies, population studies of mask mandates, and so forth. But notably, like the limitations of Galileo’s telescopic observations that he had to hand-wave away, the CDC did not report (and has never adequately responded to) the most important experiment in connection to this issue, the DANMASK-19 random controlled trial (RCT), the largest and most robust trial of its kind, which showed no statistically significant effect, which not coincidentally was also what was reported by every RCT conducted on face masks against respiratory viruses up to 2019, a point that has been made by various researchers ever since community masking was first proposed.
The CDC had a great hypothesis – that we could hinder the spread of SARS-CoV-2 with community deployment of face masks. It was plausible. It had evidence in its favor. But this particular theory could not absorb all the evidence. Indeed, it could not even absorb all the evidence they had chosen to single out, since the effects detected in mask mandate studies only seem to have occurred when the mandates were first enacted (presumably attributable to abrupt changes in footfall). If this were not the case, we would see enormous disparity in infection rates based on where community masking was or was not deployed. The very computer models used hypothetically to justify these interventions at the start now make it perfectly clear that they cannot possibly have worked as intended or else the infection rate data would show enormous disparity between places that did or did not have widespread masking. Rather, we see the same general shape of the infection curves everywhere, irrespective of which interventions were deployed. Apart from early border closures, which kicked the can down the road by a year, we only slowed or accelerated the virus by a few months at best regardless of our chosen mitigations. Some countries may indeed have saved lives by doing so, but many that chose to aim for slowing down have inadvertently inflicted major health harms just from monopolizing medical support networks to tackle merely one cause of death, even without considering the mental health harms and economic damage incurred.
Other scientists had other theories, of course including the theory that the face masks could not possibly work because aerosols were too small to be blocked by most face masks, and these theories too faced the very same challenge – could they absorb all the evidence? It is only now, in the latter half of 2021, that I am finally seeing constellations of theoretical frameworks capable of explaining all the evidence that needs to be accounted for, such as low rates of infection in healthcare staff with certain combinations of protective equipment, and the remarkably similar shape of all infection curves. Yet these newer understandings necessarily involve multiple theories. They have to because multiple phenomena had to be explained, including the transmission dynamics of the virus (which was only even broadly settled as aerosols rather than micro-droplets earlier this year), and the relationships between collective immunity and new viral variants (which has only just started to take clear theoretical shape in recent months). The scientific community has learned an enormous amount about respiratory viruses in the last year and a half, so much indeed that those of us who kept reading the research without reaching a firm conclusion tend now to see the prematurely authorized 2020 interventions as all too akin to asserting that the Earth doesn’t move.
It is not that the CDC failed to act as scientists. They were very much acting in the tradition of Galileo’s modern science. But this is already a very old-fashioned form of science. And to make matters worse, the CDC had the opposite misfortune to Galileo in that the contemporary equivalent to the 17th-century church – the Trusted News Initiative championed by Google, Facebook, Twitter, BBC, Reuters, the Associated Press and others – have censored and blocked the counter-discussions in public, thus enormously reducing the scientific power of the CDC at the very time that it needed this assistance. You don’t help scientists learn the truth by shielding them from counterarguments – you largely destroy the most important source of that vital pressure to keep searching for the truth.
It is precisely because effective theories must be able to absorb all the evidence that attempts to resolve questions about lockdowns or community masking in 2020 were hopeless, and authoritatively enforcing interpretations that were constructed quite early in the research process (while evidence was still scant) is impossible to defend as good scientific practice. Rather, we politicized these topics so much and so early that there was almost no possibility of adequate theory construction from that point onwards. Against Galileo, institutional authority delayed the resolution of the knowledge on trial, and put Galileo’s life and freedom at risk. Today, we astonishingly find ourselves back in the very same situation, but with everyone’s lives and freedom exposed to terrible risks.
Abandoning Modern Science?
We sometimes forget that the modern era began roughly five centuries ago – using ‘modern’ as a synonym for ‘now’ has frustratingly obscured just how dated the modern paradigm for science has become! It is not a timeless method, nor even a method as such, but rather a metaphysical understanding of the scientific research processes. We may urgently need a replacement better suited to the conditions of research we face today, which are not at all like Galileo’s valiant attempt to justify heliocentrism without having quite the correct model to do so.
While I cannot say if we are up to the task of such a shift in our understanding of the work of the sciences, the last two years have clearly shown the immense danger of remaining caught in archaic conceptions of scientific process. We need an understanding of scientific work that delays conclusions, rather than rushing rashly towards early best guesses. The chemist and philosopher of science Isabelle Stengers has argued passionately for our need for slow science, a process that resists the artificial pressure for rapid conclusions accelerated by patent races and research grant applications. But her case can be made even more robust, since what Stengers calls slow science is also strong science. If we need answers quickly, we have to accept that the sciences are often the wrong tool for the job, since they specialize in the gradual and careful construction of answers to the kinds of questions each scientific field is best suited to, which takes time and – crucially! – disagreements.
This is why politics – where answers are demanded immediately and certainty is expected or required – is the conceptual enemy of the sciences, which are always impeded and distorted by trying to provide definitive answers too swiftly. Politicians make terrible scientists and scientists make terrible politicians – at least in the political climate today, although we might draw the same conclusion from Galileo, whose political sensibilities were clearly rather poor, even if his intuition for good theories was admirable. Skipping the arguments may be politically expedient but it is simply not helpful to any research project. Worse, to prevent disagreements is to gut the scientific process of its one immortal method, the one it inherited from Aristotle and the other natural philosophers of both the East and the West, that of discourse itself.
What makes Galileo a scientific hero cannot be that he ‘got it right.’ For a start, he didn’t get it right. Rather, he had an intuition about the true state of affairs and then argued passionately for it, using both the evidence that others had gathered and the observations he himself produced with his telescope and his mathematics. That willingness to both do the work and to have the discussion – to not only research but also to debate and persuade – ought to be central to what we might deem meritorious in Galileo’s story if our purpose is to contrast him to an institutional authority stifling free discussion.
But there is another lesson we might take from Galileo. As his critics point out, he was vain and arrogant, and lacking in humility – behavioral qualities still rather common among those drawn to the sciences, even today. If he had been more open to cooperation, he might have engaged more productively with the theories of his Protestant contemporary, Johannes Kepler, who already had answers to key problems in Galileo’s cosmology. Kepler had correctly realized the moon was responsible for tidal effects on Earth; Galileo believed the sun was responsible, and dismissed the moon hypothesis as “childish” and “occult.” But Kepler was right, not only about this but also about the need to use ellipses, not circles, to correctly model the motions of the planets. If Kepler and Galileo had been able to cooperate more directly (although there were some rather lovely fanboyish letters exchanged between them!), the case for heliocentrism might have been made far more successfully in their time – and while there is still the problem of resistance from the institutional authorities to overcome, the discussion could have been so very different if it had not been reduced to individuals working in isolation.
Thus we find ourselves having come full circle today, where once again we have scientific problems (the handling of a serious respiratory pandemic), and once again institutional authority (health agencies like the CDC backed by media censorship) provides an aggravating barrier to understanding, and politicizes research to the point of abject nonsense in the process. Then as now, cooperation between researchers is the optimal way to leverage all the skills and knowledge available. It is precisely this cybernetic enhancement of our individual powers that can make the sciences today so much more effective than in Galileo’s time. At least, they are when we do not block productive cooperation by censoring disagreements and excluding the most important objections from the debate.
We can defend the actions of Google, Facebook, the BBC, and the rest of the Trusted News Initiative, I suppose, by claiming good motives. After all, the whopping lies coming out of the Trump White House were their stated reason for taking a stand against disinformation in the first place. But in the sciences, we simply do not know what is or is not disinformation until we have had the full and open discourse in order to see which theories can absorb all of the evidence. Otherwise, we are back to the methods of the 17th-century church – who, Paul Feyerabend would impishly have reminded us, also had motives that were from their perspective noble: to save immortal souls from eternal torment. If we do not wish to condemn contemporary scientific work to a new dark age of politically-distorted nonsensicality, we might well consider whether the question of Galileo is less about the astronomer and more about the scientific damage that can be wrought by well-intentioned censorship.