ArticlePDF Available

Knowing letter names and learning letter sounds: A causal connection

Authors:

Abstract and Figures

Two experiments tested the common assumption that knowing the letter names helps children learn basic letter-sound (grapheme-phoneme) relation because most names contain the relevant sounds. In Experiment 1 (n=45), children in an experimental group learned English letter names for letter-like symbols. Some of these names contained the corresponding letter sounds, whereas others did not. Following training, children were taught the sounds of these same "letters." Control children learned the same six letters, but with meaningful real-word labels unrelated to the sounds learned in the criterion letter-sound phase. Differences between children in the experimental and control groups indicated that letter-name knowledge had a significant impact on letter-sound learning. Furthermore, letters with names containing the relevant sound facilitated letter-sound learning, but not letters with unrelated names. The benefit of letter-name knowledge was found to depend, in part, on skill at isolating phonemes in spoken syllables. A second experiment (n=20) replicated the name-to-sound facilitation effect with a new sample of kindergarteners who participated in a fully within-subject design in which all children learned meaningless pseudoword names for letters and with phoneme class equated across related and unrelated conditions.
Content may be subject to copyright.
J. Experimental Child Psychology 88 (2004) 213–233
www.elsevier.com/locate/jecp
0022-0965/$ - see front matter 2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.jecp.2004.03.005
Knowing letter names and learning letter
sounds: A causal connection
David L. Share¤
Department of Learning Disabilities, Faculty of Education, University of Haifa,
Mt. Carmel 31905, Haifa, Israel
Received 1 August 2003; revised 7 March 2004
Available online 20 April 2004
Abstract
Two experiments tested the common assumption that knowing the letter names helps chil-
dren learn basic letter-sound (grapheme–phoneme) relation because most names contain the
relevant sounds. In Experiment 1 (nD45), children in an experimental group learned English
letter names for letter-like symbols. Some of these names contained the corresponding letter
sounds, whereas others did not. Following training, children were taught the sounds of these
same “letters.” Control children learned the same six letters, but with meaningful real-word
labels unrelated to the sounds learned in the criterion letter-sound phase. DiVerences between
children in the experimental and control groups indicated that letter-name knowledge had a
signiWcant impact on letter-sound learning. Furthermore, letters with names containing the rel-
evant sound facilitated letter-sound learning, but not letters with unrelated names. The beneWt
of letter-name knowledge was found to depend, in part, on skill at isolating phonemes in spo-
ken syllables. A second experiment (nD20) replicated the name-to-sound facilitation eVect
with a new sample of kindergarteners who participated in a fully within-subject design in
which all children learned meaningless pseudoword names for letters and with phoneme class
equated across related and unrelated conditions.
2004 Elsevier Inc. All rights reserved.
Keywords: Letter names; Letter sounds; Alphabetic skills; Phonemic awareness; Literacy acquisition;
Reading development
¤Fax: +972-4-824-0911.
E-mail address: dshare@construct.haifa.ac.il.
214 D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233
Introduction
Preschoolers’ knowledge of the names of printed letters has long been known to
be among the strongest predictors of future success in learning to read (e.g., Bond &
Dykstra, 1967; Chall, 1967; Durrell, 1958; Share, Jorm, Maclean, & Matthews, 1984;
Tunmer, Herriman, & Nesdale, 1988; Wilson & Flemming, 1938). The strength of
this relation typically exceeds that of IQ (Share et al., 1984; Stanovich, Cunning-
ham, & Feeman, 1984). The nature and signiWcance of this association, however,
continue to be the subject of both speculation and controversy (e.g., Adams, 1990;
Durrell, 1980; Ehri, 1983; Gibson & Levin, 1975; Mates & Strommen, 1995/1996;
Venezky, 1975). An understanding of this relation has important educational impli-
cations as there is considerable debate over the value of teaching preschoolers the
names of letters (see, e.g., Durrell, 1980; Ehri, 1983; Feitelson, 1988; Mates & Strom-
men, 1995/1996).
There are a number of possible explanations for this relation, all or some of which
may be partly correct. The most common account assumes that knowledge of letter
names simply reXects the educational environment prior to school entry. Because
parents and preschool/kindergarten teachers rightly or wrongly believe that teaching
children letter names helps prepare them for school, letter-name knowledge may sim-
ply be a proxy for preschool and home background factors known to play an impor-
tant role in promoting literacy (Share, Jorm, Maclean, Matthews, & Waterman, 1983;
White, 1982). By this account, there is nothing to be gained by a child’s mastery of
the letter names as such. Raw correlation coeYcients, however, indicate that letter
name knowledge accounts for around twice as much variance in early reading as
measures of socioeconomic status or more “dynamic” educational processes in the
home, such as parents’ story-reading activities (Share et al., 1983).
A second account of the relation between letter names and early reading views
letter-name knowledge as a proxy for variance associated with phonological mem-
ory processes—the ability to learn and recall speech-based (phonological) informa-
tion (Share, 1995; Share et al., 1984). It is now well established that a majority of
variance in early reading skill is associated with the immediate, short-term, or long-
term recall of phonological material, such as spoken pseudowords (Adams, 1990;
Brady, 1986; Shankweiler & Liberman, 1989; Share, 1995; Snowling, 1991; Stano-
vich, 1986, 2000; Wagner & Torgesen, 1987). Furthermore, there is strong evidence
that phonological processes play a causal role in reading acquisition (Wagner &
Torgesen, 1987), although the precise role of these processes is still not fully under-
stood (see, e.g., Adams, 1990; Shankweiler & Liberman, 1989; Share & Stanovich,
1995). According to what might be termed the “phonological proxy” hypothesis,
letter names, like pseudowords, are essentially meaningless phonological strings;
memorizing these labels should, therefore, rely heavily on basic phonological mem-
ory processes. Thus, phonological memory may represent a common cause for
what might be a “spurious” relation between letter-name knowledge and early
reading ability.
A third explanation for the predictive strength of letter names—the “visual
familiarity” hypothesis—proposes that letter-name knowledge may tap a child’s
D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233 215
familiarity with letter shapes. Ehri (1986) has suggested that the ability to recognize,
distinguish, and recall the form and orientation of a large number of visually confus-
able letter-symbols is a considerable achievement for a preschool or kindergarten
child. Letter-name knowledge may provide an index of the extent to which a child
has mastered this perceptual learning task.
It has also been proposed that knowing the names of letters makes a direct contri-
bution to early reading by helping young children appreciate that writing represents
spoken language rather than directly reXecting meaning. Because letter names some-
times match sequences of phonemes heard in words (e.g., farm—R, or deaf—F),
whole syllables (candy—D), or even whole spoken words (e.g., UDyou), knowing let-
ter names can sensitize children to the fundamental phonological nature of writing.
Both naturalistic (Gentry, 1982; Read, 1976; Treiman, 1993) and experimental stud-
ies (De Abreu & Cardoso-Martins, 1998; Levin, Patel, Margalit, & Barad, 2002;
McBride-Chang & Treiman, 2003; Treiman, 1994; Treiman & Rodriguez, 1999; Trei-
man & TincoV, 1997; Treiman, TincoV, & Richmond-Welty, 1996) have demons-
trated the inXuence of letter names on early attempts at reading and writing. For
example, Treiman et al. (1996) found that preschool children were better at identify-
ing the Wrst letter of the spoken word beach that contains the name of the letter B,
than the word bone that does not. Treiman et al. (1996) also found that the corres-
pondence between a spoken word and a letter name can sometimes mislead children
into identifying the letter Y as the Wrst letter in the word wife, or C as the initial letter
of the word seem (see also Levin et al., 2002).
This study focused on a related but distinct account of the role of letter-name
knowledge in early reading. Because knowledge of letter names typically develops
earlier than letter-sound knowledge (Mason, 1980; McBride-Chang, 1999; Trei-
man et al., 1996; Worden & Boettcher, 1990), and because letter names normally
contain the relevant sounds, knowing the names of letters may make it easier for
children to master the letter-sound (grapheme–phoneme) relation necessary for
eYcient decoding (Durrell, 1980; Ehri, 1986; McBride-Chang, 1999; Read, 1986;
Stuart & Coltheart, 1988; Treiman & Kessler, 2003). Support for this “name-to-
sound facilitation” hypothesis comes from a study by Treiman, Weatherston, and
Berch (1994), who found that some children approaching the end of kindergarten
spelled words with the initial /w/ sound using the letter Y (e.g., war—YR). When
directly questioned, many kindergartners said that the letter Y makes the sound
/w/. More generally, both preschoolers and kindergartners were more successful
spelling phonemes such as /b/ and /l/ that are heard in the names of the corre-
sponding letters than phonemes that are not part of the letter name (/h/ and /g/).
Further support for the name-to-sound facilitation hypothesis was reported by
Treiman, TincoV, Rodriguez, Mouzaki, and Francis (1998). Pooling the data from
three large-scale US surveys of letter-name and letter-sound knowledge, Treiman
et al. (1998, Study 1) found that children were better at identifying the sounds of
letters when the letter’s name began with that sound, such as b, than when it was at
the end, such as f, or when it was not in the name at all (e.g., h). In a follow-up
training study, preschoolers with good letter-name knowledge but poor knowledge
of letter sounds were directly taught the sounds of letters whose names they
216 D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233
already knew (Treiman et al., 1998, Study 2). Letter-sound learning was best with
letters with sounds at the beginning of their names (e.g., b), poorer for sounds at
the end of names such as l, and worst for letters whose sounds are not in the names
at all (e.g., w).
The following two experiments directly tested the “name-to-sound facilitation”
hypothesis in a sample of cognitively normal kindergarteners who were not familiar
with either the names or the sounds of English letters. In Experiment 1 there were
two groups of kindergarten children: An experimental and a control group. The
experimental group Wrst learned a representative set of (English) letter names for arti-
Wcial letter-like symbols. Following training, all children were taught the sounds for
these same “letters.” Ease of learning the letter-sound correspondences after having
learnt the letter names provided a fairly direct test of the hypothesized causal connec-
tion between letter-name and letter-sound knowledge. To check the possibility that
increased visual familiarity alone may account for any training eVect, a control group
learned the same set of symbols, but used phonologically unrelated labels—that is,
semantically meaningful names rather than pseudoword names. For example, a child
who learns that the Wgure is called ESS should, according to the name-to-sound
facilitation hypothesis, Wnd it easier to learn that this same symbol has the sound /s/
than a child taught that the same Wgure is named BRIDGE, a phonologically unre-
lated label. If letter-name knowledge only taps children’s familiarity with the visual
form of letters, there should be no signiWcant diVerences between the experimental
and control groups in criterion letter-sound learning once instructional time is
equated. In fact, any diVerence should favor the control group because real-word
names should be easier to learn, and should more readily provide a verbal label for
each shape.
In addition to the between-group test of the name-to-sound facilitation hypothe-
sis, Experiment 1 also included a within-subject test (among children in the experi-
mental condition) of this hypothesis by comparing the learning of letters whose
names contained the relevant letter sound with letters whose names did not contain
the relevant sound.
In Experiment 1 also examined whether phonemic analysis mediates the rela-
tion between letter-name and letter-sound learning. A number of authors (e.g.,
Read, 1986; Treiman et al., 1998; Treiman et al., 1994; Venezky, 1975) have sug-
gested that the ability to exploit the connection between a letter’s name and its
sound may depend on a child’s skill at phonemic segmentation, an ability known
to vary widely among preschoolers (Adams, 1990; Brady & Shankweiler, 1991).
Thus, knowing the name of the letter S can help a child learn its sound /s/ only if
s/he recognizes that the string ESS contains the phoneme /s/. If this “segmenta-
tion” hypothesis is correct, then phonemic analysis should correlate positively
with letter-sound learning in the experimental group. That is, children better able
to isolate sounds in spoken syllables should beneWt more from learning phonolog-
ically related letter names. In contrast, pretest phonemic analysis should be unre-
lated to posttest letter-sound learning in the control group because criterion
letter-sound learning does not depend on the ability to analyze previously learned
letter names.
D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233 217
Experiment 1
Method
Participants
As most normal English-speaking preschool children are familiar with many of
the letter names (Mason, 1980), this study was carried out in a population unfamiliar
with English letter names—Israeli kindergartners—for whom the selected names and
sounds constituted phonologically legal but unfamiliar strings.
Participants were recruited from two kindergartens located in an area of average
socioeconomic status in Haifa. All children who neither were new immigrants nor
had suspected learning disabilities were tested on a battery of pretest measures
for the purposes of selecting two matched groups of children. Following the adminis-
tration of these pretest measures, 24 children (12 pairs) in one kindergarten and 22
(11 pairs) in the other kindergarten were matched as closely as possible on relevant
background, and pretest measures (see below). Within each kindergarten, members
of each pair were then randomly assigned to experimental (nD24), or to control
treatments (nD22). Children in each treatment group were matched both within and
across kindergartens. Following training, one child in the experimental group was
lost to the sample due to illness, leaving 23 experimental, and 22 control children.
Although a “within-setting” design in the case of experimental training studies
introduces the possibility of contamination, whereby children in diVerent treatments
exchange information, this design was preferred because: (i) it is not susceptible to
between-group (between-setting) eVects that are common in educational settings and
that are often beyond an experimenter’s control; and (ii) any contamination tends to
dilute training eVects and consequently reinforces the validity of any obtained diVer-
ences between groups.
Pretest measures
Discrimination/production of target sounds. All children were Wrst checked to ensure
that they were able to discriminate and produce all the target sounds to be used in
criterion letter-sound learning. Each child was simply asked to repeat a sound spoken
by an examiner. No child failed on any of these items.
Peabody Picture Vocabulary Test—Hebrew adaptation. This task was administered
using standard procedures (Solberg & Nevo, 1979).
Hebrew letter names. Children were asked to supply the names of 12 consonant let-
ters printed in large type (approximately 1cm in size) and arranged in a Wxed nonal-
phabetical order. Test-retest (3-week) reliability obtained in a separate study of
representative kindergarten children was .82 (nD34).
Initial phoneme isolation. This test required a child to isolate the initial stop conso-
nant (“beginning sound”) in 12 pseudoword CV syllables (e.g., /ka/ /de/). Two
218 D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233
practice items were Wrst given with feedback to ensure that the child understood the
task. No feedback was given on the 12 test items. One point was awarded for each
correct response.
Final phoneme isolation. This test required isolation of the Wnal continuant phoneme
(“end sound”) in 12 VC pseudosyllables (e.g., /iv/ /an/). Two practice items with feed-
back preceded the test items. Again, no feedback was given on test trials. A combined
phoneme isolation score was created by summing raw scores on the above two tasks
which correlated .52in this sample.
Procedure
Pretest measures. Administration of the pretests was carried out in two separate
sessions with tests administered in the following Wxed order: Discrimination/pro-
duction of target sounds, Wnal phoneme isolation, Peabody Picture Vocabulary
(session 1), Hebrew letter names, and initial phoneme isolation (session 2). All
testing was carried out individually with sessions on diVerent (mostly alternate)
days.
Training. Children were trained in pairs which were formed by random assignment
within group. Training was carried out over ten 10-min sessions with two to three ses-
sions per week spanning alternate days. Training time was carefully monitored to
ensure that training times were matched for experimental and control groups. The
training phase lasted approximately 6 weeks. All training was carried out by a trained
psychologist and native Hebrew speaker who was not informed of the purpose of the
study. The task was introduced by telling the children that they were going to learn
some “secret symbols from another planet.” Children in both experimental and con-
trol groups appeared to be highly and equally motivated to come and to work with
the “visitor” who was continually pestered by children asking to be next in turn.
There was no case of noncompliance.
Experimental (letter-names) group. This group was taught the associations
between six letter names and six visual forms taken from Vellutino and Scanlon
(1987) and they are reproduced here in Fig. 1. (Real English letter forms were not
taught because Latin letters are relatively common in Israel, particularly in commer-
cial settings.) Each form, which was approximately 3 £2 in. in size, was presented on
5£4 in. cards, with one form per card. Each child was taught the associations
between these forms and the names of six (English) letters. Prior pilot work had indi-
cated the suitability of teaching this number of associations within the allotted time-
scale of the research.
The six letter names taught to the experimental group were chosen after consider-
ation of the relation between the 26 English letter names and the corresponding
sounds, which revealed that letter names fairly consistently fall into four major
groupings: Consonant–vowel (CV) names containing initial stops (B, D, K, J, P, T),
vowel–consonant (VC) names containing Wnal continuants (F, L, M, N, R, S), vowel
letter names representing the corresponding long vowel sounds (A, E, I, O, U), and a
D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233 219
miscellaneous group comprising letter names either unrelated to letter sounds (H, W)
or containing the relevant letter sound (or sounds) in a way that diverges somewhat
from the above categories (Y, Q, X). Only four letters are clear exceptions to this
scheme: Both C and G, which exist in both “soft” (continuant and aVricate) and
“hard” (stop) forms, assume names in the CV initial stop pattern, and the continu-
ants V and Z, which also assume the CV initial stop pattern rather than the Wnal
continuant pattern.
Two letters were chosen from the initial stop group (B, T), and two from the Wnal
continuant group (S, M). Two additional letters were selected to represent letters
whose names do not include the corresponding letter sound. One came from the
vowel group (E) and one from the unrelated consonant group (H). Note that because
the short vowel // was the letter sound taught in the criterion learning for the letter-
like form labeled E, the letters E and H provided two exemplars of letter names that
do not contain the relevant letter sound. The decision as to which letters to select
from each category was based on several criteria: The two phonemes within a given
category must diVer in place of articulation and must be phonologically legal in
Hebrew, and the letter names must not form meaningful Hebrew words. Because the
names for the letters B and M are both real words in Hebrew, these were slightly
altered to /b/ and /m/.
Associations between shapes and names were taught with a variety of traditional
kindergarten tabletop activities. These included drawing the “letters” followed by
naming, pasting colored thread onto the letter shapes accompanied by letter naming,
identifying letter names following guided tracing with eyes closed, identifying a miss-
ing letter removed by a partner from the learned set, arranging cards in the order
verbalized by the partner, cutting out shapes with accompanying naming, and color-
ing in.
In the Wrst session, children were introduced to two letters. In subsequent sessions,
an additional letter was added. In this way, the set of letters learned in each
session gradually expanded from two to six, with each session including all
previously learned letters to ensure retention. Because this training system implies
Fig. 1. Visual forms employed in Experiments 1 and 2 (Vellutino & Scanlon, 1987).
220 D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233
that letters introduced in earlier sessions received greater exposure than letters intro-
duced in later sessions, relative exposure to the three diVerent letter types (stops, con-
tinuants, and unrelated) was equated by creating three subgroups (by random
assignment) within the experimental group. All children were introduced to the same
symbols at the same time and in the same order, but each subgroup learned a diVer-
ent name for this symbol with each of the three names belonging to one of the three
letter types. This meant that each subgroup learned the letter names in a diVerent
order, but that each of the three letter types was exposed for an identical number of
sessions within the experimental group as a whole. This arrangement also controlled
for any possible diVerences between the letter-like shapes in memorability/
confusability because each shape was associated with each of the three diVerent types
of letter names.
At the beginning of each session, all previously taught letters were tested for reten-
tion. This provided an ongoing measure of learning mastery.
Control group. Instead of letter names, the control group learned meaningful real-
word Hebrew names for the symbols. These were concrete objects suggested by the
visual forms. For example, two parentheses placed back-to-back as follows was
called gesher (meaning bridge). In a similar manner, each of the six symbols was
assigned a name: ish (man), degel (Xag), pe (mouth), ogen (anchor), and vav (hook). In
the control group, all children learned the same name-shape pairings.
Criterion learning of letter-sound correspondences
Posttest learning of letter-sound correspondences was carried out in three separate
sessions each involving a series of paired-associate learning trials. Three phonemes
were taught in the Wrst session, the three remaining phonemes were taught in the sec-
ond session, and all six phonemes were taught in the Wnal session. At the end of each
session, each letter-sound correspondence was tested three times giving a total possi-
ble learning score of 36 (3£3C3£3C6£3).
The experimenter explained that the child had learned the “names” of the shapes
and now was going to learn their “sounds.” No reference was made to the connection
or otherwise between names and sounds either here or at any stage during the study.1
Moreover, no letter was explicitly referred to by its name in this phase of the study.
Letters were presented one at a time together with their phonemes and the children
were instructed to remember these sounds. Each of the Wrst two sessions began by
introducing the new letter-sound associations. Next, each of the three letters was pre-
sented three times in a Wxed random order. Corrective feedback was given in the
event that a child failed to supply the correct response within 6s or supplied an incor-
rect response. In the Wnal session, all six letters were tested three times each in a Wxed
random order with corrective feedback.
Because the three subgroups within the experimental group learned diVerent
sounds for the various symbols, children in the control group were split into three
1 Israeli kindergarteners are not taught letter sounds—these are introduced only in Grade 1 when for-
mal reading instruction begins. More details about the literacy practices of typical Israeli kindergartens
can be found in Share and Gur (1999).
D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233 221
corresponding subgroups with each subgroup learning the same letters in the same
order as the experimental children.
Results
Pretest scores for the experimental and control groups appear in Table 1 together
with letter-learning scores indicating the extent to which children had mastered the
names of the six “letters” taught during training. In each group, the data from all
three (order-of-letter-learning) subgroups were pooled because there were no signiW-
cant subgroup main eVects or interactions. For the same reason, kindergarten site
was also ignored.
It can be seen from Table 1 that the experimental and control groups were well
matched on all pretest measures. T tests indicated that there were no signiWcant
diVerences between groups on any of the pretest measures. Only one t value (age)
exceeded unity, t(43) D1.56, pD.13, and here the control group enjoyed a slight
(3-month) advantage.
It is also clear from Table 1 that the control group achieved a signiWcantly
greater level of mastery (30.5/32D95%) than the experimental group (24.4/32D
76%) in learning the letter names assigned during training, t(43) D6.46, p0.05.
It is not surprising that kindergartners appear to Wnd it easier to associate
meaningful names with objects or graphic symbols than meaningless pseudoword
names.
Criterion learning of the letter sounds appears in Table 2. Averaged over all
six letters, there was a strong and signiWcant diVerence in favor of the experi-
mental group which scored approximately 50% higher than the control group in spite
of the fact that the experimental group achieved only 76% mastery of the letter
names.
Correlations between individual letter-name learning scores recorded during
training and posttest letter-sound learning reinforced these results. Letter-name mas-
tery correlated signiWcantly, r(21) D.52, pD.005, with criterion letter-sound learning
in the experimental group but not in the control group, r(20) D.07, although the
latter coeYcient is no doubt attenuated by range restriction.
Table 1
Pretest and letter name learning scores for experimental and control groups in Experiment 1
Pretest measures Experimental group
(nD23)
Control group
(n D22)
MSD Range MSD Range
Age in months 62.3 (6.0) 49–74 65.3 (5.97) 52–74
Gender 9 girls, 14 boys 8 girls, 14 boys
Knowledge of Hebrew letter names (max D12) 5.6 (4.02) 0–12 5.9 (3.74) 1–12
Peabody vocabulary (raw score) 47.8 (7.71) 34–59 47.5 (8.75) 31–71
Initial sound isolation (max D12) 6.0 (3.10) 1–12 5.6 (3.58) 0–12
Final sound isolation (max D12) 6.7 (3.66) 0–12 6.8 (3.50) 2–12
Letter mastery in training (max D32) 24.4 (4.26) 14–31 30.5 (1.26) 27–32
222 D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233
Overall, there was a clear advantage in knowing the names of letters when it
came to learning letter-sound correspondences. Furthermore, this advantage cannot
simply be attributed to greater visual familiarity with the letter shapes because
both the experimental and control groups’ exposure times were equated during
training.
Relevant versus irrelevant letter names
If the advantage in knowing the names of letters accrues because most names con-
tain the relevant letter sounds and thereby supply a recall cue, then it should be easier
to learn sounds for letters whose names contain these sounds compared to letter
names that do not. The results in Table 2 indicate that the experimental group
learned signiWcantly more letter sounds for “related” letters (those with names con-
taining their sounds: B, T, S, M) than the control group. In contrast, the diVerence
between the two groups in learning sounds for letters with “unrelated” names did not
reach signiWcance, F(1, 43) D3.47, pD.069.
Table 2 also shows that the advantage in learning the sounds of letters whose
names contain these sounds appears to be equally great for initial (CV) stops and
Wnal (VC) continuants. This Wnding contrasts with the English-language advantage
reported by Treiman et al. (1998) and by McBride-Chang (1999) for initial stops. The
present data, however, are quite consistent with both: (i) the pretest Wnding that the
tasks of isolating initial stops and Wnal continuants are of comparable diYculty for
Hebrew-speaking prereaders, t(45) D1.82, pD.075; and (ii) the fact that among this
same population, there appear to be no intrinsic diVerences in diYculty learning
either stop or continuant correspondences as borne out by the fact that the control
group’s posttest learning scores for both types of phonemes were identical. In fact,
among controls, all three sets of sounds were virtually identical. Thus, diVerences
between related and unrelated letters in the experimental group can be attributed to
letter-name knowledge per se.
The interaction between group and letter type was directly tested in a multivariate
analysis of variance (MANOVA) with a planned contrast representing the diVerence
between related and unrelated letters. A signiWcant group by letter-type interaction,
T
a
bl
e
2
Posttest letter-sound learning scores for experimental and control groups in Experiment 1
aBecause each letter-sound correspondence was tested 6 times, the maximum possible score was 36.
*p0.001.
Criterion variables Experimental group (nD23) Control group (nD2) F(1,43)
MSDRange MSDRange
All soundsa19.04 (6.58) 9–35 12.55 (3.62) 6–20 16.62*
Related sounds
(initial and Wnal)
6.09 (3.54) 1–12 3.14 (1.18) 0–7 12.25*
Initial sounds 2.87 (2.12) 0–6 1.59 (1.26) 0–3 5.9*
Final sounds 3.17 (1.72) 1–6 1.59 (1.22) 0–5 12.29*
Unrelated sounds 2.43 (1.62) 0–6 1.64 (1.22) 0–4 3.47 n.s.
D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233 223
F(1, 43) D6.57, pD.014, conWrmed that the experimental group, but not the control
group, beneWted more from learning the related letters than from learning the unre-
lated letters.
Correlations between individual letter-name learning scores in the training phase
and criterion letter-sound learning reinforced the group results. For children in the
experimental group, letter-name learning during training correlated strongly,
r(21) D.62, pD.001, with letter-sound learning for related letters but only weakly,
r(21) D.38, pD.036, for unrelated letters. In the control group, both these correlations
were unambiguously nonsigniWcant (related letters, r(20) .03; unrelated letters,
r(20) D.06). It might be speculated that the correlation between letter-name learning
and criterion letter-sound learning for unrelated letters (rD.38) suggests that even
the “unrelated” letters may provide some general information regarding phoneme
class (e.g., front vowel, E, fricative consonant, H,) that may be useful in letter-sound
learning.
Do the beneWts of letter-name knowledge depend on phonemic analysis?
As discussed previously, it is commonly assumed that the ability to take advantage
of the connection between a letter’s name and its sound implies an ability to isolate
the relevant sound in the name. If this is correct, then pretest phoneme segmentation
skill should predict criterion letter-sound learning in the experimental group for
which letter-name knowledge was found to assist letter-sound learning. By contrast,
pretest phoneme segmentation scores should be unrelated to criterion letter-sound
learning in the control group because the latter does not depend on the ability to ana-
lyze previously learned letter names. The alternative hypothesis, according to which
the beneWts of letter-name knowledge derive not from recognition of speciWc pho-
neme identity but from overall phonetic similarity, predicts that phoneme segmenta-
tion will be uncorrelated with letter-sound learning in both experimental and control
groups.
Correlations were accordingly computed for experimental and control groups
between criterion letter-sound learning scores and individual pretest measures of
initial, Wnal, and summed composite phoneme isolation. This latter composite score
was designed to provide a more reliable general measure of phonemic segmentation
skill.
In the control group all three correlations were negligible, r0s(20) D.13,.11, and .01,
and nonsigniWcant as predicted. In the experimental group, in contrast, two of the
three correlations were signiWcant (initial-sound isolation, r(21) D.37, p0.05; Wnal-
sound isolation r(21) D.29, n.s.; and combined phoneme isolation, r(21) D.36,
p0.05, one-tailed). To check whether these correlations reXect a speciWc link
between phoneme analysis and letter-sound learning rather than simply a common
source such as general verbal ability, correlations were recomputed with Peabody
Picture Vocabulary scores partialed out. In the experimental group, partial correla-
tions were r(20) D.44, p0.05, for combined phoneme isolation; r(20) D.47, p0.05,
for initial-sound isolation; and r(20) D.33 , n.s., for Wnal-sound isolation. For con-
trols, these correlations remained negligible, r0s(19) D.01, .11, and ¡.12. Overall, these
analyses support the argument that the ability to beneWt from letter-name knowledge
224 D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233
when learning letter sounds depends on the ability to extract the component sounds
in a name.2
Letter-sound learning errors
Close to 90% of the errors were either nonresponses (41%) or an incorrect letter
sound from the set being taught (48%). An additional 8% of the errors were pho-
nemes not drawn from those taught (inventions). These errors were mostly phonemes
similar in terms of place or manner of articulation (e.g., /sh/ instead of /s/ or /p/
instead of /b/). Only 1% of the errors were letter names rather than letter sounds and
these were divided evenly between two experimental and two control children. All
these error types occurred with similar frequency in both the experimental and con-
trol groups. One error type, however, appeared only in the experimental group,
namely, supplying one of the component phonemes in the letter’s name, but the
wrong one. Of the 23 children in the experimental group, six committed this error by
supplying the initial vowel // for the letter “OM” whose sound was /m/. This would
appear to be fairly direct evidence of what might be termed a “name-segmentation”
strategy whereby a child isolates a component sound in a known letter name and
oVers this as a candidate letter sound. This would seem to be an experimental ana-
logue of the naturally occurring Wnding of Treiman et al. (1994) that some kinder-
gartners report that the letter Y makes the sound /w/. The frequency of 6/23 is almost
certainly an underestimate because only this one error (supplying the initial vowel for
the letter M) could be interpreted unequivocally as evidence of a name-segmentation
strategy. In all other cases, the accompanying vowel was either another letter name
(E) or another letter sound (//), each of which was conservatively assigned to the
latter categories.
Discussion
A number of Wndings converged on the conclusion that knowing the letter names
helps preschool children learn the corresponding sounds. First, there was a substan-
tial advantage in letter-sound learning for children taught letter names compared to
matched children who learned unrelated labels for the same symbols. Second, letter-
sound correspondences for letters with names that contained the relevant phonemes
2 The contribution of Hebrew letter names to these individual diVerences analyses was also checked.
Although Hebrew letter-name knowledge, somewhat surprisingly, was not found to be a reliable predictor
of either letter-name or letter-sound learning in either the experimental or control groups, partialing out
letter names in the analysis of the contribution of phonemic awareness to letter-sound learning in the ex-
perimental group did reduce the correlations from (unpartialed) .37, .29, and .36 to (partialed) .27, .19, and
.26 for initial, Wnal, and combined phoneme segmentation—the latter three coeYcients were all nonsigniW-
cant. The simplest interpretation of these data (but by no means the only one) is that Hebrew letter-name
knowledge did not have a direct impact on letter-sound learning in training, but contributed indirectly ow-
ing to its close relation with phonemic awareness. This conclusion, however, should be qualiWed: (i) by the
fact that the Hebrew letter names measure was only designed as a “background” measure for equating
groups, not as an independent variable; and (ii) by important diVerences between Hebrew and English let-
ter names (see Levin et al., 2002).
D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233 225
were learned more easily than correspondences for letters that did not include the rel-
evant phonemes. Third, the degree to which letter names had been mastered during
training predicted criterion letter-sound learning in the experimental group, but not
in the control group. Finally, there was some direct evidence in the form of errors
among experimental but not control children pointing to explicit analysis of sounds
in letter names in the course of criterion letter-sound learning.
The magnitude of the letter-name eVect was large and not merely signiWcant.
Overall, letter-sound learning scores in the experimental group were around 50%
higher than in the control group. For the four “related” letters—those with names
containing the relevant letter sound—the experimental group mean was twice the
control mean. Although scores for the “unrelated” letters were a little higher in the
experimental group than in the control group, this diVerence was not signiWcant.
Nonetheless, the fact that letter-name learning scores in the experimental group were
signiWcant, albeit modest predictors of “unrelated” letter-sound learning (rD.38),
suggests that even the two unrelated letters in this study (E and H) may have pro-
vided some general information about phoneme class that may have assisted letter-
sound learning. Thus, even letters whose names are only poorly related to letter
sounds (Q, X, Y, H) may not be entirely worthless with regard to learning grapheme–
phoneme correspondences. The letter W is the sole instance of a name entirely unre-
lated (phonemically) to its English letter sound.
It might be argued that the poor performance of the control group was because
these children initially learned associations that were nonarbitrary (for example the
word bridge was associated with a bridge-like form) thereby inducing an associative
strategy incompatible with the arbitrary associations required in the criterion phase.
This issue was directly examined in a second experiment that eliminated this poten-
tial confound by testing the name-to-sound facilitation hypothesis in a fully within-
subject design in which all children learned only meaningless monosyllabic letter
names that either did or did not include the sounds taught in the criterion letter-
sound phase. This second study also equated phoneme class across the two condi-
tions: In Experiment 1, the comparison between related and unrelated letter names in
the experimental group used names from diVerent phoneme classes.
Experiment 2
Method
In this second experiment, a new sample of 20 children was drawn from two
additional kindergartens in a similar neighborhood. As in Experiment 1, all children
participating in this second experiment were able to discriminate and pronounce
all the letter names and sounds correctly and were unfamiliar with English letter
names.
All children in this study learned the same four letter-like shapes, representing the
consonantal letters B, T, S, and M. Each child learned two related letter names
(one stop and one continuant) and two unrelated letter names (stop and continuant).
226 D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233
For example, the name of the symbol representing the letter S was either learned as
the related name /s/ or as the unrelated name /f/. Ten children learned the related
names for B (/b/) and S (/s/), and the unrelated names for T (/gi/) and M (/l/), while
10 children learned the unrelated names for B (/di/) and S (/f/) and the related names
for T (/ti/) and M (/m/).
Letter names were taught to pairs of children over eight sessions using the same
tabletop activities as in Experiment 1. Two letters were taught in each of the Wrst six
sessions, with related and unrelated pairs in alternating sessions, followed by two fur-
ther sessions with all four letters. Criterion letter-sound correspondences were taught
and tested over four sessions using the same procedure (paired-associate learning
with corrective feedback) as in Experiment 1.
Results
Once again, there was a signiWcant advantage in learning letter sounds,
t(19) D2.14, p0.05, when children were already familiar with a letter name contain-
ing that sound (Table 3). The magnitude of this diVerence (50 versus 39%) was very
similar to the diVerence obtained in the experimental group in Experiment 1 between
related and unrelated items (51 versus 41%), and indicates that the name-to-sound
facilitation eVect is robust and not attributable to either incompatible associative
learning strategies or diVerences in phoneme class. As in Experiment 1, evidence for
an explicit name-segmentation strategy was revealed in responses that included the
vowel component of a letter name. Four children supplied the vowel phoneme //
rather than the consonantal phoneme; another three children supplied the vowel //.
Because training times were equated for both experimental and control children in
Experiment 1 and for the related and unrelated conditions in Experiment 2, the
advantage of knowing letter names cannot be attributed simply to greater familiarity
with their graphic forms. Of course, the present data do not exclude the possibility
that general letter-name knowledge also reXects visual familiarity with letter shapes,
as this factor was held constant in both studies. An additional no-treatment control
group with no exposure to letter forms prior to criterion letter-sound learning would
be needed to assess the independent contribution of visual familiarity.
Table 3
Background characteristics and criterion letter-sound learning in Experiment 2
Variable M(SD) Range
Background characteristics
Age in months 65.2 (3.79) 59–72
Gender 10 boys, 10 girls
Hebrew letter name knowledge (max D12) 4.9 (3.73) 0–12
Letter name mastery in training (max D18) 11.0 (3.56) 3–18
Criterion letter-sound learning
Related letter sounds (max D24) 12.1 (5.11) 1–21
Unrelated letter sounds (maxD24) 9.4 (6.64) 0–20
D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233 227
General Discussion
The present pair of studies demonstrate that letter names provide more than just
verbal labels for these forms. As with personal names, letter names supply convenient
verbal labels that uniquely identify each letter, and that are important if a child is to
understand the language of literacy. These labels also fulWll the essential function of
distinguishing critical from noncritical features of letters. For example, b and d have
the same forms but their names make clear that orientation is critical; a and a have
diVerent forms, but are functionally equivalent as are uppercase and lowercase let-
ters. Because phonemes are abstractions representing families of phonetic sounds, so
are letter identities abstract categories of (often) nonidentical graphic forms. Perhaps
the chief service rendered by letter names is in drawing a child’s attention to the
abstract symbolic nature of letters. Because letter names sometimes map sequences of
phonemes heard in spoken words (e.g., OK), they may also help prereaders under-
stand the connection between print and speech (Levin et al., 2002; Read, 1975; Trei-
man, 1993; Treiman & Kessler, 2003).
In addition to these considerations, this study suggests that the choice of a name
may also be important in early reading by virtue of the cues it can provide to assist
letter-sound learning.3 As Shakespeare’s Juliet surmises, most names of everyday
objects such as roses may well be arbitrary, and interchangeable. Printed letter
names, in contrast, may be an important exception to this rule.
This study also showed that the advantage of knowing letter names depends, in
part, on the ability to isolate phonemes in spoken syllables. Correlations between
phonemic analysis and criterion letter-sound learning were generally signiWcant in
the experimental group but consistently nonsigniWcant for controls. These results
point to a speciWc link rather than a common general ability factor, as correlations
for the experimental group remained signiWcant when vocabulary was partialed out.
Evidence for the use of an explicit name-segmentation strategy solely in the experi-
mental group in Experiment 1 reinforced the link between phonemic analysis and the
letter-name eVect. One weakness, however, with this strategy is that, in the case of
English at least, it fails to specify reliably which of the component phonemes in a
name is the critical one. For orthographies that adhere to the acrophonic principle,
such as Hebrew and Arabic, this is unproblematic because a letter’s sound is invari-
ably the Wrst phoneme in its name. In contrast, English letter names, as noted previ-
ously, are not entirely capricious. With the exception of W and Y, the target sound of
consonant names is consistently the sole consonant (or consonants in the case of X)
whether in initial or Wnal position.
3 Although it may seem odd to speak of the “choice” of a letter name given that letter names are rela-
tively Wxed features of a speciWc culture, letter names, like letter shapes, do evolve over time (see Naveh,
1982). At times of social and cultural upheaval, these changes can be quite dramatic, such as the change
from complex multisyllabic names for Russian letters to simple English-style monosyllabic names at the
time of the Bolshevik Revolution (Leikin, personal communication, 2002). On a smaller scale, consider
also the US zee compared to the British zed.
228 D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233
The present sample of non-English speaking children provided an excellent con-
trol for pre-existing letter-name knowledge typical among English-speaking popula-
tions. But to what extent can results obtained with non-English speakers be
generalized to the English language? As noted already, all letter names and sounds in
this study were phonologically legal strings. The letter-like forms were drawn from a
set used successfully in English-language studies of reading. The available Hebrew-
language data (Levin et al., 2002; Shatil & Share, 2003) suggest that general levels of
Hebrew letter-name knowledge among Israeli kindergarten children are similar to
Wndings obtained with English-speaking samples at the same age (e.g., Share et al.,
1984; Treiman et al., 1998). Other data, such as phonemic awareness training, point
to important parallels between beginning reading in Hebrew and in English (Bentin
& Leshem, 1993; Kozminsky & Kozminsky, 1995). As in North America, the over-
whelming majority of Israeli kindergarten children are nonreaders, with formal read-
ing instruction commencing in Grade 1 when children are age 6. Thus, there appears
to be no a priori reason why the present results should diVer systematically from
results obtained with English speakers, although only direct evidence can conWrm
this.
In one regard, however, there may be a fundamental diVerence between English
and Hebrew. McBride-Chang (1999) and Treiman et al. (1998) both reported that
English-speaking preschoolers are more successful learning the sounds of letters with
sounds at the beginning of the letter’s name (D) than the sounds of letters with
sounds at the end of their names (L). Treiman and Kessler (2003) argue that this
reXects the perceptual salience of syllable onsets as opposed to codas (see also Trei-
man, 1992) and possibly the fact that consonant–vowel letter names are more com-
mon in English than vowel–consonant names. The present study with Israeli
kindergartners, however, did not Wnd any diVerence between these two types of items;
in fact, there was a nonsigniWcant advantage in the opposite direction, namely, for
sound-Wnal (VC) letter names (see experimental group’s data, Table 1). Furthermore,
children in Experiment 1 found it equally easy to isolate initial sounds and Wnal
sounds with a nonsigniWcant trend once again favoring the Wnal phoneme or coda.
These Wndings point to possible psycholinguistic diVerences between Hebrew and
English in phonological structure. This possibility was recently examined by Blum
(2001) who found that native Hebrew-speaking preschoolers and older Grade 2
skilled readers alike found it easier to divide spoken CVC syllables into CV plus Wnal
C (coda) units than into onsets and rimes (C + VC). Blum (2001) concluded that
onset/rime units might not be the indigenous subsyllabic units in Semitic languages.
If letter-name knowledge promotes early reading by virtue of the help it provides
in learning grapheme–phoneme correspondences, why not simply replace the letter
names with the letter sounds? After all, a letter sound is also a verbal label. There
would seem to be several objections to this idea.
First, in many languages, two or even more letters have the same sound. Consider
the problem of the letters K and C; both have the sound /k/. If we assign the letter C
its alternate sound /s/ what becomes of the letter S? Many orthographies have alter-
nate graphemes representing the same sound (see Harris & Hatano, 1999). For exam-
ple, in Hebrew, approximately one third of all letters are homophonous, that is,
D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233 229
represent phonemes that can be represented with alternate letters (Share & Levin,
1999). The English letter C also illustrates another problem—that of letters with two
or more phonemic values. Not only do many letters have multiple sounds, but these
phonemic values are often shared with other letters or digraphs. Thus, letter names
would no longer provide unique identiWers for the letters, creating considerable
potential for confusion between letters and the sounds they represent. Almost all
English letters, both vowels and consonants, can appear as silent letters (approxi-
mately half the alphabet), can represent more than one sound (about one third of let-
ters), or have sounds that can be represented by other graphemes or digraphs (about
one third). Consider the case of the letter S, now to be renamed “s.” The letter C
often assumes this sound, as do other graphemes and grapheme-combinations (e.g.,
SC, PS, SS, etc.). Furthermore, the grapheme S itself does not always assume the /s/
sound; sometimes it falls silent (ISLAND) or marks other phonemes (e.g., PRISON,
SURE, MEASURE). In fact, almost all English letters, with the possible exceptions
of N and V, admit one of more of these complications. The point here is not to
bemoan English spelling-sound inconsistency, but to suggest that labeling letters by
their sounds would, in many situations, create considerable confusion for young
learners and almost certainly require constant and cumbersome qualiWcation by
speakers referring to the “s” letter versus the “s” sound.
Another argument for retaining letter names is that letter names may be easier to
learn than letter sounds because names are primarily syllabic and hence more percep-
tually salient—that is, word-like rather than isolated phonemes (Liberman, Cooper,
Shankweiler, & Studdert-Kennedy, 1967). Treiman and TincoV (1997) found that
young children’s knowledge of letter names (e.g., /ti/) sometimes caused them to spell
a complete consonant–vowel sequence (revi versus zevit) with that single letter,
thereby suggesting a special role for the syllable in early spelling development. For
the same reason, it has often been proposed that alphabetic writing Wrst be intro-
duced syllabically to young children (see for example, Rozin & Gleitman, 1977). By
this account then, letter names, by virtue of their syllabic and hence more word-like
nature, may be both easier to learn, and have greater value than letter sounds in sen-
sitizing children to sublexical speech units in the early stages of literacy. These con-
siderations point to a need for a unique label for letters in addition to their sounds.
What then might be the optimal form for these labels? The present study suggests
that, Wrst and foremost, the ideal letter names should include the corresponding letter
sound or at least its most common sound. Second, the name should include a mini-
mum of additional phonemes (ideally one) and be located in a position that is easiest
to isolate. The work of Goswami and Bryant (1990) and Treiman (1992) on the onset/
rime breakup of spoken syllables suggests that consonants may be easier to isolate in
initial rather than Wnal position, at least in English. In the case of stops, there is a fur-
ther advantage for the initial position in that stops cannot be pronounced without
appending a vocalic element, usually schwa. There may also be a problem with
voiced stops in Wnal position that seem diYcult to distinguish in everyday speech
from unvoiced stops (/p/:/b/, /t/:/d/, /g/:/k/). These alternations appear to be more eas-
ily identiWed in initial rather than in Wnal position (Treiman & Kessler, 2003). The
voiced forms, when in Wnal position, seem to require an eVort on the part of the
230 D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233
speaker to emphasize voicing. In light of these speculations, the names for letters rep-
resenting stop consonants would appear to be well suited to their task.
Is the syllable-Wnal pattern also appropriate for the names of continuant pho-
nemes? In this regard, it is noteworthy that unlike stop consonants that cannot be
pronounced in isolation without adding a subsequent vocal element such as schwa,
continuants such as /l/ and /m/ can be pronounced, and indeed “stretched,” in isola-
tion. This implies that VC-style names may be more appropriate for continuants than
for stops, although it remains to be seen if continuants are more easily isolated/iden-
tiWed by preschoolers in Wnal rather than initial position. And what of vowel names?
Perhaps these too are near optimal. The long-vowel name has no consonantal ele-
ment that might be confused with a consonant name, rather, it signals a diVerent class
of letters. Furthermore, the long-vowel sound is probably the most frequently occur-
ring sound assumed by these letters after the short sound, which is not a phonologi-
cally legal entity in English (see Treiman & Kessler (2003) for a discussion of the
legality criterion in letter names). If these speculations are not unfounded, it could be
argued that the general system of English letter names (W and Y excepted) may be
surprisingly well adapted to the task of providing both unique identiWers for English
graphemes and aids to letter-sound learning.
In conclusion, this study succeeded in identifying at least one reason behind the
predictive strength of a preschooler’s letter-name knowledge. There are clearly other
factors both cognitive and noncognitive that contribute to this relationship. As dis-
cussed at the beginning of this article, phonological memory could be one candidate
source of variance. It seems reasonable to assume that another would derive from the
fact that letter-name knowledge almost certainly taps educational/environmental fac-
tors in the home likely to have a direct and ongoing impact on literacy learning
(Snow, Burns, & GriYn, 1998). To investigate these factors it will be necessary to
undertake more qualitative investigations of the contexts within which children
acquire letter-name knowledge and the uses to which this knowledge is put. It is
doubtful that any one reason alone “explains” why letter-name knowledge is such a
powerful predictor of early reading. Herein may lie its remarkable strength as a mul-
tidimensional measure of knowledge important in its own right and as an indirect
index of additional cognitive and noncognitive factors.
Acknowledgments
The author thanks Smadar Reifer and Rivi Aspler for their invaluable assistance
in carrying out this study.
Appendix A
Key to phonetic notation: // as in bet, // as in (British) mop, /c/ as in sofa.
D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233 231
References
Adams, M. J. (1990). Beginning to read. Cambridge, MA: Bradford.
Bentin, S., & Leshem, H. (1993). On the interaction between phonological awareness and reading acquisi-
tion: It’s a two-way street. Annals of Dyslexia, 43, 125–148.
Blum, P. (2001). Syllable splitting in Hebrew: Orthographic and phonological inXuences [in Hebrew].
Unpublished master’s thesis, University of Haifa, Haifa, Israel.
Bond, G. K., & Dykstra, R. (1967). The cooperative research program in Wrst grade reading instruction.
Reading Research Quarterly, 2, 5–142.
Brady, S. (1986). Short-term memory, phonological processing and reading ability. Annals of Dyslexia, 36,
138–153.
Brady, S. A., & Shankweiler, D. P. (Eds.). (1991). Phonological processes in literacy: A tribute to Isabelle Y.
Liberman. Hillsdale, NJ: Erlbaum.
Chall, J. S. (1967). Learning to read: The great debate. New York: McGraw-Hill.
De Abreu, M. D., & Cardoso-Martins, C. (1998). Alphabetic access route in beginning reading acquisition
in Portuguese: The role of letter-name knowledge. Reading and Writing, 10, 85–104.
Durrell, D. D. (1958). Success in 1st grade reading. Journal of Education, 140, 1–6.
Durrell, D. D. (1980). Letter name values in reading and spelling. Reading Research Quarterly, 16, 159–163.
Ehri, L. C. (1983). Critique of Wve studies related to letter name knowledge and learning to read. In L. M.
Gentile, M. L. Kamil, & J. S. Blanchard (Eds.), Reading research revisited (pp. 143–153). Columbus,
OH: Charles E. Merrill Publishing Company.
Ehri, L. C. (1986). Sources of diYculty in learning to spell and read. Advances in Developmental and Behav-
ioral Pediatrics, 7, 121–195.
Feitelson, D. (1988). Facts and fads in beginning reading: A cross-linguistic perspective. Norwood, NJ:
Ablex.
Gentry, J. R. (1982). An analysis of developmental spelling in GNYS AT WRK. The Reading Teacher, 36,
192–200.
Gibson, E. J., & Levin, H. (1975). The psychology of reading. Cambridge, MA: MIT Press.
Goswami, U., & Bryant, P. (1990). Phonological skills and learning to read. Hove, East Sussex: Erlbaum.
Harris, M., & Hatano, G. (Eds.). (1999). Learning to read and write: A cross-linguistic perspective.
Cambridge: Cambridge University Press.
Kozminsky, L., & Kozminsky, E. (1995). The eVects of early phonological awareness training on reading
success. Learning and Instruction, 5, 187–201.
Levin, I., Patel, S., Margalit, T., & Barad, N. (2002). Letter names: EVect on letter saying, spelling, and
word recognition in Hebrew. Applied Psycholinguistics, 23, 269–300.
Liberman, A. M., Cooper, F. S., Shankweiler, D., & Studdert-Kennedy, M. (1967). Perception of the speech
code. Psychological Review, 74, 431–461.
Mason, J. M. (1980). When do children begin to read: An exploration of four year old children’s letter and
word reading competencies. Reading Research Quarterly, 15, 203–227.
Mates, B. F., & Strommen, L. (1995/1996). Why Ernie can’t read: Sesame street and literacy. The Reading
Teacher, 49, 300–306.
McBride-Chang, C. (1999). The ABC’s of the ABC: The development of letter-name and letter-sound
knowledge. Merrill–Palmer Quarterly, 45, 285–308.
McBride-Chang, C., & Treiman, R. (2003). Hong Kong Chinese kindergartners learn to read English ana-
lytically. Psychological Science, 14, 138–143.
Naveh, J. (1982). Early history of the alphabet: An introduction to West Semitic epigraphy and palaeography.
Jerusalem: The Magnum Press.
Read, C. (1986). Children’s creative spelling. London: Routledge & Kegan Paul.
Rozin, P., & Gleitman, L. R. (1977). The structure and acquisition of reading II: The reading process and the
acquisition of the alphabetic principle. In A. S. Reber & D. L. Scarborough (Eds.), Toward a psychology
of reading: The proceedings of the CUNY conference (pp. 49–65). Hillsdale, NJ: Lawrence Erlbaum.
Shankweiler, D., & Liberman, I. Y. (1989). Phonology and reading disability: Solving the reading puzzle.
Ann Arbor, MI: University of Michigan Press.
232 D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233
Share, D. L. (1995). Phonological recoding and self-teaching: Sine qua non of reading acquisition. Cogni-
tion, 55, 151–218.
Share, D. L., & Gur, T. (1999). How reading begins: A study of preschoolers’ print identiWcation strategies.
Cognition and Instruction, 17, 177–213.
Share, D. L., Jorm, A. F., Maclean, R., Matthews, R., & Waterman, B. (1983). Early reading achievement,
oral language ability, and a child’s home background. Australian Psychologist, 18, 75–89.
Share, D. L., Jorm, A. F., Maclean, R., & Matthews, R. (1984). Sources of individual diVerences in reading
acquisition. Journal of Educational Psychology, 76, 1309–1324.
Share, D. L., & Levin, I. (1999). Learning to read and write in Hebrew. In M. Harris & G. Hatano (Eds.),
Learning to read and write: A cross-linguistic perspective (pp. 89–111). Cambridge: Cambridge Univer-
sity Press.
Share, D. L., & Stanovich, K. E. (1995). Cognitive processes in early reading development: Accommodat-
ing individual diVerences into a model of acquisition. Issues in Education, 1, 1–57.
Shatil, E., & Share, D. L. (2003). Cognitive antecedents of early reading ability: A test of the modularity
hypothesis. Journal of Experimental Child Psychology, 86, 1–31.
Snow, C. E., Burns, M. S., & GriYn, P. (Eds.). (1998). Preventing reading diYculties in young children.
Washington, DC: National Academy Press.
Snowling, M. J. (1991). Developmental reading disorders. Journal of Child Psychology and Psychiatry, 32,
49–77.
Solberg, S., & Nevo, B. (1979). Preliminary steps towards an Israeli standardization of the Peabody Test (in
Hebrew). Megamot, 3, 407–413.
Stanovich, K. E. (1986). Matthew eVects in reading: Some consequences of individual diVerences in the
acquisition of literacy. Reading Research Quarterly, 4, 360–406.
Stanovich, K. E. (2000). Progress in understanding reading: ScientiWc foundations and new frontiers. New
York: Guilford Press.
Stanovich, K. E., Cunningham, A. E., & Feeman, D. J. (1984). Intelligence, cognitive skills, and early read-
ing progress. Reading Research Quarterly, 19, 278–303.
Stuart, M., & Coltheart, M. (1988). Does reading develop in a sequence of stages? Cognition, 30, 139–181.
Treiman, R. (1992). The role of intrasyllabic units in learning to read and spell. In P. B. Gough, L. C. Ehri,
& R. Treiman (Eds.), Reading acquisition (pp. 65–106). Hillsdale, NJ: Erlbaum.
Treiman, R. (1993). Beginning to spell: A study of Wrst-grade children. New York: Oxford University Press.
Treiman, R. (1994). Use of consonant letter names in beginning spelling. Developmental Psychology, 30,
567–580.
Treiman, R., & Kessler, B. (2003). The role of letter names in the acquisition of literacy. In R. Kail
(Ed.), Advances in child development and behavior (Vol. 31, pp. 105–135). San Diego, CA: Academic
Press.
Treiman, R., & Rodriguez, K. (1999). Young children use letter names in learning to read words. Psycho-
logical Science, 10, 334–338.
Treiman, R., & TincoV, R. (1997). The fragility of the alphabetic principle: Children’s knowledge of letter
names can cause them to spell syllabically rather than alphabetically. Journal of Experimental Child
Psychology, 64, 425–449.
Treiman, R., TincoV, R., & Richmond-Welty, E. D. (1996). Letter names help children to connect print and
speech. Developmental Psychology, 32, 505–514.
Treiman, R., TincoV, R., Rodriguez, K., Mouzaki, A., & Francis, D. J. (1998). The foundations of literacy:
Learning the sounds of letters. Child Development, 69, 1524–1540.
Treiman, R., Weatherston, S., & Berch, D. (1994). The role of letter names in children’s learning of pho-
neme–grapheme relations. Applied Psycholinguistics, 15, 97–122.
Tunmer, W. E., Herriman, M. L., & Nesdale, A. B. (1988). Metalinguistic abilities and beginning reading.
Reading Research Quarterly, 23, 134–158.
Vellutino, F. R., & Scanlon, D. M. (1987). Phonological coding, phonological awareness and reading abil-
ity: Evidence from a longitudinal and experimental study. Merrill–Palmer Quarterly, 33, 321–363.
Venezky, R. (1975). The curious role of letter names in reading instruction. Visible Language, 9, 7–23.
Wagner, R. K., & Torgesen, J. K. (1987). The nature of phonological processing and its causal role in the
acquisition of reading skills. Psychological Bulletin, 101, 192–212.
D.L. Share / Journal of Experimental Child Psychology 88 (2004) 213–233 233
White, K. R. (1982). The relation between socioeconomic status and academic achievement. Psychological
Bulletin, 91, 461–481.
Wilson, F. T., & Flemming, C. W. (1938). Correlations of reading progress with other abilities and traits in
Grade 1. The Journal of Genetic Psychology, 53, 33–52.
Worden, P. E., & Boettcher, W. (1990). Young children’s acquisition of alphabet knowledge. Journal of
Reading Behavior, 22, 277–295.
... For younger students, use premade lists with target words for them to trace or highlight. (Cunningham, 2006;Cunningham et al., 2002;Share, 2004a). Subsequent encounters no longer necessitate laborious decoding, as the printed letter strings are now familiar to the reader. ...
... Print context and exposure to print are important aspects of the self-teaching hypothesis (Li & Wang, 2022;Nation et al., 2006;Ricketts et al., 2011;Share, 2004aShare, , 2004bTamura et al., 2017). The context describes how target words or nonwords are associated with meaning in text, whereas exposure refers to the frequency at which target items are encountered (Share, 2004a). ...
... Print context and exposure to print are important aspects of the self-teaching hypothesis (Li & Wang, 2022;Nation et al., 2006;Ricketts et al., 2011;Share, 2004aShare, , 2004bTamura et al., 2017). The context describes how target words or nonwords are associated with meaning in text, whereas exposure refers to the frequency at which target items are encountered (Share, 2004a). The contributions of context and exposure on children's learning vary by age. ...
Article
Full-text available
Purpose Print knowledge is a powerful predictor of later reading abilities, which are crucial to children's academic success. Children with speech sound disorders (SSDs) are at risk for literacy difficulties. Speech sound therapy is an opportune time to address not only speech sound production but also the connections between speech sounds and their associated letters and letter patterns. This tutorial aims to provide a rich source of evidence to support the use of print and print-referencing in speech sound therapy sessions. We include feasible suggestions for speech-language pathologists (SLPs) to facilitate print-rich experiences for children with SSDs. Incorporating orthography into therapy sessions, through activities such as shared book reading, allows SLPs to address speech sound errors while supporting children's literacy development. Conclusions This tutorial provides an overview of the risk of literacy deficits in children with SSDs and describes the theoretical underpinnings of why incorporating print is a powerful learning tool. We expand on the research behind print referencing in preschool classrooms. Finally, we provide detailed examples, including book lists and therapy ideas for a variety of ages, to help SLPs add print referencing to their therapy toolkit.
... Letter knowledge is one of the strongest predictors of word decoding and spelling from an early age. Previous studies have repeatedly shown that knowledge of the names or sounds of letters in the alphabet helps children isolate specific sounds in the language, understand the relation with the written units that represent them (e.g., see Cardoso-Martins et al., 2011;Share, 2004), and learn letter-sound associations (Treiman & Kessler, 2003). Because letter knowledge supports children's grapheme-phoneme mapping ability, it becomes the driving force of decoding and spelling accuracy during the early years of reading instruction (Piasta & Wagner, 2010;Thompson et al., 2015). ...
... First, letter knowledge can make the child aware of phonemes as critical linguistic units facilitating early mastery of the sublexical grapheme-phoneme mapping strategy sufficient to read and spell any word (Suárez-Coalla et al., 2014). Second, letter knowledge can facilitate the encoding of letters within words and the storage of accurate word representations necessary for accurate reading and spelling (Share, 2004;Suárez-Coalla et al., 2016). ...
... • Letter knowledge and word identification skills predict third-grade children's word spelling outcomes • Word identification skills are uniquely associated to children's word decoding • Receptive vocabulary supports word identification skills • Letter knowledge and word identification skills make independent contributions to reading and spelling in children learning in transparent orthographies • Spanish children in grade three rely more on word-level knowledge (i.e., word identification skills and vocabulary) than on their letter knowledge for both reading and spelling knowledge facilitated learning letter-sound mappings involved in accurate decoding of words (Share, 2004). Because mastering letter-sound mappings is sufficient to read and spell accurately in Spanish, children's letter knowledge may play a long-lasting role in both decoding and spelling outcomes in this orthography. ...
Article
Full-text available
Letter knowledge and word identification are key skills for reading and spelling. Letter knowledge facilitates the application of sublexical letter-sound mappings to decode words. With reading experience, word identification becomes a key lexical skill to support decoding. In transparent orthographies, however, letter knowledge might be an enduring predictor of decoding and spelling, even in children with some reading experience. This study investigated the association of children's sublexical (letter knowledge) and lexical skills (word identification and vocabulary) with word decoding and spelling accuracy in Spanish, which is a transparent orthography. The sample consisted of 117 Spanish-speaking children, aged 8 to 10. Results revealed that (1) letter knowledge and word identification were independently associated with childre's word spelling; (2) word identification was uniquely associated with word decoding; and (3) children's vocabulary level was associated with word identification. The implications of these findings were examined within the framework of reading models and the characteristics of a transparent orthography.
... Organon, Porto Alegre, v. 38, n. 76, jul/dez. 2023 (TREIMAN et al.,1996;TREIMAN et al., 1998;TOGERSEN, 2002), muito embora essa questão seja objeto de debate na literatura, incluindo sua relação com a consciência fonológica (SHARE, 2004 ...
Article
Full-text available
Este artigo focaliza a aquisição da escrita, especificamente o conhecimento sobre o alfabeto, na fase anterior à alfabetização formal que ocorre a partir do 1º ano do Ensino Fundamental, com base em dados obtidos através da observação de crianças do 5º período da Educação Infantil de uma escola pública. A metodologia utilizada foi a da observação participante. O objetivo foi o de descrever e analisar o processo das crianças aprendizes de apropriação do conhecimento da escrita nas atividades propostas pela professora. O comportamento observado indicou que o conhecimento ainda incipiente sobre o alfabeto do português brasileiro, de acordo com as convenções ortográficas estabelecidas, leva a estratégias de memorização típicas desta fase, referida como pré-alfabetização ou holográfica (Ehri; Snowling, 1998).
... Another concern regarding the poor performance of the wholesyllable group in the Sargiani et al. (2022) study, is that all the children participating in the study were familiar with the (Brazilian) Portuguese alphabet and already knew all the names of the 10 consonantal letters and 5 vowel letters used in training. Because knowing letter names helps children learn the corresponding letter sounds (Share, 2004), the phoneme group would have enjoyed a significant advantage in learning the letter-sound correspondences because all the names of the 15 letters taught included the letter sounds (e.g., letter ⟨s⟩, name /εsε/, sound /s/). The phoneme-group, but not the syllable group, thus had the advantage of being trained not only on fewer units but on units that were congruent with their prior literacy experiences. ...
Article
Full-text available
This study addressed four research questions: (1) Does teaching using syllables or using phonemes lead to better progress in beginning reading and spelling? (2) Does the effectiveness of syllabic or phonemic instruction depend on children’s preferences for these units as predicted by Ziegler and Goswami’s (2005) “availability” hypothesis? (3) Do children taught via syllabic consonant–vowel (CV) units spontaneously develop insight into the phonemic basis of an alphabetic writing system, and (4) Do individual differences in reading and spelling gains in phoneme-based instruction depend more on working memory, short-term memory, and Rapid Automatized Naming (RAN) owing to the greater number of units that must be rapidly retrieved and processed? To test these hypotheses, 104 preliterate preschool children were taught to read and spell using an unfamiliar script. Across 14 training sessions, children were taught using either whole CV units, phoneme units, or demiphoneme units. Retention and generalization were evaluated during training and 1 week later. Our results showed that CV-based teaching was found to be significantly and substantially more effective in terms of reading accuracy and speed than teaching via phonemes or demiphonemes. Ziegler and Goswami’s (2005) availability hypothesis was not supported: All groups learned more easily with CV-based instruction regardless of their preferred phonological unit. In addition, some children taught solely via whole-syllable units showed evidence of spontaneously induced insight into the phonemic basis of alphabetic writing. Finally, working-memory, short-term memory, and Rapid Automatized Naming predicted learning via phonemes but not via CV units. We discuss the implications for beginning reading instruction in different languages and writing systems.
... The child's response to the first four words received feedback. In line with research into children's learning of Hebrew, where the consonant-vowel is a natural unit [81][82][83], responses were scored according to a scale used in previous studies: (0) incorrect answer (e.g., for the word 'sir' the child says 'lid' or 'I don't know'); (1) subsyllabic (e.g., for the word 'sir' the child says 'si'); or (2) phoneme (e.g., for the word 'sir' the child says 's'). The average score across the 17 words served as the total phonological awareness score. ...
Article
Full-text available
This study explores the stability and progress of parents' literacy beliefs and home literacy activities and their relationships with their children's early literacy skills in their last year of preschool. Participants were 50 preschool children (M = 61.44 months) and their parents. Data collection sessions occurred in the family home in the fall and spring, with six months between them. At each time point, parents completed questionnaires regarding their beliefs relating to children's literacy development, parents' role in supporting literacy development, and the frequency of home literacy activities. We also evaluated the children's early literacy skills (letter names, letter sounds, word writing, phonological awareness, and motivation for literacy activities). Results revealed overall stability in parents' beliefs between the two time points, an increase in home literacy activities, progress in children's early literacy skills, and greater motivation to engage in early literacy activities. We found a positive relationship between parents' beliefs and home literacy activities in the fall with their children's early literacy skills in the spring. Further, the progress in parents' literacy beliefs between the fall and the spring correlated with their children's progress in early literacy skills, controlling for parents' education and children's age. This study highlights the importance of promoting parents' literacy beliefs and home literacy activities.
Chapter
The use of printed words to capture language is one of the most remarkable inventions of humankind, and learning to read them is one of the most remarkable achievements of individuals. In recent decades, how we learn to read and understand printed text has been studied intensely in genetics, education, psychology, and cognitive science, and both the volume of research papers and breadth of the topics they examine have increased exponentially. Theories of Reading Development collects within a single volume state-of-the-art descriptions of important theories of reading development and disabilities. The included chapters focus on multiple aspects of reading development and are written by leading experts in the field. Each chapter is an independent theoretical review of the topic to which the authors have made a significant contribution and can be enjoyed on its own, or in relation to others in the book. The volume is written for professionals, graduate students, and researchers in education, psychology, and cognitive neuroscience. It can be used either as a core or as a supplementary text in senior undergraduate and graduate education and psychology courses focusing on reading development.
Chapter
This chapter reports findings and provides discussion of both qualitative and quantitative analyses of the data collected from the teaching materials for phonological decoding, and the actual classroom teaching of phonological decoding.
Chapter
This chapter provides a review of literature on the notion of phonological decoding, the learning and the teaching of phonological decoding in L1 and L2 contexts, with the aim of establishing an understanding of the theoretical background and empirical research that underpin the present study.
Chapter
The most authoritative resource for students and researchers, The Cambridge Handbook of Child Language has been thoroughly updated and extended. Enhancements include new chapters on the acquisition of words, processing deficits in children with specific language impairments, and language in children with Williams syndrome, new authors for the bilingualism and autism chapters, a refocused discourse chapter on written narratives, and a new section on reading and reading disorders, cementing the handbook's position as the best study of the subject available. In a wide-ranging survey, language development is traced from prelinguistic infancy to adolescence in typical and atypical contexts; the material is intuitively grouped into six thematic sections, enabling readers to easily find specific in-depth information. With topics as varied as statistical learning, bilingualism, and the neurobiology of reading disorders, this multidisciplinary Handbook is an essential reference for students and researchers in linguistics, psychology, cognitive science, speech pathology, education and anthropology.
Article
Full-text available
Letter Name Knowledge is among the most vital emergent literacy skills that predict future reading and writing among pre-primary children, and teachers are central to learning. Much has been researched on the factors influencing letter name acquisition, including the child's characteristics in the home and school environment. Less research has been done on pre-primary teachers' views on the factors they encounter in enhancing letter name skills in children. This study investigated pre-primary teachers' challenges in enhancing letter-name knowledge among pre-primary learners in the Shinyanga District Council, Shinyanga Region. The study identified challenges pre-primary teachers encounter in enhancing letter-name knowledge. The study also assessed the challenges pre-primary teachers face in enhancing the knowledge of letter names for pre-primary children with the guidance of the Whole-Child Approach. A qualitative approach and phenomenological design were utilised for the study.20 (to the point of saturation) pre-primary teachers were involved in the interview for data collection. The obtained data were analyzed thematically. The findings identified the following challenges: pre-primary children's use of mother tongue, inadequate teaching and learning facilities and instructional materials, lack of teachers' professionalism, large class size and instructional burden, children absenteeism and implementation of pre-primary curriculum. In conclusion, the identified challenges influence pre-primary teachers' effectiveness in enhancing letter name knowledge acquisition for public pre-primary children in the Shinyanga region. The study recommends that the Ministry of Science and Technology (MoEST) create a mother tongue policy for pre-primary children for quality, fair and holistic comprehensive curricula provision in rural areas.
Chapter
Full-text available
Around the world, children embark on learning to read in their home language or writing system. But does their specific language, and how it is written, make a difference to how they learn? How is learning to read English similar to or different from learning in other languages? Is reading alphabetic writing a different challenge from reading syllabic or logographic writing? Learning to Read across Languages and Writing Systems examines these questions across seventeen languages representing the world's different major writing systems. Each chapter highlights the key features of a specific language, exploring research on learning to read, spell, and comprehend it, and on implications for education. The editors' introduction describes the global spread of reading and provides a theoretical framework, including operating principles for learning to read. The editors' final chapter draws conclusions about cross-linguistic universal trends, and the challenges posed by specific languages and writing systems.
Book
In this study, originally published in 1986, Professor Charles Read examines the ways in which pre-school and primary children create spellings - or misspellings, as they appear to be. He focuses on the hidden phonetic bases for some frequent patterns in young children's spelling, both in and out of school. Professor Read examines children's spelling in other languages (Dutch, French, Spanish) as well as in various dialects of English, in order to see the influence of other sets of speech sounds and other standard spelling systems. Overall, the evidence suggests that children are very much affected by phonetic characteristics, as they tend to spell alike certain classes of speech sound which are indeed phonetically similar. In devising spellings, children can be remarkably independent and inventive in an activity which for adults is anything but creative, and Professor Read suggests ways in which educators can build upon this creativity.
Article
A careful analysis of 10 Sesame Street episodes reveals that the literacy messages conveyed on the program do not reflect current understanding of the process of becoming literate.
Article
Leaders of political parties often have to choose between conflicting objectives, such as influence on policy, control of the government, and support among the voters. This book examines the behaviour of political parties in situations where they experience conflict between two or more important objectives. The volume contains a theoretical introduction and case studies of party leaders in Germany, Italy, France and Spain as well as six smaller European democracies. Each case focuses on the behaviour of one of several parties in situations of goal conflict, such as the 'historic compromise' in Italy, the 1982 Wende in West Germany, the making of the new Swedish constitution in the 1970s, and the termination of the Austrian 'black-red' grand coalition. In their conclusions, the editors discuss how such leadership decisions can be understood and examine the causes of different choices among party leaders.