If you follow education matters you know that the home environment in very early years are vital. One aspect of that home environment is the language infants and toddlers hear at home.

The groundbreaking work of Hart & Risley (1995; replicated by others, e.g. Huttenlocher et al, 2010) showed that socio-economic status of the parents is correlated with vast differences in the amount and complexity of language that children hear at home. 

But what aspect of this speech is important? Does speech need to be directed to children? Perhaps all that’s needed is for children to be in the presence of this more complex language. After all, we know that children do not learn language via instruction; they learn it by observation.

Three studies published in the last couple of years build a convincing case that parents should, indeed, talk to their children. Talking in the presence of their children (but to others) does not confer the same vocabulary benefit.

In the most recent study (Weisleder & Fernald, 2013), experimenters tested 29 Spanish-learning infants at age 19 months. The children wore a small device that made an audio recording of all speech to which the child was exposed. The audio recordings were analyzed by software meant to differentiate speech directed toward the child versus speech audible to the child, but directed to others. A subset of recordings was coded by human observers to ensure the accuracy of the software.

Recordings of a full day’s speech were analyzed and the results showed a huge range in child-directed speech; caregivers in one family spoke over 12,000 words to the child whereas in another family that figure was just 670 words. The amount of child-directed speech as not significantly correlated (r = .17) with the amount of overheard speech.

At 24 months the productive vocabulary of the children was measured by asking the parents to judge words that they believed their child understood and words that their child used.

Of greatest interest, the amount of child-directed speech at 19 months was correlated (r = .57) with vocabulary at 24 months. The amount of overheard speech at 19 months was not (r = .25).

The sample size in this study is limited and there were some quirky features. (E.g., the software sorting “child-directed” vs. overherd speech is good, but not perfect.) But my confidence in the conclusion is bolstered by reports of the same finding from another lab, investigating speakers of other languages: English (Schneidman et al, 2013) and Yucatec (Schneidman & Goldin-Meadow, 2012).

Why must speech be directed to the child?

Weisleder & Fernald administered another task at 19 months meant to measure word processing efficiency. They speculated that the effect of child-directed speech on vocabulary was mediated through efficiency—something like, for example, the speed and accuracy with which the particular phonemes of the child’s language are processed.

This doesn’t fully explain the difference between child-directed and overheard speech. The obvious hypothesis is that other cues (e.g. eye gaze direction) prompt greater attention to speech that is child-directed, and that attention is necessary to build efficiency.

More details will have to await further research. For now, we can say with greater confidence “talk to your children” not just “talk in the presence of your children.”

References

Hart, B. M., & Risley, T. R. (1995). Meaningful differences in the everyday experience of young American children. Baltimore, MD: Brookes.

Huttenlocher, J., Waterfall, H., Vasilyeva, M., Vevea, J., & Hedges, L. V. (2010). Sources of variability in children’s language growth. Cognitive Psychology, 61, 343–365.

Shneidman, L. A., Arroyo, M. E., Levine, S., & Goldin-Meadow, S. (2013). What counts as effective input for word learning? Journal of Child Language, 40, 672–686.

Shneidman, L. A., & Goldin-Meadow, S. (2012). Language input and acquisition in a Mayan village: How important is directed speech? Developmental Science, 15, 659–673.

Weisleder, A. & Fernald, A. (2013). Talking to children matters: Early language experience strengthens processing and builds vocabulary. Psychological Science, DOI: 10.1177/0956797613488145

This is a negative finding, so I'll keep it brief.

How do kids acquire new vocabulary? This process is poorly understood.

An influential theory has been that the phonological loop in working memory provides essential support. The phonological loop is like a little tape loop lasting perhaps two seconds; it allows you to keep active a sound you hear.

The idea is that a new unfamiliar word can be placed on the loop for practice and to keep it around while the surrounding context helps you figure out the meaning.

If so, you'd predict that the larger the capacity of the phonological loop and the greater the fidelity with which it "records" the better children will be able to learn new vocabulary.

The efficacy of the phonological loop is measured by having kids repeat nonsense words. Initially they are short--tozzy--but they increase in length to pose greater challenge to the phonological loop--liddynappish.

Several studies have shown correlations between phonological loop capacity and vocabulary size in children (for a review, see Melby-Lervag & Lervag, 2012).

The problem: it could be that having a big vocabulary makes the phonological loop test easier, because it makes it more likely that some of the nonsense words remind you of a word you already know. (And so you have the semantics of that word helping you remember the to-be-remembered word.) Indeed, even proponents of the hypothesis argue that's what happens when kids get older.

What you really need is a study that measures phonological loop capacity at time 1, and finds that it predicts vocabulary size at time 2. There is one such study (Gathercole et al, 1992) but it used a statistical analysis (cross-lagged correlation) that is now considered less than ideal.

A new study (Melby-Lervag et al, 2012) used probably the best methodology of any used to date. It was a longitudinal study that tested nonword repetition ability and vocabulary once each year between the ages of 3 and 7.

They used a different statistical technique--simplex models--to assess causal relationships. They found that both nonword repetition and vocabulary show growth, both show stability across children, and both are moderately correlated, but there was no evidence that one influenced the growth of the other over time.

The group then reanalyzed the Gathercole et al (1992) data and found the same pattern.

This is one depressing paper. Something we thought we knew--the phonological loop contributes to vocabulary learning--may well be wrong.

If anyone is working on a remediation program for young children that centers on improving the working of the phonological loop, it's probably time to rethink that idea.




Gathercole, S. E., Willis, C., Emslie, H., & Baddeley, A. (1992). Phonological memory and vocabulary development during the early school years: A longitudinal study. Developmental Psychology, 28, 887–898.

Melby-Lervåg, M., & Lervåg, A. (2012). Oral language skills mod-erate nonword repetition skills in children with dyslexia: A meta-analysis of the role of nonword repetition skills in dyslexia. Scientific Studies of Reading, 16, 1–34.

Melby-Lervåg, M., & Lervåg, A., Lyster, S-A H., Klem, M., Hagtvet, B., & Hulme, C. (in press). Nonword-repetition ability does not appear to be a causal influence on children's vocabulary development. Psychological Science.