Articulating What Infants Attune to in Native Speech

Abstract

To become language users, infants must embrace the integrality of speech perception and production. That they do so, and quite rapidly, is implied by the native-language attunement they achieve in each domain by 6-12 months. Yet research has most often addressed one or the other domain, rarely how they interrelate. Moreover, mainstream assumptions that perception relies on acoustic patterns whereas production involves motor patterns entail that the infant would have to translate incommensurable information to grasp the perception-production relationship. We posit the more parsimonious view that both domains depend on commensurate articulatory information. Our proposed framework combines principles of the Perceptual Assimilation Model (PAM) and Articulatory Phonology (AP). According to PAM, infants attune to articulatory information in native speech and detect similarities of nonnative phones to native articulatory patterns. The AP premise that gestures of the speech organs are the basic elements of phonology offers articulatory similarity metrics while satisfying the requirement that phonological information be discrete and contrastive: (a) distinct articulatory organs produce vocal tract constrictions and (b) phonological contrasts recruit different articulators and/or constrictions of a given articulator that differ in degree or location. Various lines of research suggest young children perceive articulatory information, which guides their productions: discrimination of between- versus within-organ contrasts, simulations of attunement to language-specific articulatory distributions, multimodal speech perception, oral/vocal imitation, and perceptual effects of articulator activation or suppression. We conclude that articulatory gesture information serves as the foundation for developmental integrality of speech perception and production.

Figure 1.

Schematic diagram of the Perceptual Assimilation Model (PAM; Best, ; Best & Tyler, 2007), illustrating an adult's native language phonological space, in which the conical “islands” represent native consonant categories that have been delineated and sharpened by experience with perceiving and producing native speech, and the major predicted patterns of perceptual assimilation of nonnative consonant contrasts to the native phonological system. Pairs of black circles represent nonnative consonant contrasts, with the various predicted contrast assimilation patterns indicated by arrows and labels.

Figure 2.

Schematics of the three dimensions of articulatory gestures as proposed for the revised Perceptual Assimilation Model with Articulatory Organ Hypothesis (PAM-AOH): (A) articulatory geometry (modeled after Browman & Goldstein, 1989, 1992); (B) articulatory organ hierarchy (active articulators and their nested nodes), which is an unfolded, straightened version of A; (C) articulatory actions (constriction degrees) represented along a straightened side view of the vocal tract's ventral (lower surface: active articulators) and dorsal (upper surface: passive articulators/locations) surfaces.

Figure 3.

(A) Frequency distribution of tongue tip (TT) constriction location along hard palate (normalized to 0–1), as measured by electromagnetic articulometry (EMA), for all coronal stops in a natural spoken English passage of ∼1,000 words produced by an adult female native speaker (L. M. Goldstein et al., 2008); (B) corresponding distribution for all coronal stops in a natural spoken Hindi passage of ∼6,000 words produced by an adult female native speaker (L. M. Goldstein et al., 2008).

Figure 4.

Simulation of “blank slate” infant (lower left) attunement to English “parent” input (top center) based on articulatory data from a native English speaker, which shows a unimodal frequency distribution of alveolar Tongue Tip (TT) constriction locations along the hard palate (normalized 0–1). The time series (lower portion of diagram) indicates successive 2,500-iteration steps in the 10,000-iteration simulation.

Figure 5.

Simulation of “blank slate” infant (lower left) attunement to Hindi “parent” input (top center) based on articulatory data from a native Hindi speaker, which shows a bimodal frequency distribution of dental versus retroflex Tongue Tip (TT) constriction locations along the hard palate (normalized 0–1). The time series (lower portion of diagram) indicates successive 2,500-iteration steps in the 10,000-iteration simulation.

Figure 6.

Simulation of an adult second language (L2) learner of Hindi showing native-language (L1) Single Category (SC) assimilation (lower left) and L2 attunement (time series) to input from an idealized Hindi “teacher” (top center). The L2 learner is an idealized speaker of English with a well-established unimodal distribution of English coronal stops centered at alveolar position, which does not line up with either Hindi mode. The time series shows the same simulation steps as in Figure 5.

Figure 7.

Simulation of a different adult L2 Hindi learner showing initial L1 Category Goodness difference (CG) assimilation (lower left) and L2 attunement (time series) to input from the same idealized Hindi “teacher” (top center) and simulation steps as in Figure 6. The second language (L2) learner is an idealized native language (L1) speaker of Spanish with a well-established unimodal distribution of Spanish coronal stops centered at dental position, which does line up with one of the two Hindi modes (dental).