Forming a grammatically correct sentence may seem to require advanced cognitive skills, but it turns out that our creative language capacity might rely on a less sophisticated system than is commonly thought. A recent study suggests that our ability to construct sentences may arise from procedural memory—the same simple memory system that lets our dogs learn to sit on command.

Scientists distinguish between procedural memory, which is relevant for learning skills such as how to swim, and declarative memory, which stores knowledge, including facts and memories of events, such as one’s birthday, says Victor S. Ferreira of the University of California, San Diego. To find out which system is at work when we form sentences, Ferreira and his team exploited a phenomenon called syntactic persistence—speakers tend to use the same grammatical pattern they have used or heard in previous sentences.

The researchers tested four healthy individuals and four amnesic patients. The amnesiacs’ procedural memory was intact, meaning that they could learn skills with repetitive practice, but their declarative memory abilities were damaged, leaving them unable to memorize new facts. First, all participants heard and repeated a sentence. Then they saw an unrelated picture and had to describe it. Finally, participants heard another sentence that was either identical to the original sentence or subtly changed in its meaning or grammatical structure, or both.

Both groups tended to use the grammatical rules of the prime sentence when describing the picture; amnesic patients, however, did not remember that they had seen the sentence before. The fact that they still used its syntactic structure is surprising because it suggests that the procedural memory system is responsible for putting grammatical sentences together, Ferreira says. He adds that the findings also shed light on our understanding of procedural memory itself, which was thought to be restricted to specific experiences and motor skills. This study shows it is also able to support abstract knowledge, making it “more powerful than previously thought,” he explains.

Note: This article was originally printed with the title, "Memory for Grammar".

4 Comments

1. 1. MikeB 05:21 PM 4/21/09

   Without examples of the sentences used in the experiment this article tells us very little. Apparently the subjects used grammatical structures they had previously heard, but maybe those structures were optimal -- or only -- forms to describe the event, in which case little of note was revealed. How many ways are there to say "the cat sits in the window"?
   
   In fact, the finding that amnesiacs used similar constructions to non-amnesiacs would seem to indicate that both groups were using common grammatical rules rather than memory to form the sentences. Isn't that what they are doing when forming sentences about completely new experiences? Why isn't that an equally valid explanation here?
   
   And finally, the authors of the study seem to think that a general rule of human language formation can be extrapolated from an experiment apparently conducted in one language (not identified but probably English).
   
   This article, I'd have to say, is pretty lame.
   
   

   Reply | Report Abuse | Link to this

2. 2. ironChicken in reply to MikeB 05:17 PM 4/22/09

   the cat is sitting in the window
   a cat is sitting in a window
   there's a cat sitting in the window
   there's a window with a cat sitting in it
   in a window there sits a cat

   Reply | Report Abuse | Link to this

3. 3. kheiner 05:14 PM 12/28/09

   good job, iron chicken!
   so how many languages can you speak, MikeB?

   Reply | Report Abuse | Link to this

4. 4. kheiner 05:15 PM 12/28/09

   good job, ironChicken!
   How many languages do you speak, MikeB?

   Reply | Report Abuse | Link to this