The Architecture of Cognitive Compromise

4–6 minutes

Abortion, Ontological Grammar, and the Limits of Civil Discourse

When federal abortion protections were withdrawn in the United States, many observers treated the event as a policy reversal, a judicial shift, or a partisan victory. Those are surface descriptions. They are not wrong. They are simply too thin.

What was exposed was not a failure of dialogue. It was the collision of ontological grammars.

Audio: NotebookLM summary podcast.

1. Thick Concepts and the Illusion of Neutral Ground

book cover

Bernard Williams famously distinguished between ‘thin’ moral terms (good, bad, right) and ‘thick’ ones (cruel, courageous, treacherous), where description and evaluation are fused.

Abortion is not a thin concept. It is thick all the way down.

For one framework, the operative grammar is something like:

  • Life begins at conception.
  • The foetus is a rights-bearing entity.
  • Termination is morally equivalent to killing.

For another:

  • Personhood is socially and biologically emergent.
  • Bodily autonomy is foundational.
  • Forced gestation is coercive.

Notice that these are not competing policies. They are competing ontological commitments about what exists, what counts as a person, and what kind of being a pregnant body is.

Argument across this divide does not merely contest conclusions. It contests the background conditions under which reasons register as reasons.

This is not ‘people see the world differently’. It is: people parse reality through grammars that do not commute.

2. Ontological Grammar: Where Deliberation Stops

By ‘ontological grammar’, I do not mean syntax in the Saussurean or Chomskyan sense. I mean the pre-reflective substrate that structures what appears salient, real, morally charged, or negligible.

We deliberate within grammars. We do not deliberate our way into them.

Liberal Enlightenment optimism assumes that if disagreement persists, more information, better reasoning, or improved empathy will close the gap. But if the dispute concerns the very ontology of ‘life’, ‘person’, or ‘rights’, no amount of fact-sharing resolves the issue. The disagreement is upstream of facts.

The closure of federal abortion access did not prove that one side reasoned better. It demonstrated that institutional containment had failed.

3. Biopower and the Management of Bodies

Michel Foucault gives us a crucial lens: biopower. Modern states do not merely govern territory; they administer life. Birth rates, mortality, sexuality, health – these become objects of policy.

Abortion sits directly inside this matrix.

A state that restricts abortion is not only expressing moral judgment. It is reallocating control over reproductive capacity. It is asserting a claim over which bodies count, which futures are permitted, and which biological processes are subject to regulation.

The conflict is therefore not purely ethical. It is biopolitical.

And what appears as ‘civil discourse’ around abortion is often possible only so long as institutional frameworks keep both grammars partially satisfied. When federal protections existed, they acted as a stabilising superstructure. Remove that, and the ontological conflict becomes naked.

4. Habitus and the Illusion of Reasoned Consensus

Pierre Bourdieu would remind us that our dispositions are not self-authored. Habitus sedimented through family, religion, class, and institutional life shapes what feels obvious, outrageous, or unthinkable.

People do not wake up one morning and choose an abortion ontology.

They inherit it. It becomes embodied common sense.

Thus, when someone says, ‘Surely we can agree that making a person feel whole is more important than ideological purity’, they are already speaking from within a grammar that prioritises individual authenticity and psychological coherence. That priority is not universal. It is historically situated.

Compromise is not achieved by stepping outside habitus. It is achieved when institutional and social conditions allow divergent grammars to coexist without totalising one another.

5. The Popperian Threshold

Karl Popper warned of the ‘paradox of tolerance‘: unlimited tolerance may enable intolerant forces to eliminate tolerance itself.

In particularly virulent climates, appeals to compromise are heard not as gestures of goodwill but as tactical weakness.

When one faction succeeds in unilaterally redefining the legal status of abortion at a federal level, it is not merely participating in discourse. It is altering the biopolitical infrastructure. Once altered, the range of permissible disagreement narrows.

Civil discourse, then, is not a natural equilibrium. It is a managed condition sustained by institutional design, social trust, and shared legibility.

NB: Popper’s paradox of tolerance is often invoked as a moral axiom. But it is better understood as a self-protective clause internal to liberal ontology. It presupposes a shared commitment to rational exchange. When that commitment erodes, the paradox does not resolve disagreement; it merely marks the point at which biopower intervenes to preserve a regime.

6. Why This Is Not Just ‘People Disagree’

The lay intuition – ‘people see the world differently’ – is descriptively correct and analytically useless.

What the ontological grammar model adds is structure:

  • Disagreements cluster around thick concepts.
  • Thick concepts fuse description and evaluation.
  • Frameworks determine what counts as a reason.
  • Institutions temporarily stabilise incompatible grammars.
  • When stabilisation weakens, conflict appears irreconcilable.

Abortion is not uniquely polarising because people are irrational. It is polarising because it touches ontological primitives: life, personhood, autonomy, and obligation.

In such cases, ‘compromise’ is not achieved by discovering a middle truth. It is achieved – if at all – by constructing a legal and institutional arrangement that both grammars can grudgingly inhabit.

NotebookLM Infographic

7. The Uncomfortable Conclusion

The Enlightenment story tells us that disagreement is a surface phenomenon, curable by better reasoning.

The ontological grammar story tells us something harsher: some disagreements are not resolvable through language because they are about the conditions under which language binds.

This does not entail quietism. It entails clarity.

Civil discourse is not proof that grammars have converged. It is evidence that power, institutions, and habitus have aligned sufficiently to prevent rupture.

When that alignment shifts, the illusion of shared ontology evaporates.

And what we are left with is not failed reasoning – but exposed foundations.


I planned to use prostitution and anti-natalism as other cases for elucidation, but I see this has already grown long. I’ll reserve these are others for another day and time.

I Need a Break

5–7 minutes

More precisely, I need less sleep and longer days – preferably twice as long. I’ve been writing almost non-stop for the better part of a week: fourteen- to sixteen-hour days, fuelled by irritation and the stubborn belief that if I just keep reading, something will finally click into place.

I’m not complaining. This is a virtuous cycle.
Reading leads to writing. Writing demands more reading. Eventually, the loop closes into something that looks suspiciously like progress.

Audio: Short NotebookLM summary podcast on this topic.

Still, there’s a bottleneck.

Because some of this work – the work I’m most excited about – I’m deliberately not publishing yet. Journals, bless their glacial hearts, don’t much care for prior publication. So ideas sit in limbo for six to eighteen months, locked in a room like argumentative houseplants, slowly growing sideways.

From the perspective of someone who thinks in public, this is maddening.

Now add AI to the mix.

This is where things get dangerous.

I’ll feed ChatGPT a thesis, a skeletal structure, notes, and references. I ask what I’m missing. It obliges – often helpfully – by pointing me toward adjacent thinkers and relevant literature, complete with page numbers. From there, I verify, hunt down the sources, skim, read, discard, or integrate.

And every so often, I stumble across something that makes me swear out loud.

This week, it was Bernard Williams.

I’ve cited Ethics and the Limits of Philosophy before. But this time, I actually sat down and read it properly. Which immediately prompted the thought:

Why didn’t I read this sooner?

Williams dismantles moral objectivity with the calm precision of someone who knows the Enlightenment project has already lost – he just hasn’t told everyone yet. Thick and thin moral concepts, locality, non-extensibility, the collapse of universal moral reason at scale – yes, yes, yes. He published this in 1985. Fine. I’ll survive.

But then I went further.

Williams shows that morality fails between people at scale.
I argue that it fails within a single person over time.

That became my second paper.

And this is where things went off the rails.

Because in the course of writing that paper, I dipped into Hart’s The Concept of Law and Endicott’s Vagueness in Law. These are not fringe polemics. These are law textbooks. For law students. People allegedly trained to parse language for a living.

And what I found was… astonishing.

Let me paraphrase the admissions:

Image: When the law is vague, judicial decisions may be unconstrained by the law.

Endicott: “By upsetting the standard view of adjudication, the book reaches conclusions that some people find horrible: when the law is vague, judicial decision- making will in some cases be unconstrained by the law. It is impossible in principle for judges always to treat like cases alike. Predictability in the law is to some extent unattainable. Moreover, I argue in Chapter 9,2 that vagueness cannot be eliminated from law. These conclusions might seem to imply that the rule of law is, at least to some extent, conceptually impossible.”

Image: Vagueness is inevitable. Deal with it.

Endicott: “Secondly, I do not claim that vagueness is a purely linguistic feature of law. And the book relies on no claim about the relation between law and language. These points must be stressed, because vagueness is commonly thought of as a linguistic phenomenon. And. indeed, most of the discussion in the book concerns the vagueness of linguistic expressions. But the indeterminacy claim is not just a claim about language (so I argue in Chapter 3.12). So. for example, the claim in Chapter 6 that general evaluative and normative expressions are necessarily vague is not just a claim about the word ‘good’ and the word ‘right1: it is a claim about any linguistic expression in which we could conceivably express general evaluative and normative judgments. It therefore includes a claim about what is good and what is right.”

Image: Whether law is morally valuable to a community is not my concern. Justice and the rule of law may be political virtues — or not. I don’t defend them here.

Endicott: “Disputes between legal positivists and natural law theorists have concerned not only the relation between law and adjudication, but also the relation between law and morality. Here I take no general position on the intrinsic moral value of law. I do rely on the claims that law can be valuable to a community, and that justice and the rule of law are two ideals which a com- munity can intelligibly pursue as political virtues. Even those claims are controversial (Kelsen and some of the theorists discussed in Chapter 2 have controverted them ). But I do not defend them here. This work aims to show that the indeterminacy claim does nothing to threaten the pursuit of justice and the rule of law. Those ideals cannot be well understood if we try to make them depend on determinacy in the requirements of the law.”

Say what?

Read together – not even uncharitably – the message is clear:

Law is indeterminate.
Indeterminacy is unavoidable.
And whether law is good, just, or valuable is… optional.

The subtext isn’t even hiding.

Law is a power structure first.
If it happens to align with justice, fairness, or communal value, well, lovely. A bonus. Champagne all round.

This does not sit well with a sceptical cynic.

What really broke me, though, wasn’t the argument itself. Philosophers make grim claims all the time. What broke me was the silence around it.

How does this pass under the radar?

How do cohorts of law students – drilled in textual analysis, trained to read footnotes like tea leaves – not trip over this elephant stampede? How do they graduate believing they’re upholding inalienable rights, rather than participating in a managed system of coercion that occasionally behaves itself?

Self-preservation, I suppose.
Wilful ignorance.
Professional cosplay.

I’ve seen this before.

As an economist, ask the wrong foundational question, and you’re instantly radioactive. Persona non grata. Careers don’t end with explosions — they end with polite silence and no invitations.

I probably should have committed to heterodox philosophy from the start.
Or stayed a musician.

I remember leaving graduate school, putting on a suit, and feeling like I was wearing a costume. Cosplay, before we had the word. “Business professional” as a role, not an identity.

I’ve always felt intellectually capable of doing whatever I set out to do. My temperament, however, has never agreed to play along.

Which is perhaps why diagnosing ontologies comes so naturally. Once you see the scaffolding, you can’t unsee it – whether it’s metaphysics, jurisprudence, or a corporate department pretending it has a mission.

Then David Graeber came along with Bullshit Jobs, and I remember thinking:
Thank God. It’s not just me.

So yes. I need a break.

I need sleep.
I need silence.
I need to stop reading law books that accidentally admit they’re about power and then act surprised when someone notices.

Mostly, I need to type:

WTAF?

And then go outside.