Acquisition and developmental descriptions of consonants and vowels

This package provides dataframes of information about the consonants and vowels in American English. The following datasets collect acquisition (acq) features which (try to) characterize the expected acquisition or speech-motor difficulty of speech sounds. See also data_features_consonants.

Usage

data_acq_consonants

Format

An object of class tbl_df (inherits from tbl, data.frame) with 24 rows and 10 columns.

Details

Crowe and McLeod (2020) norms for English consonant acquisition

Crowe and McLeod (2020, below as the cm2020_ variables) provides a systematic review and summary statistics for age of acquisition norms for English consonants. They scoured the literature of acquisition ages for individual consonants and computed summary statistics on them. They considered just accuracy of sounds when produced in single words. Their sources include a mix of a journal articles and norms for articulation assessments. They do not weight statistics from individual studies by sample size or sampling procedure.

I prepared the Crowe and McLeod (2020) data by copying the relevant numbers from their Table 2 making the following changes: 1) rounding mean and SD values to 1 decimal point (3 days for ages in months), 2) dropping /ʍ/, 3) using /r/, /g/, /tʃ/, /dʒ/ for IPA characters instead of the specialized characters used in the article.

The early 8, middle 8 and late 8 (Shriberg, 1993)

The English consonants are often broken down into three developmental classes, based on Shriberg (1993):

Early 8: m b j n w d p h
Middle 8: t ŋ k g f v tʃ dʒ,
Late 8: ʃ θ s z ð l r ʒ

This classification is included as the s93_eights column.

From these names alone, we might interpret these classes such that sounds in the Early 8 would be acquired before the ones in the Middle 8, and likewise that the Middle 8 would be acquired before the Late 8. But these classes were not created by examining patterns of typical consonant acquisition.

For some context, Shriberg (1993) introduces the Early 8, Middle 8, and Late 8 data by describing the following panel of the article's Figure 7:

About which, Shriberg (1993) says: "The values for this trend, which is a profile of consonant mastery, were taken from a group of 64 3- to 6-year-old speech-delayed children Shriberg, Kwiatkowski, & Gruber, 1992). Severity of involvement of the 24 English consonants is represented as the percentage correct for each consonant sorted in decreasing order from left to right. Notice that the most obvious breaks in this function allow for a division of the 24 consonants into three groups of eight sounds termed the Early-8, averaging over 75% correct, the Middle-8, averaging 25%-75% correct, and the Late-8, including consonants averaging less than 25% correct in continuous conversational speech (/ʒ/ is infrequently represented in young, speech-delayed children's spontaneous conversational speech)."

So, there were 64 3–6-year-old children with speech delays, and consonant sounds were divided into three classes based on how often these children produced the sounds correctly on average in a conversational speech sample. This classification is not so much a measure of the relative ordering of speech sound development as it is the relative difficulty of these sounds for children with a speech delay of unknown origin. It would be more appropriate to replace the levels of Early/Middle/Late with Easy/Medium/Hard.

Consonant acquisition features

data_acq_consonants provides the following features:

knitr::kable(data_acq_consonants)

phone	cmubet	wiscbet	cm2020_90_age_mean	cm2020_90_age_sd	cm2020_90_age_min	cm2020_90_age_max	cm2020_90_num_studies	cm2020_90_stage	s93_eights
p	P	p	33.2	6.9	24	48	12	early	early
b	B	b	31.4	7.8	24	48	13	early	early
t	T	t	38.5	9.2	24	60	13	early	middle
d	D	d	35.7	6.7	24	48	13	early	early
k	K	k	37.7	7.3	24	48	13	early	middle
g	G	g	36.8	6.6	24	48	13	early	middle
tʃ	CH	tsh	53.5	10.7	36	72	12	middle	middle
dʒ	JH	dzh	51.0	11.8	36	72	13	middle	middle
m	M	m	33.2	6.7	24	48	13	early	early
n	N	n	33.1	7.4	24	48	13	early	early
ŋ	NG	ng	40.3	10.8	24	55	10	early	middle
f	F	f	38.3	6.3	24	48	13	early	middle
v	V	v	50.8	10.8	36	66	12	middle	middle
θ	TH	th	77.0	7.4	72	96	10	late	late
ð	DH	dh	69.0	11.3	54	96	12	late	late
s	S	s	51.3	16.3	24	84	12	middle	late
z	Z	z	56.8	14.3	30	84	11	middle	late
ʃ	SH	sh	55.0	10.5	36	72	12	middle	late
ʒ	ZH	zh	70.7	12.2	60	84	3	late	late
h	HH	h	35.0	7.0	24	48	13	early	early
l	L	l	53.8	10.4	24	60	12	middle	late
r	R	r	66.6	18.6	30	96	12	late	late
w	W	w	35.2	6.8	24	48	13	early	early
j	Y	j	45.8	11.0	30	60	13	early	early

Description of each column:

phone: phone in IPA
cmubet: phone in the CMU alphabet
wiscbet: phone in an older system used by our lab
cm2020_90_age_mean, cm2020_90_age_sd, cm2020_90_age_min, cm2020_90_age_max: Age of acquisition statistics reported by Crowe & McLeod (2020). Statistics are the mean, SD, min and max age (in months) when children reached 90% accuracy on a consonant.
cm2020_90_num_studies: Number of studies used by Crowe & McLeod (2020) to compute the corresponding statistics.
cm2020_90_stage: Developmental stage assigned to the consonant by Crowe & McLeod (2020). Sounds with an age_mean before 48 months are early, before 60 months are middle, and of 60 or older are late.
s93_eights: Developmental stage of Shriberg (1993)—that is, the early 8, middle 8 and late 8 consonants.

References

Crowe, K., & McLeod, S. (2020). Children’s English Consonant Acquisition in the United States: A Review. American Journal of Speech-Language Pathology, 29(4), 2155–2169. https://doi.org/10.1044/2020_AJSLP-19-00168

Shriberg, L. D. (1993). Four New Speech and Prosody-Voice Measures for Genetics Research and Other Studies in Developmental Phonological Disorders. Journal of Speech, Language, and Hearing Research, 36(1), 105–140. https://doi.org/10.1044/jshr.3601.105