Two cultures, one world. A working stance for someone who refuses to choose between rigor and meaning, and who would like to defend that refusal at a dinner party without raising their voice.
Two ways of knowing have organised themselves over the last four centuries into separate disciplines, separate institutions, and largely separate kinds of people. The first treats the world as a system of mechanisms to be modelled, predicted, and verified against measurement. Physics, chemistry, the parts of biology that look like physics, the parts of psychology that look like statistics. The second treats the world as a field of meanings to be interpreted, lived, and resonated with. Literature, history, depth psychology, the contemplative traditions, the parts of philosophy that take ethics and beauty seriously.
Each tradition, at its best, is rigorous on its own terms. Each tradition, at its worst, mistakes itself for the only legitimate way to know anything. When the first overreaches it becomes scientism, the claim that anything that cannot be empirically falsified is meaningless. When the second overreaches it becomes a soft relativism, the claim that empirical method is just one cultural perspective among many and need not constrain anyone. Both overreaches are familiar; both are category errors. They consist of taking a real, powerful method that works in one domain and pretending it works everywhere.
The position that holds both registers without collapsing into either has a name. It is called epistemic pluralism. It says, roughly, that there are different kinds of questions in the world, that each kind has its own methods and standards of rigor, and that the mature practitioner knows which tool fits which job. This piece is an attempt to lay it out carefully enough that the position is not just a slogan but a working stance you can carry into your own thinking and into hard conversations.
There is a compact way to feel the idea before the argument starts. Look at the Mona Lisa. It is, demonstrably, a particular arrangement of pigment molecules on a particular weave of poplar wood, oxidising at a measurable rate, reflecting light at wavelengths a spectrometer can record. Every word of that sentence is true. It is also, demonstrably, a portrait of a woman smiling in a way that has held the attention of human beings for five centuries. Every word of that sentence is true. Neither description reduces to the other. You cannot derive the smile from the chemistry. You cannot derive the chemistry from the smile. They are two simultaneously valid registers of description sitting on top of the same object, and the object is unbothered by this. It just keeps being what it is.
If that example lands, you already have most of the position. The rest of this primer unpacks it carefully enough that the next time the question arises, in your own head or in someone else's challenge, you have the moves.
The cleanest map of the territory comes from late 19th century German philosophy of science, mostly Wilhelm Dilthey. He noticed that when natural scientists were doing their best work, they were doing one specific kind of thing, and when historians and philologists and psychologists of meaning were doing their best work, they were doing a different specific thing. He called the first Erklären, explanation, and the second Verstehen, understanding. The natural sciences he grouped as Naturwissenschaften. The sciences of meaning he grouped as Geisteswissenschaften, sciences of spirit, a phrase that sounds embarrassing in English but means something quite specific in the original.
The split is not "objective versus subjective." It is not "real versus fluffy." It is not "hard versus soft." Those framings are the propaganda of one side. The actual distinction is methodological. Two registers, each rigorous in its own way, each appropriate to a different kind of question.
The crucial move, the one that takes a while to actually settle in the body, is this: neither column is a defective version of the other. A literary critic is not a failed neuroscientist. A neuroscientist is not a failed literary critic. They have picked different objects out of the world and brought different methods to them, and both methods, when done well, produce real knowledge in the form appropriate to their object.
If the right column makes you nervous, notice the nervousness. We were trained to be nervous. The 20th century did a thorough job of teaching educated people to treat the left column as the only column that counts. The history of how that happened, which is the subject of Piece 2, matters for our purposes only insofar as it explains why you sometimes feel ashamed of the things you actually believe. You should not be. You are operating in two registers because the world has two registers in it.
Long before Dilthey, Aristotle had already given us the cleanest version of the pluralist move. In the Physics and the Metaphysics, he argues that a complete account of any thing requires answering four different questions, and that each question picks out a different kind of cause.
Take a wooden table. Why is it as it is?
Aristotle's claim is that you do not know the table until you can answer all four. Drop one and your account is incomplete. The modern scientific revolution did something specific with this scheme. It kept material and efficient causes, which are the ones that lend themselves to mathematical description, and it bracketed formal and final causes as merely subjective, or worse, as theological residue from a worldview the new physics was trying to leave behind.
This was, in narrow terms, a brilliant move. Bracketing teleology let Galileo and Newton get traction on planetary motion without having to ask what the planets were for. The whole architecture of modern physics is built on that bracketing. We owe it everything from antibiotics to satellites.
But here is the pluralist's quiet observation. Those two causes did not stop existing. They got bracketed. The question "what is this for" did not become a meaningless question; it became a question we no longer had institutional permission to ask in scientific contexts. The question "what form does this exhibit, of what kind is it, what does it belong to" did not become incoherent; it became something we now do badly, in registers we have half-forgotten, because we trained ourselves out of speaking that language.
A modern physicalist will tell you that a human being is a particular arrangement of carbon, hydrogen, oxygen, nitrogen, and a handful of trace elements, undergoing thermodynamically defined processes, propagated by gene-driven reproduction. Every word true. None of it tells you what a human is for, what form a good human life takes, or what shape a particular person is striving toward. Those questions did not disappear. We just stopped pretending they were answerable inside the bracket.
That single sentence, written around 340 BCE, is the entire pluralist position in compressed form. Expect that amount of exactness in each kind which the nature of the particular subject admits. Do not demand falsifiability from poetry. Do not accept Barnum-effect hand-waving from a physicist. Calibrate your standards to your object. This is not relativism. It is the opposite of relativism. It is the demand that you take each subject seriously enough to learn its proper standard.
Now I want to make the central argument as cleanly as I can, because the rest of this piece is essentially commentary on it.
Consider any sufficiently rich object in the world. A painting, a dream, a person, a piece of music, a conversation between two friends, a national history. Pick whichever. I will keep using the Mona Lisa because it is the simplest version of the case.
The Mona Lisa is a particular arrangement of atoms. This is not a metaphor. It is a literal, exhaustive physical description. Given perfect knowledge of the position and momentum of every particle in the painting, you would have, in one sense, said everything physics has to say about it. You would know its mass, its energy content, its electromagnetic absorption spectrum, the precise way it will decay over the next ten thousand years. You would have, in the language of physics, a complete account.
The Mona Lisa is also a portrait of a woman, painted by a particular man in a particular century, with a smile that has provoked specific responses in specific human nervous systems for five hundred years, that has become a symbol of art itself, that has been the subject of theft and forgery and pilgrimage. This is not a metaphor either. It is a literal, exhaustive account of what the painting means. Given perfect knowledge of its history, its iconography, its place in the human imagination, you would have, in one sense, said everything art history and hermeneutics have to say about it. You would have, in that register, a complete account.
Neither account reduces to the other. This is the point. Take the atomic description and try to derive from it the fact that the smile holds attention. You cannot. Nothing in the wavelength of reflected light entails the response of a human nervous system encountering it as a portrait of someone alive. The response runs through a whole stack of context, history, embodied recognition, narrative framing, none of which is contained in the physical description, even in principle.
Now go the other direction. Take the art-historical account and try to derive from it the rate at which the varnish will yellow over the next century. You cannot. The hermeneutic account is silent about chemistry. Not because it is impoverished, but because it is in a different register.
This is what philosophers of science call multi-level description, and the technical term for the relationship between the levels is mutual irreducibility. Two genuinely valid descriptions, both true of the same object, neither derivable from the other. The object is not split. There are not two Mona Lisas, one for the physicists and one for the curators. There is one Mona Lisa, and it admits of two descriptions, and the descriptions do not collapse into each other.
Once you have the move, you can apply it everywhere. Apply it to a dream. The dream is a pattern of neural activity in REM phase, involving the pontine reticular formation, the limbic system, the visual cortex, measurable on an EEG. Every word true. The dream is also, on Jung's account or Freud's or Hillman's, a symbolic communication from the unconscious, structured by personal complexes and collective archetypes, addressed to the dreamer, available for interpretation by methods that have produced clinical results for over a century. Every word also true. Neither reduces. The neurology does not predict the meaning of the snake that appeared on the night your father called. The meaning of the snake does not predict the EEG trace. Both are real. Both are about the same dream.
Apply it to a person. You are a body, a particular metabolism, a set of brain states, a result of evolutionary pressures and developmental contingencies. You are also a self, a narrative, someone with a history and a future, capable of being loved, capable of meaning what you say, capable of growth. Both true. Neither reducible.
If you find yourself across the table from the tech person from the opening scene, this is the section you want loaded. They will likely respect three things: contemporary credentials, neuroscience specifically, and arguments that the smartest hard scientists themselves take seriously. All three are available.
Iain McGilchrist is a working psychiatrist, an All Souls fellow at Oxford, the author of two large books on hemispheric asymmetry: The Master and His Emissary (2009) and the truly massive The Matter With Things (2021). He is not a New Age figure. He is a careful neuroscientist who has spent his career on the question of why the brain is bilaterally divided at all, given how metabolically expensive that division is.
His answer, condensed brutally: the two hemispheres attend to the world in fundamentally different ways. The left hemisphere is good at narrow focus, abstraction, decontextualisation, manipulation of static representations. It is the hemisphere of "this is a tool, used for that, in this category." The right hemisphere is good at broad context, gestalt, embodied recognition, novelty, the apprehension of wholes. It is the hemisphere of "this is a face I have not seen before, in this particular situation, alive."
McGilchrist's strong claim is that healthy cognition requires both, with the right hemisphere taking the lead (it is closer to the actual world) and the left serving as its useful but limited servant. His historical claim is that Western modernity has progressively inverted this relationship, letting the left hemisphere dominate. The result is a culture that is excellent at the manipulation of decontextualised symbols, and increasingly bad at recognising what is in front of it.
You do not have to swallow the historical thesis whole to take the neurological one seriously. The hemispheric data is robust, accumulated over a century of split-brain research, stroke studies, and imaging. The relevance for our purposes is this: the brain itself, the very organ the scientism enthusiast is appealing to, is built around two different ways of knowing. Insisting that only one of them counts is not even consistent with the neuroscience.
Mark Solms is a neuropsychologist who has, almost single-handedly, rebuilt the scientific case that dreams carry meaning. His book The Hidden Spring (2021) is the main source, with earlier groundwork in The Brain and the Inner World (2002). For the dream-skeptic specifically, Solms is your trump card, because he is fighting on their territory and winning.
The popular skeptic line, the one your dinner companion likely absorbed, comes from Allan Hobson and Robert McCarley's 1977 activation-synthesis hypothesis. It says dreams are essentially noise, generated by random pontine activation during REM sleep, which the cortex then dresses up with images. On this account, dreaming is not a meaningful psychological process at all. It is a side effect.
Solms's work demolished this in two main moves. First, he showed that REM sleep and dreaming are dissociable. Patients with damage to the pons (which generates REM) still dream. Patients with damage to specific forebrain structures (the dorsal striatum and parts of the white matter of the frontal lobes) lose dreaming even though REM sleep continues normally. Dreams therefore cannot be reducible to REM activation; they are generated by a separate, motivationally driven system.
Second, he identified the SEEKING system, a dopaminergic circuit centred on the mesolimbic pathway, as the engine of dreaming. SEEKING is the basic motivational system that drives all goal-directed behaviour. Dreams are produced by it, structured by it, and on Solms's account, function in something like the way Freud said they did: as expressions of unresolved wishes and motivational tensions, processed in a regressive hallucinatory mode while the prefrontal cortex is offline.
You do not have to be a Freudian to use this. The point is narrower. The leading neuropsychologist on dreams, working in hard scientific journals, has shown that dreams are motivationally driven products of a goal-seeking system, not random noise. The activation-synthesis dismissal is twenty-five years out of date. Whatever you do with your dream journal, the cleanest version of the scientific consensus is now on your side.
The third witness is the largest, and is by now so well established that even committed physicalists feel obliged to address it. It is what David Chalmers called, in 1995, the Hard Problem of Consciousness. The cleanest earlier statement is Thomas Nagel's 1974 essay What Is It Like to Be a Bat?.
The argument is short. Take any complete physical description of a brain. List every neuron, every synapse, every neurotransmitter concentration, every electrochemical event. Nothing in that description tells you what it is like to be the system being described. The "what it is like" part, the qualia, the felt redness of red, the felt pain of pain, the felt strangeness of dreaming, sits outside the physical description in principle, not just in practice. No matter how complete the third-person account, the first-person fact of experience is not in it.
Chalmers's contribution was to distinguish the easy problems of consciousness, which are the various functional problems of how the brain integrates information, attends, reports, and so on, from the hard problem, which is why any of that functional activity is accompanied by subjective experience at all. The easy problems will yield to neuroscience. The hard problem is not obviously even the right kind of question for neuroscience to handle.
Physicalists have not solved this. They have proposed strategies, eliminativism, illusionism, higher-order theories, integrated information theory, and these are interesting, but the strongest of them concede that subjective experience is genuinely unaccounted for in the standard physical inventory. This is the hardest of hard sciences openly admitting its own ceiling at the most important place: the very fact of being a someone who is doing the science.
These three witnesses are not auxiliaries. They are the modern, credentialled, contemporary case that the territory of meaning is real, that meaning-rich registers like dream interpretation are scientifically defensible, and that even the strongest version of physicalism concedes a ceiling at the place where subjective experience starts. If you are arguing with a neuroscience-respecting skeptic, you can stand on this ground without retreating an inch.
Dreams are the test case for this whole position because they are the place where the two cultures collide most directly. They are biological events. They are also meaning events. They lend themselves to almost every form of bad reasoning. They are also a domain in which careful work has produced real results for over a century. If pluralism does not survive the dream case, it does not survive.
I want to give the skeptic their strongest version first. Not the lazy version. The strongest one.
The skeptic should grant Solms on the neuropsychology and still mount a serious case against interpretation. It runs roughly like this.
All of this is real. It is not the sneering version of the case; it is the serious version. If you cannot answer it, you do not get to defend dream interpretation in good faith. So let us lay it next to the other side.
Here is what I think is the strongest version of the case for, told as I would actually argue it at the table.
First, the neuropsychology. Solms has shown that dreams are not random. They are products of a goal-seeking motivational system, processed in a state where critical evaluation is offline. That is exactly the kind of state in which unresolved psychological material would surface if there were unresolved psychological material to surface. The mechanism does not prove interpretation works, but it removes the dismissal that there is nothing to interpret.
Second, the cross-cultural data. Jungians make a lot of this; the skeptic discounts it; the truth sits in between. There are recurring motifs across cultures (initiation, descent, dismemberment, return, marriage of opposites, the trickster, the wise old figure) that show up in dreams, myths, and folklore in ways that are not fully explained by diffusion or coincidence. The strongest version of this case is not Jung's most cosmic readings of it; it is the more cautious work in comparative mythology and folklore, plus the recurring patterns documented in dream content studies. Something is going on at the level of the human psyche that is partly shared across cultures.
Third, clinical outcomes. Dream-focused therapeutic work, in both psychoanalytic and Jungian traditions, has produced documented therapeutic results across a long period. The evidence is not as clean as for SSRIs in major depression; it is also not as nonexistent as the skeptic implies. The relevant literature includes process-outcome studies on dream-focused therapy, single-case studies with long follow-up, and clinical observations from practitioners who have run thousands of cases. Treat it the way you would treat the case for any other psychological intervention: real but mixed, worth more than no evidence and less than a randomised drug trial.
Fourth, the phenomenological argument. This is the one that lives closest to the bone. When an interpretation hits, something moves in the dreamer. There is a recognition. The body changes. A pattern that had been opaque clarifies. This is not nothing. It is, in fact, what success looks like in any depth-psychological work. It is also the part the skeptic cannot evaluate without participating, which is itself relevant: the data here is first-person, and refusing to look at first-person data does not make the data go away.
Fifth, the pragmatic test. William James proposed that the truth of an idea is partly visible in its consequences when held. Try keeping a dream journal seriously for six months. Track what happens. The empirical result, in my experience and in the experience of many others who have tried, is that life gets richer. Patterns become visible. Decisions become clearer. Stuck places start to move. This is not a knockdown argument, but it is data, and the cost of acquiring it is low.
Sixth, Jung's own empirical work. The Jung who is caricatured as a mystic is the one who, at the Burghölzli, ran the word-association studies that gave us the modern concept of the complex, who developed type theory through painstaking observation, who insisted on grounding his theoretical constructs in clinical material. He was an empirical psychiatrist before he was anything else. The mysticism, where it appears, sits on top of decades of clinical observation, not in place of it.
Finally, the observation-versus-interpretation distinction. This is the move that, if you make it cleanly, defangs almost all the skeptic's objections. The dream itself is the datum. You wrote it down. The fact that you dreamt of a snake on the night your father called is a verifiable observation. Interpretations of that observation are a separate layer, and they should be held lightly, tested against further material, revised, abandoned when they stop generating. The Barnum effect attacks bad interpretation. It does not attack the observation. The unfalsifiability charge attacks interpretation held dogmatically. It does not attack interpretation held provisionally.
One of the lazier moves in the scientism playbook is the implication that the humanities have no quality standards, that anything goes, that interpretation is just storytelling. This is false in a way the speaker would notice immediately if they spent ten minutes in a serious literary studies seminar. There are standards. They are different from falsifiability. They are not absent.
Here are the working criteria I have come to use, drawn from how good readers, critics, therapists, and philosophers in the right column actually evaluate work.
These criteria are real. People are trained in them for years. Bad humanities work is recognisable by failing them in identifiable ways, just as bad physics is recognisable by failing different criteria. The reason the criteria are not the same is that the object is not the same. Expect that amount of exactness in each kind which the nature of the particular subject admits.
It will be tempting to treat pluralism as a polite middle-ground position, the diplomatic option for people who do not want to upset anyone. That is the wrong picture. Pluralism is the strong position. The two opposing positions are weaker, and they are weaker in symmetric ways.
The first opposing position is scientism: the view that the methods of natural science are the only legitimate route to knowledge, and that anything that does not yield to them is either reducible to something that does, or is not knowledge at all. The second opposing position is the one I will call wooism: the view that the methods of natural science are oppressive, partial, culturally contingent, and that the response is to reject epistemics entirely in favour of intuition, vibes, lived experience as self-validating, and so on. Both positions are mistakes. They are the same mistake in opposite directions: the failure to recognise that there are genuinely different kinds of questions in the world.
Pluralism is the position that both of these are confused. The scientist working on the gravitational constant is doing something real and is right to use falsifiability. The Jungian analyst working on a recurring dream is also doing something real, and is right to use coherence, fidelity, and resonance. Neither has the right to colonise the other's domain. The scientist who dismisses dream work because it is not falsifiable is committing the same error as the Jungian who dismisses physics because it is not embodied. Both are reaching across the line and grabbing the wrong tool.
The pluralist stance lets you stay sharp in both registers. You do not get muddier; you get cleaner. You can read Popper on falsification and Hillman on soul in the same week and find that they are addressing different questions and that both of them are right within their question. You can be the person at the table who, when the conversation goes one direction, says "wait, that is a meaning-question, you cannot test it that way" and, when it goes the other direction, says "wait, that is a mechanism-question, you cannot just have a feeling about it." Both moves come from the same stance.
This is the most useful section. The rest of the piece is theory; this section is practice. I want to lay out, as concretely as I can, what to do when the conversation actually happens.
The first move is always concession. Find the part of the skeptic's case that is right, and give it to them out loud. This costs you nothing and it disarms the conversation. The skeptic is usually expecting defensiveness; concession reorients them.
"You are right that a lot of dream interpretation is garbage. The Barnum effect is real. I have seen self-styled dream readers tell people anything they want to hear. If you are arguing against that, I am with you."
If you cannot honestly concede, you are probably defending the wrong thing. Do not try to defend astrology with pluralism. Do not try to defend homeopathy with pluralism. Pluralism is not a license to wave through any claim that wears the right cultural clothes. The point is that meaning-work, done well, in the right register, with appropriate humility, produces real knowledge of a real kind. Bad meaning-work is still bad. Defend the careful version.
Once you have conceded what is true, you make the central move. Try to keep it concrete and short.
"Using science epistemics on meaning-questions is like using a multimeter to evaluate a poem. The multimeter is great, the volts are real, but it cannot tell you whether the poem is any good. The poem is also real. They are different questions, with different tools."
Or, with the Mona Lisa: "Sure, the painting is just atoms. It is also a portrait that has held human attention for five centuries. Both true. The atomic description does not contain the smile. The smile does not contain the chemistry. They are two real descriptions of the same object, in two registers, and neither reduces to the other."
The reason these work is that they do not ask the skeptic to abandon their commitments. They show, by example, that the skeptic already operates in both registers. Nobody actually believes the Mona Lisa is only atoms. The argument is just to make that operative belief explicit.
Different audiences need different arguments. Pick one. Do not stack them; that signals desperation.
This is the move that prevents the conversation from collapsing into "well it is just your opinion." It is not. The observation is the dream itself, written down. The observation is the painting, hanging in the Louvre. The observation is the symptom, the text, the historical event. Those are not opinion; they are data, in the register appropriate to them. Interpretations sit on top, and they should be held with care, tested against further material, revised when they stop working.
The skeptic's attacks on interpretation are mostly fair. The skeptic's attacks on the observations themselves are mostly category errors. Keep the two separate in your own head, and you will be able to keep them separate at the table.
You do not owe anyone conversion. Make the argument once, cleanly, with whichever move you have chosen. If it lands, beautiful, the conversation goes somewhere. If it does not, change the subject. Do not escalate. Trying to convince a committed hard reductionist to take meaning seriously is itself a category error: you are using argument to do what only experience can do. Some people will need a dream of their own to clarify something they cannot clarify any other way. Argument cannot reach there.
The temptation will be to keep pushing because you want to be understood. Notice that wanting, and let it pass. The position is not weakened by their failure to grasp it. You can hold it whether or not the table comes with you.
I want to close with a quieter voice. Most of this piece has been argumentative because most of this piece is intended as armament. The position is under attack from the scientism side, and from the wooism side, and from the general cultural exhaustion that finds it easier to pick one register and shrug about the other. The pluralist stance has to be defended; defending it is part of the work.
But the deeper reason to hold the position is not to win arguments. It is so that you can live a richer life. The whole point of having access to both registers is that you get to be a person who knows what falsifiability is and what a dream might be saying. You get to read Popper in the morning and Hillman in the evening and feel them as collaborators on the same project, which is the slow project of being a human being who pays attention to what is.
Alan Watts liked to point out that the map is not the territory. The Tao Te Ching says the same thing more sharply: the way that can be named is not the eternal way. Both columns of the Dilthey distinction are maps. Neither is the territory. The territory is the thing itself, the actual life you are in the middle of, with the actual people you love and the actual problems you face and the actual death waiting on the other end of all of it. The two registers are two ways of being faithful to that territory. One is faithful by measuring it. The other is faithful by listening to it. We need both, because the territory is the kind of thing that responds to both.
Some truths cannot be measured. They live in a register that measurement does not reach. This is not a failure of measurement; it is a feature of the truths. Other truths cannot be merely felt. They live in a register that requires the discipline of test. This is not a coldness of the truths; it is a feature of the discipline. The mature person holds both. The mature culture cultivates both. The mature stance, when it is challenged from either side, knows what to give up and what to hold.
Hold both. Defend both. Use the right tool for the right question. When you do not know which question you are in, slow down and find out. That is the whole position. The rest is practice.
Organised by intended use rather than by alphabet. Each entry is a real book, real author, with a one-sentence note on what it gives you.