BRUCE CRONNELL

SWRL Educational Research and Development

ABSTRACT This article presents an analysis of the errors found in a set of writing samples produced by third- and sixth-grade black children attending an inner city school in a low-income neighborhood. The analysis focuses on those errors that may be influenced by Black English oral usage. Hypercorrections are also noted. A few error types (primarily morphological) seem to be most influenced by Black English speech patterns.

Research indicates that oral language forms may have some influence on the written texts of students who do not speak standard English. The influences seem most strong--or at least most obvious--at the level of surface features of written text. Nonstandard pronunciations may affect spelling. Nonstandard grammatical features may also appear in writing, but such effects are harder to identify and are less well documented.

A major non-standard dialect in the United States is Black English--that form of English commonly spoken by low-income blacks. Black English (BE) is characterized by the presence of a number of phonological and grammatical features ( Berdan, 1981; Dillard, 1975; Fasold & Wolfram, 1970; Labov, 1972). The effects of BE on the spelling of elementary school students (as reflected on spelling tests) has been documented ( Carney, 1979; Groff, 1978; Kligman & Cronnell, 1974; O'Neal & Trabasso, 1976; Sullivan, 1971). However, most studies of the writing of BE speakers used older students ( D. G. Briggs, 1969; O. D. Briggs, 1969; Collins, 1971; Goppert, 1975; Weaver, 1974; Wolfram & Whiteman, 1971 ). The present study was undertaken to investigate Black English influences on the writing of elementary school students.

Although BE influences may be found in the writing of black students, this is not to say that any writing problems are due only--or even primarily--to dialect. Instead, this is a study of how dialect features might affect students' writing.

The study used a set of writing samples produced by third- and sixth-grade students as part of an end-of-year assessment in a large school district in the metropolitan Los Angeles area. The third-grade students wrote a story about a drawing of a monkey and an elephant on roller skates at a starting line. The sixth-grade students wrote a letter to convince a friend to watch a favorite television program.

The writing samples analyzed were from one school in a low-income, predominantly black, inner city area. Papers from students identified by teachers as black were included for analysis. The papers were from 99 students in five third-grade classes and 68 students in three sixth-grade classes. No data were obtained on students' actual speech. However, the school is in an area where many BE speakers live; moreover, previous studies at the school had found that most students spoke Black English. Thus, when Black English features are found in the writing of these students, it is extremely likely that these same features are found in their speech.

All errors--defined as deviations from standard formal written English--were compiled, with the following exceptions: 1) capitalization and punctuation errors; 2)

for the sixth-grade samples, errors in letter format and errors in the spelling of proper names from television programs. The errors were classified in three ways: possible Black English influence (e.g., "lef" for "left"); possible hypercorrection from Black English (e.g., "lefted" for "left"); or, no obvious Black English influence or hypercorrection (e.g., "loft" for "left").

The classification of errors was generally conservative; that is, errors were classified as BE influenced or as hypercorrections only when strong evidence suggested the possibility. However, many of the errors so classified may not actually be the result of dialect influence or hypercorrection: All writers tend to make some errors no matter what their dialect, especially under testing conditions ( Wing & Baddeley, 1980).

Moreover, BE-related errors are not necessarily unique to speakers of Black English since BE shares some features with other dialects of English. In addition, research (e.g., Whiteman, 1981) has consistently shown that many dialect-related errors also appear (although less frequently) in the texts of writers who do not speak the dialect. For example, writers who always use third person singular -s in their speech may sometimes omit it in writing. Thus, all results reported here must be viewed cautiously.

In the writing samples from 99 third-grade students, 550 errors were identified (an average of 5.6 per student); of these, 39% were considered to have possible influences from Black English (212, an average of 2.1 per student). In the writing samples from 68 sixth-grade students, 437 errors were identified (an average of 6.4 per student); of these, 33% were considered to have possible influences from Black English (144, an average of 2.1 per student).

Those errors that were possibly influenced by Black English oral patterns were sorted into five categories: verbs, nouns, syntax (excluding verb and noun use), consonants, vowels. In general, the relative frequency of specific error types was similar across both grade levels (see Table 1). However, sixth graders tended to have more variety in their errors (see "other" categories), probably because the writing-sample prompt allowed more options for variations.

Verbs. The largest category of errors was related to verb formation and use. A major problem was the lack of third person singular -s or -es in the present tense; e.g., the elephant look rather than the elephant looks. (Only a few corresponding hypercorrections were found--that is, an -s form where only the simple verb was needed.)

Another major problem was the lack of -ed on past tense verbs and on past participles. Third graders had

particular difficulties with named, as in the monkey is name Bounce. (Third graders frequently had trouble with tense in general, often shifting between present and past; consequently, sometimes the missing verb suffix could be interpreted as either -s or -ed. Sixth graders were more consistent in their use of tense.)

Use of the verb be also created some problems, particularly for third graders, who frequently used was for were (and occasionally vice versa) and is for are. Such errors were less common for sixth graders. A few students at both grade levels also omitted is or was in some sentences. Only one case of invariant be was found (Mork be talking), suggesting perhaps that students are aware of highly stigmatized forms and avoid them in writing.

Other verb errors can generally be placed in three minor categories: 1) misuse of perfect tenses (either past instead of participle form, e.g., had came; or use of past perfect when only the simple past was needed, e.g., had landed instead of landed); 2) inappropriate tense in modals (e.g., can for could, and vice versa); and 3), got for has or has got.

Nouns. Problems with nouns involved the use of suffixes (as did many problems with verbs): lack of the plural -s and lack of possessive -s. Moreover, the addition of an unnecessary -s or -s (especially among third graders) was a common hypercorrection.

Syntax. The major syntactic problem (excluding noun and verb errors) was the Black English use of a for an, with an occasional hypercorrection of an for a. A number of sixth graders used the preposition-less structure come over my house (instead of come over to my house).

Four syntactic features that are usually noted as particularly distinctive of Black English occurred only rarely: 1) inversion in indirect questions (e.g., asked did he want instead of asked if he wanted); 2) multiple negation (actually only double negation in these papers; e.g., don't have nothing); 3) it's for there's; 4) repeated subject (e.g., Wilona she). However, the rarity of the first three features could be the result of students not needing to use such constructions in these writing samples.

Consonants. The major error with consonant spellings was the omission of the last consonant in final consonant clusters; e.g., port for pond. Among sixth graders (particularly one student who made half of such errors), omission of the final consonant was most frequent in and, which was spelled an and in. Since and is such a common word, it is surprising that students misspelled it--even if their misspellings reflect the most common pronunciation for all speakers of English. Hypercorrections sometimes included the addition of an unnecessary final consonant; e.g., find for fine.

Some single final consonants were also missing in the writing samples; e.g., loo for look. Post-vocalic r, l, and n (whether or not in final position) were sometimes missing; e.g., mothe (for mother), aways (for always), dow (for down). Occasionally, th was spelled with t or d; e.g., tink, den (for then); however, voicing contrasts were not necessarily preserved in the spelling; e.g., ten for then. Since "law" does not occur in Black English (or in a large number of other English dialects), it is not surprising to find cases of w instead of wh (e.g., wen for when), but it is somewhat surprising that all these cases occurred in the sixth-grade samples and with the highfrequency words when and where. A few hypercorrections also reflected these consonant problems (e.g., were spelled where).

Vowels. Vowel misspellings that could be attributed to dialect pronunciations were rare. Moreover, because English vowel spellings are particular problems for all elementary school students, vowel errors cannot be clearly attributed to dialect even when they match dialect pronunciations. One vowel misspelling that suggests Black English influence did occur with some slight frequency: i for e before nasals (e.g., thin for then); however, this pronunciation is common in many dialects, especially in Southern California.

The data reported in this article suggest that a significant portion of the writing errors made by these thirdand sixth-grade black children can be attributed to Black English influence, either directly or through hypercorrections. However, dialect isn't the only--or the predominant--source of error for these students, and the correction of surface errors is not the only--or even the most important--goal of writing instruction.

In addition, the relative presence or absence of errors was not necessarily related to the overall quality of the writing samples. Some papers with few or no errors were short, simple, and dull. Some papers with many errors were lengthy, complex, and interesting. In the former cases, the students apparently played safe and ventured little. In the latter cases, the students, in their attempts at more complex writing, took risks and sometimes had problems. However, these risk-takers probably have more potential as writers, even though they may make more errors as they push their writing skills to the limits.

This particular study has a number of limitations, which suggest directions for needed additional research. Comparison groups were not used. Thus, although the study provides information about one dialect group, it does not indicate whether other dialect groups might also make the same errors. However, other research does suggest that error frequency, at least, differs among dialect groups ( Whiteman 1981).

Speech data were not available for the students whose compositions were analyzed, although there is no reason to believe that they did not speak Black English.

The students whose writings were used in this study were from the metropolitan Los Angeles area. Since dialects vary regionally, somewhat different results may be found if black students in other parts of the country are studied. Additionally, writings studied were at only two grade levels; different error patterns might well be found at other grade levels.

The writing samples studied were on different topics, which made different linguistic demands on the writers. Consequently, the data for the two grade levels cannot be compared.

The specific topics for the writing samples clearly influenced the structures used by the students. Different topics could elicit different syntactic structures, thus producing different error types and different raw frequencies of error types.

Writing samples were produced under test conditions; under more natural conditions, the results might be different. In particular, students might proofread and edit their writing more carefully and perhaps produce fewer errors.

The study looked only at errors, not at what was correct. Therefore, the frequency of an error type may be

related to the frequency of the construction in the writing samples and may not reflect the relative difficulty of the construction for black third- and sixthgrade writers. For example, more errors may have been found on third person singular -s than on plural -s because students used more third person singular forms than they did plurals. Errors were analyzed out of context. However, specific linguistic environments may affect the likelihood that an error will occur. Analysis of context would, of course, need to include the study of correct usage as well as errors.

Nonetheless, despite these limitations, the present study does provide additional knowledge about the dialect-related writing problems of black elementaryschool students.

Berdan, R. "Introduction to black English". In B. Cronnell (Ed.), The writing needs of linguistically different students. Los Alamitos, Calif.: SWRL Educational Research and Development, 1981 .

Briggs, D. G. Deviations from standard English in papers of selected Alabama negro high school students. Doctoral dissertation, University of Alabama, 1969. (University Microfilms No. 69-6528)

Briggs, O. D. A study of deviations from standard English in papers of negro freshmen at an Alabama college. Doctoral dissertation, University of Alabama, 1969. (University Microfilms No. 69-6529)

Carney, S. The relationship of black non-standard dialect features to a core spelling vocabulary. Union, N.J.: Kean College, 1979 . (ERIC Document Reproduction Service No. ED 170 700)

Collins, J. J. "Deviations from standard English in written compositions of disadvantaged college freshmen and regular admissions students at Glassboro State College". Dissertation Abstracts International, 1971 , 32, 749-A.

Dillard, J. (Ed.). Perspectives on black English. The Hague: Mouton, 1975 .

Fasold, R. W., & Wolfram, W. "Some linguistic features of negro dialect". In R. W. Fasold & R. W. Shuy (Eds.), Teaching standard English in the inner city. Washington, D. C.: Center for Applied Linguistics, 1970 .

Goppert, N. "The black English vernacular in the writing of young adults from Dayton, Ohio". Dissertation Abstracts International, 1975 , 36, 2180-A.

Groff, P. "Children's spelling of features of black English". Research in the Teaching of English, 1978 , 12, 21-28.

Kligman, D., & Cronnell, B. Black English and spelling. 1974 . (ERIC Document Reproduction Service No. ED 108 234)

Labov, W. Language in the inner city. Philadelphia: University of Pennsylvania Press, 1972 .

O'Neal, V. & Trabasso, T. "Is there a correspondence between sound and spelling? Some implications for black English speakers". In D. S. Harrison & T. Trabasso (Eds.), Black English: A seminar. Hillsdale, N.J.: Lawrence Erlbaum, 1976 .

Sullivan, R. E. A comparison of certain relationships among selected phonological differences and spelling deviations for a group of negro and a group of white second grade children. Unpublished doctoral dissertation, University of Texas at Austin, 1971 .

Weaver, C. Black dialect? or black face? 1974 . (ERIC Document Reproduction Service No. ED 091 713)

Whiteman, M. F. "Dialect influence in writing". In M. F. Farr (Ed.), Writing: The nature, development, and teaching of written communication (Volume 1: Variation in Writing). Hillsdale, N.J.: Lawrence Erlbaum, 1981 .

Wing, A. M. & Baddeley, A. D. "Spelling errors in handwriting: A corpus and a distributional analysis". In U. Frith (Ed.), Cognitive processes in spelling. London: Academic Press, 1980 .

Wolfram, W. & Whiteman, M. "The role of dialect interference in composition". Florida FL Reporter, 1971 , 9, 34-38; 59.