Children's books, brain development, and language acquisition

Ralf Thiede

• Publisher: Routledge
• Year: 2018
• Print ISBN:
  9780367583323
  9780815361893
• Edition: 1
• eBook ISBN:
  9781351113977
• Topic: Languages & Linguistics
• Subtopic: Psycolinguistics
• Index: Languages & Linguistics

This book correlates English-speaking children's brain development and acquisition of language with the linguistic input that comes from children's books. Drawing from the most current research on the developing brain, the author demonstrates how language acquisition is exclusively interactive, and highlights the benefit that accrues when that interaction includes the exploratory language play found in early childhood literature. Through discussions of specific domains of grammar, the relation of these domains to children's literature through scaffolding, and the resultant linguistic and cognitive advantages for the child, this volume offers an innovative approach to early brain maturation.

Part I
The Nurture of Nature

Contributions to Brain Maturation Beyond Adult Speech

1 Delightful Sounds

Phonetics to Phonology

Boom Boom

Splatt Splatt

Tick Tick Tock

Sizzle Sizzle

Blurp Blurp

Knock Knock Knock

A Slurp and a Whisper

and a Fish Kiss, too.

Mr. Brown can do it.

How about YOU?

Mr. Brown Can MOO! Can You?
Dr. Seuss

How could one possibly read a word like slurp to a baby and expect her to keep a straight face? S-l-u-r-r-p … it is a funny-sounding word. And then there is Dr. Seuss’ visceral groove. The rhythms of language play—a slúrp and a whísper and a físh kíss tóo—pát-a-cake, pát-a-cake, bá-kér’s-mán—áshes, áshes, and áll fáll down—become body memories, recalled in adult songs and poetry: when we fínd oursélves in the pláce júst ríght (‘Simple Gifts,’ Elder Joseph).

How we experience the earliest sounds and melodies and rhythms of human speech affects speech processing throughout life. Their perception draws on the left and the right hemispheres, on auditory cortex and premotor cortex all at once (Friederici 2011:1376, 1360). In speech planning, sounds and melodies of speech are primary (cf. Fromkin 1971:49). Rhythms and frequencies guide their acquisition.

As we saw in the introduction, well-meaning parents have sought to accelerate the intellectual development of their babies (p. 10). Baby Einstein, a multimillion-dollar industry (Aigner-Clark 2001), has promoted early learning of vocabulary, colors, and shapes over creative and playful engagement with sounds, making ‘Dr. Seuss’s Book of Wonderful Noises’ look unnecessary. The urge to turn progeny into prodigy marks the return to an adult construction of the child. This chapter will show, however, that books that go splat! also develop an infant’s brain.

With sound as the central component of speech processing, Baby Einstein needs an iconic linguistic wunderbaby companion—and there are two candidates, depending on what one assumes the relationship between sound and meaning to be. On the one hand, one could argue that infants discover meaning through sounds, noticing patterns and making semantic connections. Cognitive scientists are steering their research in this direction, and their scientific vessel’s galleon baby should be Baby Aristotle. On the other hand, one could argue that infants assign meanings to sounds, using grammar to make the connections. That would be a job for Baby Chomsky. Let us compare the positions.

Meaning From Sound: Aristotle

Aristotle viewed language as σημαντικός τις ψόφος, ‘meaning from sound’ (‘peri psyches’; cf. Long 2011:84). Thus, Baby Aristotle’s task is to make chaotic sounds meaningful: first by taking inventory and discovering patterns, then by associating those patterns with contexts. In cognitive science, that is the process referred to as ‘baby taking statistics.’ Infants note distributions and estimate transitional probabilities. In doing so, they exhibit a talent for patterns. Experiments that expose infants to artificial languages to see how they extract such patterns have shown, as Julie Sedivy put it, just ‘how efficient tiny children can be in sucking out information from even quite small amounts of exposure to a language’ (2014:225). Babies excel at acquiring the sounds and meanings of languages because of a generic skill: They can gather information and make predictions based on likelihoods. This is the quick and flexible talent for association that we will encounter again for the lexicon, in Chapter 2, in fast mapping (p. 62).

To conceive how an infant extracts information from minimal data, cognitive scientists apply Bayesian statistics, a procedure honed in theoretical chemistry and physics. That approach tests hypotheses from best available data. As better information becomes available, hypotheses and parameters are successively optimized to yield better models. In chemistry and physics, such approximations are necessary if direct measurements are out of reach —usually because the object studied is too small or too large or too complex to experiment on directly. Cognitive scientists face such a constraint: They cannot ‘look into the brain’ and understand from it what it is doing (cf. Chomsky 1988:6) and must hypothesize what the brain might be doing at a particular moment. Hence their attraction to Bayesian modeling.

Using Bayesian statistics to develop models of language acquisition also goes well with the premise that language learning is no different from any kind of learning. Cognitive scientists assume that language acquisition is based on cognition (learning from understanding). They do not start with the notion that humans are endowed with language-specific faculties. Learning is the same across all domains —it is ‘domain general.’ Humans are good at extracting patterns, at generalizing, at making predictions based upon those generalizations—and language is no different. No innate ‘domain-specific’ faculty is needed for language, because language is not separate from cognition. When infants find meaning from sound, they do so with a learning ability that humans have taken to an extraordinary level: ‘stochastic’ thinking, i.e. extracting patterns from unpredictable data. Babies figure out what the sounds of speech mean.

Meaning With Sound: Chomsky

Noam Chomsky, in contrast, sees language as ‘meaning with sound’ (2011). Grammar connects the two (Chomsky & Halle 1968:3). To be sure, Chomsky does not disavow statistical modeling: ‘The role of statistical reasoning and other cognitive processes in language acquisition, potentially at least, should be a significant area of research’ (2011). However, to claim that language can be discovered solely with domain-general strategies—taking statistics, extrapolating, pattern finding, associating—amounts, for Chomsky, to saying that there is no such thing as ‘language.’ But clearly there is. Language is an efficient computational system (ibid.). And that computational system has at least two inherent properties that Bayesian statistics cannot capture, let alone explain.

The first unpredicted property that makes language acquisition efficient is that babies do much more than extract information from minimal data: They are able to draw good conclusions even from insufficient data. Chomsky dubbed this phenomenon (how something could be acquired without being fully determined by evidence) ‘Plato’s Problem,’ because Plato was similarly puzzled by how humans learn the first thing if learning connects new things to known things (see p. 59).

The second unexplained property follows from the first: Children produce forms that never occurred in the data they encountered (‘I am brooming the floor’). They are creative, not just imitative. Their ability transcends ‘sucking out information.’ Some systemic knowledge is ‘triggered’ by input, so there must be some linguistic (domain-specific) endowment.

Acquiring the sounds of a language involves an endowment that lets children formulate and apply accurate rules, not just statistically optimal ones. Without postulating innate principles from which such accurate rules follow, we have no theory of why they reliably avoid errors that are statistically possible, nor why they proceed in well-defined stages of acquisition. Unlike cognitive scientists, linguists must ‘go beyond hand-waiving about analogy and training and pattern formation and so on’ (Chomsky 2011).

Chomsky argues that by modeling (associationist, connectionist, Bayesian, etc.), cognitive science has not produced, and will never produce, an explanatory theory of why babies make such sure-footed progress and avoid pitfalls. A simple example may serve to illustrate his claim—the planning involved in articulating the phoneme /l/. In English, the sound is different in the words lip and elk. Its location changes in anticipation of the next phoneme (/i/ is in the front, /k/ in the back), a so-called ‘coarticulation’ effect. In contrast, a German baby does not experiment with variants of /l/, as the /l/ in Lippe (‘lip’) and welk (‘wilted’) is frontal each time. Coarticulation thus is already rule governed (cf. Whalen 1990) —and so is, as we shall see below, the grouping of different speech sounds into a single abstract phoneme such as /l/.

Exclusive reliance on statistics in the research of language acquisition is essentially a way of avoiding theories that can be verified or falsified. Such an approach may perpetuate grant money, but the ‘lack of interest in theory and explanation’ (Chomsky 1965:193) draws the same condemnation Noam Chomsky leveled at behaviorist B. F. Skinner (1957) in 1959: ‘play acting at science’ (1959:39).

A Dual-Track Approach to the Acquisition of Sounds

The contrary approaches of Aristotle and Chomsky divide linguists and cognitive scientists, seemingly preventing Chomsky’s envisioned integration of linguistics as a subdiscipline of ‘cognitive psychology’ (1984:1). Here is the litmus test: Are babies (/are brains) simply good at modeling, i.e. bottom-up processing from sound to meaning? Or are they genetically endowed with faculties to abstract and make rules that follow from inborn principles of language, i.e. top-down processing to connect meaning to sound? My answer is ‘yes.’ Both. An Aristotle and a Chomsky in every infant.

The brain has it both ways, taking statistics on patterns and imposing productive, creative rules. Infants mature into the sound systems of their native languages in stages: patterns followed by rules. Their perception proceeds from modeling all sounds they hear based on how they are physically produced (phonetics) to systematizing them with feature-based rules (phonology). Their production of sounds follows anatomic growth: lowering of the larynx, growing teeth, developing increasingly fine-motor control. In addition, the emergence of spoken sounds reflects their frequency in the speech of adults. Substitutions along the way, however, also show rules at work. For example, one would have to analyze phonetic features to decide that sherry is a reasonable substitute for cherry. The brain uses patterns and rules, Aristotle’s bottom-up approach and Chomsky’s top-down approach. Stochastic processing characterizes the initial state, and rule formation comes on line as additional, language-specific pathways mature.

The sounds of language make sense only because sense lives in an enriched environment of meaningful situational context. Without that added complexity, no association of sound and meaning ‘makes sense.’ Failure to interact with babies causes their failure to thrive linguistically if not physically; studies of the results of isolation in Romanian orphanages, such as Windsor et al. (2007), have documented the cascading damage resulting from linguistic deprivation: developmental mispronunciations, small vocabulary, minimal syntax. Only if they were placed in foster care did these children’s language development improve, though not to the level of that of peers who grew up in an intact family (1377–78).

Children’s books that are read aloud and interactively...

Table of contents

• Cover
• Half Title
• Series
• Title
• Copyright
• Dedication
• Contents
• List of Figures
• List of Tables
• Acknowledgments
• Permissions
• A Book With Too Many Variables: Some Introduction …
• Part I The Nurture of Nature: Contributions to Brain Maturation Beyond Adult Speech 21
• Part II The Nature of Nurture: Concerted Cultivation Toward Adult Speech 123
• Children’s Books Cited (by Author)
• Index