Accurate distinction between language differences and language disorders within differences is critical in the promotion of equity in education. Although this section of this book focuses on African American English (AAE), the principles of cultural sensitivity, including respect for non-mainstream American English (NMAE) apply to work with speakers of all NMAEs.

One important addition to the evaluation process when working with speakers of NMAEs is a measure of dialect density. Dialect densityDialect density refers to the proportion of dialectal features used by a speaker. Three methods used to quantify dialect density are listener judgment, type-based counts, and token-based counts (Oetting & Macdonald, 2002). In the listener judgment method, the listener rates the speaker’s dialect using a likert scale. Type-based counts count the number of dialectal features a speaker uses, regardless of the number of times the feature is used. Token-based counts use the number of times dialectal features are used. Three main token-based approaches include 1) number of utterances containing dialectal features divided by total number of utterances, 2) number of times dialectal features were produced divided by total number of words, and 3) number of times dialectal features were produced divided by total number of utterances.

Lee-James and Washington (2018) advocate for a strengths-based perspective to working with bilingual and bidialectal speakers. A strengths-based perspective focuses on what the child does bring to the learning process, rather than focusing on the child’s deficits. Complex syntax is one strength of speakers of AAE. Children with greater dialect density produce more complex syntax, suggesting that the production of dialectal features is a strength. 

Given the predictive value of positive family history as an indicator of language disorder, Oetting and colleagues (2016) suggest gathering information about a child’s immediate and extended family members’ history of speech, language, reading, writing, spelling, learning, stuttering and hearing difficulties. Universal screenings in preschool and kindergarten are useful for identifying children needing a comprehensive language evaluation or increased supports and progress monitoring, as in a Response to Intervention (RtI) or Multitiered Systems of Support (MTSS) model (Oetting et al.). Using more than one type of screening tool, including one that utilizes teacher ratings, is necessary in order to ensure accurate identification of children needing an evaluation or increased supports (Hendricks & Adlof, 2017, Oetting et al., 2016). Sentence recall tasks have the potential to be useful in differentiating children with disorders within dialects when credit is given for non mainstream productions that do not differentiate children with and without disorders within a dialect, such as was  for were and zero-marking of -s (Oetting et al., 2016).

Standardized Tests

The use of scoring modifications on standardized tests that have been normed on speakers of MAE should be performed with caution. Hendricks and Adlof (2017) revealed that, although giving credit for dialectally acceptable productions on sentence repetition tasks decreased over-identification of children speaking NMAE dialects as having DLD, under-identification increased using these modifications. Oetting et al. (2019) developed dialect-informed probes, so termed because the verbs and content were selected to discourage zero-marking in AAE. These probes targeted irregular and regular past tense, habitual and non-habitual verbal -s, is and are, and was and were. Table 2 lists the verbs targeted in each probe.

Table 2. Verbs targeted by Oetting et al. (2019, p. 3447).

Oetting et al. (2019) compared a modified scoring approach with a strategic scoring approach. A modified scoring approach counted as dialectally appropriate all MAE and NMAE overt forms and zero forms as dialectally appropriate. This did not penalize children when they used NMAE forms, but it did not result in detection DLD. A strategic scoring approach counted MAE and NMAE overt forms as marked and zero forms as unmarked. Examples of NMAE overt forms include drinked, drunk, seen, fount, had play, and had played (Oetting et al., 2019, p. 3444). Zero forms are dialectally appropriate in AAE and Southern White English (SWE), but children with DLD have been shown to produce a higher rate of zero-marking. The strategic scoring approach had better diagnostic accuracy than an unmodified or modified approach. This approach would be beneficial to include in clinical practice when evaluating speakers of NMAE.

One set of standardized tests that has utility in identifying disorder within dialect is the Diagnostic Evaluation of Language Variation (DELV; Seymour, Roeper, & de Villiers, 2018). The DELV includes a screening test (DELV-ST), a norm-referenced test (DELV-NR), and a criterion-referenced test (DELV-CR). The DELV focuses on non-contrastive features of English; that is, features shared by all dialects of English. Deficits in non-contrastive features can be indicative of a disorder, rather than a difference due to dialect. The DELV assesses syntax using passive voice constructions, articles, and wh- questions. It evaluates pragmatics via role-taking, question-asking, and questions about a short narrative. The DELV assesses semantics in the areas of verb contrast, preposition contrast, quantifiers, and fast-mapping. Consonant cluster production in sentences provides information on phonology. The DELV-NR has excellent diagnostic accuracy, with  sensitivity and specificity at 0.95 and 0.93 respectively at 1 standard deviation below the mean, and 0.85 and 0.93 respectively at 1.5 standard deviations below the mean (Seymour et al., 2018).

Language Sampling

Language sampling is considered the “gold standard” for assessment of language ability. When assessing the language ability of NMAE speakers, it is important to ensure that the method of analysis does not penalize speakers for using dialectal features. Washington (2011) provides a list of AAE features that can be coded in Systematic Analysis of Language Transcripts (SALT: Miller) software, which provides computerized analysis of language samples. The Child Language Assessment Project’s (CLASP) ongoing work endeavors to identify measures that reliably distinguish AAE speakers with and without language impairment.

Developmental Sentence Scoring (DSS, Lee & Canter, 1974) is one measure that demonstrates the ability to do this, and can be scored automatically using freely available Computerized Language ANalysis (CLAN) software (Overton et al., 2021). The DSS scores the following grammatical constructions: indefinite pronouns, personal pronouns, main verbs, secondary verbs, negatives, conjunctions, interrogative reversals, and wh- questions. Norms are provided for ages three years through six years, 11 months. The information gleaned from the DSS can be useful in developing intervention goals.

The Index of Productive Syntax (IPSyn; Scarborough, 1990) is another measure of syntactic complexity that can be automatically scored using CLAN and does not demonstrate bias against speakers of AAE (Horton-Ikard, 2010; Oetting, 2010, Overton et al., 2021; Stockman et al., 2016). The IPSyn provides information about use of noun phrases, verb phrases, questions, negation, and sentence structure. Although it may not reliably discriminate between AAE speakers with and without developmental language disorder (DLD), it can be useful in the development of goals for intervention (Overton et al., 2021).

Another potential dialect-neutral method of evaluating the language of NMAE speakers is the assessment of rare vocabulary in spoken narratives. The Wordlist for Expressive Rare Vocabulary Evaluation (WERVE, Mahurin Smith et al., 2015) accurately distinguishes gifted African American children from typically developing African American children (Mills et al., 2017), and WERVE scores are not correlated with dialect density. Although more research is needed into the validity of this tool in distinguishing bidialectal children with and without language disorders, these results are encouraging.

For young children (i.e., three-year-olds), a minimal competence core of morphosyntax (MCC-MS; Stockman et al., 2013) has been shown to differentiate speakers of AAE with and without DLD (Stockman et al., 2013; Stockman et al., 2016). Although this was designed for use with young children, it is included in this text because it could be of use with older children whose language ability is similar to that of a younger child. If an older fails the MCC-MS, it would indicate a need for further testing. Noting the structures that are missing from a child’s repertoire could be helpful in identifying intervention goals.

To pass the MCC-MS, in a language sample of 50 utterances, a child must produce

• MLU of at least 2.70
• Elaborated Sentences With Subject + Verb + Complement Plus One orMore Lexical Modifiers of Any Type in 51% of simple sentences
• 4 multiclausal sentences
• 4 sentences with any 2 types of bound morpheme inflections
• 4 sentences with any 2 types of modifiers
• 4 sentences each with subjects and objects, each set with 2 different exemplars of personal pronouns
• 4 sentences each with noun and prepositional phrases 4 sentences with any type of complement combination
• 4 sentences with subject + verb + complement form 4 sentences with 1 of 3 permutations shown
• 4 single-word or phrasal answers to questions

(Stockman et al., 2013, p. 45)

In addition to sampling oral language, a comprehensive assessment should include evaluation of written language. Density of AAE-specific forms in writing does not predict scores on literacy assessments, suggesting that focusing on dialect-neutral forms, which are consistent across GAE and AAE, may be ideal in order to avoid penalizing dialect use in writing assessments (Fitton et al., 2021).

References

Fitton, L., Johnson, L., Wood, C., Schatschneider, C., & Hart, S. (2021). Language variation in the writing of African American Students: Factors predicting reading achievement. American Journal of Speech-Language Pathology, 30, 2653-2867.

Hendricks, A., & Adlof, S. (2017). Language assessment with children who speak nonmainstream dialects: Examining the effects of scoring modifications in norm-referenced assessment.  Language, Speech, and Hearing Services in Schools, 48, 168-182.

Horton-Ikard, R. (2010). Language sample analysis with children who speak non-mainstream dialects of English. Perspectives on Language Learning and Education, 17(1), 16–23. https://doi.org/10.1044/lle17.1.16

Lee, L, & Canter, S. (1974). Developmental sentence scoring: A clinical procedure for estimating syntactic development in children’s spontaneous speech.  Journal of Hearing and Speech Disorders, 36(3), 315-340.

Lee-James, R., & Washington, J.A. (2018). Language skills of bidialectal and bilingual children: Considering a strengths-based perspective. Topics in Language Disorders, 38, 5-26.

Mahurin Smith, J., Dethorne, L., & Petrill, S. (2015). From aardvark to ziggurat: A new tool for passing children’s use of rare vocabulary. Clinical Linguistics & Phonetics, 29(6), 441-454.

Mills, M., Maturin-Smith, J., & Steele, S. (2017). Does rare vocabulary use distinguish giftedness from typical development? A study of school-age African American narrators. American Journal of Speech-Language Pathology, 26, 511-523.

Oetting, J. B., Berry, J. R., Gregory, K. D., Rivière, A. M., & McDonald, J. (2019). Specific language impairment in African American English and Southern White English:

Measures of tense and agreement with dialect-informed probes and strategic scoring. Journal of Speech, Language, and Hearing Research, 62(9), 3443-3461.

Seymour, H. N., Roeper, T. W., & de Villiers, J. (2018). Diagnostic evaluation of language variation – Norm referenced. Sun Prairie: Ventris Learning.

Stockman, I. J., Guillory, B., Seibert, M., & Boult, J. (2013). Toward validation of a minimal competence core of morphosyntax for African American children. American Journal of Speech-Language Pathology, 22, 40–56.

Stockman, I., Newkirk-Turner, B., Swartzlander, E., & Morris, L. (2016). Comparison of African American children’s performances on a minimal competence core for morphosyntax and the index of productive syntax. American Journal of Speech-Language Pathology, 25, 80-96. DOI: 10.1044/2015_AJSLP-14-0207

Washington, J. (2011). The dialect features of AAE and their importance in LSA. In J. Miller, K. Andriacchi, & N. Nockerts (Eds.), Assessing language production using SALT software, (pp. 113-1124. SALT Software, LLC.