cac.ophony.org

I just finished running a psycholinguistic experiment for my dissertation research. I am working in the field of sentence processing, which looks at how one ‘parses’ the sentences they hear and how to resolve potential ambiguity they might encounter. My project is on English and I am targeting native speakers of English, more precisely those of American English. The participants are undergraduates at Queens Collee who are enrolled in Psychology 101, and they were doing this for a course credit.

This is how the experiment goes. You read a mini dialogue, proceeding from one ‘frame’ to another with the button press, and your task is to choose the answer choice that best fits the dialogue in the third frame. For example,

<start dialogue>
1. Until Frank got the fancy job
that he was just bragging about,
<proceed>
2. how much money was he making?
<proceed>
3. He earned far less than you. OR He earned more than anyone.
<choose Left or Right>

Obviously the right hand answer would best fit the dialogue if you carefully
read and understood the content of the whole dialogue.
I won’t go into details of what we are interested in testing, but it suffices here to know that although students tend to think that they are only being tested for the question-answering accuracy, we also measure how long it took to move through the dialogue, and compare the ‘reading times’ across different constructions.

After running the experiment, we evaluated their performance, since if the participants are reading too slow or making too many errors, we need to discard the data. And an interesting thing I noticed was that most of the data that we had to throw away come from people who were recent immigrants from other places where English is spoken but that English is a dialect that is not American English (for example, Jamaica or Trinidad). They were all ‘prescreened’ by the research management system as native speakers of English. But it fascinated me to know that it showed such a systematic difference in performance between people who are early immigrants (came to the US in their infancy (0-6)) and those who came later (9 or later, when according to a theory the learning of a second language will become dramatically more difficult).
As I interacted with them as the experimenter, they speak good English and they look and behave reasonably smart - it’s not like they forged their background information or they are not part of the smart bunch. I can see them contribute good ideas to the class and write good papers. But in the world of milliseconds in an intricate reading experiment like this, the difference shows up. Their reading speed was almost as if they were second language speakers of English, though highly proficient ones.

Although we are not looking to research on this issue, I brought this up in the meeting I had with one of my advisors and we agreed that the dialectal difference between American English and their English is so huge that we just can’t view them as the kind of ‘native speakers of English’ that we are targeting for.
It has definitely proven something to me. The students who speak certain dialects of English, though treated the same as other American native speakers, they might have more in common with advanced ESL population than we think. And, from my experience of interacting with them, it is definitely not the matter of intellect, or their ‘broken’ English. Though this point may be obvious to some people, I feel like I truly learned something about the variety of Englishes.