Dog Might Provide Clues on How Language Is Acquired -

A. Katz amnfn at
Tue Jan 18 20:04:55 UTC 2011

I don't think there is any doubt that dogs who live with humans comprehend 
much of what is said around them. I've known dogs who were bilingual who 
understood things that were said, when other humans in the room who did 
not speak the language in question didn't understand. The problem with all 
this evidence? It's anecdotal.

There is the opposite problem with the experiment described here.
In this type of set-up, all the evidence is predictable and rote.
The experiment described in the NYTimes article did not teach the dog to 
comprehend by rote: the dog was trained to obey by rote.

The problem with this kind of experiment is that it doesn't seem to 
distinguish comprehension from obedience. The way the dog proves 
comprehension is by fetching an item. (The assumption being, if he didn't 
fetch it, we have no proof of comprehension.)

As an aside, all the words for "objects" in the experiment are really 
being used as commands. Sue Savage-Rumbaugh made a very big point that if 
a chimp uses "apple" to ask for an apple, this does not mean that he 
understands that apple also refers to an apple in a different kind of 

It's easy to get chimps in experimental settings to ask for things, and 
it's easy to get dogs (of certain breeds) to fetch things, but what is not easy 
is to prove comprehension when uncoupled from obedience or someone asking 
someone else to do something.

I don't doubt that the dog in question understands the words. I'm just 
saying there is something wrong with our standards of proof.


On Tue, 18 Jan 2011, Tom Givon wrote:

> Before we all go wild patting ourselves on the shoulder for being able to 
> learn a 2nd (and 3rd, and 4th) language, I thought the enclosed link might 
> furnish some amusement, maybe even perspective. The SL-learner in question is 
> native speaker of Canine, a language many of us have been struggling to 
> acquire for many years with, alas, less-than-perfect results. Cheers,  TG

More information about the Funknet mailing list