Stupid canine lexical acquisition claims

« previous post | next post »

Dogs as intelligent as two-year-old children, says a headline in the Daily Telegraph, a newspaper that is marketed to people of a conservative disposition and their dogs. And in case you did not quite understand the headline, they say it again in the subhead: "Dogs are as intelligent as the average two-year-old child, according to research by animal psychologists." It is bylined "By Richard Gray, Science Correspondent". (Science Correspondent! He almost certainly has a Master's degree, possibly in Science!)

Research conducted at Language Log Plaza has shown a somewhat different result. Dogs are not as bright linguistically as a human two-year-old. But what is true is that dogs have the same general intelligence and ability to detect bullshit as the average Science Correspondent for the Daily Telegraph or BBC News.

The details Mr. Gray reports are that "Researchers have found that dogs are capable of understanding up to 250 words and gestures, can count up to five and can perform simple mathematical calculations." And in case you (or your dog) did not quite understand that, it is repeated in more detail for you:

"The average dog is about as bright linguistically as a human two-year-old," said Professor Stanley Coren, a leading expert on canine intelligence at the University of British Columbia in Vancouver who has carried out the work.

"This means they can understand about 165 words, signs and signals. Those in the top 20 per cent were able to understand as many as 250 words and signals, which is about the same as a two and a half year old.

"Obviously we are not going to be able to sit down and have a conversation with a dog, but like a two-year-old, they show that they can understand words and gestures."

The evidence of understanding words comes from experiments in which a border collie was trained to go and fetch a ball when "Ball!" was shouted at it, and so on for other medium-sized fetchables. (Ah, border collies. Long-time Language Log readers will recall that we have been here before.)

And the evidence of mathematical calculations was that a trained border collie can work out the square root of an arbitrary integer written on a chalkboard, to an accuracy of at least three decimal places.

No it wasn't. I said that just to see if you were paying attention. If you're a Daily Telegraph reader you probably believed me. The evidence was from differential gaze experiments: if you drop three doggie treats behind a screen and then surreptitiously remove one or two before shifting the screen, a dog looks a little bit longer at the remaining treat or treats than it does when all three are still there like they should have been. They are capable of noticing, in other words, when slightly weird shit is going down.

If this all satisfies you — if you now think border collies can understand the meaning of lexical items and do mental arithmetic — then Professor Coren has won his game of spoof the public. But it has left me wondering whether I will be reading stories about lexical item acquisition in dogs and other stupid fake pet communication tricks until my dying day, or whether one day we will wake up on a bright new morning and Science Correspondents will have realized that they don't have to just paraphrase the press release put out at the APA convention, they can ask a few penetrating questions about what it means to understand the meaning of a word. (Like, could a dog understand an adverb, such as "surreptitiously"? Why is it always nouns and verbs triggering trained physical actions like fetching? I understand the noun "turd", but if you say it to me I don't run to try and find one.)

[Hat tip: Brian Davies.]

(I have left comments open below, but if you are a dog, please say so. On the Internet, nobody knows these things.)


  1. stormboy said,

    August 12, 2009 @ 7:52 am

    What a fine tradition of journalism we have in the UK!

    I'm not a dog but have two. Like all doting dog owners I know my dogs are highly intelligent (whatever intelligence means in the canine context), but even I think that comparing dogs to two-year-old humans is daft.

    My friends – many of whom are Brazilian – are amazed that the dogs understand both English and Portuguese. But dogs are masters of reading situations/context and anticipate extremely well. I can say just about anything to my dogs in any language in a given situation and they will always respond as I expect them to, if I use the expected intonation (the word choice itself doesn't seem to matter) and body language.

    Everyone else seems to think that I have extremely talented, bilingual dogs and that I'm just too sceptical…

  2. Paul said,

    August 12, 2009 @ 8:00 am

    A dog we had once went mad while we were preparing some Chinese food, thereby revealing a problem with his vowel inventory: he completely failed to distinguish [ɒ] from [ɔː] in the context [w___k]. On reflection, perhaps we should have called in a speech therapist.

  3. Simon Holloway said,

    August 12, 2009 @ 8:18 am

    I have had hours of fun (in total, not at a time) saying, "bad girl" in a loving voice and shouting "walk!" as though the poor thing had done something wrong. The engendered confusion would seem to indicate that, in addition to acquiring information from my tone, she can also become familiarised with the sounds of individual words and even the quality of individual consonants ("talk", "pork" and "work" all produce the same effect). While I might be driving her mad, I would have to be mad myself were I to believe her comprehension was equal to that of a human child.

  4. Mark P said,

    August 12, 2009 @ 8:20 am

    I had a dog that liked to roll in bullshit. Apparently that makes her about as intelligent as a
    Daily Telegraph science correspondent.

  5. peter mcburney said,

    August 12, 2009 @ 8:35 am

    "they can ask a few penetrating questions about what it means to understand the meaning of a word."

    What does it mean, precisely, for a human being to understand the meaning of a word? I ask this as a computer scientist trying to engineer intelligent machine-to-machine (M2M) communication. So far the only explanation of understanding that I have found which make sense to me is essentially pragmatic and discourse-related, eg, a person understands the meaning of a word when he or she uses it in a communicative interaction where a hearer (or reader) correctly infers the intended communication of the person speaking (or writing). But such a definition is difficult to operationalize for M2M communications, not least because everyday human conversation (and writing) is full of practices where determination of meaning is not straighforward, eg puns, irony, sarcasm, feints, metaphor, analogies, allusions, etc.

    [(myl) Some discussion of these issues, specifically in the context of the border-collie experiments, can be found in some older LL posts: "The strange, new sight", 6/11/2004; "Canine intelligence", 6/12/2004; "Signs or symbols? Words or tools?", 6/15/2004. The last one, and the earlier work cited in it, is especially relevant.]

  6. Ginger Yellow said,

    August 12, 2009 @ 8:35 am

    There's a reason August is called the silly season.

  7. Amy B said,

    August 12, 2009 @ 8:41 am

    As stormboy said, dogs have an uncanny ability to read humans, emotions, and contexts. The fact that they can do this without language is what's amazing to me. I suspect that if we had a closely related dog species with some linguistic skills, however crude, we'd be more in awe of the dogs without it, because they manage to understand and adapt to our world without the ability to think in words.

    P.S. My old Border collie, Dougal, was truly a doggie genius. Just thought I'd get that on the record.

  8. Dan T. said,

    August 12, 2009 @ 8:57 am

    Isn't it also the "Dog Days"?

  9. anon said,

    August 12, 2009 @ 8:58 am

    Many years ago, some Army combat engineers tried to teach dogs to sniff out explosives. They taught the dogs to bark once for a land mine, twice for an unexploded bomb, and so forth. Unfortunately, when the dogs encountered explosives for which they had not been taught barking and growling patterns, the dogs ignored the explosives, and some of the dogs and their trainers were blown to bits.

    The engineers concluded that the dogs' ability to detect explosives depended on the words for explosives in the dogs' language.

    This is called the Sapper-Woof hypothesis.

  10. Aaron Davies said,

    August 12, 2009 @ 9:07 am

    @anon: bravo. a perfectly rendered shaggy-dog story.

  11. John Cowan's nonexistent dog said,

    August 12, 2009 @ 9:10 am

    Arf! Arf! Woof-woof! Woof-woof-woof! Arf!

  12. Brett said,

    August 12, 2009 @ 9:37 am

    What I find especially galling is the gaze time data. To compare that ability to children aged two is ridiculous, since children substantially younger exhibit the same ability to recognize when an object has been surreptitiously taken away from a hidden group. In fact, testing dogs this way was almost certainly inspired by the experiments on infants.

  13. Janice Huth Byer said,

    August 12, 2009 @ 9:56 am

    Research has reportedly shown most infants, months before they utter their first word, understand whole sentences. That means either dogs are no smarter than newborns, or we humans are idiots to imagine their ability to ape our singular faculty for language is, in any way, a valid marker of their intelligence. I'm not just saying that, because I can't smell my way out of a paper bag.

  14. ajoubert said,

    August 12, 2009 @ 10:08 am

    Geoff, you omit the obvious: Dogs may learn "lexical items", but they don't have any grammar at all. Even a mentally retarded two-year-old has grammar. The good scientist in this story should have consulted any of the eminent linguists at his own university before claiming that dogs had the linguistic intelligence of a two-year-old child.

    Perhaps you have tired of repeating such claims, but not all of LL's readers are "old hats".

  15. Robin Queen said,

    August 12, 2009 @ 10:42 am

    It's kind of interesting that no one has replicated the Rico study, which is the primary (only?) source of (non-anecdotal) support for the claims about dogs "knowing words." There are (IMO) interesting questions to be asked about the development of the language ability in humans and about linguistic functionalism more generally by looking at various abilities in those species that don't have linguistic abilities. For instance, the issue of rapid mapping that the published study on Rico addressed is intriguing but it is not what seems to get presented by science reporters (or Stanley Coren). It also hasn't been shown to be a generalized ability in "dogs," even in the singular context of fetching, as far as I know.

    Like Amy B's, Dougal, my Border Collies are unquestionably geniuses in all sorts of ways. I think they might even understand "surreptitiously" since they are consummate at stealing food off the counters when I'm not looking. They just don't know there's a word for that ;-)

  16. Emily Lilly said,

    August 12, 2009 @ 11:10 am

    I came by to make essentially the same comment as Brett, Janice and Ajoubert combined:

    The counting ability is equal to young infants (3-6 months of age); my son was adding and subtracting (within the first twenty numbers, sometimes less accurately than others) when he was two.

    Comprehensive vocabulary size is not the same thing as language ability, as you know; while the average 2-year-old can _produce_ 250 words (some children do way more than that), they can understand _many_ more than that! Also, understanding solitary nouns and verbs does not in any way show language prowess, as there is no grammar ability.

    Did they even use two-word combos on the dogs?

    How did that researcher get a license to talk directly to the press about language ability when he clearly knows nothing about human language development? Bad researcher, bad! No biscuit!

  17. Stephen Jones said,

    August 12, 2009 @ 11:17 am

    I have found that I can always tire out a decent -sized dog running until it is about six months old. Therefore physically humans have the development level of a six-month-old dog.

  18. Dougal Stanton said,

    August 12, 2009 @ 11:52 am

    Well I'm not a dog but I do share my name with a few.

    The presentation mentioned is probably this one at APA's Convention 09. Nothing there to go on, just a bio for an invited address.

    It is unlikely that the science correspondent was in the audience or has even heard/read the content of the presentation. Until we can get newspapers to provide meaningful references I wouldn't put much faith in what they report.

    I encourage everyone who cares to keep complaining to major outlets like BBC to cite papers properly. If that happened maybe the behaviour will trickle down. Or I'm just chasing my own tail…

  19. Picky said,

    August 12, 2009 @ 11:54 am

    But Richard Gray is just a journalist. He grabs morsels from the feast that he thinks his masters might want him to serve up to his readers (his Daily Telegraph readers, remember).

    Prof Coren is apparently a genuine professor (admittedly only of psychology) in a genuine university. Can we hurl some of the broken bottles at academia, or is that too close to home?

    [I am certainly prepared to hurl the first bottle. Though before I call Professor Coren's work "bogus" I'd better consult my lawyers; remember what happened and is still happening to poor Simon Singh. —GKP]

  20. Chud said,

    August 12, 2009 @ 12:01 pm

    Cavemen figured out that they would be better off cooperating with dogs, as they could help with getting food and protection and warmth. That's why they developed bigger craniums: so they could talk to dogs. But once you've got that big cranium, you can't really stop the language development. And that's how we got this crazy Babel-world we live in. Thanks, dogs.

  21. Mr Fnortner said,

    August 12, 2009 @ 12:05 pm

    Paul, your dog did not realize that one must learn to wok before one can fry. Again, it's the two-year-old thing.

  22. Mark P said,

    August 12, 2009 @ 12:11 pm

    Does a dog understand "sit" if he consistently responds in the way he was taught when he hears that word? What if he responds in exactly the same way to fit, bit, lit, quit, snit, slit, pit or any number of other words that sound similar?

  23. Martin Hauser said,

    August 12, 2009 @ 12:55 pm

    As a teenager, I had a friend who kept boasting how his dog would "understand" words without hardly any training. To illustrate his point, he said "attack" and the dog quickly started growling and baring her teeth. We were having peanut butter sandwiches at the time, so I shouted "peanut butter" imitating my friend's tone and body stance and sure enough the dog started growling and baring teeth again. I never had to hear about the dog's talent for language again.

    [(myl) The experiments with Rico the border collie were controlled so as to prevent cueing of this sort, or of the kind mentioned below by Melvyn Quince. But Rico seems to have been a sort of canine savant.]

  24. Melvyn Quince said,

    August 12, 2009 @ 12:56 pm

    I have a dog which I carefully taught to sit when I said "Sit." I carefully verified that he knew how to shit (it was all too clear, in fact). I explained to him (by ostension) what shit was. I then performed the crucial experiment of pointing at his rear end when he was standing up and saying "Shit!", in order to determine whether he would sit or shit. He sat.

  25. Ceiswyn said,

    August 12, 2009 @ 1:27 pm

    It's been some time since my degree, but I seem to recall that 'counting' numbers of five or less actually uses a different mechanism than counting numbers above five. Specifically, you don't need to count them at all; you recognise at a glance that there are three, or four, or one, without having to do any counting at all.

    [What you're recollecting, Ceiswyn, is the literature on subitization. I didn't mention it because I figured that anyone who had done Psych 101 would know it, and the others would rather have more poop jokes. Subitizing is instant perception of numerosity: spotting a trio of matchsticks on the floor or three musicians on a platform doesn't need counting, you just see that it's a trio, not a duo or a quartet. Up to about 5 everyone can do this; animals can do it too, for very low numbers. Some humans manage it for up to 7; certain gifted autistic people can do it for astonishingly high numbers. The important thing is (do I really need to say this?) it is not mathematical calculation. It is the fact that Mr. Gray reports evidence of mathematical calculations that shows he is a credulous doofus. —GKP]

  26. Bobbie said,

    August 12, 2009 @ 1:31 pm

    Best use for the Daily Telegraph: Place the newspaper on the floor and encourage your house-broken dog to "use" the paper!

  27. Sili said,

    August 12, 2009 @ 2:27 pm

    Dear professor Pullum,

    Please tell me you'll be going to TAM (The Amazing Meeting) London. Not that I got a ticket for myself, but I'd love to see you hang out with Ben Goldacre (and James Randi for that matter).

    (Nought to say about dogs; I'm a cat person. I think I can safely say that Dummkatz does not understand a word of what is said to him. Nor does he consistently come when he is whistled for.)

  28. Faldone said,

    August 12, 2009 @ 2:36 pm

    I have observed both border collies and two-year old children, and, while the border collies can run rings around two-year old children in herding sheep the children will run rings around the border collies in language production. The pair of children I knew even knew how to play with language to their parents' consternation. They consistently left initial S's off the beginning of words that started with S followed by one or more consonants. Their parents tried all reasonable means to break them of this habit, assuming it to be a deficiency in their learning. I caught them one time accidentally putting the S on a word. They looked at each other with a kind of "Oh. Oops" look. I've never seen a border collie do anything linguistically comparable.

  29. Rover said,

    August 12, 2009 @ 3:00 pm

    "I understand the noun 'turd', but if you say it to me I don't run to try and find one."

    Yay, more for me!

  30. Matthew Walenski said,

    August 12, 2009 @ 4:05 pm


    Dogs can be trained to urinate on command, but not to defecate on command. I am not suggesting that this is why your experiment failed, but after you've trained a dog to urinate (usually by associating some random word with the process), you could see if a very similar word produced the same response or not (presumably outside, or maybe even on your favorite neighbor's lawn).

  31. Ron said,

    August 12, 2009 @ 5:32 pm

    Not sure I understand the relevance of the Telegraph's political leanings. Bullshit is bullshit, whether in the Telegraph or the Guardian, no?

    [Wow. Someone who imagines that everything I write in Language Log posts is relevant sure hasn't been reading Language Log very long, have they?

    (Certainly, there is nothing especially elevated or perceptive about the science reporting on linguistic topics in other newspapers. I was just explaining to American readers what the Telegraph is: a moderately high-quality large-format newspaper with clear conservative leanings. And let me add that it is to be admired in some respects: its recent extraordinary campaign to publicize the self-indulgent uses that Members of Parliament had been making of their parliamentary expenses accounts was a very good example of how a free press can change the world; and they didn't by any means let Conservative Party MPs off lightly. Or their dogs.) —GKP]

  32. Simon Cauchi said,

    August 12, 2009 @ 6:41 pm

    My border collie puppy, 10 months old, doesn't understand language very well, just the commands we have taught him, but he has decided views about public art. There's a sculpture in the local botanic garden that upsets him. Every time we go past it, he barks at it. He likes to watch television with us and sometimes barks during emotional scenes, but rather than human drama he much prefers watching the dog programmes. He used to go round behind the television set looking for the dogs, but has stopped doing that. So does he now understand that the TV is showing him a representation, not reality?

  33. Amy Stoller said,

    August 12, 2009 @ 11:44 pm

    "Dogs are as intelligent as the average two-year-old child, according to research by animal psychologists."

    I didn't know there were any animal psychologists. Don't you have to be a human being to get a degree?

    I have thoroughly enjoyed this post and all the comments. And so has my dog, Ricky.

  34. Bob Ray said,

    August 13, 2009 @ 6:34 am

    I used to convince people that my fairly unintelligent Brittany understood English, and especially "left" and "right" by saying things like, "give me your left paw," or "shake hands with your right paw" to her.

    She always got it right because I always asked for the paw she had less weight on. No one was the slighest bit skeptical about my claim.

  35. Kenny Easwaran said,

    August 13, 2009 @ 8:21 am

    Paul in 2:

    I'm surprised there's no British article pointing to your dog as an example of the claim that dogs have the linguistic intelligence of the average American! (I had to read your post a few times to figure out which word was supposed to have which vowel – I was trying to stick [ɔː] into 'wok' and it seemed totally crazy.)

  36. Stephen Jones said,

    August 13, 2009 @ 9:38 am

    or maybe even on your favorite neighbor's lawn

    Urine is an excellent nitrogen fertilizer. I always make a point of going out of the house and pissing on my own lawn even if the bathroom's nearer.

    I always maintain parks should have signs saying "Please piss on the grass."

  37. ShadowFox said,

    August 13, 2009 @ 9:41 am

    The fine thing about British science reporting–BBC and The Times are hardly more capable than the Torygraph–is that it is well balanced by British libel laws, equally incapable to distinguish fact from bullshit. A fitting response to this kind of reporting is for a parents of a precocious two-year old to sue the paper and the reporter for libel. Given the track record of British courts in this matter, they may well win.

  38. Sandra Wilde said,

    August 13, 2009 @ 11:48 pm

    Has anybody seen the defense of Kanzi's "language acquisition" in the latest issue of Skeptic magazine? I'd love for someone who really knows their stuff to write a good letter to the editor. The gist of it is that Kanzi has acquired syntax because he can follow commands like Give Rose a hug. I feel like the magazine should change its name to Credulous.

    Also, 3 uses of "beg the question" in the issue (different authors) misuse it.

  39. Spinoza said,

    August 16, 2009 @ 8:16 am

    Yo dawg, ever heard of the science correspondent who wrote an article on eskimo dogs? Apparently, Eskimos have 43 words for snow, but their dogs understand only 21 of them.

  40. Mark P said,

    August 18, 2009 @ 9:09 am

    Last weekend I had a long conversation with my two dogs about this. One of them is young, so he didn't understand many of the words I used. He looked quizzical, which I interpret to mean that he was trying to infer meaning from context. I have warned him about the danger of doing that, but, well, dogs will be dogs. The other was pretty bored by the whole thing; he is rather mature, not to say elderly, and is more interested in talking about his aches and pains and sometimes about hot dogs. But both agreed that they understand far more words than the cats do.

  41. Forrest said,

    August 18, 2009 @ 7:35 pm

    Obviously, the problem with this story is that they interviewed dogs, rather than cats. While it's arguably true that dogs are as intelligent as two-year-old humans, cats are a more intelligent species, and comparable to thirty-year-olds.

    Unfortunately, measuring the linguistic ability of a typical American cat isn't very easy. As Whorf and Sapir showed, cats think in German. Between this, and their generalized indifference toward human affairs, it's understandable that the researchers settled for dogs.

    Even a highly intelligent cat has trouble with grammar, however. If you ask Miss Parsley "Who did you play with outside?" she'll ignore you; if you ask her "With whom did you play outside?" you'll get a confused stare.

  42. Miss Cellany said,

    February 2, 2013 @ 8:09 pm

    Dogs do understand the difference between nouns and verbs. "Ball" was not shouted at the dog, the dog was asked to "paw" "nose" or "take" different toys to prove she knew that nouns are a reference to an object and not just words associated with a specific behaviour. Other experiments were done with the same dog such as those proving that the dog could learn by inferential reasoning by exclusion (she could discern that a command to fetch a novel toy that she hadn't seen or heard of before applied to the only unfamiliar object in the group of familiar toys that she already knew the names of) She was more accurate at many of these experiments than 2-3 year old human children, and remembered the names of the toys better than the researchers themselves did (they had to write the names on the toys as they kept forgetting them). Now of course this doesn't mean that dogs are MORE intelligent than children or even AS intelligent (intelligence is more about the ability to learn than what an individual knows, and children clearly have a far greater potential to learn than dogs) but it does show that dogs are not 'stupid beasts' and they have some mental abilities comparable to human children between 2-3. Blame stupid readers for not understanding the distinction, the papers were just desperate to get their attention with shocking headlines – and if you are surprised by the papers doing so then you aren't so bright yourself.

RSS feed for comments on this post