A U-shaped curve of development

We've finished up First Language Acquisition in my LA class and are moving on to Second Language Acquisition. As part of our study of L1, each of us had to do a child language observation. We had to choose a child between the ages of 0-7 (by age 7, most aspects of the basic framework of spoken language have fully developed) and then choose a specific feature of language to investigate. We did this by either replicating a language elicitation device from the literature or coming up with something on our own. Then we had to write up the entire experiment, complete with results analysis. It was required that we film our interaction with the child.

Of course I chose Magdalena - I have easy access to her and she is already accustomed to my presence, so I figured an observation would be cake. Unfortunately, it didn't turn out that way. I had a hard time getting her to produce the language samples I needed. Specifically, I was testing her progress along what is called a U-shaped curve of development.

Something interesting that young kids do during the course of figuring out their native language is make progress by taking a few steps backwards. At an early age, they can produce correct irregular past-tense verb forms such as

went

, without understanding that it is an irregular past-tense verb form. They're just slapping together unanalyzed (but correct) chunks of language that they've heard their interlocutors use.

After a while, however - right around Magdalena's age and a little older - the kids have a "wait a second..." moment when they realize that verbs in English are supposed to form the past tense by adding +ed at the end. So then they produce incorrect (but painstakingly analyzed) language such as

goed,

because they've overgeneralized the +ed past-tense verb ending to apply to ALL verbs in English.

A little later, they'll figure out that certain verbs have an irregular form that needs to be used, and they swing up the other side of the U-shaped curve of development and produce

went

again. This time, however, they're producing it correctly on purpose. They're not just imitating what they've heard others say. They've analyzed the situation and figured out that the

went

form needs to block the incorrect

goed

form.

Pretty neat, huh? I was all ready for Magdalena to bust out all kinds of incorrect verb+ed forms, like holded, bringed, finded, drawed, etc., because that's what stage she is approaching, age-wise. But when it came down to actually filming the experiment, she did weird things like say, "Yesterday, the boy go to school." I could have sworn she doesn't really talk like that in real life. However, as I transcribed the things she was saying, I was shocked at how kid-like her speech really is. I never realized how much I took into account body language and tone and inference and context when interpreting her speech. Put on paper, kids' language really is quite patchy, unclear, and riddled with errors.

(It's not just kids, though. In

The Language Instinct

, Steven Pinker talks about how one of the most damning sentences Nixon produced on the Watergate recordings was, "For your immediate thing you've got no choice with Hunt but the hundred and twenty or whatever it is." That sentence doesn't make sense on paper, does it? But with intonation and pauses and context added in, it was completely clear to to the person he was talking to.)

Anyway, I was sad that Magdalena ended up not being a textbook illustration of the U-shaped curve of development, but I concluded in my report that she's probably just starting to move away from the correct, unanalyzed forms like

went

and is on her way to the incorrect, analyzed

goed

. My experiment with her happened to catch a snapshot of a few days of her language development, that's all.

As a reality check, here's what it's like working with a 3.5-year-old in a language experiment. This was even after I'd changed the original elicitation technique to try to hold her attention more. Sigh.