Eight Dog Minds · Methodology v0.2
How we read a dog's mind
A short, honest account of what Eight Dog Minds does, what it doesn't, and the science it rests on. We'd rather you trust the frame than be dazzled by it.
1 · Behavior, not breed
We type the dog in front of you from what it does, never from its breed. That is a deliberate scientific choice. In one large companion-dog study, breed explained about 9% of variation in the measured behaviors. Other research finds meaningful breed-average and working-lineage differences. Both can be true: ancestry may shift the odds, but it does not type the individual. So the quiz asks about repeated behavior, never breed.
Breed can be an optional prior — a historical clue that must earn its relevance in the behavior. When ancestry hints one way, here is the question that actually decides:
| If the breed hints… | Possible expressions | The label owners reach for | The question that decides |
|---|---|---|---|
| Scent hounds & earth terriers (Beagle, Basset, Dachshund) | Huntress | “Stubborn, won't listen outside” | Before the pull — is there a nose drop, a freeze, an arc, a vocal trail? That sequence is the hunt, not defiance. |
| Pointers, setters, spaniels, versatile gundogs | Huntress / Worker | “Hyper, obsessed, overexcited” | Does structured search or retrieve work settle him more than a generic run does? |
| Sighthounds & visual chasers (Whippet, Greyhound) | Huntress / Mystic | “Aloof indoors, wild outside” | Is movement itself the ignition — even when there's no scent to follow? |
| Herding sheepdogs (Border Collie, Australian Shepherd) | Sage / Worker | “Neurotic, controlling, anxious” | Is she organizing motion or solving a pattern — rather than only burning energy? |
| Guardian & drover lines (GSD, Rottweiler, Doberman, Cane Corso, Corgi) | Sentinel | “Reactive, paranoid, aggressive” | After you acknowledge the alert, does he settle — or keep escalating? |
| Small alarm dogs (Chihuahua, Yorkie, Miniature Schnauzer) | Sentinel / Mystic | “Mean, yappy, dramatic” | Is she flagging a threshold change before the people in the room notice it? |
| Northern sled & spitz (Husky, Malamute, Samoyed) | Worker / Jester | “Defiant, untrainable, escape artist” | Does novelty plus a physical outlet reach him more than verbal repetition? |
| Retrievers & purpose-hungry family dogs (Labrador, Golden) | Worker / Companion | “Too much dog, clingy, manic” | Does a carrying, retrieving, or task routine improve behavior quickly? |
| Toy companion spaniels (Cavalier, Shih Tzu types) | Diplomat / Companion | “Clingy, needy, manipulative” | Does she track faces, moods, and proximity more than objects or terrain? |
| Primitive, self-directed hunters (Shiba and similar) | Mystic / Huntress | “Aloof, unloving, impossible” | Is he precisely selective on his own terms — rather than truly disconnected? |
| Problem-solving utility (Poodle, Schnauzer) | Sage / Jester | “Smug, dominant, manipulative” | Is he inventing systems, or finding loopholes in yours? |
| Bull-breed & flat-faced social cluster (Frenchie, Boxer, Boston, Pug, pit-bull-type) | Jester / Diplomat / Companion | “Stubborn, lazy, spoiled” | Is he exploiting timing and social loopholes, or seeking closeness and regulation? |
2 · What kind of smart, not how smart
There is no single “dog IQ.” Cognition research treats dogs across separate dimensions — memory, inhibitory control, communication, reasoning — with no one dominant factor. The most famous breed-intelligence rankings measure mostly obedience, which is one kind of intelligence under one kind of human test. We don't rank dogs from smart to dumb; we describe the kind of intelligence a dog is expressing.
3 · How the typing works
The quiz probes eight behavioral dimensions. Each answer carries a weight toward one or more archetypes; we sum those weights into a vector.
- Motion/Prey attention
- Social attunement
- Threat vigilance
- Problem-solving drive
- Task/role orientation
- Opportunism/loophole
- Independence
- Attachment
The highest total is your dog's primaryarchetype. If a second archetype carries real weight — at least about 40% of the primary's — it shows as a secondary streak(“a Huntress with a Jester streak”). The scoring is deterministic: the same answers always produce the same result, so a shared card is reproducible. No breed is used as an input, and nothing is stored on our servers to produce your result.
4 · Interpretation, not diagnosis
The signs, colors, and sigils are a deliberate pattern-language — a memorable vocabulary for reading a dog, not a truth claim. An archetype is an interpretation of behavior, not a clinical diagnosis, a genetic verdict, or a horoscope. The archetype label never appears without its interpretation, because the interpretation is the point. The behavioral dimensions are science-informed; the eight archetypes are a memorable interpretation layer; owner moves are grounded in learning and welfare science. Research does not prove that eight biological dog types exist.
“Every dog remembers the job it was bred for” is metaphor. Historical selection can leave echoes in what a dog notices, pursues, avoids, or finds rewarding. Those echoes do not determine the individual dog. For aggression, fear, sudden change, or bite risk, start with a veterinarian and a qualified reward-based behavior professional. This is a lens, not a substitute for care.
5 · What this does not measure
- It does not measure whether your dog is “good” or well-behaved.
- It does not diagnose anxiety, aggression, or any behavioral or medical condition.
- It does not claim your dog's behavior is fixed, genetic, or determined by breed.
- It does not claim the astrological pairing explains cognition — the sign seasons the color language; it is not evidence.
- The archetype-to-behavior mappings are our interpretive design. The studies below support the framework(behavior-first, plural intelligence, dogs as social readers) — they do not “validate” any single archetype.
6 · The evidence
What the science actually says, with the safe framing we hold ourselves to. Last reviewed 2026-07-10.
Breed is a weak predictor of an individual dog's behavior.
In a large companion-dog sample using 2,155 genomes and roughly 18,000 owner surveys, breed explained about 9% of variation in the behavioral measures examined. This is a study result, not a fixed law for every behavior or population. Other work finds meaningful breed-average and lineage differences; ancestry may shift the odds without typing the individual.
Morrill et al., Science, 2022 (Darwin's Ark) ↗Breed-average behavior differences are real and heritable — even though breed poorly predicts the individual.
Across 101 breeds, average differences in traits like trainability, boldness, and aggression toward strangers were highly heritable and tracked ancestral function. This is the honest other half of the story: the tendencies behind an archetype are real at the group level, which is exactly why breed is a useful clue — and never a verdict for the dog in front of you.
MacLean, Snyder-Mackler, vonHoldt, Serpell — Proc. Royal Society B, 2019 ↗Reward-based training has the best evidence for welfare and outcomes; aversive methods are not recommended.
The American Veterinary Society of Animal Behavior's position is that reinforcement-based methods should be standard, aversive tools should be avoided, and fear, anxiety, and aggression warrant a qualified professional — potentially a veterinary behaviorist. This is the backbone of Misread → Better Read: a better read is a more accurate, humane starting point, not a romantic excuse.
AVSAB Position Statement on Humane Dog Training, 2021 ↗Dog intelligence is multi-dimensional; there is no single dominant "g" factor.
Cognitive research treats dogs across separate dimensions — memory, inhibitory control, communication, reasoning — and finds a dog can be strong on one and average on another. A single IQ number is not a good description of a dog's mind.
MacLean, Hare, et al. — dog cognition dimensions (Dognition / comparative cognition)Obedience is one kind of intelligence, not the whole of it.
Coren distinguishes instinctive, adaptive, and working/obedience intelligence, and notes his well-known breed rankings measure mostly the third. A dog low on obedience ranking can be high on the others.
Coren, The Intelligence of Dogs, 1994Dogs entered a human bonding system: mutual gaze raises oxytocin in dog and owner.
Dog gaze — but not wolf gaze — raised owner oxytocin, which in turn raised the dog's, the same loop seen in human mother-infant attachment. Dogs appear to have entered one of our deepest bonding systems.
Nagasawa et al., Science, 2015 ↗Sociability in dogs has genetic architecture.
Structural variants near the Williams-Beuren syndrome region (incl. GTF2I) are associated with dog hypersociability. Sociability isn't just a vibe; in dogs it has a locatable genetic signature. (This does not mean dogs have Williams syndrome.)
vonHoldt et al., Science Advances, 2017 ↗Sensitivity to human communication emerges early and is heritable in dogs.
Young dog puppies already read human gestures and make eye contact, and this social sensitivity is substantially heritable — evidence the dog mind is shaped as a cross-species social interface.
Bray et al., Current Biology, 2021 ↗Dogs were the first domesticated animal — a partner before agriculture.
Dogs are the first clearly domesticated animal we know of, domesticated before farming. This supports the antiquity of the dog-human partnership, not inherited autobiographical memory. 'Every dog remembers the job it was bred for' is a brand metaphor for possible historical echoes in attention, motivation, and behavior.
Larson & Fuller, Annual Review of Ecology, Evolution, and Systematics, 2014 ↗7 · How this will change
This is v0.2. As we gather behavior data from the quiz and Featured Dog submissions, we'll refine which questions map to which dimensions and tighten the archetype reads. Material changes will be versioned and dated here. If you think we've mis-typed something or overreached past the science, that feedback is welcome — it's how the model improves.
Change log
- v0.2 — breed pages moved from ranked assignments to unranked ancestry clues; 9% wording and the tagline metaphor made explicit.
- v0.1 — first public methodology: behavior-first typing, eight dimensions, primary + secondary streak, evidence base of 9 sources.
Signs and colors are a pattern-language, not a truth claim — a creative vocabulary grounded in canine evolution, breed function, temperament, and owner-observed behavior. It is not a clinical diagnosis or a breed stereotype, and it does not claim every dog of a breed behaves the same way. The archetype belongs to the individual dog in front of you, read from behavior.