This article evaluates the ''Frames, then Content'' hypothesis for spe
ech acquisition, which states that much of the patterning of babbling
is a direct result of production of syllabic ''Frames'' by means of rh
ythmic mandibular oscillation, with relatively little of the intrasyll
abic and intersyllabic ''Content'' of the syllable-like cycles under m
andible-independent control. Analysis was based on a phonetically tran
scribed corpus of 6,659 utterances of 6 normally developing infants ob
tained from one-hour weekly audio-recordings over a 4-6 month period.
Intrasyllabic predictions were that front vowels would preferentially
co-occur with front (alveolar) consonants, back vowels with back (vela
r) consonants, and central vowels with labial consonants, with the lat
ter effect presumably resulting from mandibular oscillation alone. Int
ersyllabic predictions were for more variegation in tongue height for
vowels than in front-back tongue movement, and for consonant manner ch
anges to predominate over place changes (related primarily to mandibul
ar oscillation). All 30 individual predictions from both hypotheses we
re confirmed, leading to a conception of the articulatory basis of bab
bling as ''Frame Dominance.''