The discovery

The Agency Lab

Your arm moves without you telling it to. You can't move it when you will it. Same brain damage, same limb. For decades the assumption was that one drove the other. A systematic study found no such link — the severity of one doesn’t track the other.

Alien limb phenomenon is one of neurology's strangest symptoms: your own limb acts like it belongs to someone else. It grabs, reaches, interferes with your intended movements — autonomous, unintended, unsettling. Apraxia is the opposite: you intend a movement but your brain can't execute it, so your limb stays silent. Both show up in corticobasal syndrome, both involve the same damaged brain regions, and for decades the field assumed one caused the other: apraxia creates the motor failure that manifests as alien limb.

That assumption was never systematically tested. Until now. A structured cohort study using Bayesian analysis found no support for it: apraxia severity did not predict alien limb. Same damage, same patient, same limb — but in this cohort the two didn’t track each other. They behave like parallel failures, not a causal chain.

Within this syndrome, that hints at something bigger — that the brain may run “execute my intention” and “is this action mine?” as separate systems. But that bigger reading is the lab’s, not the paper’s: the study itself stays inside corticobasal syndrome.

The study · Lewis-Smith, Wolpe, Ghosh & Rowe · Cambridge · 2019

Sample: 30 corticobasal-syndrome patients + 25 progressive-supranuclear-palsy (Richardson) disease controls; 28/30 (93%) apraxic, 25/30 (83%) with one or more alien-limb phenomena. Method: a structured alien-limb questionnaire (ownership, control and unwanted-movement phenomena), standardized apraxia testing, regression + Bayesian model comparison. Finding: no significant association between apraxia severity and alien limb; Bayesian comparison found apraxia severity unlikely to be a positive predictor — the assumed causal link wasn’t supported.

A plausible, untested assumption — that these two motor failures are causally linked — met its first systematic test, and the data didn’t back it. One careful cohort isn’t a demolition; it’s a crack worth chasing. Same lesion, seemingly different mechanisms.

Two forms of motor breakdown

Expected vs. Found

What the field was built on. What the data showed instead.

EXPECTED — for decades

Causally linked

Apraxia is the primary motor failure — the inability to execute a willed movement. Alien limb is the downstream consequence, a secondary manifestation of that same apraxia. Severity of apraxia predicts severity of alien limb. They're two faces of the same breakdown.

FOUND — Bayesian analysis

Functionally independent

Both arise from damage to the same brain regions (supplementary motor area, prefrontal, parietal cortex), but they break different systems. Apraxia prevents intended action. Alien limb produces unintended action. In this cohort, one doesn’t track the other — the assumed dependence isn’t there.

The phenomenology

Apraxia: "I want to reach for the cup. My brain understands the goal. But the command never leaves my cortex. My hand stays still. I can see what I want to do, but the motor execution is blocked."

Alien limb: "I didn't tell my arm to do that. It reached on its own. It grabbed the other person's coffee cup. It interferes with what my other hand is trying to do. It acts like it doesn't belong to me."

The field assumed: Apraxia is the root cause; alien limb is what happens when apraxia goes wrong. One broken system producing two symptoms.

The data showed: Same brain lesion, different systems. Intention-to-execution (apraxia) is independent from intention-monitoring (alien limb). A patient can be severely apraxic with minimal alien limb, or vice versa.

The reframe: The brain doesn't control voluntary movement as a single system. It has parallel channels: one that says "execute this intention," another that says "is this action mine or not?" When both break in the same lesion, they break independently.

How assumptions hide

The Assumption Moment

An intuitive idea that seemed airtight. Nobody tested it because it seemed obvious.

Before 2019 (the assumption era)

Alien limb has been documented since 1908 — Goldstein's patient. By 2019, both apraxia and alien limb were well-known to occur in CBS. The logic was straightforward and anatomically plausible:

Both involve motor control breakdown
Both localize to supplementary motor area and prefrontal cortex
They co-occur constantly in the same patients and same limbs
Therefore: apraxia must drive alien limb, or they're expressions of a single failure

What was never done: Quantitative measurement. Structured cohort. Statistical test of the causal hypothesis. Everything prior was scattered case reports and small clinical series. The assumption was never challenged because it felt inevitable.

1908–2018

Alien limb documented since 1908 (Goldstein). The idea that apraxia drives it in CBS was a later, plausible suggestion — never put to a systematic, quantitative test.

2019

First structured cohort: Wolpe et al. assemble 30 CBS patients, develop standardized questionnaire for alien limb phenomena, measure apraxia severity standardly, run Bayesian regression.

Result

The assumption breaks: No evidence of causal link. Apraxia severity is a poor predictor of alien limb occurrence. Bayesian odds strongly favor dissociation. The assumed link isn’t supported — and no one had asked the question with a cohort and a statistic before.

Why this happened

The assumption was not stupid — it was intuitive. Motor breakdown + motor breakdown in the same lesion = related phenomena. But intuition isn't proof, and case reports can't answer causal questions. You need a cohort, a measurement instrument, and statistical power. Until someone builds that, the assumption stands unchallenged, even if it's wrong.

Your position

Where do you stand?

Lock your position before you scroll to the answer. No answer key. No right or wrong here — only calibration.

What's your read on the Wolpe finding?

Crank — the finding is wrong

One cohort of 30 patients doesn't overturn 110 years of clinical observation. The field already knows apraxia and alien limb are related. This is a measurement quirk, not a paradigm shift.

Provisional — check, don't bless

Interesting finding. Wait for independent replication in another CBS cohort. Wolpe et al. is careful work, but one study on a rare disease needs confirmation.

Copernicus — the assumption was wrong

The field assumed causal link for 110 years. Systematic data shows dissociation. The brain doesn't control voluntary action as a single system — it has independent channels. This rewrites how we think about agency.

How confident are you?

50%

What would move you?

Independent replication in another CBS cohort

Brain imaging showing separate neural correlates for apraxia vs. alien limb

A targeted intervention that separates the two phenomena clinically

Archival re-analysis of prior case reports showing historical cases also dissociate

YOUR COMMIT · locked

The pattern

Sledgehammer Room

Papers that stand on a stool and kick out the legs. Same structure across unrelated fields.

What belongs in the Sledgehammer Room

A paper that takes a load-bearing assumption of its own field and demonstrates — with systematic data — that the assumption is wrong. Not edge cases. Not refinements. Direct contradiction of something the whole field is built on.

The pattern

1. The assumption: Obvious, intuitive, built into decades of case reports and clinical wisdom. Nobody tests it because it seems inevitable.

2. The systematic study: Someone assembles a cohort, develops measurement instruments, and asks the question statistically. Usually they find what everyone expects.

3. The surprise: The data contradicts the assumption. Not dramatically — carefully, quantitatively, with Bayesian odds or confidence intervals showing the assumption was backwards.

4. The reckoning: The field now has to ask: what else are we wrong about because we never tested it?

Why this matters: Paradigm shifts don't come from mystery. They come from asking obvious questions with rigorous methods. The field wasn't stupid to assume apraxia and alien limb were linked. It just never ran the experiment.

Companions in the Sledgehammer Room

The Continuum (knee osteoarthritis): Expected: multiple distinct diseases. Found: signs of one continuous condition.

The Swing (cosmology): Expected: an isotropic universe. Found: a contested claim of large-scale anisotropy (still unsettled).

Three different fields. Same shape. The assumption gets challenged by data. The field learns it was building on something wrong.

Sources & an honest read

The paper: Lewis-Smith DJ, Wolpe N, Ghosh BCP, Rowe JB. “Alien limb in the corticobasal syndrome: phenomenological characteristics and relationship to apraxia.” Journal of Neurology (2019). doi.org/10.1007/s00415-019-09672-8 · open full text (PMC).

What the paper supports: apraxia severity does not predict alien limb — the assumed causal link isn’t there in this cohort. What is the lab’s framing (not the paper’s): the “paradigm shift,” the “Sledgehammer Room,” and the leap to “how agency works” in general. The honest size of it: one cohort of 30 in a rare disease, not yet independently replicated — by this wing’s own four-question test it has cleared the first question and still owes the other three. Promising, not proven.