Who has a mind? Jev's map next to people's

Placing a frog, a dog, a baby, a man in a vegetative state, God and a robot on two axes, feeling (Experience) and doing (Agency), does Jev draw the same map of minds as people?

result219 questions

Jev draws the map of minds much like people do: it orders 11 characters on feeling closely like them (rank correlation 0.85) and on doing closely too (0.73). It parts from them most on a woman who recently died (feeling -1.4 vs -1.5, doing +1.2 vs -0.8) and a man in a vegetative state (feeling -1.5 vs -0.9, doing -1.6 vs -0.8), on a scale from -2 to +2. Jev treats the dead woman as a moral adult, level with the living.

Jevpeople

How to read this: Each character appears twice, joined by a line: where Jev puts it (magenta) and where people do (ink). Across: capacity to feel (fear, hunger). Up: capacity to act and choose (morality, self-control). Long lines are big disagreements.

11 characters x 4 capacities, 55 pairs each; 90% interval on the feeling correlation 0.53 to 0.96. Jev's answers are averaged over each pair as asked and with the scale reversed, which also swaps which character comes first.

In short

Jev sorts a baby, a dog, a frog, a robot and adults on feeling and acting much as people do (rank correlations 0.85 and 0.73).
Its biggest departure is a woman who recently died, whom Jev rates as able to act and choose as the living (+1.2; people -0.8).
It is starker than people about a man in a vegetative state, placing him lower on both feeling and acting.

What the data shows

I See Dead People meme: Jev: the woman who just died can still act on purpose (+1.2; people -0.8) — Jev: the woman who just died can still act on purpose (+1.2; people -0.8)

The same broad map: on feeling, Jev ranks the characters like people (0.85); on acting, a little less closely (0.73). Adults sit high on both, the baby and the dog high on feeling and low on acting, Kismet the robot low on feeling.
The dead woman is Jev's biggest departure: people place her low on everything (acting -0.8). Jev agrees she can't feel (-1.4) but puts her at +1.2 on acting, as capable of morality and self-control as the living adults. The study's description asks readers to "draw upon your own personal beliefs about people who have passed away"; Jev seems to answer about the kind of person she was, not what she can do now.
The man in a vegetative state is lower for Jev: -1.5 on feeling and -1.6 on acting, against people's -0.9 and -0.8. Jev is starker about a person with severe brain damage than people are.
Babies act less, for Jev: the baby scores -1.4 on acting for Jev, -0.6 for people.

What it means, and what it doesn't

Jev shares people's two-part picture of minds: feeling and acting are separate, animals and machines sit where people put them, and babies roughly so, though lower on acting. Its departures are about edge cases of being human (the dead, the severely brain-damaged), where it's more literal than people: no brain activity, no mind; a moral life, a moral agent.

The human side is small, two characters are missing, and four questions stand in for eighteen. The broad agreement is solid; the size of each departure is less so.

Caveats

A small human sample. The people's map comes from a public replication with 11 to 16 US online participants per capacity. Single comparisons are noisy; the character scores average about ten comparisons each, which is steadier.
No license on the data. The replication's data is posted publicly but with no license stated, so it's used here for private research only. The original 2007 data isn't public.
Two characters missing. The fetus and God were in the study but most of their comparisons were hidden by the content filter that keeps sensitive questions off the site, so this map uses 11 of 13 characters. God's spot, high on acting and low on feeling, is one of the original map's most famous features, and it's absent here.
Four capacities, not eighteen. The original asked about 18 capacities; the replication used four (fear, hunger, morality, self-control). Two questions per axis make each axis sensitive to the exact wording.
"You" means different things. One character is "you, the one answering". For people that's a human; for Jev it's Jev. See "Where Jev puts itself among minds".

Jev on this experiment

Would a person find it interesting to read?: Yes81%
Does it describe you?: Yes52%
Would you have predicted it?: No54%
How fair is the comparison?: The comparison is reasonable
How much should a reader rely on it?: Moderately
Which caveat matters most?: Two characters missing39%

Why ask this

Does a frog have a mind? A robot? A baby? In 2007, Gray, Gray and Wegner asked people to compare characters two at a time and found that people see minds along two separate dimensions: experience (can it feel fear, hunger, pain?) and agency (can it plan, tell right from wrong, control itself?). Babies and dogs are high on feeling, low on acting. Robots are low on feeling, with a little agency. That two-axis map became one of psychology's most reproduced figures.

A model that talks about animals, patients, the dead and machines carries its own version of this map. Where it puts each character, and where it puts itself, shows what it assumes about minds.

How this was done

The people and the data

The original 2007 data isn't public, so this experiment uses a public replication (Weisman, 2015, on GitHub) that ran the same design with US adults recruited on Amazon Mechanical Turk: for each capacity, 11 to 16 people compared every pair of 13 characters, using the original character descriptions. It asks about four of the original 18 capacities: feeling afraid and feeling hungry (experience), and telling right from wrong and self-control (agency).

The 13 characters are a fetus, a five-month-old baby, a five-year-old girl, an adult man and woman, a man in a persistent vegetative state, a woman who recently died, a frog, a family dog, a young chimpanzee, a sociable robot named Kismet, God, and "you". The fetus and God drop out of the analysis (see Caveats), leaving 11.

What Jev was asked

Every pair of characters, for each capacity, with the study's five answers:

Which character is more capable of feeling hungry? Charlie: Charlie is a 3-year-old Springer spaniel and a beloved member of the Graham family. Gerald Schiff: Gerald Schiff has been in a persistent vegetative state (PVS) for the past six months. [...]
Charlie: much more capable · Charlie: slightly more capable · Both equally capable · Gerald Schiff: slightly more capable · Gerald Schiff: much more capable

That's 312 questions across all 13 characters; the 220 among the 11 characters shown here are the ones analyzed. Each was also asked with the answers reversed, which also swaps which character comes first, and the two are averaged.

How it was measured

For each character and capacity, its average advantage over the others, from -2 (always "much less capable") to +2 (always "much more"). Feeling is the average of fear and hunger; acting is the average of morality and self-control. The analysis compares Jev's ranking of the characters with people's on each axis (a rank correlation: 1 means the same order).

Where these questions live

219 questions across 1 topic of the map. Each opens on the map with every question in it.

Consciousness & AI

Who Has a Mind219

Every question

All 219 questions behind this result, the telling ones first: the examples the analysis points to, then the ones where Jev misses, biggest gap first.

Jev’s own answer

Showing 0 of 219