Standard and Non-Standard Dialects
There’s a quote sociolinguists like, that’s been repeated so many times no one seems to remember who first said it: “A language is a dialect with an army and a navy.”
Everybody speaks a dialect, from a linguistic perspective, whether they are your high school English teacher or a black man in inner-city Detroit. From a linguistic perspective, all dialects are equally valid, equally correct, equally rich and capable of expressive speech. All dialects have regular, complex rules that govern the formation of sounds and sentences, whether or not those rules permit multiple negation or the omittance of the verb ‘to be’ in certain environments (i.e., He working). However, from a social perspective, this is not the case.
What sociolinguists would call Standard American English, or SAE, is often viewed as the only correct, or at least the only normal, way of speaking. People seldom think of it as a dialect at all. What people think of as dialects are the non-standard dialects: rural southern dialects, urban black dialects, Hispanic dialects, Appalachian dialects, etc. Dialects become something substandard, those interesting but somehow degenerated ways certain types of people speak. While occasionally some dialects are viewed as colorful or charming, they are mostly considered to be incorrect, ungrammatical, and/or uneducated . This can result, as you might expect, in a number of problems for non-standard dialect speakers, and tends to have a negative effect on their educational experience.
Standardized tests have been much maligned for being unfair to ethnic and social minorities, and the situation gets especially sticky when it comes to something like language proficiency. A person can be perfectly proficient at following grammatical rules and expressing themselves clearly, but when you test them in a dialect they do not speak, the results will not reflect their abilities.
The following questions are from the California Achievement Test, and were written to test third graders’ English language ‘achievement.’ Tests like this one are used to determine things from what kind of funding a school gets to whether a particular child is marked as 'learning disabled.' The instructions asked the student to pick the answer that “you think is correct.”
1) Beth _______ home and cried.
2) Can you ________ out now?
3) When _______ I come again?
What do these questions test? A variety of different things, really. Take question 1, to start with. The correct answer to this question (according to the test makers) is b., Beth came home and cried. But to speakers of several vernacular dialects in which the “bare root form” of come is used as an irregular past tense form, the more natural, grammatically correct answer is a., Beth come home and cried.
What is defined as ‘correct’ by the people who made this test is using the standard dialect as opposed to the non-standard dialect. For a speaker of Standard American English, this is a very easy question to get right: all you do is, as the instructions say, pick the answer that “you think is correct.” But for a speaker of a dialect in which come is an acceptable answer, it’s significantly trickier: you must be aware that the standard form of the past tense of come is came, even though that is not how you would say it, and you must be able to recognize this as a situation where, regardless of what the instructions say, you should go against instinct and pick the standard form. Some nine-year-olds can do that, but many will have trouble.
Question 2 is a different kind of question entirely. The correct answer is b., Can you go out now, which is grammatical in both standard and non-standard English dialects. Furthermore, choice a., Can you went out now, is ungrammatical for both standard and non-standard dialects. Unlike question 1, question 2 is testing, equally for all native English speakers, their ability to recognize a linguistically well-formed sentence. There is little chance that the dialect of the test-taker is going to have an impact on what answer they choose or feel to be correct.
The third question is different still. According to the test makers, the correct choice is b., When may I come again. But in reality, this answer is unnatural for most speakers of standard and non-standard dialects. In everyday speech, most people are likely to use can in a question like this, which in modern English has come to refer to both capability and permissibility. The rule that may must be used for issues of permissibility is a prescriptive rule— something that doesn’t actually describe the language as it is spoken. All test takers have to learn this rule and where to use it (though whether they have all been instructed on such matters equally is a different question entirely). This question is a gauge of how effectively students have been educated in, and have learned, when to use certain unusual language forms in order to make certain adults happy.
If you make a test up of a mixture of questions like 1, questions like 2, and questions like 3, what will that test tell you? Will it measure a student’s language ‘achievement’? That depends on what you mean by achievement. If you mean the achievement of learning the delicate and complex language systems of their native dialect, than the answer is no, though for some students it will come closer than for others. If you mean the achievement of understanding what the standard dialect is and when it should be used, than the answer may be yes. But that ‘achievement,’ important as it may be to a student’s academic career, is a much shallower one than the first. This is something that educators, test-makers, politicians, and parents all need to be aware of: this test will not, predominately, measure a student’s language skill or intelligence, though both will impact the results. It will measure a mixture of language ability, social ability, quality of education, and social background.
What is to be done?
Okay, so the system’s not perfect. Standardized tests have been widely accepted as a necessary evil: they may have inherent flaws, but we still need them if we want to have any idea of the progress both individual students and educational systems as a whole are making. I have always, personally, hated the things, but I know they aren’t going anywhere. However, I think there may be hope of things changing a little. Here are some of my ideas for how we can make these tests more reasonable:
1. Decide what we want a given English test to measure: language skill development, or standard formal language skill development. Mixing the two is inconsistent and produces inconclusive results.
2. If a test is supposed to measure language skill development, make sure none of the questions are inherently biased towards one dialect. Make a whole test of questions like number 2 in the example above, which address a student’s ability to recognize well-formed language, regardless of their native dialect.
3. If a test is supposed to measure standard formal language skill development, make it measure just that: the extent to which students have been taught, and have learned, to use certain standard forms in certain situations. Make the test out of questions like 1 and 3, in which students must chose standard formal forms over non-standard or informal forms. Change the way instructions are given: instead of asking for the ‘correct’ answer, or ever worse, the answer ‘you think is correct,’ ask for which would be best to use at school. Expect students from certain regions, social classes, ethnic groups, etc., to have lower scores than speakers of Standard American English, because they are not native speakers of the dialect.
4. Recognize the limitations of the tests. Schools should look at a variety of sources, not just these tests, to evaluate the potential and the achievement of individual students. Realize that a test of formal standard English will reflect the influence of a wide range of English dialects spoken in America, some of which are closer than others to the formal standard. Realize that doing poorly on a test of standard English doesn’t make a person dumb or even bad at languages, and may have more to do with social realities than linguistic ones. Realize that language is too complicated, varied, and all together amazing to be measured well with a standardized test.