The Early

Catastrophe

The 30 Million Word Gap by Age 3

By Betty Hart and Todd R. Risley

uring the 1960's War on Poverty, we were among added to the dictionaries of all the children and an abrupt

D the many researchers, psychologists, and educators

who brought our knowledge of child development
to the front line in an optimistic effort to intervene early to
acceleration in their cumulative vocabulary growth curves.
But just as in other early intervention programs, the in
creases were temporary.
forestall the terrible effects that poverty was having on some We found we could easily increase the size of the chil
children's academic growth. We were also among the many dren's vocabularies by teaching them new words. But we
who saw that our results, however promising at the start, could not accelerate the rate of vocabulary growth so that it
washed out fairly early and fairly completely as children would continue beyond direcr teaching; we could not
aged. change the developmental trajectory. However many new
In one planned intervention in Kansas City, Kans., we words we taught the children in the preschool, it was clear
used our experience with clinical language in tervention to that a year later, when the children were in kindergarten, the
design a half-day program for the Turner House Preschool, effects of the boost in vocabulary resources would have
located in the impoverished Juniper Gardens area of the city. washed our. The children's developmental trajectories of vo
Most interventions of the time used a variety of methods cabulary growth would continue to point to vocabulary sizes
and then measured results with IQ tests, but ours focused in the future that were increasingly discrepant from those of
on building the everyday language the children were using, the professors' children. We saw increasing disparity between
then evaluating the growth of that language. In addition, the extremes-the fast vocabulary growth of the professors'
our study included not juSt poor children from Turner children and the slow vocabulary growth of the Turner
House, but also a group of University of Kansas professors' House children. The gap seemed to foreshadow the findings
children against whom we could measure the Turner House from other studies that in high school many children from
children's progress. families in poverty lack the vocabulary used in advanced
All the children in the program eagerly engaged with the textbooks.
wide variety of new materials and language-intensive activi Rather than concede to the unmalleable forces of hered
ties introduced in the preschool. The spontaneous speech ity, we decided that we would undertake research that would
data we collected showed a spurr of new vocabulary words allow us to understand the disparate developmental trajecto
Betty Hart is professor ofHuman Development at the Univer ries we saw. We realized that if we were to understand how
and when differences in developmental trajectories began,
sity ofKansas and senior scientist at the Schiefelbusch Institute
for Life Span Studies. Todd R. Risley is professor in the Depart we needed to see what was happening to children at home at
ment ofPsychology at the University ofAlaska Anchorage and the very beginning of their vocabulary growth.
director ofAlaska's Autism Intensive Early Intervention Project.
The two have collaborated on research projects for more than

W
e undertook 2 1/2 years of observing 42 families
35 years. This article is excerpted with permission from Mean for an hour each month to learn aboU[ what typi
ingful Differences in the Everyday Experiences of Young cally went on in homes with 1- and 2-year-old
American Children, 1995, Brookes; www.brookespublish children learning to talk. The data showed us that ordinary
ing.com; 1-800-638-3775; $29.00. families differ immensely in the amount of experience with
4AMERICAN EDUCATOR SPRING 2003
language and interaction they regularly provide their chil
dren and that differences in children's experience are Eighty-six percent to 98 percent
strongly linked to children's language accomplishments at
age 3. Our goal in the longitudinal study was to discover of the words recorded in each
what was happening in children's early experience that could
account for the intractable difference in rates of vocabulary child's vocabulary consisted
growth we saw among 4-year-olds.
of words also recorded
Methodology in their parents' vocabularies.
Our ambition was to record "everything" that went on in
children's homes-everything that was done by the children,
to them, and atound them. Because we were committed to
undertaking the labor involved in observing, tape recording,
and transcribing, and because we did not know exactly
which aspects of children's cumulative experience were con
tributing to establishing rates of vocabulary growth, the
more information we could get each time we were in the
home the more we could potentially learn.
We decided to start when the children were 7-9 months
old so we would have time for the families to adapt to obser
vation before the children actually began talking. We fol
lowed the children until they turned three years old.
The first families we recruited to participate in the study
came from personal contacts: friends who had babies and
families who had had children in the Turner House
Preschool. We then used birth announcements to send de
scriptions of the study to families with children of the de
sired age. In recruiting from birth announcements, we had
[wo priorities. The first priority was to obtain a range in de

mographics, and the second was stability-we needed fami

lies likely to remain in the longitudinal study for several
years. Recruiting from birth announcements allowed us to
preselect families . We looked up each potential family in the
city directory and listed those with such signs of permanence
as owning their home and having a telephone. We listed
families by sex of child and address because demographic
status could be reliably associated with area of residence in
this city at that time. Then we sent recruiting letters selec
tively in order to maintain the gender balance and the repre
sentation of socioeconomic strata.
Our final sample consisted of 42 families who remained
in the study from beginning ro end. From each of these fam
ilies, we have almost 2 1/ 2 years or more of sequential
monthly hour-long observations. On the basis of occupa
tion, 13 of the families were upper socioeconomic status
(5E5), 10 were middle 5E5 , 13 were lower 5E5, and six were
on welfare. There were African-American families in each
5E5 category, in numbers roughly reflecting local job alloca
tions. One African-American family was upper 5E5, three
were middle, seven were lower, and six families were on wel
fare. Of the 42 children, 17 were African American and 23
were girls. Eleven children were the first born to the family,
18 were second children, and 13 were third or later-born
children.

What We Found
Before children can take charge of their own experience and
begin to spend time with peers in social groups outside the
home, almost everything they learn comes from their fami
6AMERICAN EDUCATOR SPRING 2003
lies, to whom society has assigned the task of soc ializing coded to check the reliability of the coding, after each file
children. We were not surprised to see the 42 children turn had been checked one more time and the accuracy of each
out to be like their parents; we had no t full y realized, how aspect verified, and after the data analysis programs had fi
ever, the implications of those similari ties for the children's nally been run to produce frequency counts and dictionary
futures. lists for each observation, we had an immense numeric
We observed the 42 children grow more like their par database that required 23 million bytes of computer file
ents in stature and activity levels, in vocabul ary resources, space. We were flllally read y to begin asking what it all
and in lan guage and interaction styles. Despite the consid meant.
erable range in vocabulary size among the children, 86 per It took six years of painstaking effort before we saw the
cent to 98 percent of the words recorded in each child's vo first results of the longitudinal research. And then we were
cabulary consisted of words also recorded in their parents' astonished at the differences the data revealed (see the graph
vocabularies. By the age of 34-36 months , the children below).
were also talking and using numbers of differen t words
very similar to the averages of their parents (see the table
below). Children'S Vocabulary Differs Greatly
By the time the children were 3 years old, trends in Across Income Groups
amount of talk, vocabulary growth, and style of interaction 1200 13 higher
SES children
were well established and clearly suggested widening gaps to " (profeSSional)

come. Even patterns of parenting were already observable 1000

among the children . When we listened to the children, we '"
'E
0
seemed to hear their parents speaking; when we watched the 3
~
23 middle/lower
800 SES children
children play at parenting their dolls, we seemed to see the '"
'5
.0 (workingclass)
futures of their own children. ~
0
>
Ql
> 600
Families' Language and Use
~
'5 6 children from
Differ Across Income Groups
~ families on welfare
0
400 ,
Families ,
12 Professional 23 Working.class 6 Welfare
Measures and scores Paren! Child Pateur Child Paten! Child 200

Pretest score' 41 31 14
Recorded vocabulary
Size 2,176 1,116 1,498 749 974 525
Average utterances ~ 10 12 14 16 18 20 22 24 26 28 30 32 34 36
per hour" 487 310 301 223 176 168 ~ Age of child in monlhs
Average d ifFerem ~
382 297 25 1 216 167 149 "'
words per hour
~

L
'When we began the longitudinal study, we asked rhe parents to complete a vocab u ike the children in the Turner House Preschool, the
lary ptetest. At the first observa tion each paren t was asked to complete a fotm abo
three year old children from families on welfare not
stracted from the Peabody Picture Vocabu lary Test (PPVT ). We gave each parent a
list of 46 vocabulary words and a seties of pictures (fou r options per vocabulary only had smaller vocabularies than did children of the
word) and asked the pa ten t to write beside each word the number of the picture same age in professional families, but they were also add ing
rhar corresponded ro th e wrirren wo rd . Parenr performance on [he resr was highly
words more slowly. Projecting the developmental trajectory
correlated with years of ed ucation (r = .5 7).
'Parent u[(erances and different words were averaged over 13-36 months of child of the welfare children's vocabulary growth curves, we could
age. Child utterances and different wotds we re ave raged for the four observarions . see an ever-widening gap similar to the on e we saw berween
when the child ren we re 3336 month s old .
the Turner House children and the professors' children in
1967.
We now had answers to our 20-year-old questions . We While we were immersed in collecting and processing
had observed, recorded , and analyzed more than 1,300 the data, our thoughts were concerned only with the next
hours of casual interactions berween parents and their lan utterance to be transcribed or coded. While we were ob
guage-learning children. We had dissembled these interac serving in the homes, though we were aware that the fami
tions into several dozen molecular features that could be reli lies were very different in lifestyles, they were all similarly
ably coded and counted. We had examined the correlations engaged in the fundamental task of raising a child. All the
berween the quantities of each of those features and several families nurtured their children and played and talked with
outcome measures relating to children's languageaccom them . They all disciplined their children and taught them
plishments. . good manners and how to dress and toilet themselves.
After all 1,318 observations had been entered into the They provided their children with much the same toys and
computer and checked for accuracy against the raw data, talked to them about much the same things. Though dif
after every word had been checked for speJling and coded ferent in personality and skill level s, the children all
and checked for its part of speech, after every utterance had learned to talk and to be socially appropriate members of
been coded for syntax and discourse function and every code the family with all the basic skills needed for preschool
checked for accuracy, after random samples had been re- entry.

SPRING 2003 AMERICAN FEDERATION OF TEACHERS 7

Test Performance in Third Grade Follows amount of experience were so durable over the more than
two years of observations that they provide the best basis we
Accomplishments at Age 3
currently have for estimating children's actual life experience.
We wondered whether the differences we saw at age 3 would
A linear extrapolation from the averages in the observa
be washed out, like the effects of a preschool intervention, as
tional data to a 100-hour week (given a 14-hour waking
the children's experience broadened to a wider community
day) shows the average child in the professional families
of competent speakers. Like the parents we observed, we
with 215,000 words of language experience, the average
wondered how much difference children's early experiences
child in a working-class family provided with 125,000
would actually make. Could we, or parents, predict how a
words, and the average child in a welfare family with 62,000
child would do in school from what the parent was doing
words of language experience. In a 5,200-hour year, the
when the child was 2 years old?
amount would be 11 .2 million words for a child in a profes
Fortune provided us with Dale Walker, who recruited 29
sional family, 6.5 million words for a child in a working
of the 42 families to participate in a study of their children's
class family, and 3.2 million words for a child in a welfare
school performance in the third grade, when the children
family. In four years of such experience, an average child in a
were nine to 10 years old.
professional family would have accumulated experience with
We were awestruck at how well our measures of accom almost 45 million words, an average child in a working-class
plishments at age 3 predicted measures of language skill at family would have accumulated experience with 26 million
age 9-10. From our preschool data we had been confident words, and an average child in a welfare family would have
that the rate of vocabulary growth would predict later per accumulated experience with 13 million words. By age 4,
formance in school; we saw that it did . For the 29 children the average child in a welfare family might have 13 million
observed when they were 1-2 years old, the rate of vocabu fewer words of cumulative experience than the average child
lary growth at age 3 was strongly associated with scores at in a working-class family. This linear extrapolation is shown
age 9-10 on both the Peabody Picture Vocabulary Test-Re
in the graph below.
vised (PPVT-R) of receptive vocabulary (r = .58) and the
Test of Language Development-2: Intermediate (TOLD)
The Number of Words Addressed to Children

(r = .74) and its subtests (listening, speaking, semantics,

Differs Across Income Groups

syntax).
Vocabulary use at age 3 was equally predictive of measures 50 million Professional

of language skill at age 9-10. Vocabulary use at age 3 was

strongly associated with scores on both the PPVT-R ~ 40 million
(r = .57) and the TOLD (r = .72). Vocabulary use at age 3 i'!
"0
"0 Workingclass
was also strongly associated with reading comprehension '"
<n

scores on the Comprehensive Test of Basic Skills (CTBS/U) 1:'

o
~
(r= .56). .~
;;;
:; Welfare
E
::J
The 30 Million Word Gap By Age 3 "
"0
OJ 10millio
;;;
All parent-child research is based on the assumption that the
data (laboratory or field) reflect what people typically do. In ~
w
most studies, there are as many reasons that the averages o 12 24 36 48
would be higher than reponed as there are that they would Age of child in months
be lower. But all researchers caution against extrapolating
their findings to people and circumstances they did not in But the children's language experience did not differ just
clude. Our data provide us, however, a first approximation in terms of the number and quality of words heard. We can
to the absolute magnitude of children's early experience, a extrapolate similarly the relative differences the data showed
basis sufficient for estimating the actual size of the interven in children's hourly experience with parent affirmatives (en
tion task needed to provide equal experience and, thus, couraging words) and prohibitions. The average child in a
equal opportunities to children living in poverty. We depend professional family was accumulating 32 affirmatives and
on future studies to refIne this estimate. five prohibitions per hour, a ratio of 6 encouragements to 1
Because the goal of an intervention would be to equalize discouragement. The average child in a working-class fam
children's early experience, we need to estimate the amount ily was accumulating 12 affirmatives and seven prohibitions
of experience childten of different SES groups might bring per hour, a ratio of 2 encouragements to 1 discouragement.
to an intervention that began in preschool at age 4. We base The average child in a welfare family, though, was accumu
our estimate on the remarkable differences our data showed lating five affirmatives and 11 prohibitions per hour, a ratio
in the relative amounts of children's early experience: Simply of 1 encouragement to 2 discouragements. In a 5,200-hour
in words heard, the average child on welfare was having half year, that would be 166,000 encouragements to 26,000 dis
as much experience per hour (616 words per hour) as the av couragements in a professional family, 62,000 encourage
erage working-class child (1,251 words per hour) and less ments to 36,000 discouragements in a working-class family,
than one-third that of the average child in a professional and 26,000 encouragements to 57,000 discouragements in
family (2,153 words per hour). These relative differences in a welfare family.
8AMERICAN EDUCATOR SPRING 2003
 Extrapolated [Q the first four years of life, the average
In four years, an average child in a child in a professional family would have accumulated
560,000 more instances of encouraging feedback than dis
professional family would couraging feedback, and an average child in a working-class
family would have accumulated 100,000 more encourage
accumulate experience with almost menrs than discouragemenrs. But an average child in a wel
fare family would have accumulated 125,000 more instances
45 million words, an average child in of prohibitions than encouragemenrs. By the age of 4, the
average child in a welfare family might have had 144,000
a working-class family 26 million fewer encouragemenrs and 84,000 more discouragemenrs of
words, and an average child in a his or her behavior than the average child in a working-class
family.
welfare family 13 million words, Extrapolating the relative differences in children's hourly
experience allows us [Q estimate children's cumulative experi
ence in the first four years of life and so glimpse the size of
the problem facing inrervenrion. Whatever the inaccuracy of
our estimates, it is not by an order of magnitude such that
60,000 words becomes 6,000 or 600,000. Even if our esti
mates of children's experience are [00 high by half, the dif
ferences between children by age 4 in amounrs of cumula
tive experience are so great that even the best of intervention
programs could only hope [0 keep the children in families
on welfare from falling still further behind the children In
the working-class families.

The Importance of Early Years Experience

We learned from the longitudinal data that the problem of
skill differences among children at the time of school entry
is bigger, more inrractable, and more important than we had
thought. So much is happening to children during their first
three years at home, at a time when they are especially mal
leable and uniquely dependent on the family for virtually all
their experience, that by age 3, an intervention must address
not just a lack of knowledge or skill, but an entire general
approach [0 experience.
Cognitively, experience is sequential: Experiences in in
fancy establish habits of seeking, noticing, and incorporating
new and more complex experiences, as well as schemas for
categorizing and thinking about experiences. Neurologically,
infancy is a critical period because cortical developmenr is
influenced by the amounr of central nervous system activity
stimulated by experience. Behaviorally, infancy is a unique
time of helplessness when nearly all of children's experience
is mediated by adults in one-to-one interactions permeated
with affect. Once children become independent and can
speak for themselves, they gain access to more opportunities
for experience. But the amount and diversity of children's
past experience influences which new opportunities for ex
perience they notice and choose.
Estimating, as we did, the magnitude of the differences in
children's cumulative experience before the age of 3 gives an
indication of how big the problem is . Estimating the hours
of inrervenrion needed [0 equalize children's early experience
makes clear the enormity of the effort that would be re
quired to change children's lives. And the longer the effort is
put off, the less possible the change becomes. We see why
our brief, intense efforts during the War on Poverty did not
succeed. But we also see the risk to our nation and its chil
dren that makes intervenrion more urgenr than ever. 0
SPRING 2003 AMERICAN FEDERATION OF TEACHERS 9