How Many Animals Can You Name in a Minute? The Science of Rare Answers

A healthy adult names about 13 to 22 animals in one minute, depending mostly on age. In the best known set of norms, people in their twenties to fifties averaged around 20 to 21, people in their seventies about 16, and people in their late eighties and nineties about 13 to 14. Those numbers come from a clinical test, and the test is far more interesting than the count it produces.

Because it is not only how many animals you name that researchers look at. It is which ones. The ordinary animals that everybody reaches first (cat, dog, lion) tell a clinician very little. The odd ones, the axolotl and the pangolin, tell them a lot. That same idea, scoring the answer nobody else gave, is the whole of our Krillion game.

The task is as plain as it sounds. A tester says something like 'Tell me all the animals you can think of', starts a stopwatch, and writes down every animal you say for 60 seconds. Your score is the number of different animals, with repeats and non-animals left out. Psychologists call it a semantic fluency or category fluency test, because you are generating words from a category of meaning. Its sibling is phonemic fluency, where you name words starting with a letter such as F, A or S.

It turns up in some of the most widely used assessments in neurology. The Boston Diagnostic Aphasia Examination, a standard battery for aphasia, the language disorders that often follow a stroke, includes animal naming in its oral expression section; there the examiner offers 'dog' as an example to get things going. The CERAD battery, built in the late 1980s by the Consortium to Establish a Registry for Alzheimer's Disease, opens with the same one-minute animal task with no example given. A 2004 study in Neurology by Canning and colleagues found that fewer than 15 animals was about 20 times more likely in a person with Alzheimer's disease than in a healthy control, which is why the test is popular as a quick screen in memory clinics.

The norms most often quoted come from Tom Tombaugh and colleagues, published in 1999 in Archives of Clinical Neuropsychology. They tested community-dwelling adults with no cognitive impairment, 735 of whom did the one-minute animal task. Their average scores by age:

• 16 to 19: 21.5
• 20 to 29: 19.9
• 30 to 39: 21.5
• 40 to 49: 20.7
• 50 to 59: 20.1
• 60 to 69: 17.6
• 70 to 79: 16.1
• 80 to 89: 14.3
• 90 to 95: 13.0

The spread inside each group was about 4 to 5 animals either way, so a 40-year-old naming 16 is still well within the normal range. Schooling mattered too: people with 8 years of education or fewer averaged 13.9, and those with 13 to 16 years averaged 19.0. In that study age explained about 23% of the variation in animal scores and education about 14%, while sex explained less than 1%. Notice the curve: scores hold steady from the late teens to the late fifties and only then begin to fall.

Watch anyone do this test and you will hear bursts. Cow, pig, sheep, goat, then a pause, then lion, tiger, leopard. In 1997 Angela Troyer, Morris Moscovitch and Gordon Winocur, at the Rotman Research Institute in Toronto, gave those bursts a name and a scoring method that is still the standard. A cluster is a run of animals from one subcategory (farm animals, pets, African animals, primates, birds, insects). A switch is a jump to a new subcategory.

Both matter. In their sample of 54 older and 41 younger adults, the number of animals named rose with cluster size and with the number of switches alike. Younger adults named more animals mainly because they switched more often; the two age groups built clusters of the same size. The authors linked clustering to the temporal lobes, where word knowledge is stored, and switching to the frontal lobes, which steer the search. A good one-minute run is a balance: mine a cluster until it slows, then jump before you stall.

A count of 18 hides a lot. One person's 18 might be the first 18 animals anybody thinks of; another's might include a quokka and a marabou stork. Researchers measure this with two properties of the words: frequency (how common the word is in ordinary language) and typicality (how good an example of the category it is; a dog is a very typical animal, a sea cucumber is not).

In a 2014 study in the Journal of Alzheimer's Disease, M. G. Vita and colleagues at the Catholic University of the Sacred Heart in Rome compared the animals named by healthy older adults, people with amnesic mild cognitive impairment and people with Alzheimer's disease. Both patient groups named more typical animals than the healthy controls, and among the people with mild impairment, those who named the most typical animals were the ones more likely to go on to develop dementia over the following two years. Later meta-analyses have found the same direction for word frequency: as semantic memory thins, people fall back on the commonest words. Put the other way round, a healthy, well-stocked memory is one that reaches past the obvious.

Please read your own score lightly. A real assessment is given by a trained clinician under set rules, compared with norms that match your age, education and language, and interpreted alongside many other tests and your history. Tiredness, anxiety, a second language, or simply never having thought much about animals can all move a score by several points. A low number at the kitchen table means nothing on its own, and a game, ours included, is not a medical test of any kind. If you are worried about your memory or someone else's, talk to a doctor.

You need a timer and a friend to write, or a phone that records audio.

1. Start a 60-second timer and say every animal you can think of. Do not stop to check spelling or whether something counts.
2. Stop at 60 seconds. Write the list out from the recording.
3. Score one point for each different animal. Repeats score nothing. Clinical rules vary on details; many accept a general name and a specific one (dog and terrier), but not a word plus its plural.
4. Compare your total with the age table above, remembering the spread of 4 to 5 either way.
5. Now give yourself a second score. Cross out every animal that a typical eight-year-old would say in their first ten seconds: cat, dog, horse, cow, pig, lion, tiger, elephant, bear, monkey, bird, fish and the like. Count what is left. That is your rarity score.
6. Look at your clusters. Draw a line wherever you jumped to a new group. More short, deep clusters usually mean a higher total.

Most people find the second score sobering. A total of 20 often hides only five or six animals that survive the cull.

That second score is the idea behind Krillion. A prompt such as 'Name a bird of prey' is a fluency test for one answer, with 25 seconds on the clock. The game checks your answer against a bank of everything people say and scores you by how few gave the same thing, from Plankton at 4 points for the answer everyone gives to One in a Krillion at 100 for the one almost nobody does.

The animals at the front of the line are the same ones the clinicians count. Our analysis of Krillion's prompt banks found that Cat sits in the bottom tier on 16 different prompts, Fish on 16, Dog on 15 and Bee on 13. Animal prompts accept a median of 124 different answers, and across all prompts more than 40% of accepted answers sit in the top two tiers. The rare animals are there; the crowd just does not reach them in time.

So the clustering trick from the test transfers directly: do not answer with the first animal that arrives, use the first few seconds to jump to a cluster most people never open (deep sea, Australian marsupials, birds of a single country), and pick from there. If you want to know why that kind of search works, our piece on working memory and puzzle games goes further.

Time your 60-second count first, then try seven prompts of the same skill in Krillion Unlimited. It is free, runs in the browser and has no daily cap, so you can see whether your rarity score climbs as you learn which clusters the crowd leaves alone.

Tombaugh, T. N., Kozak, J. and Rees, L., Normative Data Stratified by Age and Education for Two Measures of Verbal Fluency: FAS and Animal Naming (Archives of Clinical Neuropsychology, 1999).

Troyer, A. K., Moscovitch, M. and Winocur, G., Clustering and Switching as Two Components of Verbal Fluency (Neuropsychology, 1997).

Vita, M. G. et al., Typicality of words produced on a semantic fluency task in amnesic mild cognitive impairment (Journal of Alzheimer's Disease, 2014).

Canning, S. J. D. et al., Diagnostic utility of abbreviated fluency measures in Alzheimer disease and vascular dementia (Neurology, 2004).