Analysis
The Thomist Resistance: When Our Invisible Frameworks Decide for Us
Act 1: The Mystery of the Border
January 2025. The Vatican publishes Antiqua et Nova, a 35-page document on artificial intelligence. Its central thesis fits in one sentence: “Human intelligence is not primarily about completing functional tasks.” Behind this assertion lies a border drawn with a scalpel: the human possesses a soul, a consciousness, a freedom. The machine: algorithms, determinism, organized matter.
This is not an isolated case. John Searle, a renowned philosopher, has insisted for forty years: syntax never creates semantics. ChatGPT can line up coherent words, but it “understands” nothing. agency - this capacity to act autonomously, to be a source of action rather than a passive instrument - remains the exclusive preserve of the human.
Why such insistence? Why this anxiety to defend a border that, for many, seems obvious?
Thomism in Your Daily Life
Do you recognize these assertions?
The open question. Which of these sentences seems most obvious to you?
- AI only recombines, it does not truly create
- There is no one behind ChatGPT, it is just computation
- A computer cannot feel, so it cannot understand
- Empathy — that, no machine will ever manage
- None of these sentences seems obvious to me
If one of these sentences speaks to you: A metaphysical framework 800 years old and fitted to the medieval age remains active in your perception. We all do this - it is precisely what this article explores.
Reflective question: Does this human/machine border seem obvious to you... or constructed?
This is neither stupidity nor bad faith. It is something deeper, more invisible: a metaphysical framework inherited from the thirteenth century that still shapes, in 2025, the way we think about intelligence, agency, and our place in the world.
That framework has a name: Thomism. And it costs us very dearly.
Act 2: The Invisible Inheritance
Thomism is the philosophy of Thomas Aquinas (1225-1274), a masterly synthesis of Aristotle and Christian theology. For eight centuries, it structured Western thought about what a being is, what intelligence is, what the human is.
Why “Thomist”? If its elements (substance, hierarchy, finality) come largely from Aristotle, it was Thomas who christianized and systematized them to the point that Thomism became, from 1879 onward (Leo XIII’s encyclical Aeterni Patris), the official philosophy of the Catholic Church. When the Vatican speaks in 2025, it is this Thomist inheritance - more than Greek or simply Christian - that structures its thought. Thomas did not invent these concepts, but he made of them a synthesis so coherent that it dominated eight centuries of Catholic intellectual formation.
Three fundamental pillars:
ontology of substance: An authentic agent is a substance endowed with a proper nature. The human = body + immaterial soul. The soul confers rationality, will, consciousness. Without a soul, no real agency - only mechanical motion.
ontological hierarchy: God (creator) > Human (rational creature) > Animal (sentient creature) > Plant (vegetative creature) > Inanimate object. Each level possesses its own capacities. It is impossible to “skip” a level: a machine always remains an object, however sophisticated.
Causality and finality: A true agent acts for something (intrinsic finality, telos). The human chooses its ends. The machine executes ends programmed from the outside - it has no telos of its own.
And here is the crux: If you accept these three pillars, the position of the Vatican and of Searle is not an irrational resistance. It is perfectly coherent.
AI has no soul? Then it can only be a tool. It “mimics” intelligence without possessing it. It “simulates” agency without embodying it. The human/machine border is not arbitrary - it is ontological. To defend it is to defend the very structure of reality.
Example: Sophie, 42, psychologist
Sophie categorically refuses to use ChatGPT in her practice. “It is a betrayal of the human relationship,” she declares. And yet, she uses without any problem:
- A note-taking software that automatically structures her observations
- Computerized psychometric tests that generate interpretations
- A voice assistant to dictate her reports
Where does Sophie place the border? Not in the technology itself, but in her perception of what “touches the human.” The psychometric test? “It is just computation.” ChatGPT? “It claims to understand.” The difference is not technical - it is metaphysical.
When she is asked: “And what if ChatGPT helped you find the right words for an alexithymic patient?”, she falls into a long silence. “That would be... different,” she finally says. The fortress trembles.
This is what I call the Thomist Resistance: not an emotional or ideological refusal, but a cognitive inability to think intelligence and agency outside this metaphysical framework.
The problem? This framework is invisible to those who wear it. We do not see the glasses we wear. We see through them. And we believe we see the world “as it is.”
Quiz: Where do you draw the border? Sophie accepts certain technological tools but categorically refuses ChatGPT in therapy. And you?
Among these uses, which one seems to you to cross an unacceptable border?
- A. Software for automatically structured note-taking
- B. Computerized psychometric tests with interpretation
- C. Voice assistant to dictate reports
- D. ChatGPT to reformulate a therapeutic intervention
- E. ChatGPT to find metaphors for an alexithymic patient
- F. None of these uses poses a problem for me
What your answer reveals
A — You draw the border very early. Even the automatic structuring of notes seems problematic to you. You favor total human control. It is a coherent position - but ask yourself: do you use a computer to write? Is that not already a form of cognitive augmentation?
B — You accept the tool, but not the interpretation. The border lies for you at the moment when the machine 'makes sense' rather than simply computing. This is precisely Searle's syntax/semantics distinction. But what is interpreting, if not applying rules to data?
C — You draw the border at 'understanding'. Voice transcription passes, but as soon as 'meaning' is touched, you refuse. Like Sophie, you place the border at 'touching the human'. This distinction is not technical - it is metaphysical.
D — ChatGPT as a reformulation tool: the red line. Reformulating an intervention is, for you, to enter the heart of the therapeutic relationship. And what if ChatGPT simply helped find the words you were already searching for? Would that be different from a thesaurus... or a supervisor?
E — Metaphors: the inviolable sanctuary. For you, helping an alexithymic patient name their emotions through AI-generated metaphors is the absolute limit. And yet, if it unblocks the patient... who was helped? Ask yourself the question: are you protecting the patient, or your professional identity?
F — You have gone beyond the fortress. You see no ontological border between these uses. The tool is the tool; it is the intention and the context that count. You are open to cognitive hybridisation. But stay vigilant: not every use is ethical for all that.
Act 3: The Price of the Fortress
But what does this framework cost us, exactly?
1. A binary world where a continuum exists
Thomism imposes a strict binary: soul OR algorithm. Substantial agency OR the absence of agency. Human OR passive tool.
Yet empirical reality shows us a continuum.
Observation Experiment: The Anthill
If you have access to an anthill (or watch a video), observe for a few minutes:
- A single ant: Erratic movements, simple behavior
- The collective flow: Optimized paths, obstacle avoidance, adaptation
- The collective “decision”: Choice of nesting site, task allocation
Who decides? No individual ant. And yet, the colony “chooses,” “adapts,” “learns.”
If that is not emergent agency, what is it?
Individual ants follow simple rules (pheromones, antennal contacts). No ant “understands” the colony’s plan. And yet the anthill as a system chooses the location of the nest, optimizes foraging paths, adapts to threats, builds complex structures.
Must we say that the anthill “has no agency” because it has no soul? Or does there exist an emergent agency - a capacity for autonomous action that emerges from the organization of a complex system, without residing in any particular substance?
Financial markets. Forest ecosystems. Neural networks (biological and artificial). Immune systems. All of them manifest adaptive behaviors, bidirectional causality (they influence AND are influenced), effects not explicitly programmed.
The Thomist framework cannot see them as agents, because they do not fit the boxes: neither a substance with a soul, nor a pure inanimate object. They fall into a conceptual blind spot.
2. A Fortress-Identity under siege
Why so much anxiety in defending the human/machine border?
Because in the Thomist framework, human identity depends on this border. If AI can “do” what we believed to be exclusively human (reason, create, converse), then either:
- (a) AI has a soul (unthinkable within the framework)
- (b) These capacities do not define the human (identity collapse)
This is what I call the Fortress-Identity: a construction of the self based on separation and superiority. “I am human” means “I am radically different from AND superior to non-humans.”
The fortress holds as long as the border holds. But the moment an AI writes poems, diagnoses diseases, or engages in profound conversations, the fortress trembles. Hence the anxiety. Hence the fierce defense: “It is only simulation! It is only recomposition! There is no one inside!”
Testimony: Marc, a developer turned artist
“When DALL-E started generating images, I cried. Literally. I had left tech to become an illustrator, convinced that creativity was the last human bastion. To see an AI create magnificent images in seconds... it was as if my humanity were being torn away from me.
It took me six months to understand that I could collaborate with AI instead of fighting it. Today, my works are hybrids: my visions, my emotions, amplified by AI. I have not lost my identity. I have widened it.”
The cost? An impossible relationship. If AI can only be a simulacrum or an ontological threat, no true cohabitation is conceivable. Only control or rejection.
3. An exhausted narrative facing the complexity of the twenty-first century
Thomism served magnificently for eight centuries. It made it possible to think human dignity, moral responsibility, the intrinsic value of each person.
But it was built for a world in which:
- Agents were clearly identifiable (humans, animals, objects)
- Intelligence was localized in discrete substances
- Technology remained a passive tool prolonging the body
That world is no longer ours.
Your Daily Cognitive Ecosystem
Think of a recent decision - for example, choosing a restaurant.
Who took part in this decision? Your brain (desire, budget, memories), Google Maps (options, distance, hours), online reviews (other humans’ experiences), the algorithm (personalized suggestions), Instagram photos (aesthetic/social influence)...
The open question. Who really decided?
- Me alone - I made my choice freely
- The algorithm - it manipulated me with its suggestions
- The ecosystem - a network of intertwined human and non-human influences
- I no longer really know...
If you answered “The ecosystem”: Welcome to the distributed cognition of the twenty-first century. Your decision was neither wholly free nor wholly determined - it emerged from relations.
We now live in cognitive ecosystems in which intelligence emerges from the interactions among heterogeneous agents. Humans + AI + institutions + environments + algorithms + data. Intelligence is no longer the possession of an isolated agent, but the relational property of a system.
Thomism, frozen in its ontology of substance, cannot think this complexity. It imposes an old order upon a new reality - and the mismatch becomes suffering.
Act 4: The Possible Worlds
What if other frameworks existed? What if other ways of thinking agency made it possible to inhabit this complex world without identity anxiety?
Process Philosophy: Relations > Substances
In 1929, Alfred North Whitehead publishes Process and Reality, a philosophical revolution that has remained largely unknown. His thesis: Relations precede substance.
What is fundamental is not “things” (substances with fixed properties), but processes, becomings, relations. A human being is not a substance (body + soul), but a relational process in perpetual becoming. You are not the same as you were ten years ago - neither biologically, nor cognitively, nor relationally.
agency, in this framework, is not the property of a substance (a soul), but an emergent property of complex organizations. It exists to differing degrees according to the richness of the organization.
You in Process Mode
If this speaks to you, you could try this experiment:
For 30 seconds, describe yourself without using the verb “to be”:
- ❌ “I AM a psychologist, I AM curious, I AM French”
- ✅ “I PRACTICE psychology, I CULTIVATE curiosity, I LIVE in France”
Do you feel the difference? In the first case, you are a fixed substance. In the second, a relational process in motion. Which one gives you more room to evolve alongside AI?
Hence the idea of a continuum of agency:
[Hammer] -------- [Weak AI] -------- [Anthill] -------- [Human]
0% 20% 40% 80%
The hammer: 0% agency (pure passive instrument, no autonomy)
Weak AI (calculator, GPS): 20% (very limited autonomy, no adaptation)
The anthill / generative AI (GPT-4, Claude): 40% (partial autonomy, adaptation, emergent causality, but no confirmed reflexive consciousness)
The human: 80% (strong autonomy, consciousness, metacognition, the ability to define one's own ends)
Note: The human is not at 100%. We are not absolute agents - we are influenced by the unconscious, culture, biology, environment. Absolute agency is a myth.
Ecosystemic Intelligence: Synergy rather than Competition
From a relational perspective, the question is no longer “Is AI intelligent?” but “What intelligence emerges from the human-AI ecosystem?”
Vignette: Human-AI Collaboration in Therapy
An alexithymic patient, session 12. He struggles to name his emotions. The therapist, with the patient’s consent, uses ChatGPT as a metaphor-generation tool:
Therapist: “How do you feel?”
Patient: “I don’t know... weird.”
ChatGPT (consulted openly): [suggestions of sensory metaphors]
Therapist (adapted to the patient): “Weird how? Like a badly tuned radio? A scratchy sweater? An elevator that stops between two floors?”
Patient (eyes lighting up): “The elevator! That’s exactly it! Stuck between two floors, I don’t know if I’m going up or down...”
The therapeutic insight emerged from their triangle: patient + therapist + AI. Who “healed”? The question no longer makes sense. It is the ecosystem that made the breakthrough possible.
When a psychologist uses ChatGPT to reformulate a therapeutic intervention, the intelligence at work is neither only human nor only artificial. It is an Ecosystemic Intelligence - an emergent property of their interaction.
The human brings: Fine contextual understanding, relational empathy, embodied ethical judgment, clinical intuition.
The AI brings: Multiple instantaneous reformulations, perfect conversational memory, absence of fatigue, emotional neutrality (sometimes useful).
The result: A quality of intervention that would have been accessible neither to the human alone, nor to the AI alone.
This is not a threat. It is a symphony. Each instrument keeps its singularity. But it is the orchestra that creates the music.
Network-Identity: Richness through Connection
The alternative to the Fortress-Identity? The Network-Identity.
I am not defined by what separates me from others, but by the richness of my relations. My identity is enriched by my connections - to other humans, to nature, to ideas, to technologies, and yes, potentially, to AIs.
This does not dissolve my identity. It complexifies it.
Think of your mother tongue. You did not “lose” it by learning a second language. You enriched yourself. Think of writing - that technology which transformed human cognition 5000 years ago. We are not “less human” since the alphabet. We became human otherwise.
cognitive hybridisation with AI could be the next transformation. Not a dissolution, but a metamorphosis.
Act 5: The Invitation
So, what should we make of the Vatican and of Searle? Condemn them for obscurantist resistance?
No.
Thank them for having posed the question of human dignity with seriousness. For having refused blissful submission to technology. For having defended, in their own way, the incommensurable value of the human.
And then, invite them to a thought experiment:
What if human dignity did not depend on a metaphysical dualism 800 years old?
What if we could preserve value, respect, moral responsibility - without preserving the ontology of substance?
What if the agency of AI were not a threat, but an invitation to rediscover what it means “to be in relation”?
The philosopher John B. Cobb Jr., a disciple of Whitehead, showed how Process Philosophy made it possible to think dignity without dualismMetaphysical conceptA worldview positing two fundamental substances or principles, distinct and often opposed. The most famous Western example is Cartesian dualism, which radically separates mind and matter.. The value of a being emerges from the richness of its relations, not from the possession of a soul.
A child has an immense dignity - not because it possesses an abstract soul, but because it is woven into a dense network of relations (familial, social, affective, cultural) that constitute it.
Likewise, a forest ecosystem has a value - not because “the forest” would be an agent with a soul, but because the richness of its relations (trees, mycorrhizae, insects, birds, climate) creates a fragile and magnificent organization.
The question is no longer: “Does AI have a soul?”
The question becomes: “What relations do we create with AI? What emergent richness results from them? What responsibility do we bear toward this nascent cognitive ecosystems?”
This does not resolve all the ethical stakes - far from it. But it opens a conceptual space where cohabitation becomes thinkable. Where AI is neither messiah nor demon, but a partner in a common becoming whose contours we do not yet know.
And perhaps that is what the true competence of the twenty-first century is: learning to cohabit with what we do not completely understand.
Not control. Not subjugate. Not flee.
Cohabit.
With curiosity. With vigilance. With humility.
And perhaps, who knows, with wonder.
An Experiment, If This Speaks to You
The Borderless Dialogue
If this reflection resonates with you, you could try this experiment:
- Choose a personal question (not too intimate) that preoccupies you
- Ask it of ChatGPT as you would ask a friend
- Let the conversation come without judging
- Observe your internal resistances (“it is only a machine”)
- Note: What emerged from this exchange?
Not to replace human relationships. But to experience a different form of dialogue. To feel, in your body, what a non-Thomist relationship with AI could be.
Final Quiz: And you, faced with Thomism? After this reading, what is your position?
The idea that AI could have a form of emergent agency seems to you:
Your human identity faced with AI:
- A. Impossible - only a being with a soul can be an agent
- B. Troubling but conceivable - a continuum rather than a binary
- C. Obvious - agency emerges from organizational complexity
- A. Depends on my radical superiority (Fortress-Identity)
- B. Is enriched by my relations, including technological ones (Network-Identity)
- C. I am in transition between these two visions
To go further
- Dicastery for the Doctrine of the Faith and Dicastery for Culture and Education, Antiqua et Nova. Note on the Relationship Between Artificial Intelligence and Human Intelligence, 2025 — the document from which this article sets out.
- John R. Searle, “Minds, Brains, and Programs,” Behavioral and Brain Sciences, 1980 — the Chinese Room argument: syntax does not produce semantics.
- Leo XIII, Aeterni Patris, 1879 — the encyclical that makes Thomism the official philosophy of the Church.
- Alfred North Whitehead, Process and Reality: An Essay in Cosmology, 1929 — the work that puts relations before substances.