Keywords
African American English - assessment - treatment - DLD
Learning Outcomes: As a result of this activity, the reader will be able to:
-
Describe and give an example of a dialect-general overt form, dialect-specific overt
form, and zero form for past tense within AAE.
-
Describe the formula for calculating a child's percent of overt forms for plural subject
auxiliary BE past (e.g., “the puppets were/was building a tower”) within AAE using
dialect-informed, strategic scoring.
-
Explain differences between an AAE-speaking child with and without DLD in terms of
T/A form productivity and sensitivity to linguistic context.
Speech-language pathologists assess, treat, and advocate for individuals who present
with a variety of speech and language disorders within linguistically diverse communities.
A linguistically diverse community includes not only monolingual and multilingual
speakers of various languages but also mono(dia)lectal and multi(dia)lectal speakers
of English. Unfortunately, although several dialects of English are spoken in the
United States, General American English (GAE) remains the primary dialect affirmed,
taught, and tested in school.
Perhaps early studies of English language acquisition based on GAE are understandable
given the dialects of the research pioneers, such as Roger Brown who studied Adam,
Eve, and Sarah. However, the field has progressed, and there is now a growing number
of multilectal researchers who are passionate about developing materials and practices
that are dialect-appropriate for all children. Dialect discovery worksheets are tools
created from these efforts. Clinicians can use dialect discovery worksheets to learn
about the various types of overt forms and zero forms children produce to express
grammatical structure within their dialects and the relative frequencies (i.e., productivity)
of these form types overall and by clausal context. A focus on grammatical productivity
is important for clinical practice because across dialects of English, children with
language impairments often present with a less productive grammatical system than
their same dialect-speaking, typically developing peers (Hendricks & Adlof, 2020;
Oetting et al., 2019; 2021; 2022; Rice, 2003).
The goal of this study is to demonstrate the clinical utility of a dialect discovery
worksheet focused on tense and agreement (T/A) structures using archival data from
four kindergartners who spoke a rural southern variety of African American English
(AAE). As background, we describe the conceptual framework and empirical evidence
supporting the worksheets.
The Disorder Within Dialects Framework
Dialect discovery worksheets were created from the disorder within dialects framework
(Oetting, 2018; 2024; Oetting et al., 2016; 2022; Oetting & Gregory-Martin, in press).
This framework is cross-linguistic and can be applied to any linguistic variety (whether
described as a language or a dialect), including AAE. Key to the framework is the
consideration of a child's entire linguistic system (Green, 2011). For T/A, this means
examining all overt forms and zero forms a child uses to express T/A. The disorder
within dialects approach is also focused on individual differences within a dialect
and is less concerned about differences between dialects. Given this, the disorder
within dialects approach leads to questions that differ from those asked when guided
by the traditional, dialect versus disorder approach. Whereas the dialect versus disorder
approach focuses on whether a child's utterance or response is reflective of a dialect
or a disorder, we ask whether a child demonstrates a productive repertoire of dialect-appropriate
T/A forms (i.e., able to express different types of T/A forms with a diverse set of
verbs) and presents form use that is sensitive to linguistic context in ways that
are dialect-appropriate and age-appropriate relative to same dialect-speaking peers.
Guided by the disorder within dialects framework, members from our research lab have
conducted several studies focused on T/A structures across dialects of English, including
rural and urban dialects of AAE and rural dialects of Southern White English, Cajun
English, and Gullah/Geechee-influenced AAE (e.g., Berry & Oetting, 2017; Cleveland
& Oetting, 2013; Oetting & Garrity, 2006). These structures include auxiliary forms
of BE in the present and past tense (e.g., “is, are, was, were, Ø”), regular and irregular
past tense (e.g., “mowed, threw, throwed, had show, mowØ”), and regular verbal “-s”
(e.g., “shows, showØ”). Within these examples and as will be discussed later, zero
forms are denoted by the symbol “Ø.” We include these zero forms and several other
types of dialect-appropriate overt forms in these examples to illustrate the larger
repertoire of forms found in AAE than in GAE. Many other dialects of English spoken
in and outside of the United States also contain a larger repertoire of zero forms
and overt forms than found in GAE (c.f., Britian, 2024; Hickey, 2014). We note this
to advocate for greater recognition of the English dialect variation that exists worldwide
and to encourage more studies of childhood developmental language disorder (DLD) within
these dialects of English.
Across dialects of English, there is evidence that children with DLD are delayed in
their development of T/A compared to their same-age and same dialect-speaking typically
developing (TD) peers (for AAE and SWE, see Garrity & Oetting, 2010; Oetting, 2019;
Oetting et al., 2016; 2019, 2021; Oetting & Garrity, 2006; Oetting & McDonald, 2001;
for past tense data in AAE, see Seymour et al., 1998; for a mix of dialects, see Hendricks
& Adlof, 2020; for a review of many studies conducted in GAE, see Rice, 2003). To
illustrate, it is useful to consider the published study by Oetting et al. (2019)
from which the current four cases were extracted. The AAE group included 70 kindergartners,
with half classified as DLD and the other half as TD, and their data were elicited
from probes targeting T/A forms. When the data from all probes were combined, the
DLD group's dialect-appropriate overt forms averaged 43% (SD = 22%) compared to the
TD group's 71% (SD = 20%).
AAE-speaking children with DLD have also been found to benefit from language interventions
that allow forms of T/A structures to serve as targets when form productivity is limited.
As an example, Smith and Bellon-Harn (2015) completed a treatment study involving
10 AAE-speaking preschoolers with DLD. The treatment involved storybook reading, and
forms of auxiliary “is” and “are” were two of many grammatical structures targeted
with expansions, cloze procedures, and modeling. At baseline, the preschoolers produced
few clauses (n = 160) to support an auxiliary form, and within these clauses, very low percentages
of dialect-appropriate overt forms (“is” = 6% and “are” = 7%). Following treatment,
the preschoolers increased their total auxiliary clauses to 384 and their average
percentages of dialect-appropriate overt forms to 47% in “is” clausal contexts and
21% in “are” clausal contexts. As shown by this study, targeting T/A structures, such
as auxiliaries, led to important gains for the AAE-speaking children with DLD.
It is also noteworthy that as the preschoolers with DLD increased their use of auxiliary
clauses and auxiliary forms, their percentage of overt forms was higher for “is” contexts
than for “are” contexts (47 vs. 21). Adults and TD children who speak AAE show this
same sensitivity to linguistic context, producing higher percentages of overt forms
for “is” contexts than for “are” contexts (Blake, 1997; Oetting et al., 2019; 2021;
Rickford et al., 1991; Roy et al., 2013; Wyatt, 1996). Thus, as the preschoolers with
DLD developed a more productive auxiliary system, their auxiliary forms also showed
sensitivity to linguistic context as is appropriate for their dialect of AAE.
Labels to Describe the Large Repertoire of T/A Forms Within AAE
As mentioned earlier, AAE has a large repertoire of surface forms to express T/A.
As an example, speakers of AAE can produce multiple surface forms when expressing
the past tense of the verb “find,” including “found, fount, founded, finded, had found,
had find, and findØ.” Over the years, researchers and clinicians have grappled with
how to talk about the larger repertoire of surface forms for AAE as compared to GAE.
Seymour et al. (1998) introduced the labels, noncontrastive and contrastive, while
we and others have used the labels, mainstream and nonmainstream (e.g., Hendricks
& Adlof, 2020; Oetting et al., 2019; Terry et al., 2012) or mainstream and AAE (e.g.,
Byrd & Brown, 2021; Washington et al., 2018). With all three sets of labels, the first
in the pair (i.e., noncontrastive and mainstream) refers to AAE forms shared with
GAE and the second (i.e., contrastive, nonmainstream, and AAE) refers to forms found
in AAE but not in GAE.
As authors, we have recently moved away from these labels because they require a reference
to GAE to be meaningful. These labels also divide the grammar of AAE into two components
(those shared and not shared with GAE), which makes it difficult to study how children
make use of their entire linguistic system to communicate with others and succeed
in school. Finally, the label, nonmainstream has a negative connotation to us and
some of our students and colleagues. Indeed, the Merriam-Webster dictionary defines
nonmainstream as an adjective that means “not having, reflecting, or being compatible
with the prevailing attitudes and values of a society or group: not belonging…” (www.merriam-webster.com/dictionary/nonmainstream).
As an alternative, we have adopted three terms that do not require a reference to
GAE. The use of three labels rather than two also adds scientific precision to practice
and facilitates the study of AAE as a complex linguistic system. The three terms are:
dialect-general overt forms, dialect-specific overt forms, and zero forms. Descriptions
and examples of these forms are as follows.
Dialect-General Overt Forms
These forms are morphemes one can perceive as a listener, and they are found in many
dialects of English. Examples of AAE dialect-general overt forms documented in the
current work include: “The girl ‘is’ painting; The ladies ‘are’ sneezing; The bear
‘was’ touching; The puppets ‘were’ building; The tiger mowed; The boy ‘threw’; The
girl ‘dries’.”
Dialect-Specific Overt Forms
These forms are morphemes one can also perceive as a listener, but they are found
in only some dialects of English. Examples of AAE dialect-specific overt forms documented
in the current work include the use of “is” and “was” with third-person plural subjects
(e.g., “They's throwing, They was building”) and for past tense, preterite “had” (e.g.,
“had threw, had throw”) and regularized forms (e.g., “throwed, threwed”). Dialect-specific
overt forms also exist for verbal “-s” (e.g., use with first-, second-, or third-person
plural subjects, such as “I says, You says, They says”).
Zero Forms
Zero forms are morphemes not perceived by the listener. Examples of AAE zero forms
documented in the current work include: “The girl Ø painting; The ladies Ø sneezing;
The bear Ø touching; The puppets Ø building; The tiger mowØ; The boy throwØ; The girl
dryØ.”
Without careful study, zero forms are often misunderstood (Oetting et al., 2022).
When very young AAE-speaking TD children are beginning to produce grammatical morphemes,
a zero T/A form likely represents an omission, with the “Ø” form lacking grammatical
content. As an example, consider AAE speaker CH, whose language samples were studied
by Newkirk-Turner et al. (2016). At 1 year and 8 months of age, she produced 179 utterances,
an MLU in morphemes of 2.64, five clauses that could support an auxiliary BE form,
and within these clauses, three different auxiliary forms (i.e., “I'm, I'ma, Ø”).
In addition, her clauses that supported these auxiliary BE forms were limited to declaratives.
However, as AAE-speaking TD children age, their linguistic systems become more complex,
and their T/A zero forms begin to encode grammatical content just as their overt forms.
To illustrate, consider AAE speaker CW, whose language samples were also studied by
Newkirk-Turner et al. (2016). At 4 years and 3 months of age, he produced 235 utterances,
an MLU in morphemes of 5.30, 36 auxiliary BE clauses, and eight different auxiliary
forms (i.e., “I'm, I'ma, Ø, ain't, it's, was, they is, they're”). Across his language
samples, his clauses supporting auxiliary BE forms also diversified to include not
only declaratives but also questions, negation, and ellipses. Moreover, his use of
dialect-general overt, dialect-specific overt, and zero forms varied by linguistic
context in ways that were consistent with a more mature version of AAE. Specifically,
he produced higher percentages of dialect-general overt forms and dialect-specific
overt forms for “am” and “was” clausal contexts than for “is” clausal contexts, and
higher percentages of overt forms for all three of these clausal contexts than for
“are” clausal contexts. CW's sensitivity to linguistic context when producing his
overt and zero forms provides evidence of grammatical encoding for both form types.
In other words, whereas 1 year and 8-month-old CH's zero forms reflected grammatical
omissions, 4 years and 3-month-old CW's zero forms reflected phonetically silent and
grammatically rich synonyms to his overt forms.
Other studies have also shown AAE-speaking TD children's use of T/A forms to be sensitive
to linguistic context. Roy et al. (2013) examined auxiliary BE forms as produced within
language samples by 24 AAE-speaking preschoolers, aged 4 to 6 years. Consistent with
CW's data, the 4-year-olds produced dialect-general and dialect-specific overt forms
58% of the time, and the 6-year-olds produced these forms 64% of the time. In addition,
their use of overt forms and zero forms varied by linguistic context, yielding a high
percentage (>90%) of overt forms for “was, were, and am” clausal contexts, a lower
percentage (59%) for “is” clausal contexts, and the lowest percentage (27%) for “are”
clausal contexts. As was found for 4-years and 3-month-old CW, these AAE-speaking
TD children were producing syntactically rich zero forms and demonstrating dialect-appropriate
sensitivity to linguistic context when producing their overt forms and zero forms.
Dialect-Informed, Strategic Scoring
A final concept that is important for measures of grammar within a dialect is how
percentages of overt form use are calculated. All the reviewed studies calculated
percentages of overt form use by combining a child's number of dialect-general overt
forms and dialect-specific overt forms and dividing this sum by the child's total
number of dialect-general overt, dialect-specific overt, and zero forms. Calculating
percentages in this way reflects what we refer to as dialect-informed, strategic scoring.
With this scoring approach, the full repertoire of T/A forms (dialect-general overt,
dialect-specific overt, and zero) within a child's dialect is viewed as appropriate,
and the focus is on the relative frequencies of overt forms and zero forms.
Elsewhere we have compared strategic scoring to two other scoring systems, referred
to as unmodified and modified (Oetting et al., 2019; 2021; see also Girolamo et al.,
2024; Hendricks et al., 2023). Unmodified scoring is implemented when a test designed
for GAE is given to a child who speaks AAE, and the child's responses are not scored
with modification for the child's dialect; this scoring system does not recognize
the child's AAE-appropriate responses. Modified scoring systems are often encouraged
within manuals of tests designed for GAE. With these scoring systems, clinicians are
encouraged to score as correct any response that is dialect-appropriate. Unfortunately,
this scoring system does not consider important individual differences in the relative
frequencies at which different types of dialect-appropriate forms are produced by
children who vary in age (younger vs. older) or language ability (DLD vs. TD). Given
strategic scoring's focus on children's relative frequencies of dialect-appropriate
forms, it is sensitive to both developmental change and childhood DLD.
We now turn to the four case studies to demonstrate the usefulness of dialect discovery
worksheets for clinical practice. Dialect discovery worksheets can be created for
any aspect of a child's grammar system, but here we crafted the worksheets to examine
the participants' repertoire of overt and zero T/A forms and their sensitivity to
linguistic context when producing their forms. Consistent with the disorder within
dialects framework, clinicians can use results from the worksheets to determine if
a child's use of forms to express T/A is productive (i.e., able to express different
types of T/A forms with a diverse set of verbs) and whether the child's form use shows
sensitivity to linguistic context in ways that are expected for the dialect being
learned.
Methods
Participants
Data for the four participants were from Oetting et al. (2019); a subset of data from
two of these participants was also presented as part of a learner activity within
Oetting and Gregory-Martin (in press). We refer to these four participants as Tyson
and Tia and Devin and Destiny. Tyson and Tia were classified as TD, and Devin and
Destiny were classified as DLD. DLD is a clinical condition that, in the original
study, was referred to as specific language impairment; as with other studies of childhood
SLI, the profile for those with SLI present low language abilities with age-appropriate
nonverbal abilities and articulation skills and without other clinical conditions,
such as autism, Down syndrome, etc.
Tyson and Devin were male and Tia and Destiny were female. All were enrolled in a
public kindergarten in the rural south. Their ages spanned 61 to 68 months, and their
mothers' highest level of education spanned 9 to 17 years, with 12 reflecting completion
of high school. The four children were selected for this case study by first identifying
an African American male and female in the DLD group in the datafile and then finding
an African American child in the TD group of similar gender and age. Race was reported
by the children's caregivers. Also, per caregiver report, the four children were not
identified as Hispanic or bilingual.
Tyson and Tia did not have a family history of speech or language impairments, and
neither received services from a speech-language pathologist. Devin and Destiny had
a positive family history of speech and language impairment, although only Destiny
was receiving services from a speech-language pathologist. Finally, the dialects of
all four participants were perceived to be AAE by three independent listeners during
a blinded listener judgment task. The dialects of all four participants were also
classified as reflecting strong variation from mainstream English as measured by the
Language Variation subtest of the “Diagnostic Evaluation of Language Variation: Screening
Test” (DELV-ST; Seymour et al., 2003). This screener does not classify children's
dialects as AAE, GAE, or any other named dialect but instead classifies the children's
dialects as either mainstream, some variation from the mainstream, or strong variation
from the mainstream. We retain the screener's dialect classification labels here to
be consistent with others who may use this screener to describe the dialects of clients
or research participants.
As part of the original study, the participants were administered a battery of assessments,
including the “Primary Test of Nonverbal Intelligence” (PTONI; Ehrler & McGhee, 2008),
Sounds-in-Words subtest of the “Goldman-Fristoe Test of Articulation-2” (GFTA-2; Goldman
& Fristoe, 2000), Risk subtest of the DELV-ST, Syntax subtest of the “Diagnostic Evaluation
of Language Variation: Norm Referenced” (DELV-NR; Seymour et al., 2005), and the “Peabody
Picture Vocabulary Test-4” (PPVT-4; Dunn, 2007).
As shown in [Table 1], all four participants' standard scores on the GFTA-2 and PTONI were above 84; these
tests have a normative mean of 100 and a standard deviation of 15. On the Risk subtest
of the DELV-ST, Tyson, and Tia, the TD participants were identified as low risk for
language impairment, whereas Devin and Destiny, the participants with DLD, were identified
as either medium to high risk or high risk for language impairment. Similarly, on
the Syntax subtest of the DELV-NR, Tyson, and Tia earned a standard score above 7,
and Devin and Destiny earned a standard score below this level; this subtest has a
mean of 10 and a standard deviation of 3. Finally, the children's PPVT-4 standard
scores ranged from 68 to 105; this test has a mean of 100 and a standard deviation
of 15. It was administered for descriptive purposes only given concerns regarding
the validity of this measure for identifying children with DLD (Gray et al., 1999;
Ukrainetz & Duncan, 2000). As shown, Tyson scored above the normative mean on the
PPVT-4, and the others scored close to or below—1 SD of the normative mean.
Table 1
Participant profiles
|
Participant
|
Tyson
|
Tia
|
Devin
|
Destiny
|
|
Clinical group
|
TD
|
TD
|
DLD
|
DLD
|
|
Gender
|
M
|
F
|
M
|
F
|
|
Age (mo)
|
68
|
61
|
64
|
63
|
|
MED
|
17
|
13
|
9
|
11
|
|
Family history
|
No
|
No
|
Yes
|
Yes
|
|
Receiving services
|
No
|
No
|
No
|
Yes
|
|
GFTA-2
|
95
|
110
|
94
|
84
|
|
PTONI
|
95
|
103
|
94
|
84
|
|
DELV-ST dialect
|
Strong variation
|
Strong variation
|
Strong variation
|
Strong variation
|
|
DELV-ST risk
|
Low-risk
|
Low-risk
|
Medium to high-risk
|
High-risk
|
|
DELV-NR
|
10
|
11
|
5
|
3
|
|
PPVT-4
|
105
|
86
|
78
|
68
|
Data
Data for the case studies were elicited using four probes presented through PowerPoint
slides on laptops. The probes were designed to elicit a child's full repertoire of
T/A forms with 64 different verbs; 16 items were included in each probe (auxiliary
BE present “is” and “are,” auxiliary BE past “was” and “were,” past tense, and verbal
“-s”). For each item, participants were shown a video of either children, adults,
or puppets completing an action. The actions were 4 seconds for auxiliary BE present
tense, past tense, and verbal “-s” items and 6 seconds for auxiliary BE past tense
items. One probe was administered each day, with the order counterbalanced across
participants. A description of the prompts and verbs targeted within the probes can
be found in Oetting et al. (2019); however, we re-present some of the information
here to explain how the participants' T/A forms were elicited.
Auxiliary BE Present
This probe included eight third-person singular present auxiliary BE (e.g., “is”)
contexts and eight third-person plural present auxiliary BE (e.g., are) contexts.
Nouns were used as subjects and combined with 16 different verbs. For each item, a
participant was given a verbal prompt accompanied by an action in a still frame (e.g.,
“The ladies seem sick. Tell me what you see”). Then, the action was played to elicit
a participant's response (e.g., “The ladies/they are/is/Ø sneezing”). As shown by
this example, the participants' dialect of AAE allowed them to produce a dialect-general
overt form, dialect-specific overt form, or zero form for these items.
Auxiliary BE Past
This probe included eight third-person singular past auxiliary BE (e.g., “was”) contexts
and eight third-person plural past auxiliary BE (e.g., “were”) contexts. Nouns were
again used as subjects and combined with 16 different verbs. For each item, the examiner
used an imperative context to introduce the target verb (e.g., “Watch the guys cut
paper”). Then the action was played while the examiner repeated the prompt twice.
During the third prompt and with the action in play, the examiner covered the screen
with paper and posed a question (e.g., “Before I covered this up, what do you remember
seeing”). This question encouraged participants to respond based on their memory of
the action (e.g., “The guys/they were/was/Ø cutting paper”). As shown by this example,
the participant's dialect of AAE allowed them to produce a dialect-general overt form,
dialect-specific overt form, or a zero form for these items.
Past Tense
This probe included eight regular, past-tense main verb contexts and eight irregular,
past-tense main verb contexts. The verbs selected for the regular contexts ended with
a vowel (e.g., “tie”), liquid (e.g., “tear”), or glide (e.g., “throw”), but those
selected for the irregular contexts were free to vary in phonetic composition. Also,
all 16 verbs were followed by “a” or “an” to increase transcription reliability. To
encourage the participants to produce verbs with past tense, four action videos were
shown sequentially in four quadrants on the screen. For each item, the examiner presented
the target verb in an imperative context while simultaneously playing the action (e.g.,
“Watch the doll ride a scooter. Watch her ride a scooter. Now she's done”). When an
action within a quadrant ended, it stayed in the still frame to reduce the memory
load for the participant. After the four actions were in a still frame, the examiner
prompted the participant to describe the completed actions using temporal adverbs
(e.g., Now you tell me, “First…; Then …; Then …; Then …”). This presentation and prompting
format encouraged participants to describe the four past events in a sequence (e.g.,
First, “she drew/drawØ/drawed/drewed/had draw/had drew/had drawed/had drewed a picture,
then…”). As shown by this example, the participants' dialect of AAE allowed them to
produce several different dialect-general overt forms, dialect-specific overt forms,
and zero forms for these items.
Verbal “-s”
This probe included eight actions within a third-person singular habitual context
and eight actions within a third-person singular nonhabitual context. We refer to
this morpheme as verbal “-s” here and elsewhere instead of third-person singular because
AAE and some other dialects of English allow the “-s” morpheme to attach to verbs
in other person contexts (e.g., “I says, they comes”). Nevertheless, the probe administered
was limited to third-person singular contexts. All verbs within this probe ended in
a glide (e.g., “chew”) or vowel (e.g., “see”) to encourage a diverse repertoire of
T/A forms. Also, negative contrast was used to introduce the target verb following
the work of Seymour et al. (2003). To use negative contrast, the examiner played the
action two times while providing prompts describing what the agent of the action was
not doing (e.g., “This is a girl. She doesn't dry a spoon. She doesn't dry a cup”).
Then, the examiner played the action a third time and asked the participant to describe
what the agent was doing (e.g., “She dries/dryØ a plate”). The habitual prompts included
the added word, “always,” in an attempt to encode an action that occurred often, whereas
the nonhabitual prompts did not, in an attempt to encode an action that could occur
once and at the moment of viewing. As shown by the examples, the participants' AAE
dialect allowed them to produce a dialect-general overt form or zero form. Since all
subjects were singular, the probe did not give the participants an opportunity to
produce dialect-specific overt verbal “-s” forms.
Coding the T/A Forms
The participants' responses during the probes were audio-recorded for later coding
in the lab. Even though the probes were designed to elicit T/A forms with specific
subjects and verbs, participants were free to respond in other ways, and all targeted
T/A forms were coded based on the participant's produced clause structure. In other
words, participants could produce a singular subject for a plural subject and vice
versa, a regular verb for an irregular one and vice versa, and the word, “always,”
to an item designed to be nonhabitual and vice versa. The participants were also free
to produce a different verb than the target if the verb was not included within an
earlier response on the day the probe was administered.
For each targeted T/A clausal context, the children's responses were coded as either
dialect-general overt, dialect-specific overt, or zero (see [Table 2] for examples of responses from the case studies presented here). As shown in the
table, the participant's use of “is” for a plural subject auxiliary present tense
context was scored as a dialect-specific overt form in the “are” column, and a child's
use of “was” for a plural subject auxiliary past tense context was scored as dialect-specific
in the “were” column.
Table 2
Examples of elicited T/A forms from the four participants
|
Dialect-general overt
|
Dialect-specific overt
|
Zero
|
|
Auxiliary BE singular subject present
|
is painting
|
–
|
Ø clapping
|
|
Auxiliary BE plural subject present
|
are giving
|
's punching
is giving
|
Ø washing
|
|
Auxiliary BE singular subject past
|
was feeding
|
–
|
Ø rocking
|
|
Auxiliary BE plural subject past
|
were building
|
was building
|
Ø building
|
|
Regular and irregular past tense
|
mowed
drew
|
teared
drawed
had dyed
had throwed
had threw
|
showØ
catchØ
|
|
Regular verbal “-s”
|
chews
buys
|
–
|
flyØ
emptyØ
|
The probes also elicited some un-scoreable responses from the participants, and those
with DLD produced more un-scorable responses than the TD children. In the original
study, these responses and others were coded as other. For dialect discovery worksheets
and the four cases highlighted here, unscorable responses were those that did not
obligate a dialect-general overt, dialect-specific overt, or zero form for the targeted
T/A clausal context. For example, if a participant produced, “they are building” during
the auxiliary BE past (“was/were”) probe, this T/A form was deemed unscorable for
auxiliary BE past. Similarly, if the participant produced, “he walks” during the past
tense probe, this T/A form was deemed unscorable for past tense.
Across all four probes, zero T/A forms also were deemed unscorable if the participant
produced a verb in isolation (e.g., “pull, build, mow, show”). In other words, for
a zero form to be scorable, participants had to produce a subject preceding the verb
(e.g., “‘The girl’ swallowØ”) and for the auxiliary BE contexts, they also had to
produce a verb with a present progressive morpheme (The tiger “Ø mowing”). When children
produced a verb in isolation, examiners prompted the participants to try the item
again with a subject (e.g., “Try it again but start with the tiger”). If a participant
provided a revised response with the required elements to confirm the clausal structure,
the revised response was scored.
The practice of not scoring verbs in isolation in studies of T/A follows work by Rice
and colleagues (for a review of works, see Rice, 2003), because a verb in isolation
may reflect the production of an action label, which does not obligate a T/A form.
As such, verbs in isolation are uninformative as to the participants' repertoire of
T/A forms. Nevertheless, several unscorable responses, especially after repeated examiner
prompting, provide information about a participant's ability to use the materials
and prompts to generate clausal structure.
Completing the Dialect Discovery Worksheets
Once a participant's responses were coded, a dialect discovery worksheet was completed
(see [Table 3]). The worksheet provides a cell to record each participant's number of each type
of T/A form (i.e., dialect-general, dialect-specific, and zero) by clausal context.
Finally, the last column in the worksheet allows the clinician to combine the participant's
number of forms across the clausal contexts. At the bottom of the worksheet is a row
for the clinician to calculate percentages of overt forms by each clausal context
and an overall percentage of overt form use across contexts. As described earlier,
strategic scoring was used to calculate percentages of overt form use. Recall that
strategic scoring combines the participant's number of dialect-general overt forms
and dialect-specific overt forms and divides this sum by the participant's total number
of dialect-general overt forms, dialect-specific overt forms, and zero forms.
Table 3
Blank dialect discovery worksheet for T/A forms
|
Child
|
Form
|
Elicited context
|
|
Is
|
Are
|
Was
|
Were
|
Past tense
|
Verbal “-s”
|
Overall
|
|
Dialect-general overt
|
|
|
|
|
|
|
|
|
Dialect-specific overt
|
|
|
|
|
|
|
|
|
Zero
|
|
|
|
|
|
|
|
|
Percent overt
|
|
|
|
|
|
|
|
Note: Is = third-person singular subject BE present; are = third-person plural subject
BE present; was = third-person singular subject BE past; were = third-person plural
subject BE past; past tense = main verb regular and irregular past tense; verbal “-s” = main
verb habitual and nonhabitual third-person singular subject present.
Results
[Table 4] provides completed dialect discovery worksheets for Tyson and Tia, the two participants
classified as TD, and [Table 5] provides this same information for Devin and Destiny, the two participants classified
as DLD. Below, we describe the results from these worksheets, focusing on the children's
form productivity and use of their various forms as a function of clausal context.
Table 4
Dialect discovery worksheet for participants Tyson and Tia
|
Child
|
Form
|
Elicited context
|
|
Is
|
Are
|
Was
|
Were
|
Past tense
|
Verbal “-s”
|
Overall
|
|
Tyson
|
Dialect-general overt
|
8
|
0
|
8
|
0
|
10
|
10
|
36
|
|
Dialect-specific overt
|
0
|
7
|
0
|
8
|
3
|
0
|
18
|
|
Zero
|
0
|
1
|
0
|
0
|
3
|
6
|
10
|
|
Percent overt (%)
|
100
|
88
|
100
|
100
|
81
|
63
|
84
|
|
Tia
|
Dialect-general overt
|
7
|
0
|
8
|
0
|
5
|
0
|
20
|
|
Dialect-specific overt
|
0
|
5
|
0
|
7
|
1
|
0
|
13
|
|
Zero
|
1
|
3
|
0
|
0
|
10
|
16
|
30
|
|
Percent overt (%)
|
88
|
63
|
100
|
100
|
38
|
0
|
52
|
Notes: Percent overt form use with strategic scoring = ([dialect-general overt forms + dialect-specific
overt forms]/[dialect-general overt forms + dialect-specific overt forms + zero forms]) × 100.
Is = third-person singular subject BE present; are = third-person plural subject BE
present; was = third-person singular subject BE past; were = third-person plural subject
BE past; past tense = main verb regular and irregular past tense; and verbal “-s” = main
verb habitual and nonhabitual third-person singular subject present.
Table 5
Dialect discovery worksheet for participants Devin and Destiny
|
Child
|
Form
|
Elicited context
|
|
Is
|
Are
|
Was
|
Were
|
Past tense
|
Verbal “-s”
|
Overall
|
|
Devin
|
Dialect-general overt
|
0
|
0
|
0
|
0
|
2
|
0
|
2
|
|
Dialect-specific overt
|
0
|
0
|
0
|
0
|
3
|
0
|
3
|
|
Zero
|
8
|
7
|
0
|
1
|
9
|
16
|
41
|
|
Percent overt (%)
|
0
|
0
|
0
|
<1
|
36
|
0
|
11
|
|
Destiny
|
Dialect-general overt
|
0
|
1
|
0
|
0
|
2
|
2
|
5
|
|
Dialect-specific overt
|
0
|
0
|
0
|
0
|
0
|
0
|
0
|
|
Zero
|
4
|
5
|
8
|
7
|
12
|
14
|
50
|
|
Percent overt (%)
|
0
|
17
|
0
|
0
|
14
|
13
|
9
|
Notes: Percent overt form use with strategic scoring = ([dialect-general overt forms + dialect-specific
overt forms]/[dialect-general overt forms + dialect-specific overt forms + zero forms]) × 100.
Is = third-person singular subject BE present; are = third-person plural subject BE
present; was = third-person singular subject BE past; were = third-person plural subject
BE past; past tense = main verb regular and irregular past tense; and verbal “-s” = main
verb habitual and nonhabitual third-person singular subject present.
Tyson
Tyson produced a scorable response for all 64 elicited contexts, of which 36 were
dialect-general overt forms, 18 were dialect-specific overt forms, and 10 were zero
forms. Recall that dialect-specific overt forms (e.g., “The puppets ‘was’ building;
The lady ‘drawed’ a picture”) are not errors but reflect the large repertoire of forms
and the productive nature of the TD system within AAE. With dialect-informed scoring,
his overall percentage of T/A overt forms was 84 ([36 + 18]/[36 + 18 + 10]) × 100.
Together, his high number of scorable T/A forms and his percentage of dialect-appropriate
overt forms indicate that Tyson presented a productive T/A system. He also produced
higher percentages of overt form use for “was” and “were” and “is” clausal contexts
than for “are” clausal contexts (“was, were, is” = 100% > “are” = 88%), and a higher
percentage of overt forms for past tense clausal contexts (81%) than for verbal “-s”
clausal contexts (63%). This pattern of form use indicates that Tyson's use of T/A
forms was sensitive to linguistic context in ways that were appropriate for his AAE
dialect.
Tia
Tia also produced many scorable T/A forms (n = 63 or 98%) for the 64 items. Of her scoreable responses, 20 involved a dialect-general
overt form, 13 involved a dialect-specific overt form, and 30 involved a zero form.
With dialect-informed, strategic scoring, her overall percentage of T/A overt forms
was 52 ([20 + 13]/[20 + 13 + 30]) × 100. She also produced higher percentages of overt
form use for “was” and “were” clausal contexts (100%) than for “is” clausal contexts
(88%), and higher percentages of overt forms for all three of these clausal contexts
than for “are” clausal contexts (63%). In addition, she produced a higher percentage
of overt forms for past-tense clausal contexts (38%) than for verbal “-s” clausal
contexts (0%). Like Tyson, Tia demonstrated a productive T/A system as evidenced by
her high number of scoreable T/A forms and her pattern of form use, which showed sensitivity
to linguistic context in ways that were appropriate for her dialect of AAE.
Devin
Devin produced a scoreable response for only 46 (72%) of the 64 items. Of these 46
T/A forms, 2 were dialect-general overt, 3 were dialect-specific overt, and 41 were
zero forms. With dialect-informed, strategic scoring, his overall percentage of overt
forms was 11 ([2 + 3]/[2 + 3 + 41]) × 100. This finding shows that Devin did not have
a productive repertoire of forms to express T/A. Instead, he relied heavily on zero
forms. Moreover, his relative percentages of overt forms and zero forms to express
auxiliaries were not yet showing sensitivity to linguistic context because he did
not show higher percentages of overt forms in “was” and “were” clausal contexts than
in “is” clausal contexts or “are” contexts. However, Devin showed a higher percentage
of overt forms in past-tense clausal contexts (36%) than in verbal “-s” clausal contexts
(0%), and this pattern of form use is appropriate for his dialect of AAE.
Destiny
Destiny produced a scoreable response for only 55 (86%) of the 64 items. Of these
55 T/A forms, 5 were dialect-general overt and 50 were zero forms. With dialect-informed,
strategic scoring, her overall percentage of overt forms was 9 ([5 + 0]/[5 + 0 + 50]) × 100.
Like Devin, Destiny did not have a productive repertoire of forms to express T/A.
Like Devin, she also relied heavily on zero forms. Finally, Destiny's relative percentages
of overt forms and zero forms did not show sensitivity to linguistic context. Destiny
did not show the expected pattern of higher percentages of overt forms in “was/were” > “is” > “are”
auxiliary clausal contexts nor did she show a significantly higher percentage of overt
forms in past tense clausal contexts (14%) than in verbal “-s” clausal contexts (13%).
Conclusion
The disorder within dialects framework supports the study, assessment, and treatment
of children with DLD within the context of their respective dialects. Dialect discovery
worksheets were created from this framework to help clinicians do this. To illustrate
the clinical usefulness of these worksheets, we created a worksheet focused on children's
T/A forms and filled it out using elicitation probe data from four participants who
spoke AAE. The AAE dialects of all four participants were classified on the DELV-ST
as showing strong variation from mainstream English, and they all attended a public
kindergarten in the rural south. In the original study, two participants were classified
as TD and two as DLD; in each of these groups, male and female participants were selected
to demonstrate that it was the participants' clinical condition and not their gender
that contributed to the results.
Results from the dialect discovery worksheets showed that the two TD participants
produced many different scoreable T/A forms when completing the elicitation probes
(i.e., 63 and 64 out of 64 possible). In addition, the participant's percentages of
overt forms were calculated as 84 and 52. These percentages differ from each other,
but both are within the range expected for AAE-speaking TD kindergartners. In the
original study by Oetting et al. (2019), the TD group's average percentage of overt
form use with strategic scoring was 71 (SD = 20%), and the range was 35 to 100.
Both TD participants also demonstrated sensitivity to linguistic context when producing
their T/A forms. Tia's patterns of overt form percentages by linguistic context were
identical to patterns documented within previous adult and child AAE studies, with
her patterns summarized as: “was” and “were” clausal contexts > “is” clausal contexts > “are”
clausal contexts; main verb past tense clausal contexts > main verb verbal “-s” clausal
contexts. In other words, within her auxiliary system and with main verbs, she produced
higher percentages of overt T/A forms expressing the past than the present. Tyson's
patterns of overt form percentages by linguistic contexts were like Tia's although
his percentages of overt forms of “is” contexts were higher than is typically observed
within AAE, with his patterns of overt forms summarized as: “was, were, is” clausal
contexts > “are” clausal contexts; main verb past tense clausal contexts > main verb
verbal “-s” clausal contexts. Several studies of adults and child studies that have
shown similar linguistic context effects for auxiliaries (Blake, 1997; Newkirk-Turner
et al., 2014; Oetting et al., 2019; 2021; Rickford et al., 1991; Roy et al., 2013;
Smith & Bellon-Harn, 2015; Wyatt, 1996), and there are a few child studies showing
higher percentages of overt forms for past tense clausal contexts than verbal “-s”
clausal contexts (Green, 2019; Oetting, 2019; Oetting et al., 2021; but see Hendricks
& Adlof, 2020 who did not find this pattern).
Compared to the participants classified as TD, Devin, and Destiny produced fewer scorable
T/A forms when completing the elicitation probes (i.e., 46 and 55 out of 64 possible).
Fewer scorable T/A forms indicate that both Devin and Destiny struggled to use the
materials and examiner prompts to produce clauses to support a T/A form. Devin and
Destiny also did not produce a diverse set of dialect-general overt forms, dialect-specific
overt forms, and zero forms. As a consequence, their percentages of overt form use
were very low at 11 and 9. Compared to the original study, these percentages are lower
than the lowest percentage produced by an AAE-speaking TD child. Also, Devin and Destiny
did not produce their T/A forms in a way that demonstrated sensitivity to linguistic
context. Instead, both DLD children relied heavily on zero forms for all T/A clausal
contexts elicited.
Implications for Clinical Practice
The four case studies presented here show how dialect discovery worksheets focused
on T/A can be used to identify the grammar strengths of children classified as TD
and the grammar weaknesses of children classified as DLD within the dialect of AAE.
Recall that only Destiny was receiving services from a speech-language pathologist.
Had a dialect discovery worksheet been available, a clinician may have been able to
use it to explore Devin's linguistic system, especially if his caregivers and teachers
raised concerns about his language abilities. Although results from a dialect discovery
worksheet in isolation would never qualify a child for services, these results could
be used to support other assessment measures. Following a convergence of evidence
approach to assessment (Castilla-Earls et al., 2020), Devin may have qualified for
services using results from a dialect discovery worksheet along with caregiver and
teacher input, classroom observation, analysis of a language sample, and a dialect-appropriate
norm-referenced test such as the DELV-NR.
For children with DLD (including Devin and Destiny), dialect discovery worksheets
can also help a clinician develop treatment goals. Based on the worksheets, Devin
and Destiny likely would have benefited from a goal focused on increasing their production
of diverse clausal structures to support T/A forms and a goal focused on increasing
their production of dialect-general overt forms, dialect-specific overt forms, and
zero forms with a diverse set of verbs within these clauses. This latter goal could
also be written with the added criterion that their T/A forms show dialect-appropriate
sensitivity to linguistic context, with higher percentages of overt forms within clausal
contexts expressing the past (“was, were,” and past tense on main verbs) compared
to clausal contexts expressing the present (i.e., “is, are,” and verbal “-s” on main
verbs).
Regarding treatment approaches, expansion, cloze procedures, and modeling within the
context of storybook reading may be ideal for young children like Devin and Destiny
who produce limited numbers of clauses to support T/A forms and showed a heavy and
dialect-inappropriate reliance on zero forms. Recall that these methods were implemented
in Smith and Bellon-Harn's (2015) treatment study of AAE-speaking preschoolers with
DLD. After treatment, increases were seen in the frequency at which the preschoolers
produced clausal structures to support auxiliaries and in their percentages of dialect-appropriate
overt forms within these clauses.
Limitations
For T/A structures and these types of goals to be incorporated into clinical practice,
clinicians will need to embrace the study, assessment, and treatment of childhood
DLD within the context of children's respective dialect(s) and become proficient in
the use of dialect discovery worksheets and dialect-informed, strategic scoring. The
authors acknowledge that clinicians have varying levels of linguistic knowledge about
grammar structures within and across dialects of English, and a certain level of knowledge
is needed to utilize the worksheets. Future plans by the authors include creating
training tutorials to facilitate clinicians' use of the worksheets. Prompting children
to produce forms for specific T/A structures from materials and learning how to code
children's forms as either dialect-general overt, dialect-specific overt, zero, or
un-scorable for the target T/A structure also takes practice. Neither of these examiner
behaviors is part of current pedagogy in the field of speech-language pathology. Also,
additional dialect discovery worksheets focused on other aspects of grammar need to
be developed as not all children with DLD show T/A weaknesses within their dialect.
Third, the current study focused on archival data from four participants who participated
in a previously published group comparison study. Although results for the four participants
mirrored those reported for the groups in the published study, dialect discovery worksheets
should be completed with more children, children of different ages, and children who
speak other varieties of AAE and other dialects of English. It would be wrong to assume
that all children perceived to speak a dialect of AAE (or any other dialect of English,
including AAE as influenced by another language, such as Spanish) will produce their
T/A forms following the same linguistic patterns documented here. Indeed, a slightly
different pattern of linguistic context effects has been documented for T/A forms
in children whose AAE dialects have been influenced by Gullah Geechee—a Creole language
spoken in South Carolina and elsewhere (Berry & Oetting, 2017). To help guide these
future works, dialect discovery worksheets could be used as shown here or as modified
to learn about these dialects and to better understand individual differences among
children within their respective dialects.
