Comprehension of complex syntactic structures in Southern varieties of American English and mainstream American English

. Research on the acquisition of complex syntactic structures in Southern English (SE) and Southern African-American English (SAAE) is near absent, though an extensive body of literature is available on phenomena such as those of auxiliary and copula BE , and 3SG -(e)s . de Villiers et al. (2011) supported that characteristics found in AAE help avoid commonly observed developmental errors in the compre - hension of wh-questions. Prior work on mainstream American English (MAE) has shown that where and what are acquired before who , how , why , which , and when . Research on passive voice revealed that children comprehend action verb passives earlier than non-action verb passives . We investigate the comprehension of wh-questions and passive voice in 222 SAAE-, SE-and MAE-speaking children, aged 2-13 and examine whether there were certain structural environments where we examined the comprehension of wh-questions or passive voice that were more chal - lenging for the three groups. The results show that SE and SAAE have comparable development with wh-questions and passive voice, with minor exceptions. They confirm findings from previous studies on both the order of acquisition of wh-ques - tions and the earlier acquisition of action passives, as well as the SAAE-speaking participants’ highly accurate performance with wh-questions, especially structures which include indirect/medial questions .

1. Introduction.Southern English (SE), Southern African-American English (SAAE) and mainstream American English (MAE) are three varieties of English that coexist in most communities in the southern United States.While a growing body of literature on African American (AAE) and SE has shown that what is misperceived as lack of grammatical knowledge, is in fact a characteristic of these linguistic varieties (Green 1995;Oetting 2015), these are still socially stigmatized varieties.A clear example showing that this is still the case comes from Mills et al. (2021), where adults rated children's language in narratives as more acceptable when the language use was closer to MAE than when it varied (specifically with AA children), especially with fictional, as opposed to personal, narratives.
Earlier work on the developmental patterns in child AAE suggested stark contrasts but also many similarities across AAE and MAE (de Villiers et al. 2011;Green 2011;Oetting 2015).Some of the main differences noted in the literature across AAE and MAE are: (a) the frequent omission but also differing uses of auxiliary and copula BE (Green 2011;Green 2002;Roy, Oetting & Moland 2013), the omission of the 3rd person singular agreement morpheme -s/-es (Amaral & Roeper 2014;Green 2016;Newkirk-Turner & Green 20162021;Schneider 1989, and (c) a variety of phonological differences, including differences with the production of some consonants in specific phonological environments, e.g. the change of stops or fricatives from dental to labio-dental, as in [briːv] instead of [briːð], as well as some vowels e.g.[mãn ] or [mae ̃n]instead of [maen], or differences in stress [ˈpoˌlis] instead of [ˌpoˈlis] (Edwards 2008;Tillery & Bailey 2008;Wyatt 1995).The majority of the existing research on AAE falls under three main categories: (a) the overall language development using a variety of diagnostics, (b) the omission of 3SG -s morpheme, (Amaral & Roeper 2014;Green 2016;Newkirk-Turner & Green 2016, 2021;Schneider 1989), and (c) the most widely studied area, the development of BE in all its possible uses, as a copula or an auxiliary (Green 2011;Green 2002;Oetting & Moland 2013;Roy, Oetting & Moland 2013).A small body of work on other linguistic phenomena, such those of negations (Coles-White 2004), wh-questions (Christodoulou, Tsimpli & Cayli 2023;de Villiers, de Villiers & Roeper 2011), whquestions and passives (Christodoulou, Tsimpli & Alquayb 2023), as well as phonetic and phonological development is available.Our current work focuses on the acquisition of two structurally complex syntactic phenomena, the comprehension of wh-questions and passive voice across SE-, SAAE-and MAE-speaking children who are born and brought up in the same communities in Northern Mississippi.
A study on the overall language development of AAE-and MAE-speaking children, aged 2;6 to 3;6) reported that while the performance of 3;6-year-old MAE-speaking children presented lower rates of non-standard speech than that of the younger MAE-speaking children, 3;6-yearold AAE speakers' performance was almost equivalent to that of the 2;6-year-old AAE-speakers.Additionally, the older AAE-speaking children presented an increase in the use of non-MAE forms, but the 3;6-year-old MAE speakers presented a decrease in the production of common developmental errors found during the early stages of language acquisition (Horton-Ikard & Weismer 2005).Research studies, presenting comparisons across MAE and AAE or AAE and SE, discussed a number of similarities and differences across the productions of their tested populations.Oetting (2015) reported similarities between AAE-and SE-speaking children on a morphological and syntactic level, e.g. the use of parallel relative clause markers (AE and what), and the omission of inflectional marking, the 3SG -s agreement morpheme, and absence of BE in environments where it is used as a copula or an auxiliary.Differences mainly concerned: (a) the rates at which AAEand SE-speaking children produce the grammatical forms that deviate from MAE (BE and 3SG-s), (b) the manner in which these grammatical forms are used, and (c) the roles those play selecting some grammar forms over others.
Studies examining the overall performance of AAE-and SE-speaking children revealed difficulties with language acquisition and increased rates or risk for a language disorder.Specifically, the results from Moland & Oetting (2021) revealed a 'fail rate' of 48% for the Washington and Craig Language Screener and a 52% high risk for a language disorder for the language disorder portion of the Diagnostic Evaluation for Language Variation-Screening Test.Similarly, a parallel study on the linguistic abilities of SE-and SAAE-speaking children reported unexpected high rates of risk for a language disorder: 37.6% of the SE-speaking children, and 60.5% of the SAAE-speaking children presented medium-high or highest risk for a language disorder (Christodoulou & Tsimpli (C&T, hereafter) 2021, under review).However, further analysis revealed no correlation across the two parts of the diagnostic, and more explicitly the variety of English and risk for language disorder.We concluded that this unexpected result seems to be due to lack of early education, which may, in turn, cause the test to "over-screen" and falsely flag some of the children as being at risk for a language disorder.Results reported by studies using a more detailed diagnostic, examining the language development on the SE, SAAE and MAE varieties of American English across syntax, semantics, pragmatics and phonology, showed that not only do the three groups perform in parallel across the three populations, but they also presented a highly accurate performance across these 4 domains (C&T 2023a, 2023b), especially with syntax and phonology.
A study on the comprehension of wh-questions in AAE-speaking children revealed higher accuracy with the comprehension multi-clause questions, specifically those including indirect/ medial questions with inversion on the lower subject-auxiliary, as sentences similar to John asked could he do it?,instead of the expected MAE form John asked if he could do it because of the AAE's characteristic marking of indirect questions via inversion in the lower clause.(de Villiers, de Villiers & Roeper 2011).They argue that this helps AAE-speaking children avoid typical errors in the comprehension of wh-questions, observed with MAE-speaking children.Previous work on the production and comprehension of wh-questions in SAAE-and SE-speaking children revealed a parallel performance across the two groups, with marginally higher accuracy scores by the SAAE group (Christodoulou, Tsimpli & Cayli 2023;Christodoulou, Tsimpli & Alquayb 2023).Both groups were far better with comprehension over production (Comprehension: SAAE= 85.5, SE=82.9;Production: SAAE= 60.1, SE=58.9).
Studies on the order of acquisition of wh-questions in MAE-speaking children reported that whwords are mostly acquired in the following order: where and what > who, how, why, which, and when (Bloom et al. 1982;Kuczay & Brannick 1979;Tyack & Ingram 1977).Tyack & Ingram (1977) studied the acquisition of Subject vs. Object wh-questions and found an advantage with the acquisition of subject wh-questions and questions which included an intransitive vs. a transitive verb.Specifically, they showed that children aged 3;0 to 5;6 presented 80% accuracy in providing an appropriate response for subject questions, 57% correct answers with object questions, 66% accuracy with intransitive verb questions, and 57% correct answers to transitive verbs.
Previous research on the comprehension of passives revealed that children comprehend active earlier than passive voice, and comprehend action verb passives earlier than non-action verb passives (Maratsos et al. 1985).Along the same lines, Fox & Grodzinsky (1998) claimed that children use the strategy of assigning an agentive thematic role from the preposition "by", which works for action-/active-verb passives (The rock star is being chased by the koala bear) and not for non-action verb passives (The boy is seen by the horse) whose external argument is often a theme or experiencer.Borer & Wexler (1987) concluded that children's ability to form argument chains that are required to move a verb's object into subject position to make passive sentences is not acquired before the age of five.
With the current study we aim to investigate the development of wh-questions and passive voice in the two Southern varieties of American English and Mainstream American English, and examine potential similarities and differences across the different structural environments where we find/test wh-questions and passive voice.An additional objective of the current study is to produce developmental trajectories for these three varieties of American English, for the overall comprehension of wh-questions and passives from early stages to full acquisition.
Based on what we already know from previous research on these populations, especially C&T (2021) and Moland & Oetting (2021), we should expect: (a) considerably low percentages of accuracy across the two environments for SAAE-and SE-speaking children, and (b) a considerable difference across SAAE-speaking and SE-speaking children, and the SAAE and MAE groups.In accordance with the results reported by C&T (2021), Horton-Ikard &Weismer (2005), andMoland &Oetting (2021), we may expect significant differences across SE and SAAE, as well as these populations and MAE.However, based on C&T (2023a; 2023b), Christodoulou, Tsimpli & Cayli (2023) and Christodoulou, Tsimpli & Alquayb (2023) we may expect a parallel performance across the three groups.Moreover, considering the results from prior work on the acquisition of wh-questions, we should expect where and what, questions to present higher rates of accuracy than how and who/what questions (Bloom et al. 1982;Kuczay & Brannick 1979;Tyack & Ingram 1977).In line with de Villiers, de Villiers & Roeper (2011), we should expect SAAE-speaking children to perform at high rates, potentially higher than SE-and MAE-speaking children, with the more complex 'double wh-questions', such as: How did the mother ask what to bake?.Concerning the comprehension of passives, we predict that action verb passives will be acquired earlier than nonaction verb passives, hence higher accuracy rates will be recorded with action verb passives (Maratsos et al. 1985).

Materials and method.
Forty-six children, speakers of Southern African American English aged 3 to 13, one hundred and thirty-nine Southern English-speaking children aged 2 to 12, and thirty-seven Mainstream American English-speaking children aged 2 to 11, from at least 7 different public and private schools located in northern Mississippi, from both urban and rural areas, participated in this study.Participants came from homes with diverse socio-economic and educational backgrounds.They were monolingual speakers of English, and were exposed to MAE, SAAE, and SE on a daily basis.Information on the three groups' chronological and mental age, as well as their raw score on the IQ test we administered is provided in Table 1 In order to evaluate the participants' performance with the comprehension of wh-questions and passive voice, we administered the Diagnostic Evaluation of Language Variation -Norm Referenced (DELV-NR) (Seymour, Roeper & de Villiers 2005), a standardized test designed to explicitly test the linguistic abilities of children speaking varieties of English that differ from MAE.The test is divided into twelve parts testing for four main linguistic domains: syntax, pragmatics, semantics and phonology.We then isolated Part 1 and Part 2, which focused on the comprehension of wh-questions and passive voice, respectively.Next, we focused on the different types of sentences/ items included in each part and divided them in different categories, depending on their structural environment.Examples 1 though 8 give a breakdown of the different categories we evaluated (Tables 1-2).
1 Categorization of the children into the three study groups was determined based on two questionnaires that parents were required to complete and return to us, alongside the consent form.Questionnaires inquired on information related to the children's and parents' linguistic background, place of birth for the child and the parent, as well as information on education, employment, etc.Some of these were examined as variables to determine whether they had any potential effects on the participants' performance.C. by-phrase (PP) passive sentences.9.The ball was rolling by the boy.

Table 3. Examples of experimental material for passive voice
Prior screening was conducted.Particularly, we administered the Diagnostic Evaluation of Language Variation -Screening Test (DELV-ST) (Seymour, Roeper & de Villiers 2003) to decide whether any of our participants' language presented a variation from mainstream, and whether any of them were at risk for a language disorder.I additionally administered a non-verbal Intelligence Quotient (IQ) test, the Raven's Coloured Progressive Matrices IQ test.We chose a non-verbal IQ test to avoid any potential linguistic effects.To eliminate the possibility of hearing challenges, two auditory tests were also administered to all participants during this session; Test 1 was an imitation production task.It included 25 items of single words.Test 2 included a combination of an imitation production and picture matching task (15 minimal pairs). 2 The screening sessions always preceded the administration of the DELV-NR, and a gap period of at least 2 and up to 4 weeks between the two sessions was ensured to avoid any potential learning effects.In an additional session, we adminis-tered numerous customized experiments, including an imitation production task and four narratives. 3 The four narratives examined, among various other linguistic phenomena, the comprehension of wh-questions.The three sessions were part of a large project, the Marie Skɫodowska Curie Action (MSCA) project, studying language development in SE-and SAAE-speaking children.The current study focuses on results collected only from the second session.
All linguistic testing was administered by highly trained native speakers of SAAE and SE varieties.More than 30 graduate and undergraduate students from the University of Mississippi were employed as research assistants to administer all experimental stimuli and assist with data coding.This was deemed necessary to prevent any bias effects from the participants' interactions with non-native speakers of their respective varieties.For the same reason, and to avoid any effects from variability across speakers/examiners, all experimental stimuli for each child were administered by the same native speaker of their respective variety.Due to reasons related to appropriateness of training, the investigator/author administered the Raven's Coloured Progressive Matrices IQ test to all children.
Two pilot studies were conducted to test how easily and efficiently our standardised and customised experimental tasks could be administered and to correct any difficulties with the material and method of administration.Pilot studies were also partly used to train our numerous research assistants who conducted data collection.The 10 children (aged 3 -7) who participated in the pilot studies were not included in the main study.All sessions were recorded.

Results.
The participants' performance with each of the tested environments was evaluated.To calculate the means, the overall number of correct responses was divided by the overall number of items per answer, for each child.Participants across the three groups presented parallel performance with the comprehension of wh-questions and passive voice (Figure 1), though SAAE-speaking children present a slight advantage with both grammatical phenomena.Both SEand SAAE-speaking children performed better with wh-questions.MAE-speaking children's performance was identical across the two structural phenomena, and the lowest across the three participant groups.

Figure 1. Comprehension of wh-questions and passive voice
The participants' performance was also evaluated per age, and plotted with 2 developmental trajectories, one for each structural phenomenon.Figure 2 shows a near identical performance with the development of wh-questions across age especially after entering the school system at age 5 or 6.MAE-speaking children do present a slight advantage before age 6, especially for ages 4 and 3 Each of the narratives included (a) a portion where participants where either telling or re-telling a story by watching a different video for each story, and (b) a portion where they needed to respond to a list of wh-questions related to each story.There were four stories of varying length in total.

Figure 2. Comprehension of wh-questions per age
Note.There was only one participant for the SAAE group for ages 12 and 13.
The slight advantage that MAE-speaking children have with the comprehension of wh-questions before age 6 is not observed with passive voice.The three groups present a comparable performance with the development of passive voice across age, with minor exceptions.(Figure 3).From age 5 to age 7 the three groups' performance is indistinguishable, but minor variations across all other ages are observed.

Figure 3. Comprehension of passives per age
An analysis of the various wh-question words examined by the current study shows that the three participant groups presented minor variations across the different types of wh-question words (Figure 4).Overall, the SAAE group presented a slight advantage across all question words, with the exception of where questions, where the SE group was better.The lowest performance across all groups was noted with the how and how/what questions, confirming previous findings on MAE, revealing a later acquisition of these wh-words, compared to where and what questions.SAAEspeaking children, however, were considerably better with how questions than the other two groups.
The next level of analysis pursued is the type of wh-questions based on the answer provided; we categorized wh-questions (or required answers) into: subject-object questions, Noun Phrase (NP) object questions, and prepositional phrase (PP) questions.Results reveal that all three Passive Voice groups present their highest performance with the seemingly more complex Subject-Object questions, and their lowest performance with NP object questions (Figure 5).The last question categorization we pursued was based on whether the wh-question included an embedded vs. non-embedded questions (Figure 6).The three groups presented a fairly parallel performance with both categorizations.Passive voice items fell under three separate categories: action passives, by-phrase passives and progressive passive.The lowest accuracy rates were reordered with progressive passives, and highest accuracy rates were noted with action passives.Results revealed a comparable performance across the three groups, with the exception of progressive passives, where MAE-speaking children's performance was considerably lower (Figure 7).The most common error for passive voice was using the picture depicting the same action in active voice.The boy found a cat with a broken leg that was lying on a table.He found a scarf and fixed the cat's leg with the scarf.(Pause.).

What did the boy fix the cat that was lying on the table with?
Answer: a ribbon.CA:3.4,SE-speaker #1 a broken leg #2 scarf (reprompt) CA:5.3,SE-speaker Part 7.1 Prime: This mom didn't know how to bake a cake.She saw a TV program about cooking, and she learned to make a lovely cake with pudding mix.(Pause.)

How did the mom learn what to bake?
Answer: A cooking program CA:9.0, AAE-speaker 4. Discussion.The main goal of the current study was to examine the acquisition of complex syntactic structures, particularly the comprehension wh-questions and passive voice in Southern English, Southern African American English and mainstream American English, three varieties of English which co-exist in most communities in the American South.An additional objective of this study was to produce developmental trajectories across different ages for these three varieties, in the two tested environments, in order to identify any potential deviations in performance across age.Even though there is a growing body of literature on the use of the copula and auxiliary BE, as well as the production of the third person singular -(e)s verbal agreement, research on complex structures in SAAE and SE is limited.The results presented by the current work show that the three groups do present parallel performance with the two structural environments, especially after the age of 5.It appears that none to their variation-specific characteristics have influenced the groups' performance with the comprehension of either wh-questions or passive voice, except for the minor advantage that SAAE appear to have with wh-questions.This is not surprising, as the diagnostic eliminated or took into consideration variety-specific differences.Identical performance from age 5 (or 6) onward for whquestions, seems to point towards effects from lack of early education for the younger children, who show a more variable performance before age 6.Note that most children in the children from the MAE group who performed much higher than the children in the SE and SAAE groups were attending a school following a structured, academic curriculum from as early as 18 months.

Conclusion.
The current work contributes to the limited research on the acquisition of complex syntactic structures by Southern English-, Southern African American English-and Mainstream American English-speaking children as young as 2-years-old and as old as 13-years-old, who are born and raised in Northern Mississippi.Parallel performance across the three groups in the variety of structural environments we tested the comprehension of wh-questions and passive voice clearly shows that the three groups are developing the two grammatical phenomena comparatively.
Results from the current study do not only add to the limited body of research on the development of the two complex syntactic structures across the three populations, but they could also help guide diagnosis and rehabilitation of language difficulties, as well as educational policies, diagnostic/assessment and rehabilitation protocols, which are critical for academic growth.This would, in turn, help prevent under-or over-diagnosis of the linguistic abilities of children.
We are currently in the process of analyzing additional data on the development of wh-questions across an imitation production task and the comprehension of wh-questions from four narratives.The analysis for these data is underway and is expected to be completed within the next 2-3 months.This will produce an additional 15,000 productions to the existing data.Additional grammatical phenomena that we are currently working on include amongst others the production and comprehension of determiners, quantifiers, consonant clusters, analysis of verbal and nominal features in free and controlled elicitation tasks, as well as data from SAAE-and SE-speaking individuals diagnosed with Down Syndrome.Upon completion of data analysis from the entire MSCA project we anticipate that we will be able to closely look at a large variety of specific linguistic phenomena and have a broader, but more informed view of the specific characteristics of each of the two Southern varieties of English and what their various stages of acquisition are.Dissemination of the current results to the greater public, not just the academic community, may also assist in social and educational inclusion and the efficient integration of children speaking the SE and SAAE varieties across primary, secondary and even tertiary education, where we do still see some signs of hypercritical, or even rather disapproving behaviours.
wh-question words: where, how what, who/what, how/what 1. Prime: This father and this baby were having lunch together.(Pause.)Who ate what?Answer: The dad ate the apple and the baby ate the banana.B. Subject-object, NP object, and PP object wh-questions I. Subject-object wh-questions 2. Prime: Here are children playing with toys.(Pause.)Who is playing with what?Answer: The girl is playing with the teddy bear and the boy is playing with the train.II.NP object wh-questions 3. Prime: This boy went fishing with a piece of string and a hook.Something pulled on the string, and he called out to his friend, "I caught a big fish."But look, it was only an old boot!(Pause.)What did the boy say he caught?Answer: A fish.III.PP object wh-questions 4. Prime: These brothers went to the circus.A clown came and tickled the little boy on the nose with a feather.He sneezed very hard, and the clown's wig blew right off!After the circus, the brothers went to buy some milk.The little boy drank his milk with a straw, but the big brother drank his straight from the carton.(Pause.)How did the boy who sneezed drink the milk?Answer: With a straw.C. Embedded and non-embedded wh-questions I. Non-embedded wh-questions 5. Prime: This boy eats different things in different ways.He eats ice cream with a fork and grapes with his fingers.(Pause.)How does the boy eat what?Answer: the boy ate the ice cream with a fork and the grapes with his fingers.II.Embedded wh-questions 6. Prime: This mom didn't know how to bake a cake.She saw a TV program about cooking, and she learned to make a lovely cake with pudding mix.(Pause.)How did the mom learn what to bake?Answer: by watching a TV program / from TV.Table 2. Examples of experimental material for wh-questions Each stimulus was accompanied by the equivalent visual stimulus.The administration of the entire diagnostic ranged between 35 and 60 minutes, depending on the age of the participant.Simple action passive sentences, 7. The elephant was pushed.B. Progressive passive sentences 8.The dog was being walked.
.0 vs. SAAE & SE = 70.0),but this lead is "neutralized" once the other two groups enter the school system.

Figure 4 .
Figure 4. Comprehension of wh-questions per wh-word

Figure 6 .
Figure 6.Comparison of embedded vs. non-embedded questions

Figure 7 .
Figure 7.Comparison of action vs. by-phrase vs. progressive passives The examples below illustrate the participants' responses to the targeted wh-questions.Examples include some instances where a participant's productions (a) matched the targeted or expected structure, and (b) did not match the target.Part 1.1 Prime: This father and this baby were having lunch together.(Pause.)Who ate what?Answer: The daddy ate the apples, and the baby ate the bananas.CA:6.3,AAE-speaker #1 They ate fruit.(reprompt) #2 The daddy ate apple and baby ate the banana CA:4.7,AAE-speaker

Table 1 .
.1This project received approval by the University of Mississippi IRB Committee (Project ID:18-035/MSCA-GF Linguistic Illusions) as well as the University of Cambridge School of Humanities and Social Sciences Ethics Committee.Participant information.