The Integral Nature of the Scientific Enterprise

1. Introduction

Today there is a crisis of faith. Experts are not trusted. Climate change is denied. Conspiracy theories abound. How should the scientific community respond? What is it about the nature of the scientific enterprise that makes it trustworthy (and how can trustworthiness be promoted?)?

What is the scientific enterprise? Of course, from one point of view this question is trivial – everyone knows what it is: it’s what’s made the modern world! And of course, the scientific enterprise is generally celebrated for its unprecedented theoretical, empirical and technological achievements over the past few centuries and the concomitant benefits that have accrued to humanity.

The scientific enterprise is a communal activity designed to obtain reliable knowledge of the natural world. We depend on each other’s expertise (since most of what we know we cannot verify for ourselves): we also depend on the general good faith of the community since fraud is hard to detect reliably, even though it is quite rare. That is, the scientific enterprise depends heavily on the personal integrity of the scientists. And wider society must be broadly convinced of the integrity of the scientific community to believe that it is being told the truth.

But what is knowledge? How do we know things?

The epistemic problem (how we know things) has been intensively discussed by philosophers¹, and the emerging consensus is that it has no analytical solution: we limp inexorably from error to error. Nevertheless, our knowledge of the natural world has proved reliable, as witnessed by the ubiquitous artefacts on which we rely. The reason for this reliability is that, just as the scientific enterprise at the communal level is built on integrity, so at the personal level it is built on the search for unity.

Truth is one. For were it not, we could know nothing! Therefore we assert the integral nature of the scientific enterprise²: that knowledge is premissed on an account of unity. For we may know truly even in the face of our limitations, since the truth the poets discern is the same sort of thing that the scientists discover – truth is one. Knowledge is personal³ and any philosophy of science must be properly metaphysical (and here by “metaphysics” we mean simply “the metanarrative of physics”)⁴.

Note that when we say “Knowledge is Personal” we do not imply that knowledge is individual: we know things together, or not at all! “No man is an island” (John Donne, 1624). Since individually we can know rather little, knowledge must be a property of the community, and scientific knowledge a property of the scientific community.

Moreover, even though all practitioners will affirm that intuition is essential to scientists to enable them not only to articulate but also even to obtain their results (for example, see the penetrating discussion of Tom McLeish, 2019), it is not considered proper to mention this in formal presentations.⁵

The new account we propose here of the scientific enterprise (premissed on a new account of unity) has not been possible until rather recently. After all, “unity” is a metaphysical idea that historically played no direct part in physics (since the scientific revolution), and one can hardly claim that one’s account of the “scientific enterprise” is satisfactory unless one can also account for how the physics is known.

But it turns out that “unity” does have a physical embodiment just as “atoms” do. The reality of atoms has only been universally accepted for just over a century, and real “unitary entities” were only established in 2022⁶. We will explain here how this and other advances in physics underpin a new account of the scientific enterprise that is holistic rather than reductionist, and allows us to speak both directly of technical unity and also (more indirectly) of moral integrity. This is a new sort of explanation that takes seriously both the physics and the metaphysics, both the analytics considered proper in scientific discourse and the poetics that are actually indispensable for any appreciation of the science.

2. Synopsis

This essay is structured to address the question of the nature of the scientific enterprise in a novel way that draws appropriate attention to the close philosophical relevance of recent developments in the physics of thermodynamics. The starting point (§3 below: “Historical Perspective …”) is Carlo Rovelli’s proposal of his “Third Way”: midway between a dogmatic acceptance of the master’s infallible pronouncements, and the frequently-found default dogmatic position of many an intelligent and inquiring (yet opinionated) person that “My view must be correct”. We show here that Rovelli’s is a false resolution of this dichotomy, but that there does exist a valid way to understand the data; and moreover that this way underlines the unity (internal integrity) of philosophical thought, whether one is a poet or a physicist.

The obvious question (§4: “Trusting Scientists”) is “Where does knowledge come from?” together with the important sequel “Who decides?” We expect experts to know stuff and that we are supposed to trust them to know best – except that manifestly they sometimes don’t, even when it matters most. And yet we have to make do with the best knowledge we have. The scientists facing the public must make an honest use of rhetorical methods, not only to counter the falsehoods of the demagogues but also because science is a human activity in which metaphysical issues are unavoidable. It is no accident that professorships in the public understanding of science have emerged in recent decades, peopled by experts with known abilities in the rhetorical arts: today’s public poets of science.

It is after all truth that matters (§5: “Realism …”), and if science is not the honest pursuit of truth then it doesn’t matter what else it is. Science is often presented as being a collection of “objective” hard-edged facts (§6: “The Illusion of Objectivity”) but this is not merely false but actually illusory. Rather, it is the search for a “unified” (coherent) view of the world (§7: “On Unity”). What is fundamental in this (§8: “What is Fundamental?”)? It turns out that a very good physical case can be made for saying that “unitary entities” are more fundamental than “fundamental particles”. That is, the programme of reductionism that has ruled since “atoms” were again seriously posited in the 17th century has reached its natural limit: it turns out that physics is now developing a powerful way of speaking about holistic approaches (also implying cognates such as wholeness and integrity).

So how legitimate is the scientific enterprise (§9: “Legitimacy …”), given that “mere facts” don’t actually prove anything at all, and we sometimes even get the “facts” themselves wrong? It turns out that the legitimacy of the scientific consensus must be established by persuading the public of its substantial truth. Legitimacy can be established only by persuasion. Science touches reality, so that the scientific truths we have grasped are also real however partial (or even fallible) they turn out to be. After all, the widgets continue to work!

As an aside, this position of insisting on the one hand of the reality of the world, and on the other that “knowledge is personal” (Polanyi, 1958) is indistinguishable from the position the quantum physicists have taken on the philosophical crises highlighted by the paradoxes of their discipline (§10: “QBism”). QBism (Quantum Bayesianism) directs our attention to basic ideas on how we think of probabilities. It turns out that the Bayesian approach to probability applies in a scale-independent way: it is not limited in any way to quantum mechanics but also applies at the cosmic scale (for which general relativity is needed). In any case, it seems that probabilities are also physical quantities obeying physical laws (Parker et al. 2025b).

One way of summarising our argument here is to draw attention to the poetics of physics (§11: “Poetry …”). A scientific argument can never be entirely self-contained (logically “complete”): it is always in some social or philosophical context whose importance may be (and sometimes will be) dominating. We then briefly discuss (§§12,13) various other issues (including some relevant aspects of Artificial Intelligence and its impact on the conduct of science) and conclude (§§14,15).

3. A Historical Perspective on Scientific Enquiry: Rovelli’s “Third Way”

Carlo Rovelli (2009, p.78) says, “Halfway between the absolute reverence of [the student for the master] and the rejection of those who hold different views, Anaximander [6th century BCE] discovered a third way … he does not hesitate to say that Thales [of Miletus] is mistaken about this or that matter, or that it is possible to do better … this narrow third way is the most extraordinary key for the development of knowledge.” Rovelli characterises the “religious” approach to knowledge by its “absolute reverence” which does not allow any modification of the master’s view, contrasting it with the “scientific” approach in which the follower is (sometimes) willing to say that the master was mistaken in this or that respect. Clearly, Rovelli’s representation of science is well-founded.

However, it seems to us that his representation of Christian attitudes is less so, even if he is correct to point to “a classic thesis” (by Geoffrey Lloyd, 2002) according to which “a scientific revolution comparable to the one in the West did not take place in Chinese civilization – despite the fact that for centuries Chinese civilization was in many ways broadly superior to the West – precisely because the master in Chinese culture was never criticised or questioned” (Rovelli op. cit. p.81). Rovelli contrasts the Pope’s view of infallibility⁷ with a scientific view which requires “faith in human beings, their being reasonable, and their honesty in searching for the truth. This kind of faith in human beings is that of the luminous humanism of the Greek cities in the 6th century BCE at the root of the extraordinary intellectual and cultural flowering of the following centuries, which continues to bear fruit in the contemporary world. But this faith in humanity is not unchallenged. Many voices rise against it: “Cursed be the man that trusteth in man …” (Jeremiah 17:5). The conflict between these two worldviews is ancient” (ibid. p.141f). It seems clear to us that Rovelli does not grasp Jeremiah’s point (see below), and we should also note that Nicholas Spencer (2023)⁸ has carefully exploded the myth that the “scientific” and the “religious” worldviews are in “conflict”. The real position is much more nuanced than Rovelli says.

The very idea of “infallibility” is heavily contested in Christianity.⁹ Indeed, infallible propositions are simply not available, neither for theologians nor for scientists! Catherine Elgin (2017, p.33) has made a careful epistemological case for the proposition that “scientific understanding is non-factive” (that is, it is “true enough”). In fact, she contends, entirely convincingly, that “the capacity to make mistakes is an epistemic achievement … far from being a concession to frailty, [fallibility] is a source of strength” (p.291).

The extract from Jeremiah that Rovelli quotes actually means the opposite of what Rovelli claims. Since Jeremiah of Jerusalem was a contemporary of Thales of Miletus, Rovelli’s reference to him in a book on Anaximander (Thales’ younger contemporary) is well-judged. But the Hebrew scholars are unusually unanimous in pointing out that the four verses (Jer.17:5-8) are a poetic unity. Cassandra Gorman (2021) has shown that the “metaphysical poets” in 17th century England were fascinated by the revelations of the new microscopes¹⁰, which made the commonplace seem outlandishly novel. She says that these poets “wrote poetry that meditated on the liberating power of atoms to dissolve and recongregate into renewed and resurrected forms. This poetic history of the irreducible particle has not been written by historians of science” (ibid. p.22). The point is that these poets explicitly recognised poems as “atoms” (that is, indivisible: their integrity must remain inviolate – each word is indispensible). Jeremiah’s poem should also be recognised as indivisible, and we therefore give it as a whole. It can be translated (we follow the authoritative translation of William McKane, 1986):

• Cursed is the man who trusts in men
  • Who leans for support on material power
  • Whose mind is alienated from Yahweh¹¹
• He is like one living destitute in the wilderness
  • He is unaware when good comes
  • He lives in parched places of the desert
  • In an uninhabited salt waste
• Blessed is the man who trusts in Yahweh
  • Who makes Yahweh his source of safety
• He is like a tree planted beside waters
  • Stretching out its roots along a stream
  • It has no fear when heat comes
  • Its leaves remain green
  • In a year of drought it has no anxiety
  • And does not cease to bear fruit Jeremiah 17:5-8

The parallelism between the first and second halves of the poem is obvious even in translation: the Psalmist quotes the poem at length (Psalm 1) and Jesus alludes to it (see John 4:14, which actually quotes the related passages Jer.2:13;17:13). Jeremiah is deprecating reliance on external things (power in this case) and lauding internal honesty in just the way Rovelli would approve: in fact Jeremiah goes on to be explicit (in v.10), “I the LORD search the heart and examine the mind”, where Biblically the mind is located in the heart¹². Whether or not one believes in God, anyone of goodwill can approve Jeremiah’s point, mutatis mutandis.

It seems clear that Rovelli’s characterisation of religious rigidity – putting dogma before reason – is false, at least for Christianity, accepting of course that European history is littered with shocking horrors perpetrated by the powerful (who chose to call themselves Christian). Luther and Calvin specifically attacked the doctrine of the Pope in just the same way that Copernicus attacked the doctrine of Ptolemy (and just as respectfully¹³). Theology must change just as science must, and for the same reasons – both must be open to new understandings. But both are also subject to the same unwelcome rigidities of thought¹⁴, again for the same reasons: it is difficult to get people to change the way they think, and even more difficult to get people to “keep their oath, even when it hurts”, as the Psalmist demands (Ps.15:4). Even though it can lead to intellectual discomfort, the pursuit of truth (with Rovelli’s admonition “it is possible to do better”¹⁵ in mind) has acted as the driver to the increase of knowledge.

Rovelli’s position appears to be ‘scientistic’ (having an exaggerated trust in the methods of science as applied solely to a “narrative” set of observations) which in this case means that he ignores (or misrepresents) the metaphysical (“metanarrative”) aspects we are drawing attention to.

4. Trusting Scientists

People (including “experts”) are rather good at getting things wrong. Hanna Metzen (2024) discusses in detail the related idea of trust as it is used in the literature, pointing out that “Epistemic trust in science has become an important issue in philosophy of science” – and, one should add, also critically important today in the politics of both Europe and the Americas. Metzen does not discuss the mutual trust present in the scientific community itself that enables new results to be approved and absorbed: it is obvious that the peer review process (for example, and on which see Tennant & Ross-Hellauer, 2020, for a valuable recent assessment) depends on the good faith of all parties. In particular, articles in the physical sciences usually report and interpret the results of experiments, and it is essential to know who did the experiments¹⁶. By publishing, the authors assert that they are telling the truth about the reported observations. It is remarkably hard and usually unfeasible to detect bad faith. Generally, I have to trust you to tell truth. Not only Truth¹⁷, but also Trust is central in this enterprise. John Hardwig (1991) says, “scientific knowledge rests on trust … in the character of scientists”. And this is the case even though bias is ubiquitous (recently extensively documented in its multiple forms by Jessica Nordell, 2021).

With the explosion of knowledge and the ubiquity today of multi-author papers¹⁸, we need effective collaboration more than ever (with the extra risk of error that implies). In collaborations the authors must all trust each other (as John Hardwig insists, 1985, 1991), even as they also must agree a common narrative¹⁹ (with the concomitant risk of groupthink). In fact, we need to “rethink expertise” (Collins & Evans, 2007). And when the wider public pick up the results, they need to be able to trust them too. How does this work?²⁰

Again, it seems that good faith is indispensable. I don’t have the expertise (or the time) to check for myself everything you say. Even were it feasible (which it is not) it would not be practical: life isn’t long enough. I can (and should) check that what you are doing is reasonable, that it conforms to your promises and also conforms to current standards (and so on), but it’s not possible to check everything. This is part of what Carlo Rovelli means when he says (as we mentioned above) “When we seek a sure foundation on which to base decisions about our actions and thoughts we find that a sure foundation does not exist” (although he was thinking mainly of personal epistemics it is true at all levels, both of individual thought and of communal effort). It might be “reasonable to believe” this or that, but in the end we can never be entirely sure that we have got it right. Where then is the truth?

Rovelli is right – we simply have to live with the possibility of being wrong. This is admittedly rather frightening. But it also allows us to participate in the “continuous critical revision of accepted knowledge … [with] the ability to explore new images of the world” (2009 p.181). This participation is in our own right, provided that we can make use of the expertise of the community that allows us to understand things that must have remained mysterious had we been alone.

Joseph Drew and others have directly addressed the issue of how to persuade the public to approve certain political actions. They draw attention to the effective practice of rhetorical methods, expanding on Aristotle’s categories of logos, ethos, pathos and kairos²¹. The same sorts of things apply to the scientific community combatting antiscientism: it is simply not sufficient to make a scientific (logical) argument! Scientists must learn to handle rhetorical methods effectively, and this is not merely a political necessity: it is also an acknowledgement of the metaphysical context of all scientific activity.

We see that science (as pursued by the individual scientist) is dependent on the scientific community, which in turn is dependent on the wider community (as the public funder, lay cheer leader, and ultimate beneficiary of the outcomes of new scientific knowledge). Without the "expertise of the scientific community" scientists are on their own and understand rather little. But if they are willing to "stand on the shoulders of giants" then they can see further. Conversely, the scientific community trusts the individual - that's how peer review works.

This is also why “trust” is a political thing: the individual and social cannot be disentangled, and the public must be persuaded to continue to support the enterprise just as the scientific community must be persuaded to accept the results of the individual²². Clearly, this phenomenon of entanglement is observed not only quantitatively in quantum systems, but qualitatively in social ones too. And this entanglement “metaphor” is more realistic than one might have thought: new physics (Parker & Jeynes 2023a) has shown that the local and non-local are commensurate, as are also the reversible and the irreversible, and the causal and the acausal (see below §6).

5. Realism and the Metaphysical Existence of Truth

What is real? How do we know? The first question is ontic, the second is epistemic. We offer a different perspective to Carlo Rovelli’s: his is scientistic²³, ignoring (or misrepresenting) the metaphysical aspects; ours is scientific, acknowledging the metaphysics (the “metanarrative of physics”²⁴, where “physics” is intended in Aristotle’s wider sense).

The point here is that Aristotle (and Plato before him) understood very well the metaphysical foundations of any physical knowledge they had. Indeed, they were explicit. Aristotle opens the Physics with a discussion of the method proper to obtaining knowledge in a scientific enquiry²⁵. Plato already spoke of “the dialectical method” as the only way to advance to first principles. For Aristotle the scientific method is necessarily metaphysical (since if the “first principles” are the beginning of physics, then the “dialectical path” to them must be a metaphysical one in precisely our sense – “metaphysical” is just ”the metanarrative of physics”); this is a conclusion reached independently by Jeynes et al. (2023) on the grounds that the meaning of the terms used in a discussion cannot generally be established by the discussion itself.

The disconcerting fact is that physics today typically concerns a variety of phenomena that need intricate experimental apparatus to observe, and often mindbogglingly complex mathematical models to interpret. Is all this stuff even real? Physicists normally ignore philosophical questions, but they all believe that the world is really there and that (on a good day) they can touch reality²⁶: indeed, no less than two Nobel-prizewinning physicists have recently published their commitments to reality (Roger Penrose 2004, and Frank Wilczek 2021). We all want to know how the world is. It is striking that where standard empiricism glosses over the idealistic foundations of how we interpret observations, Nicholas Maxwell’s “aim-oriented empiricism” approach is predicated on the metaphysical priority of unified theories (2020). We look for, and value greatly, an underlying unity in our physical theories, just as we like to hold coherent views on things in general. We will have more to say on unity (§7).

Rovelli (2007 p.177) claims that we must “choose between hiding away in empty truths or accepting the radical uncertainty of our knowledge”. But why can we not courageously affirm the “truths” to which we are committed, while also accepting (as Rovelli correctly says) that we can – and do – get things wrong?²⁷ Catherine Elgin (2017) in her book “True Enough” has explored the concept of truth from an epistemological point of view. She establishes the unsatisfactory nature of the “veritist” position, since it is fatally undermined by fallibilism (and Rovelli insists that our knowledge is necessarily “radically uncertain”: 2007, p.177). Elgin explains, “Making mistakes is an epistemic achievement” (2017, p.291). We may know, but we cannot know infallibly. Again, she explains, “To accept a contention is to be willing and able to use it as a premise in cognitively serious assertoric reasoning” (ibid. p.293). We may find out later that we were in fact wrong and then we would have to change our minds, although to do this we may have to be very brave as well as very honest. But it is a real achievement if we succeed in changing our minds!

Leon-Philip Schäfer (2024) explains that “realism … is, in the final analysis, best characterised as an alethic view which restricts itself to an idea about the objectivity of truth, rather than an epistemic view which underwrites more extensive theses concerning the accessibility of this truth for human knowledge.”²⁸ Although Schäfer is making technical (and rather limited) points, it is worth making the more general point that we are “alethic realists” in his terms: we (together with Rovelli) also think that, as he says, “realism matters because it is a presupposition of fallibilism: if truth is … determined by the facts of reality, then it is particularly difficult to access the truth and perhaps even impossible to achieve certainty in doing so”, and indeed not only would we affirm that it is hard to know reality but also we would deny that it is possible to achieve certainty about it (since reality is elusive). But truth is out there, and we are looking for it.

6. The Illusion of “Objectivity”

Karen Barad²⁹ has asserted, “The primary ontological unit is the phenomenon” (2007, p.333). This conclusion is reached following a long discussion of the meaning of quantum mechanics and the entanglements that follow from it. Barad points out that these entanglements entail a revision of what we mean by “objective”: “Knowing is not a matter of reflecting at a distance; rather, it is an active and specific practice of engagement. To know is to become entangled: objectivity requires that one takes responsibility for one’s entanglements” (ibid., Ch.7, n.1).

The EPR Paradox proposed by Einstein, Podolsky & Rosen in 1935 pointed out that entanglement of two particles apparently could persist over very large distances, meaning that a measurement on one particle would determine the result of a comparable measurement on the other, in apparent violation of the prohibition on instantaneous transmission of information. This paradox was resolved by John Bell in 1964, who established what became known as the “Bell Inequality” as a test for the existence of hidden variables in a local theory. The issue is the interpretation of reality in quantum mechanics: is quantum uncertainty³⁰ ontic or merely epistemic (as most of us would like to think)? Is Schrödinger’s cat really both alive and dead at the same time (because the states “alive” and “dead” are entangled), or is it merely that our ignorance is profound? Is knowledge objective? Are the things that science describes and explains really there? Disconcertingly, it turns out that entanglement is really real, as many “Bell Inequality” experiments have shown.³¹

We will discuss what is "fundamental” below (§8) but the EPR paradox and the consequent interest in establishing violations of the Bell Inequality are at the heart of our current understanding of reality and of our current approach to the “scientific method”. It turns out that for entangled systems we cannot use our standard notions of “objectivity” [observer independence], which are simply too crude: that very idea of “objectivity” is an approximation, and sometimes the approximation is just not good enough. The issue is locality versus non-locality, and causality versus acausality³² (and these apparently ‘simple’ dichotomies become seriously more complicated when time is complexified, see Parker & Jeynes 2023a discussed below §7). Entangled systems may have effects that are astonishingly non-local. And it may be that these are far more numerous and important than we have thought up to now. After all, we typically don’t see what we don’t expect (or what we don’t think exists)!

Barad makes the point (in the light of the quantum revolution) that our Newtonian view of the distinct knowability of objects as such (electrons, protons, etc.) is faulty. Strictly speaking, there is no such thing as “an electron” per se, since all electrons in the universe are more or less entangled together (being indistinguishable). The idea of a single electron is a convenient (and very successful) word game that enables us to set up effective models that allow us to correctly interpret certain sufficiently simple phenomena. But Barad insists: “No inherent … subject/object distinction exists” (ibid. p.114).

Faye & Jaksland (2021)³³ concentrate on Barad’s proposal of “agential realism”, commenting that her account has been very influential in the humanities and implying that it has been ignored by physicists (also critically unpicking Barad’s treatment of Niels Bohr). They point out that Barad herself has not published on agential realism in the physics or philosophy of physics literature. They assert that “agential realism [is] a theoretical framework that … aims to inform social theorizing”.

But Barad is far more ambitious than that, she wants to rewrite ontology itself: “Existence is not an individual affair” (p.ix); “mattering is an integral part of the ontology of the world … not even a moment exists on its own” (p.396). The world is really there, but only agents can know it. And the agents are themselves necessarily entangled in the world.

We are convinced that our best scientific insights can indeed help to expose the substance of reality. Therefore we regard it as axiomatic that, just as truth is one so reality must also be an integrated thing. Then our view as natural scientists of reality must be consistent with our view of it as concerned citizens (or poets). And consequently, we cannot treat insights into reality obtained from the natural sciences as “mere metaphors”, the link is stronger than that. Iris Murdoch (1967³⁴) was very definite about this: “Metaphors are not merely peripheral decoration or even useful models, they are fundamental forms of the awareness of our condition … it seems to me impossible to discuss certain kinds of concepts without the resort to metaphor, since the concepts are themselves deeply metaphorical and cannot be analysed into non-metaphorical components without a loss of substance”.

Murdoch speaks of the unanalysable (non-analytic) nature of the use of language together with its rhetorical tropes (including metaphor), since she is deeply interested in the “substance” of reality, both as a philosopher and also as a novelist. The substance of reality as exposed by our best scientific insights, is not a different thing from the substance of reality as it impinges on our social lives, even if scientists tend to have a “Blind Spot” ³⁵ when thinking about these wider issues.

Parker & Jeynes (2023a) give a more complete physics that is explicit in treating commensurately the local and the non-local, the causal and acausal, the reversible and the irreversible. They do this by systematically complexifying the entire formalism (with spacetime now utilising complex time, where the complex time plane has both “real” and “imaginary” axes³⁶) and then making full use of the mathematical power of complex analysis. The result (expressed in holographic natural units) is that the (Wick rotated) complex conjugate of the system Hamiltonian (representing the system energy) is shown to be the (complex) entropy production (which normally expresses the non-local restraints on the system).³⁷ Together, they are also found to be Hilbert transforms of each other, which (among other things) express a set of (continuum) cause-and-effect relationships across the 2-dimensional complex temporal plane. This result is seriously disturbing, since it mixes up categories that we thought we understood, as well as involving “complex time”. It also disallows the possibility of a unique cause-and-effect relationship for any particular set of phenomena. Complex time is an idea which seems very weird but which is becoming more familiar.

Essentially here we are asserting that the “subjective/objective” distinction is a category error if it is used to inform discussions of reality. Grammatical terms should not be used for ontological purposes. Rather the truth of any statement is to be found through the interplay of its syntactical (narrative and analytical) and semantic (metanarrative and poetical) components.

7. On Unity

Nicholas Maxwell (2020) insists, and we think he must be correct, that “physics accepts only unified theories”. This recalls Leszek Nowak’s conception of the unity of science, not mentioned by Maxwell but discussed in detail by Mateusz Wajzer (2024), who says (without mentioning Maxwell), “Nowak concluded … that scientific theories do not arise from the generalisation of empirical facts, and they are far from hypothetical-deductive systems ... idealisation, not abstraction, is at the core of the scientific method”. Wajzer undoubtedly is correct to emphasise the centrality of idealisation³⁸.

Unity is an indispensable metaphysical idea without which (in our view) science would be impossible, and knowledge would be unattainable. Truth must be one to mean anything at all.³⁹ The rest mass of the electron is close to 9.10^-31 kg (or 511 keV). This is true whoever or wherever you are. We can debate the uncertainties (which are very small for this parameter), and we can debate whether an individual “electron” can be thought even to exist per se (pace Barad 2007), and we can debate the difference between an interpretation we put onto an observed phenomenon and the posited entity being observed, and so on; but after all, we know the value we need to use for the rest mass of the thing we model as “an electron”.

It is disconcerting that even for the simplest possible example (the mass of a fundamental particle) such a periphrasis is required for adequate accuracy. Reality is elusive.

Do “unitary entities” exist? Gottfried Leibniz (1714) thought they did, and famously called them “monads”. But it was Henry More in the 17th century who coined the term “monad”, and it was the “metaphysical poets” who were deeply interested in the idea of “atoms”⁴⁰. We consider the archaic term “monad” to be essentially synonymous with the more modern “unitary entity” that we use here.

Using the new Quantitative Geometrical Thermodynamics (QGT: Parker & Jeynes 2019⁴¹) Parker et al. (2022) have recently demonstrated (without any quantum mechanics) that the alpha particle can be considered an atom, that is not strictly indivisible of course (since it consists of two protons and two neutrons) but indiscerpible⁴² (since it strongly resists being torn asunder due to its very high binding energy). In QGT terms an “atom” is a “unitary entity” (that is: than which exists nothing simpler) which has the fewest possible degrees of freedom (when considered at the appropriate length scale).

Thus, in its lowest energy (“ground”) state the alpha (⁴He) is unitary⁴³. It turns out that the nuclear sizes of the helium isotopes ⁶He and ⁸He can also be obtained by QGT: but these isotopes are not unitary since they are unions of the (unitary) alpha and a (unitary) neutron halo: that is, ⁶He has a 2-neutron halo (the “dineutron”) and ⁸He has a 4-neutron halo (the “tetraneutron”). Even though the tetraneutron is on its own very unstable (disintegrating almost immediately into four separate neutrons), it is curious that Parker & Jeynes can also correctly calculate its lifetime by QGT by treating it as unitary (2023b).

It is also curious that the physicists love the image of “Borromean rings” (exemplified by the ⁶He and ¹²C nuclei⁴⁴): Constantino Tsallis (2024) calls the Borromean rings “a beautiful metaphor for complex systems”. According to QGT (Parker et al. 2022) ¹²C consists of a union of three alphas, as is also indicated experimentally: see the review by Freer & Fynbo (2014), who say that “[¹²C] is believed to possess a rather unusual structure, where the dominant degrees of freedom are those of α-particle clusters rather than nucleons”.

The QGT treatment of ¹²C is as a union of a unitary alpha and a unitary di-alpha, where the identity of the di-alpha is indeterminate (that is: of the three alphas in ¹²C only two are involved, but which two is indeterminate; see Parker et al. 2022). This has a molecular parallel in the benzene molecule, which has three single bonds and three double bonds in the carbon ring, but specifically which carbon atoms have which bond is indeterminate⁴⁵.

Raphael Bousso (2002) has reviewed Gerard ’t Hooft’s idea of the “Holographic Principle” which asserts that the information⁴⁶ in a volume can be represented completely on the enclosing area. Parker & Jeynes (2021) have independently derived this “Principle” (represented by the Bekenstein-Hawking equation⁴⁷) from the entropic Liouville theorem, and applied it to unitary objects to obtain ab initio the size of the alpha particle (also related nuclei; Parker et al. 2022, who use the QGT formalism without any quantum mechanics). “Holographic” literally means “writing the whole” (that is, writing all the 3D information into 2D). It turns out, surprisingly, that the Bekenstein-Hawking equation applies not only to alpha particles (ibid.) but also to black holes (Parker & Jeynes 2023a). And both black holes and alpha particles may be treated (in QGT) as “atoms”, that is, as unitary entities (than which exist no simpler) with quantitatively correct results. Note that while the alpha is only indiscerpible, the black hole is actually indivisible (meaning, in 17th century terms, that not even God could divide it: that is, in modern terms, to speak of its division is to speak nonsense).

Berry & Dennis (2003) have looked for unity in their treatment of the case of classical crystal optics, which turns out to be much richer than expected. They say that “the physics and the underlying mathematics can appear complicated, especially in the most general case of crystals that possess natural optical activity (chirality) and dichroism (absorption) in addition to biaxial birefringence”. But they find “a geometrical approach” which “unifies and clarifies the many different phenomena that can occur”. What is interesting here is a) that their treatment is entirely classical, and b) that their “geometrical” methods treat abstractions of abstractions. Reality is elusive, and its elusiveness is intrinsic – quantum mechanics, general relativity and classical crystal optics are all independently mind-bogglingly intricate. QGT is also essentially (and abstractly) geometrical, and treats reversibility and irreversibility commensurately (Parker & Jeynes 2023a) just as Berry & Dennis can treat the general case (including dichroism which involves irreversibility). These issues are addressed at some length in the mini-Review of Jeynes (2023).

Harald Atmanspacher’s (2024) mereological review⁴⁸ of how unity is manifested in a disunited world, concentrates in particular on the distinction between “the mental and the physical”. He points out that there exists a “domain of reality” neither mental nor physical: in this domain there is a “psychophysical neutrality” between mind and matter. Physical science examples of psychophysically neutral positions are the “mathematical Platonism” of (for example) Roger Penrose (2007) and the “implicate order” of David Bohm (1980).

The ancient poet exclaimed (Psalm 133:1), “Behold, how good and joyful a thing it is: brethren, to dwell together in unity”⁴⁹, but the word “unity” is present neither in the Hebrew original nor in the ancient Greek translation (the Septuagint). It seems that “unity” (ἑνότητα) is a relatively modern idea, in Biblical texts first explicit⁵⁰ at Ephesians 4:3 (dated at 58 CE by John Robinson, 1976: “keep the unity of the Spirit in the bond of peace”)⁵¹, even though the idea of “one” (as a cardinal number) is very ancient (“Hear O Israel, the LORD your God, the LORD is one”: Deuteronomy 6:4)⁵². It is disconcerting that the idea of “unity” (the quality of one-ness) appears to have been current for barely 2 millennia. It seems that previously the idea was not so urgently needed.

Perhaps the very idea of “unity” is harder to grasp than one might think? Prior to St.Paul’s use of it in Ephesians it was used only very rarely: for example, Aristotle uses it in passing⁵³ (in a text acknowledged as “esoteric” – baffling students ever since) trying to speak coherently about substance (οὐσία), about what is fundamental, and about how change happens. “Unity” in such a context is clearly not a trivial idea! But Paul is speaking to everyone, not just the academics.

Aristotle’s physics is notoriously teleological, an aspect that physicists (and biologists) today are allergic to (for very good reason!). Today, the conventional wisdom is that the Universe is ateleological – it has no purpose. However, a unified account presupposes the variational Principles: David Hilbert famously regarded the Principle of Least Action (PLA) as fundamental to all physics, and the PLA has always seemed disturbingly teleological⁵⁴. Of course, the trouble is that purpose is obviously omnipresent in everything we do, and it appears that purpose is an essential characteristic of all living things. We have pointed out much scientific work that together seems to indicate the reality of purpose, and indeed Parker et al. (2025a) have now shown that entropic purpose may be formally defined as a valid quantity in physics. Also, Annie Crawford (2020) has carefully explained “that since teleological concepts cannot be abstracted away from biological explanations without loss of meaning and explanatory power, life is inherently teleological.”

8. What is Fundamental?

Jesus famously observed that the “wyʃe man byltt his houʃʃe on a rocke”⁵⁵ (Tyndale 1526). What is the foundation of our knowledge? We observe the sun rising in the East and setting in the West, but understand that this is an effect of the rotation of the Earth, not simply the sun riding his chariot across the sky (the obvious interpretation of the phenomenon). We observe the green leaf, and understand that the greenness comes from chlorophyll which catches the light energy, using it to fix carbon out of the air (giving us oxygen to breathe) in the wonderful process of photosynthesis, whose details are still being elucidated and which may rely on quantum mechanics (although Runeson et al. 2022 think not). What is fundamental? Frank Wilczek’s book (2021) is actually called “Fundamentals”! Is quantum mechanics fundamental? Or gravity? Or integrity (which we might call “the wisdom of wholeness”)? Or?

Michael Parker and others have shown constructively that unitary entities (than which exist no simpler) can coherently be thought to exist. The size (matter radius) of the alpha particle in its ground state is calculated correctly from thermodynamics using no quantum mechanics at all on the basis that it is unitary (Parker et al. 2022). But everyone knows that the alpha (⁴He) is made of two protons and two neutrons (2+2=4). Which description is more “fundamental”? Is it the unitary quality of the alpha, or is it its “basic” constituents (protons and neutrons)? The way this last question is posed invites the standard reductionist answer that protons and neutrons are more “fundamental” than the unitary alpha, but Parker has shown that the observed phenomenon (the nuclear size) is correctly explained by modelling the entity as unitary (with the fewest possible degrees of freedom). If the alpha is broken into its constituent particles the total system is not “simpler”! This is a surprising result which confronts our prejudices. A “holistic” approach has, in this case, been demonstrated to be more “fundamental” than a reductionist approach⁵⁶ that requires ever larger amounts of energy to be concentrated into ever smaller scales in order to observe the “fundamental” phenomenon. We should emphasise that QGT has shown that “holistic” treatments are themselves fundamental: it is not necessary to show how they “emerge” from (supposedly) “fundamental” physics (on “emergence” see for example Gil Santos, 2020).

These new results in physics result in a gestalt switch in our ideas of what counts as “fundamental”. Previously we thought that “smaller” was equivalent to “more fundamental” – on the admittedly persuasive grounds that if a house is made of bricks then the bricks are “more fundamental” than the derivative house. Reductionism rules ok! ⁵⁷ The trouble in quantum physics is that neither protons nor neutrons are “fundamental particles” since they are both made of quarks (in the Standard Model). Anyway, who knows what quarks are? or if the Standard Model is correct? After all, quantum gravity is proving frustratingly elusive. There are far more questions than are comfortable, even for physicists.

But if “more fundamental” can include such holistic ideas as “unitary entities” then perhaps the whole way we approach the scientific enterprise needs some revision? This suggestion is underlined by Parker’s subsequent resolution of the Loschmidt Paradox⁵⁸, effected by a treatment which shows how the reversible and irreversible (and the causal and acausal, the local and non-local) are in fact commensurate (Parker & Jeynes 2023a). “Local” effects correspond to our standard “cause-and-effect” ideas. But “non-local” effects correspond to holistic behaviours of systems⁵⁹ (and it should be recognised that such behaviours are also central, implicitly, in the physics of those systems). The world will remain incomprehensible to us unless we appropriate this new holistic physics.

It is interesting that Potter & Mitchell (2022) list 8 properties of living organisms that allow us to think of them validly as causative agents (without any sort of dualism): the first is “thermodynamic autonomy” (that is, the entity is in a stable low-entropy state), and the fourth is “holistic integration” (that is, pointing to the necessity of non-local effects). This account is consistent with ours.⁶⁰

Unity, non-locality, and irreversibility are all fundamental aspects of reality, but none are recognised as such by Rovelli (2009) or Wilczek (2021). Other eminent physicists are much less dogmatic: for example, Roger Penrose (2004, ch.27) has a long section on the Second Law emphasising the fundamental reality of irreversibility (never mind that the “fundamental” laws of physics are reversible as currently understood).

9. The Legitimacy of the Scientific Enterprise

What work may legitimately be regarded as “science”? The usual process is that one writes up one’s results in a “paper” which is submitted to a scholarly journal for peer review; two or more accredited scientists then review the work and report to an editor, who may decide to publish after due consideration. Is the published work now “science”? No. The publication of work is no guarantee of acceptance by the scientific community. For example, the Reviewers may not notice mistakes in the work, and even egregious mistakes may be overlooked.⁶¹

As usual, and not unexpectedly, we depend on a consensus. Science is what the scientific community says it is (although there do exist some rather severe tests: what results can the scientists show?). In the end, the widgets that the scientific community makes available to the public are what convinces the public to continue to pay the scientists’ salaries. And of course, there are also goods other than widgets: the inspiring pictures from the Hubble Telescope and its modern successors, and the wonderful TV programmes on the natural world are some examples.

What is the “scientific method”? John Randall (1940) has explored the development of the experimental scientific method in Padua in the centuries leading up to Galileo. He comments: “the fact that the seventeenth century scientists, in revolt against the humanists' appeal to the authority of the past, preferred to put their trust in "natural reason" alone, and hence cared nothing for historical continuity, has sadly misled our judgment as to the fashion in which their thought was generated” and concludes: “But the return to experience is not for the sake of certain proof: for throughout the seventeenth century it is almost impossible to find any natural scientist maintaining that a mere fact can prove any certain truth”. Because we see something does not mean that we understand it! In fact, many have noted that we typically only see what we understand!⁶² And conclusive “proof” of anything is fraught with difficulty.

On the “authority” of scientific experts, we should also point out here that as Robert Crease says (2019, p.252), “Authority is a key feature of public space … The success of totalitarianism is only possible with the erosion of authority.” And he quotes Hannah Arendt’s “Between Past and Future” (1961): “authority implies an obedience in which men retain their freedom.” The scientific enterprise depends on the scientific community acknowledging the authority of the scientific consensus: this authority is not unchallengeable of course, but it remains our baseline. St. Paul urges us that “we henceforth be no more children, tossed to and fro, and carried about with every wind of doctrine”⁶³: we ought to need adequate persuasion to change our minds! Authority is not dispensable: it is inescapable that we are social beings which implies a submission to authority. This is not a bad thing, it just is. We are free to submit and we are free to dissent.

It is inescapable that the scientific enterprise rests on integrity: we have to trust each other to tell the truth about what we observe (even if, of course, we also do whatever checks are available). For we will repeat experiments only for the most important results – experiments are expensive and repeat experiments are not usually publishable. More than that: the scientific enterprise rests on the visceral desire of the scientists to know how the particular aspect of our wonderful world under investigation really works. Science is a passionate activity – scientists want to get things right! Moreover, the legitimacy of science cannot be communicated coldly to the public. Wholeness is the key. A scientific argument clearly must be considered dispassionately to confirm its validity. But a valid argument is not necessarily a true one. We want to know the truth about the world, and obtaining knowledge is a deeper activity than merely obtaining information. I can pile up heaps of information but understand none of it!

Another reason for scientists doing their best to report results in a dispassionate (or disinterested) way, is because they have the underlying conviction that what they perceive as the truth should be recognised as such by everyone. Another reason for scientists to aspire to dispassionate reporting is that they are (usually) very well aware of their own fallibility, including the natural propensity of having such fixed ideas about how things should be that the results are distorted.

Speaking now about the scientific practice itself, that is, the getting of the results not just the reporting of them, Philippe Stamenkovic (2024)⁶⁴ discusses the “Value-Free Ideal” (VFI): the idea that science is (or should be) “value-free” (or “objective”). Of course, the original arguments for the VFI included the need to protect “public trust in science”, an issue central to our concerns here. But, as a human activity, the scientific enterprise cannot be “value-free”! Stamenkovic urges that “scientific claims” should be clearly distinguished from “scientifically informed decisions” – the former can be reasonably “value-free”, but the latter cannot be in principle since decisions are always made on the basis of their value (as assessed by the decision-maker).

Unfortunately, Stamenkovic has too optimistic a view of “scientific claims”: he asserts that “A science based on facts (further generalised in the form of laws and principles) represents the ideal of scientific inquiry”. The trouble is that interesting facts don’t exist – since if it is a “fact” then it is per se uninteresting (since everyone agrees). Nobody now worries about whether it is the Earth orbiting the Sun or vice versa, but this was only formally established as a “fact” by Bessel’s observation of stellar parallax in 1838, an observational tour de force that was only interesting as such, but not as establishing Galileo’s opinion (which had by then already long been the consensus). Thomas Kuhn’s (1962) description of the discovery of oxygen by Lavoisier (and Priestley) is an object lesson in how the very expression of “facts” is plastic and must be negotiated, and Paul Feyerabend (1975) powerfully makes comparable arguments. Surely any useful philosophical discussion should capture the fact that we only do science at all, as Robin Ince (2021) has pointed out, because we are interested.

It is true that scientists wish to establish facts (hence Stamenkovic’s insistence on “ensuring the truth of scientific knowledge”), but what eventually becomes recognised as a “fact” is still only a "good idea” while it is being established. And, as we have seen, the truth of the proposition may well be accepted before (or even long before) the “fact” itself is properly established as such. Truth is always interesting (but “facts” never are per se – until of course they start to be called into question (at which point they are no longer “facts” at all: reality is elusive).

But a “fact” (in the sense of the considered consensus view of the scientific community) is necessarily “already a social thing” since “it concerns events and circumstances in which many are involved; it is established by witnesses and depends upon testimony” (Crease, 2019 p.262; quoting Arendt, 1961). And it is this “very social character” of “facts” that can be used by science deniers “to discount them” (ibid.). We have already mentioned Richard Bauckham’s monograph on personal “eyewitness” testimony (2006) which he shows is fundamental to all knowledge. Frank et al.’s whole “Blind Spot” thesis (2024) says the same thing from an entirely different point of view. And John Hardwig was long ago insisting on a principle of testimony (1991) to establish “epistemic authority” (1985). He says, “Cooperation, not intellectual self-reliance, is the key virtue in any scientific community. But epistemic cooperation is possible only on the basis of reliance on the testimony of others” (ibid.). After all, science is a human activity, and knowledge is personal (Polanyi 1958).

For our purposes Stamenkovic’s approach appears to be insufficiently fundamental, since he seems to think that “scientific truth” is well defined (and attainable): he says, “… without [truth] there is no science.” Of course, at a practical level we can be sufficiently sure of things that we can be confident of success in proposed actions (which is his particular interest). But at a philosophical level, what sense does it make to say (for example) that Newton’s physics was “true”, in view of the non-Newtonian advances in physics starting with Maxwell’s electrodynamics? On the other hand, it is also clearly perverse to say Newtonian physics was “false”! Catherine Elgin (2019, p.1) is right to say that a “veritist” approach is not “plausible”: she insists that what she calls “felicitous falsehoods” are “epistemically valuable”.

Stamenkovic acknowledges this when he says that the scientific corpus (“the total body of scientific knowledge”) is “[representive of] our best available, most reliable (although always revisable) knowledge”. “Revisability” is the critical idea here. It turns out that “facts” may be revisable. What is considered a “fact” today” may be considered as an erroneous idea tomorrow, although the body of phenomena that had been considered to establish that “fact” remains in the record. So something must have been true about it, even if that thing was not expressed correctly. Karen Barad (2007, p.333) is surely correct to insist that it is the phenomenon that is the “primary ontological unit”.

Michael Lissack (2017) points to the “hidden assumptions” that “allow ambiguity⁶⁵ to be bracketed away at the expense of transparency” imposed as “enabling constraints” (in the form of ceteris paribus – “other things being equal”), acknowledging Jürgen Habermas (1996) who points out that “scientific activity … fragments into a number of competing viewpoints that are shot through with values”. The investigation of such enabling constraints (in the form of “uncritically examined presuppositions”) reveals “hidden assumptions … which can get in the way of the acceptance of a scientific claim”. Scientists acknowledging their hidden assumptions more transparently, as Lissack proposes, should result in the more willing acceptance by the public of scientific claims.

Stamenkovic points out that “there is a fundamental requirement to distinguish between facts and values”. Science is a human activity, done because we want to know how the world really is. This activity is “shot through with values” (as Habermas correctly says), and yet we wish to achieve a consensus view on this or that aspect of the world. This is what “facts” are. But we vigorously deny that all this is just a “matter of opinion”, since the phenomenological underpinning of an accepted “fact” (together with its formal interpretation) is a matter of public record! The “facts” may change (often in rather subtle ways) but the public record persists. Trust in scientists and the work they produce must surely require the public interplay of facts and their associated values, allowing the integrity of the work to be openly displayed. The aim is to persuade the public that reality really is being described truthfully.

10. QBism

Since we are realists who insist that “knowledge is personal” (Polanyi 1958)⁶⁶, are we not also QBists (“quantum Bayesians”)? QBism is regarded by many as a “philosophically rigorous approach to quantum mechanics” which “cuts no epistemic or ontological corners”⁶⁷. David Mermin (2012) has summarised QBism helpfully, explaining that it elegantly disposes of the “shifty split” that John Bell deplored (the discontinuity between the microscopic and macroscopic involved in making the “observation” that “collapses the wavefunction”⁶⁸), and points out that the real issue is how we think of probability. Are we “frequentist” or “Bayesian”⁶⁹? Michael Levine (2019) points out that “in many problems … there is no true model”, and we point out (again, see above) that “measurement is primarily an epistemic process” (Mari et al. 2013) and that the “Guide to the Expression of Uncertainty in Measurement” (JCGM 1995) explicitly avoids the concept of a “true value” of a measurand on the grounds that such a value is unknowable in principle, and any estimate of it must be expressed together with a “measurement uncertainty”. Although it is truth we seek we also admit our limitations (to say nothing of ontic uncertainty). It should be added that determining probabilities is not only a philosophical minefield but a technical one too: Steven Goodman (speaking to the medics, 2008) has helpfully listed various ways that people can (and do) misuse statistics to pretend to the scientific method. Howson & Urbach (1989) examine the way scientists actually appeal to probability arguments, directly addressing the charge against Bayesianism that it is "too subjective".

Leifer (2014) has provided a deeply interesting and informative review of issues associated with QBism, specifically seeking to expose the implications of choosing an ontic rather than an epistemic interpretation of the wavefunction Ψ (which is the basic entity in quantum mechanics). Leifer correctly denies that this issue is “metaphysical” (in the conventional sense of “the science of being”), insisting that it is strictly a technical (“scientific”) issue of physics, backing up this assertion with a careful mathematical exposition, and saying that “what Bell’s Theorem really shows us is that the foundations of quantum theory is a bona fide field of physics, in which questions are to be resolved by rigorous argument and experiment, rather than remaining the subject of open-ended debate.” It seems that these questions are not all resolved, however: the Pusey–Barrett–Rudolph Theorem (Pusey 2012) proved that (under certain conditions) Ψ is ontic (that is, a state of reality and not merely a state of knowledge). The consequences of this are both subtle and far-reaching, with the details beyond our scope here. However, we should recall that Karen Barad (2007) has made a persuasive case that the categories “ontic” and “epistemic” are themselves mixed, and that we should therefore speak of the “onto-epistemic”: “there is good reason to question the traditional Western philosophical belief that ontology and epistemology are distinct concerns” (ibid. p.43)⁷⁰.

Bayesian methods try to specify prior information where “frequentist” methods aim to be strictly “objective”. That is, the frequentists pretend to “objectivity” where the Bayesians admit to their beliefs on the grounds that no-one decides anything without believing something. Bradley Efron (2005) points out that “Bayesian statistics has seen a strong movement away from subjectivity and toward objective uninformative priors in the past 20 years”, but comments that “in practice, it isn’t easy to specify an uninformative prior”.

In the important case of a “Bertrand Paradox”⁷¹, Parker & Jeynes (2023c) have shown how to introduce the required uninformative prior in the form of an additional natural constraint: scale invariance. This also automatically eliminates the “shifty split” problem that has bedevilled the quantum physicists (where exactly lies the boundary between the microscopic and macroscopic worlds?). Entropy is blind to scale, and the scale-blindness of Parker’s new “quantitative geometrical thermodynamics” (QGT; Parker & Jeynes 2019) is explicitly a feature of its hyperbolic spacetime. Therefore QGT is expected to be (and is!) valid both for quantum mechanical entities (like alpha particles: Parker et al. 2022) and also for entities of general relativity (like black holes: Parker & Jeynes 2023a). This involves a length scale range of at least 35 orders of magnitude for which QGT is demonstrably valid.

David Bohm argued for an “Undivided Universe” (Bohm & Hiley 1993): in QGT Parker has shown us a uniform (undivided) treatment of both QM and GR entities. Taken with the nonlocal nature of thermodynamics and the holomorphic nature of actio-entropy (Parker & Jeynes 2023a⁷²), it is clear that QGT offers an ontological description that accords well with the Bohmian conception of wholeness (Bohm 1980). Speaking of probabilities (to the wider physics community), Mermin (2012) also says, “you had better reexamine what you wrongly may have thought you understood perfectly well about the nature of probability.” QBism revises the usual way of thinking about probability, and Parker et al. (2025b) have shown (separately) that we can “treat probability as a physical quantity grounded in hyperbolic Minkowski spacetime, and obeying all the appropriate physical laws” – that is: probability is physical.

A new understanding of the physics of probability and its Bayesian explication via a relativity of scale is now also offering important new insights into the very nature of reality itself.

11. The Poetry of Physics

It is now widely accepted that every act of understanding requires a leap of the imagination. Knowing is imaginative. To establish the terms they use for new concepts, scientists must “negotiate their meaning” (using Martin Edwardes’ phrase, 2019), which is always a metaphysical step requiring a use of language that can only be called poetic (see Jeynes et al. 2023). An imaginative leap (or an “intuitive leap”) is just the same thing as the “leap of faith” (properly the “leap by faith”⁷³) proposed by Søren Kierkegaard explicitly in his “Unscientific Postscript” (1846) and implicitly in his discussion of Abraham’s faith in “Fear and Trembling” (1843). The essential fact is that there is no disembodied or impersonal knowing. Knowledge is personal (Polanyi 1958) just as much as information is physical (Landauer 1991, 1999)⁷⁴.

No human enterprise may be formally defined. It may be described, or (better) exemplified, but no properly human activity may be entirely specified. And although the knowledge of reality may be approached variously, yet truth is one. That is, poets and scientists may variously seek knowledge of the one reality, all using their own appropriate methods, but ultimately it is still the same reality they all describe, however differently. Because reality is elusive it is hard to grasp, and we need many varieties of approach to lay hold of it.

We have said that “science is what scientists do” ⁷⁵, intending not to give a tautological and useless “definition” but to indicate the humanity of the activity. Many attempts have been made to define the “scientific method” but there remains no consensus. This is because humans want and do surprising (and undefinable) things⁷⁶. One needs to be a poet to speak adequately of “aspiration” or “inspiration”, or even “wonder” or “delight” – all ubiquitous in those captivated by science (the scientists). Those not so captivated demand to know the “use” of the (very expensive) activity, and it turns out that knowing true things about reality indeed has many material benefits, but scientists typically do not labour diligently for this knowledge because of its “usefulness” (or to get rich) but because they are moved by the poetry of the thing.

It is becoming fashionable to bracket “science” together with “poetry”. Carlo Rovelli is regularly called “the poet of science” ⁷⁷, and the astrophysicists have delighted the world with their wonderful pictures of the cosmos. There are many examples. These things should be taken seriously also by philosophers: they are not merely media froth. Reality is slippery and very hard to grasp (as Robin Ince also points out: “even the most solid and sure things you can set your watch by become malleable and slippery. This includes time itself”: Ince 2021 p.89): our methods for learning about it depend essentially on inspiration – a poetic quality par excellence. The renowned philosopher Ernst Cassirer “liked to tell the following story: once he met the great mathematician [David] Hilbert, the ‘Euclid of our time’, and asked him about one of the latter’s disciples. Hilbert answered: ‘He is all right. You know, for a mathematician he did not have enough imagination. But he has become a poet and now he is doing fine.’ Cassirer always heartily laughed, when he told this story …” (Gawronsky, 1949)⁷⁸.

The positivist attempt of the 20th century to try to tie down the “scientific method” has failed, crippled by reality itself which refuses to be tied down. In his foundational work Kurt Gödel (1931) showed it was a logical tautology that any real system cannot be known completely, proving constructively that any logical system at least as rich as arithmetic contains theorems⁷⁹ which are certainly true but nevertheless undecidable in that system. Even arithmetic, which we usually consider child’s play, is logically incomplete! It is interesting that Gödel’s Incompleteness Theorem made the first formal use of metamathematical methods, although this use was foreshadowed by Anselm’s treatment of his “Ontological Argument” in the Proslogion (1078); see the discussion by Jeynes et al. (2023) of Anselm’s poetic metanarrative, systematically overlooked by modern philosophers.

A modern view of the scientific enterprise avoiding the myth of “completeness” that Kant popularised (1786) was Charles Peirce’s view that the methodology of science starts with “abduction”. A scientific investigation involves induction from data and deduction of consequences, but the original activity of forming explanatory hypotheses (crucial to the inductive/deductive steps) is a creative (not strictly a logical) one. “The abductive suggestion comes to us like a flash. It is an act of insight, although of extremely fallible insight” (Peirce 1903, emphasis original). Peirce’s idea of abduction has been widely discussed, for example by Lucia Santaella (Braga, 2019) and Timothy Shanahan (1986). Ernan McMullin (1992) discusses abduction, but prefers retroduction (also used by Peirce): “central to retroduction … is the imaginative modification of existing concepts, or the creation of new ones quite remote from the universals or forms that might be abstracted from perceptual experience”. McMullin’s work is widely regarded as seminal, for example by Bas van Fraassen (2013), who says that he “takes us on a fabulous journey through the history of philosophy of science”, concluding “When McMullin states … that the success of retroductive inference is a contingent matter, he … does not go far enough. That the contingent conditions for the success of retroduction actually obtain could only be inferred by retroductive inference: an insight that sweeps the rug from under any assurance of epistemic safety to be found there.”

12. Machine Intelligence

We have glanced at issues of public trust in science above. However, a very modern concern is, ‘Can Artificial Intelligence (AI) be “trustworthy”?’ since one might have expected such a question to be related to the enquiry about “the integral nature of the scientific enterprise”. After all, AI is a “scientific” activity, isn’t it? “Bing” (the AI bot now supplied by Microsoft as standard) actually answers this question quite well: “Artificial intelligence is not human, so we should avoid terms like “trustworthy AI” that not only humanize AI but also imply a level of dependability that simply does not exist. However, broad adoption of AI systems will require humans to trust their output. Many AI systems to date have been black boxes, where data is fed in and results come out. Trust is a priority, but many organizations haven’t taken enough steps to ensure AI is trustworthy. Many current efforts are aimed to measure system trustworthiness, through measurements of Accuracy, Reliability, and Explainability, among other system characteristics.”

A supplementary question to Bing about the “metanarrative” of its answer also elicited a sensible response under six headings: (i) Context & Background; (ii) The Nature of AI (not human, therefore lacking human characteristics); (iii) The Dual Nature of AI (good and bad aspects); (iv) Trust and Trustworthiness; (v) Challenges and Responsibilities; (vi) Operationalizing Trustworthiness (mentioning that “practical philosophy” can “guide the development of AI”). This response is revealing from our point of view here: it appears sensible and competent (and useful), but really it hasn’t grasped what we have tried to explain here as “metanarrative” (that is, the context that enables our understanding of the narrative in question). Rather, it is, again, simply a set of facts. Could I ever compose poetry of bullet points?

In terms of the discussion here, a scientific narrative is designed to iron out ambiguities in the handling and interpretation of the observations, but such a narrative per se cannot engender understanding either in the practitioner or in the student (or any interested citizen). For an explanation of the meaning of the observations (to “get to the truth of things”) one must resort to the (metaphysical) metanarrative. We have seen how Joseph Drew (2023) has drawn attention to the real benefit available from a proper attention to Aristotle’s methods of rhetoric to persuade people to trust scientists, and James Lennox (2015) has shown how explicitly Aristotle grounded his Physics in a prior metaphysics.

We also have seen how, on the face of it, AI is capable of competent (and useful) summaries of the conventional wisdom. But clearly the boundaries of its competence are ill defined (and probably undefinable). How unknowable these boundaries are is central to the issue of its trustworthiness, and has stimulated much work helpfully reviewed recently by Karoline Reinhardt (2023), and including notably the “Ethics Guidelines for Trustworthy AI” (EU 2019) developed by the “High level working group” (the “AI-HLWG”) of over 50 experts. The AI-HLWG have proposed a pilot “Trustworthy AI Assessment List” (ibid. pp.26-31) which has 7 headings: (i) Human Agency and Oversight; (ii) Technical Robustness and Safety; (iii) Privacy and Data Governance; (iv) Transparency; (v) Diversity, Non-Discrimination and Fairness; (vi) Societal and Environmental Well-Being; (vii) Accountability. The AI-HLWG gives “Key Guidance” for developers, suppliers and users: namely to “adopt a Trustworthy AI Assessment List”, adapting it to the “specific use of the system” and insisting: “Ensuring Trustworthy AI is not about ticking boxes, but about continuously identifying requirements, evaluating solutions and ensuring improved outcomes …”. These instructions are very human and addressed to the responsible humans, deliberately using value-laden terms (“identifying”, “evaluating”, “ensuring”). This implicitly acknowledges the metanarrative that this present work is explicitly articulating. As is frequently the case with AI based on a large language model, the output is good at providing a useful narrative but it does not creatively infer the underlying metanarrative.

Fodor & Pylyshyn (1988) asserted that artificial neural networks (ANNs) cannot do "systematic compositional generalizability" in principle, and it’s worth pointing out Abigail Tulenko’s (2024) view that this generalising principle can be traced to the 11th century Persian philosopher Ibn Sina (influential in Europe where he was known as “Avicenna”). Avicenna thought that the mark distinguishing humans from animals was that humans can reason from generalized rules whereas animals can only think about particulars. And it is this human capacity for universals that AI has found remarkably hard to match, although Lake & Baroni (2023) have shown that, using “common neural networks without added symbolic machinery”, if the training is guided by the “meta-learning for compositionality” (MLC) approach the ANNs can match human performance. The critical condition for this is (as for all ANN training) to do with the training programme. Clearly training programmes using MLC can be very general, yielding remarkable results. Nevertheless, from a philosophical point of view, the fact that we can get ANNs to convincingly mimic certain human behaviours does not erase the underlying dependence of the AI system on our own intelligence and indeed the active involvement of an overarching human intervention. It remains the case that AI systems are amoral: they have behaviour but not judgment. They have no understanding of the metanarrative.

More generally, Karoline Reinhardt (2023) makes the points well that a) trust and trustworthiness are two different things (people can trust the untrustworthy and distrust the trustworthy), and also that b) trust is an ambivalent thing (we can after all “trust that the world is going to end”). She argues that “What is needed is an approach to trust that is also informed with a normative foundation and elements emphasized in the research on trust in other fields, for instance, in social sciences and the humanities, especially in the field of practical philosophy.” We draw attention to her requirement for a “normative foundation”, which we would align with our discussion of the metanarrative. She asserts that “taking the self-determination and autonomy of people seriously involves accepting that trusting is a game with an open end. It is their choice to trust, or not to trust.” That is (reprising our arguments above), unfortunately it remains necessary to persuade people to trust even those things that are trustworthy (leaving it as a possibility that malign actors will also seek to persuade people to trust the untrustworthy). And she concludes: “it might turn out that in the end what we need is not more trust in AI but rather institutionalized forms of distrust.” It is interesting that the development of AI and its recent substantial successes are stimulating us to re-visit what the scientific endeavour really is.

13. Discussion

In this work we have presented some of the integral aspects found in the highly successful adventure that is the modern scientific enterprise. In doing so, we have also underlined the contribution of new physics (including Parker and others) illuminating key aspects of the unitary nature of the enterprise, and which has also culminated in the resolution of the Loschmidt Paradox using complexified time (Parker & Jeynes 2023a). Here we briefly discuss both some further aspects of complex time and also some further aspects of knowledge being personal, including the importance of storytelling.

It is noteworthy that complex time as an emerging and increasingly powerful scientific paradigm has also been used systematically by Ivo Dinov's group in “Big Data” data science applications (Dinov & Velev 2022). Undoubtedly their use of complex time (which they call “kime”) enables them to “interrogate large, multiplex, and heterogeneous datasets … using spacekime analytics” relying on a “commonly used mathematical trick of lifting the space of a problem to identify solution paths that are not clearly visible in lower dimensional spaces”. By using complex time (“kime”) they transform 4D spacetime into a 5D “spacekime”. It is interesting that two completely independent groups have used complexified time for quite different applications, but for the same underlying mathematical reason: namely that complex analysis has extraordinary mathematical power.

In this regard we should also mention the major recent advance in quantum chromodynamics (QCD) of Deur, Brodsky & Roberts (2024: explained by Brodsky et al., 2024). These workers have also used complexification (using a 5D spacetime) and a holographic approach to calculate the strong force at “large” distances. How these various and still nominally independent applications of complex time can enable a new metaphysical⁸⁰ interpretation underpinning our understanding of reality still remains an exciting future prospect.

How does “complex time” help in our quest for unity and integrity? It is a part of the structure of reality that is only now being clarified, although time was always known to be a slippery philosophical idea: it was Augustine of Hippo in “Confessions” (XI:14; c.400 CE) who famously said, “I know very well what time is, until someone asks me!”⁸¹

Treating science as story underlines the essential truth that knowledge is personal, as Michael Polanyi (1958) observed long ago. Paul Grobstein (2005) has urged us to develop a “story of science as story telling and story revising” which “may provide a foundation for a less wrong view of science …the ultimate test of [which is] what new stories develop because of it” (emphasis original). Robert Crease has pointed out that “Stories make things appear that you cannot see or think about otherwise” (2019, p.263). Frank et al. (2024, p.ix) go further: they consider that stories “constitute a modern form of mythos … [orienting us and structuring] our understanding of the world”.

Thomas Dillern (2019) seems to be an isolated modern proponent of Polanyi’s view of personal knowledge, although Jerry Ravetz’s (1971) school was active previously in creatively developing Polanyi’s view on craftsmanship. Dillern says that “the things that are most interesting to gain knowledge about are things that … provide something meaningful to humanity, which is something the modern mechanical science view fails to accomplish on its own” and pointing out Polanyi’s insight that “tacit knowledge is the basis of all explicit knowledge … tacit powers are the ultimate faculty through which humans acquire and hold all knowledge”. We have here represented this viewpoint by insisting on the indispensable metaphysical context of our physical explanations (following Jeynes et al. 2023). ‘Polanyi’s Paradox’⁸² (David Autor, 2014) is a central and necessary idea for mediating our conception of “knowledge”. All knowledge is formally incomplete, and the idea that “scientific knowledge” is somehow different is simply mistaken. But all human knowledge (and we know no other) is metaphysical: the metanarrative is invariably indispensable (even when it is tacit), and even physical knowledge cannot be understood without its metaphysical contexts. “Grasping truth” is essentially metaphysical.

It is worth pointing out explicitly that “tacit knowledge” (described in some detail by Polanyi, 1958) is a fine cover for unrecognised assumptions being smuggled in to the interpretations of the data, so that understanding how the metanarratives work has become essential.

14. Summary

The scientific enterprise has developed out of recognition over the last four hundred years, from marvelling at microscopic wonders and exalting in the simplicity of celestial mechanics to having the power to destroy the world; from the conceit of the clockwork universe to the bafflement of quantum uncertainty; from the triumph of reason to the conundrums of reality. But it has turned out that reality is far more baffling than we had ever expected. Newton thought he could specify the scientific method, but with every advance in knowledge it seems that our account becomes less complete.

Here we point to the overriding idea of unity: you want to know how the world is? then seek a coherent account! And do it with as much diligence and honesty as you can muster. For reality is elusive, and a casual approach will find only dead ends. Moreover only a single-minded quest in good faith has any chance of finding out new things.

For it turns out that unity remains a powerful idea: even in physics “unitary entities” have been shown to successfully model things that – looked at another way – seem to be composites of other ”fundamental” particles.

15. Conclusions

The scientific enterprise seeks to gain knowledge of the natural world, where the very possibility of knowledge is premissed on an appreciation of unity. Unity is a metaphysical concept, but it is physically embodied in the alpha particle and the black hole, both of which are unitary entities (“than which exist nothing simpler”). Radically new physics now articulates these ideas.

New physics enables and informs new philosophy. The recognition that there are real examples of unitary entities which we can specify and discuss, has allowed a much more definite idea of what is fundamental (also fatally undermining a reductionist view). Unity always was a powerful idea. Now that it is underpinned by physical reality it is more powerful still.

The idea of unity is fundamental to our idea of reality itself, essential not only to the basic physics but also to the integrity with which we (both personally and socially) engage in, report, and receive the scientific enterprise. That is, it is essential at all levels: the mathematico-physical, the experimental, the scientific community, and the public for whom in the end the science is done.