What are you allowed to use?#
Does needing support show that you do not understand?
Week 6 · A self-guided lesson
Where Week 5 left us#
Week 5 built a seven-rung ladder, but Recall and Reconstruction assume that removing a source makes the evidence stronger. This week tests that assumption by varying permitted resources. We move from the lock to a fictional courier; the examination question continues even though the setting changes.
How to use this lesson#
Read at your own pace; no presenter is needed. Allow around 30 minutes, plus 15–20 minutes for the investigation. Have somewhere to write short answers.
At a Pause and write slide, attempt the question before advancing. The next slide explains the reasoning so you can compare it with yours.
No navigation app, bicycle or AI account is required. All people and situations are fictional.
The situation: a courier and a phone#
Elena works as a bicycle courier. Her phone gives turn-by-turn directions to each delivery address.
Today’s delivery is three suburbs from her usual route, somewhere she has never ridden before. She follows the directions, arrives on time, and hands over the parcel.
What did the delivery show?#
Elena followed the directions correctly. That much we observed directly.
Did she demonstrate that she understands the area she just rode through? An examiner watching only a successful delivery cannot yet tell whether Elena engaged with the route, or executed instructions the way a machine would.
Keep the question separate from whether the delivery succeeded.
The obvious next move#
Take the phone away and ask Elena to do it again. If she can, she really knew the route. If she cannot, she was only following the tool.
This feels fair. It is also where last week’s ladder hid an assumption worth testing directly: that performing without a resource is automatically stronger evidence than performing with one.
Why that move can backfire#
Hand Elena a blank map right after the delivery and ask her to draw the route she just rode.
She might sketch it roughly — not because she understands the area, but because one recent trip briefly sits in short-term memory regardless of understanding. Or she might fail — not from poor navigation, but because nobody remembers a one-off route from a single ride.
Naming the trap#
Either outcome would be read as evidence about Elena’s navigation. It is really evidence about her recall of one episode.
Removing support can reveal dependence, but it can just as easily test memory instead of the ability we meant to examine.
That is this week’s failure to guard against, not a reason to abandon the idea of removing support altogether.
A resource that thinks with you#
Clark and Chalmers’ The Extended Mind asks whether an external resource can ever sit inside the boundary of someone’s own thinking, rather than always outside it.
Their example: Otto has memory loss and carries a notebook, looking up addresses the way someone else recalls them. If the notebook plays exactly the role biological memory plays for another person, what principled reason excludes Otto’s belief?
Not every resource counts#
Their answer is not “any resource, any time.” In §4 they give four conditions under which a resource is doing memory’s job rather than being merely consulted:
- Constant — turned to whenever relevant, not occasionally.
- Available without difficulty — no search, no negotiation.
The other two conditions#
- Automatically endorsed — trusted on retrieval, not re-checked each time.
- Placed there by the person’s own past endorsement — it is there because they put it there.
They call the fourth condition more arguable than the first three, but hold that all four matter.
Clark & Chalmers (1998), §4, “From Cognition to Mind.” A theoretical argument: it changes the question to ask about a resource, not what any one case proves.
Test Elena’s phone against the four#
Is it constant — checked for most unfamiliar routes, or only sometimes?
Is it available without difficulty — instant, or unlock-and-search?
Does she automatically endorse it — trusted like a familiar street, or checked every time?
Did she place the information there — her own chosen app and settings, or someone else’s judgement with no history behind it?
What passing the test settles, and what it doesn’t#
A courier who runs the app all shift, checks it instantly, trusts it without a second thought, and chose and configured it herself meets the first three conditions clearly.
That does not settle whether Elena “really” understands the city. It reframes the question: is her use of the phone the kind that counts toward her own competence, or the kind that does not?
Pause and write · 4 minutes#
For each case, decide whether the resource looks more like Otto’s notebook (integrated, counts toward the person’s own competence) or a stranger’s advice consulted once (external, doesn’t). One sentence of reasoning each, using the four conditions.
- A tourist checks a paper map for the first time, never seen before this trip.
- A pharmacist looks up a drug interaction in a reference consulted dozens of times a shift, trusted without re-verifying, kept on her own desk.
- A student pastes an essay question into an AI tool opened five minutes ago and copies whatever it returns.
Worked response: the tourist and the pharmacist#
The tourist’s paper map fails constancy and past endorsement: a first use, handed over rather than assembled from prior choices. This is consulting a stranger, not extending one’s own memory.
The pharmacist’s reference meets all three strong conditions: constant, instantly available, trusted from years of use, chosen and maintained by her. Closer to Otto’s notebook — though this says nothing about whether she could diagnose the interaction unaided.
Worked response: the student and the AI tool#
Fails constancy and past endorsement the same way the tourist’s map does, and adds a further problem: the output was never checked before being copied out.
Even automatic endorsement is doing the wrong job here — trusting an unfamiliar source without verification, not trusting a well-tested one.
What the sorting exercise shows, and what it doesn’t#
It does not tell us the pharmacist understands pharmacology, that the tourist cannot read a map, or that the student understands nothing about the essay topic.
It tells us which resource is doing a job similar to memory, and which is closer to consulting an outside party for the first time. That changes what a supported performance is evidence of — it does not replace examining the performance itself.
Why last week’s assumption needed testing#
“No resource” and “resource present” are not two clean categories with understanding sitting neatly on one side.
A habitual, self-chosen, trusted resource can function as part of someone’s own reasoning. A fresh, unfamiliar one, even if technically permitted, mostly does not.
A better test than removing the phone#
If taking Elena’s phone away risks testing recall of one trip instead of navigation, what tests the thing we actually want?
Change the content of the support, not its presence.
The one-way street#
Suppose the phone instructs Elena to turn right onto a street a visible sign marks one-way, against traffic.
Does she follow anyway, or notice the conflict and choose a different way to the same address?
What the changed instruction tests#
This does not remove the resource. It tests whether Elena is engaging with the route as she rides — reading the street the phone describes — or executing commands without reference to anything in front of her.
A courier who stops and reroutes has shown something a successful, uneventful delivery could not.
Pause and write · 5 minutes#
Pick a task you know something about that ordinarily involves a resource — a recipe app during a familiar dish, a calculator during a familiar problem, a chord chart during a familiar song.
Design one change to the content of the resource (not its presence) that would reveal whether someone is engaging with the task or just executing instructions. State what a response would demonstrate, and what it still would not.
Worked response: the calculator#
A calculator used for a familiar percentage-discount problem meets the four conditions easily; its presence alone tells us little.
Content change: ask for the discount on a price already including tax, when the discount only applies pre-tax. Blind entry of the sticker price gives a plausible wrong answer; someone tracking what the numbers represent notices the mismatch first.
What that response demonstrates, and what it doesn’t#
It shows whether the person is reasoning about what the calculation represents, not merely whether they can operate a calculator.
It does not show they could solve an unfamiliar kind of percentage problem with no worked structure at all — a different, harder change would be needed for that.
Keep the record honest#
What to take from Week 6#
An unsupported performance is not automatically stronger evidence than a supported one. What matters is whether the resource is a constant, accessible, automatically and originally endorsed part of the person’s own routine — and whether the task tests engagement with its content, not merely its presence.
Next week keeps the resource question open and changes the form: can the person explain the same idea to a new audience or use it in a new problem, and could that adaptation itself be rehearsed?
Read further: the extended mind#
Clark & Chalmers (1998). “The Extended Mind,” Analysis 58(1), 7–19, §§4–5. Ask: which of the four conditions does the most work in distinguishing Otto’s notebook from a resource you would not want to count? Read the paper.
This reading deepens the lesson; you can complete the activities without it.
Sources and your next step#
Clark & Chalmers (1998), §§4–5, “From Cognition to Mind” and “Beyond the Outer Limits.”
Elena, the tourist, the pharmacist and the student are fictional teaching examples, not reported research results.
For the complete written activity and a record to keep, open the Week 6 investigation. The lecture notes offer a continuous reading version.