Learning of letter names and sounds and their contribution to word recognition

https://doi.org/10.1016/j.jecp.2005.08.002Get rights and content

Abstract

This study investigated knowledge of letter names and letter sounds, their learning, and their contributions to word recognition. Of 123 preschoolers examined on letter knowledge, 65 underwent training on both letter names and letter sounds in a counterbalanced order. Prior to training, children were more advanced in associating letters with their names than with their sounds and could provide the sound of a letter only if they could name it. However, children learned more easily to associate letters with sounds than with names. Training just on names improved performance on sounds, but the sounds produced were extended (CV) rather than phonemic. Learning sounds facilitated later learning of the same letters’ names, but not vice versa. Training either on names or on sounds improved word recognition and explanation of printed words. Results are discussed with reference to cognitive and societal factors affecting letter knowledge acquisition, features of the Hebrew alphabet and orthography, and educational implications.

Introduction

Letter knowledge among prekindergartners and kindergartners is one of the best predictors of reading and spelling acquisition later in school. This holds true in English, French, Dutch, Brazilian Portuguese, and Hebrew (e.g., Adams, 1990, Ball and Blachman, 1991, Bradley and Bryant, 1983, Bruck et al., 1997, Byrne and Fielding-Barnsley, 1989, de Jong and van der Leij, 1999, Muter et al., 1998). Letter knowledge as a predictor not only surpasses IQ and vocabulary (e.g., Caravolas et al., 2001, Cardoso-Martins, 1995, McBride-Chang, 1999, Shatil et al., 2000, Stuart and Coltheart, 1988) but at times also competes successfully with some tests of phonological awareness (e.g., Johnston, Anderson, & Holligan, 1996).

Letter knowledge constitutes a system of information integrated by rules governing the alphabetic code underlying the written system. It involves recognizing more than 20 shapes of letters, each in fact a family of shapes sharing prototypical features (Ehri, 1986), and grasping that these shapes belong to a distinct category of graphic notations (e.g., Brenneman et al., 1996, Callaghan, 1999, Levin and Bus, 2003, Teubal and Dockrell, 2005, Tolchinsky-Landsmann and Karmiloff-Smith, 1992). Letter knowledge includes mastering letter names and comprehending that they form a class of labels of letters. Furthermore, letter knowledge involves connecting each letter shape with its name as well as with one or more sounds for which it stands in written words.

Studies that have used letter knowledge as a predictor of reading or spelling measured it either by asking children to name letters (e.g., Bruck et al., 1997), by asking them to name letters as well as provide their sounds (e.g., Caravolas et al., 2001), or by using a combined score counting letters as known if children either named them or provided their sounds (e.g., Cardoso-Martins, 1995, Johnston et al., 1996). However, the relations between the constructs measured by naming letters and by providing letter sounds remain unclear.

Currently, many researchers and educators believe that letter knowledge should be enhanced in prekindergarten, or at the latest in kindergarten, prior to entry into school where children are formally taught to read and spell. However, no agreement exists about how this enhancement should be accomplished in terms of timing and sequencing. Reviews of curricula for kindergartners in North America (e.g., California State Board of Education, n.d., Ontario Ministry of Education, 1998) have indicated that children are expected to recognize letters, name all or most of them, and become familiar with the sounds of some of them prior to school entry. In the United Kingdom, teaching about letters starts in prekindergarten with their sounds. Letter names are not formally introduced prior to first grade (Caravolas et al., 2001; R.S. Johnston, personal communication). In Israel, the opposite situation prevails; kindergartners are taught only letter names, whereas the sounds are left to be taught in school along with learning to read and write. Curriculum developers who aim to design evidence-supported pedagogical methods must learn more about the developmental relations between naming letters and knowing their sounds and about how this knowledge affects reading and spelling acquisition.

The current study addressed five questions:

  • 1.

    Which is easier for children to learn: Letter names or letter sounds? (ease of learning)

  • 2.

    Does learning of letter names immediately transfer to performance on letter sounds and vice versa, and if so, which transfer is higher: name to sound or sound to name? (transfer)

  • 3.

    Does learning of letter names facilitate later learning of letter sounds and vice versa, and if so, which facilitation is stronger: name to sound or sound to name? (facilitation)

  • 4.

    Do letter names and letter sounds support children in word recognition, and which is more productive in this respect? (word recognition)

  • 5.

    Does the order of learning letter names and letter sounds affect word recognition after both names and sounds have been learned? (sequence)

Previous studies have reported preliminary answers to some of these questions.

It is often assumed that letter names are easier to learn than letter sounds (e.g., Evans et al., 2004, McBride-Chang, 1999, Share, 2004). This assumption derives from the frequent finding that kindergartners are more advanced at naming letters than at providing their sounds. This finding has emerged in countries with differing educational systems, orthographies, and alphabets such as Brazil, Canada, Israel, France, and the United States (Burgess and Lonigan, 1998, de Abreu and Cardoso-Martins, 1998, Ehri and Wilce, 1985, Evans et al., 2004, Hohn and Ehri, 1983, Levin and Aram, 2004, Mason, 1980, McBride-Chang, 1999, Treiman and Kessler, 2003, Treiman and Rodriguez, 1999, Treiman et al., 1996, Treiman et al., 1998, Worden and Boettcher, 1990). This developmental sequence has received both cognitive and societal explanations.

According to the cognitive explanation, children can more easily associate a graphic shape with a name that is syllabic (e.g., /εs/) than with a single phoneme (e.g., /s/), which is harder to discriminate and produce (Treiman & Kessler, 2003). This difficulty may be reinforced by the fact that lexical items rarely are single consonants; therefore, children do not gain experience with production of such sounds to help them develop the sensitivity required. Consequently, children find it natural to learn the names of letter shapes, whereas referring to letters by phonemes seems rather unnatural (McBride-Chang, 1999).

According to the societal explanation, communication with children about letters most often refers to letters by their names, not by sounds. This may occur for several reasons. First, adults naturally call letters by their names in their daily communication about letters, and they may tend to do the same with children. Second, a one-to-one correspondence exists between the shapes of letters and their names, whereas the correspondence between letters’ shapes and letters’ sounds is not as unique. Many alphabets include pairs (and rarely triads) of homophonous letters that stand for the same sound (e.g., /k/ marked by C or K in English, /t/ marked by Tet or Taf in Hebrew). Likewise, in many orthographies, a letter shape can stand for several sounds (e.g., C stands for /k/ or /s/ in English, Kaf stands for /k/ or /x/ in Hebrew). In contrast, each letter shape has only one name, which remains constant (ignoring minor variations such as Zed and Zi in English or Jud and Jod in Hebrew). Consequently, the name is a unique designator of a letter’s identity, whereas its sound is less so. Hence, when helping children to recognize written letters or words, adults may prefer to be clear and use letters’ names rather than their sounds.

According to a restricted version of this claim, adults use letter names from very early on and introduce letter sounds only later, when children are sufficiently advanced to invent spellings or to attempt decoding of words. However, British studies have found prekindergartners and kindergartners to be more advanced on letter sounds than on letter names (Caravolas et al., 2001; R.S. Johnston, personal communication [based on Johnston et al., 1996]). This surprising finding can be explained by the U.K. curriculum that instructs prekindergartners in letter sounds rather than in letter names. These data suggest that curriculum may strongly affect relative performance on letter names versus letter sounds. Curriculum may be so effective due to current intensive literacy instruction from prekindergarten onward and because pedagogical assumptions underlying curricula can influence parental communication about and teaching of letters and words. However, the seemingly different developmental sequence in the United Kingdom than in several other countries does not clarify whether it is inherently easier for children to acquire letter names than letter sounds due to cognitive reasons. In the current study, children underwent training on names and sounds of letters to compare how easily they can learn these two skills.

In alphabetic orthographies, the code underlying the spelling of words is based on grapho-phonemic mapping; each letter (and at times more than one letter) stands for a phoneme. Consequently, knowledge of letter sounds should help children to decode words. Indeed, it has been claimed that teaching letter names rather than letter sounds prior to reading is useless or even harmful (Feitelson, 1965). However, knowledge of letter names in addition to knowledge of their sounds has frequently emerged as a powerful predictor of reading or spelling. Several reasons may be offered for the predictive and possibly contributive role of letter names that, together, may explain this high prediction (see also Cardoso-Martins et al., 2002, Foulin, 2005, Share, 2004). First, parents who teach their children letter names in preschool may be the same parents who later assist their children in learning to read and write in school. If so, the link between naming letters in preschool or kindergarten and reading acquisition in school is not causal. Second, letter name knowledge may facilitate the development of phonological awareness because letters in printed words guide children to treat spoken words as segmentable into sounds and the names suggest sound values. This suggestion complements, rather than contradicts, the reverse claim that phonological processing supports the acquisition of letter knowledge (de Jong and Olson, 2004, Foulin, 2005). Phonological awareness in its own right is closely related to reading and spelling acquisition, with researchers claiming these relations to be reciprocal (e.g., Goswami, 1999, Goswami and Bryant, 1990, Perfetti, 1992, Wagner and Torgesen, 1987; but see also Castle & Coltheart, 2004). Moreover, letter knowledge and phonological awareness may interact in enhancing reading and spelling acquisition. Intervention programs that enhanced phonological awareness were more productive in promoting reading acquisition if they used letters as part of their training procedures (e.g., Wagner & Torgesen, 1987).

Third, mastery of letter names may help children to bridge the gap between spoken words and written words (Ehri, 1983, Levin et al., 2002, Treiman and Kessler, 2003, Treiman et al., 1996). In some words, one or more letter names can be heard (e.g., bead). Encounters with such printed words that are read to them may alert children to the sound value of letters. This may be a preliminary step in their understanding of the alphabetic principle, that is, that graphemes stand for phonemes. Some alphabets and written systems, however, are richer is such words. In Brazilian Portuguese, letter names are more often heard in the pronunciation of words than is the case in English (Cardoso-Martins et al., 2002, Pollo et al., 2005), whereas in Hebrew such occurrences are rare (Levin et al., 2002). Nevertheless, this factor might not be crucial, as indicated by findings that Hebrew-speaking children who hardly hear letter names in words use these names in their early attempts to read words (Levin et al., 2002) and that letter naming predicts reading–spelling acquisition in Hebrew (Shatil et al., 2000).

Finally, letter names may facilitate learning of letter sounds. In line with this view, Ehri (1983) found that children were more successful in learning the sounds of those letters whose names they already knew. Researchers have attributed name-to-sound facilitation to the fact that a letter’s name often includes a sound for which it stands (Cardoso-Martins et al., 2002, Ehri, 1986, Treiman and Kessler, 2003). In English, most letter names are composed of two phonemes, one of which comprises the relevant sound, either the first (e.g., /b/ in the letter name /bi/ for B) or the last (e.g., /l/ in the letter name /el/ for L). Yet several English letters (e.g., G, H) do not contain the predominant sound that the letters mark. Researchers have used this variation to support the idea that children’s learning of letter sounds is facilitated by the letters’ names. They showed that young children’s proficiency at providing the sounds of letters decreased from letters whose names begin with the relevant sounds, to letters whose names end with the relevant sounds, and further to letters that do not include the relevant sounds (Evans et al., 2004, McBride-Chang, 1999, Treiman and Kessler, 2003, Treiman et al., 1998). Similarly, in a training study of letter sounds, children’s improvement on letter sounds depended on whether the names of the letters included the sounds at the beginning, at the end, or not at all—decreasing from the former to the latter (Treiman et al., 1998).

The difference among letter types in name-to-sound facilitation does not guarantee that other features do not distinguish among these three types of letters. To rigorously examine the name-to-sound facilitation hypothesis, Share (2004) controlled both for letters’ features and for children’s prior letter knowledge. Israeli Hebrew-speaking children were trained to connect six graphic symbols with six names and then to connect these symbols with phonemic sounds. Share used symbols and names that do not belong to the Hebrew alphabet. Children were better at learning to connect symbols with sounds when these sounds appeared in the names that they learned for these symbols than when the sounds did not appear in the names.

Studies that support name-to-sound facilitation remain mute on the possibility of sound-to-name facilitation. It can be argued, however, that a correspondence between letters’ names and letters’ sounds can help children who know the sounds to learn to connect these letters with their names. In a longitudinal study lasting from the beginning of kindergarten to the middle of first grade (McBride-Chang, 1999), letter naming at each point in time was predicted by previous letter naming knowledge but not by previous letter sound knowledge. In contrast, letter sound knowledge was predicted both by previous letter sound knowledge and by previous letter naming knowledge. These findings suggest an asymmetry in facilitation; names help children to learn sounds, but sounds have no effect on learning names. This intriguing conclusion calls for further examination due to the inherent limitations in deriving a causal interpretation from correlational analyses. Our study, therefore, analyzed both name-to-sound and sound-to-name facilitation.

Prereaders who do not know the names of letters can be surprisingly good at recognizing certain written words such as their own names (e.g., Villaume & Wilson, 1989), names of classmates (Share & Gur, 1999), and commercial print (e.g., Masonheimer, Drum, & Ehri, 1984). These children’s reading of words is based on visual cues, be they visual elements in the letters (e.g., the tail in g cueing the reading of dog) or the context in which the words are situated (e.g., the arch cueing the reading of McDonald’s). This type of word reading is constrained to familiar words, is limited in number due to memory load, and is prone to errors.

Ehri and Wilce (1985) proposed that after children gain access to a substantial number of letters and recognize them by name or by sound, they become able to learn to recognize words phonetically. These authors examined the shift from processing visual cues to processing phonetic cues by teaching children to read two series of simplified spellings. In the phonetic series, words were spelled with letters that stood for sounds within them, as in JRF for giraffe. In the visual series, the choice of letters was arbitrary, as in WBC for giraffe, but the spellings were visually distinctive by variations in size of letters or position on the line. Prereaders who knew many letter names but few letter sounds learned the visual spellings more easily than they learned the phonetic spellings. Beginning readers who had already mastered letters by name and by sound learned the phonetic spellings more easily. The exact point in development for this shift from visual cue processing to phonetic cue processing appears to depend on prereaders’ level of letter knowledge (Scott & Ehri, 1990), the language of the participants (Cardoso-Martins et al., 2002), and the difficulty of the task (Ross, Treiman, & Bick, 2004).

The crucial question for our purposes concerns the relative difficulty in learning two types of phonetic spellings, deliberately mixed by Ehri and her colleagues in the phonetic series: name spellings that included at least one letter whose name was heard in the word, as in RM for arm, versus sound spellings where no letter name was heard in the word, as in JRF for giraffe. Prereaders with substantial knowledge of letter names and little or just some knowledge of letter sounds learned name spellings (Ross et al., 2004) or sound spellings (Cardoso-Martins et al., 2002) faster than they learned visual spellings. Moreover, children across a wide range of letter knowledge learned name spellings faster than they learned sound spellings (Treiman and Rodriguez, 1999, Treiman et al., 2001).

Studies with Hebrew support the idea that letter names contribute to word reading (Levin et al., 2002). To explain the rationale of the study, and as background to the current work, some explanation is required about the Hebrew alphabet. Letter names in Hebrew are full-sized words, mostly of a CVC or a CVCVC form. Historically, these labels derived from names of objects that were graphically symbolized in Proto-Canaanite pictographs. For instance, the letter named dalεt comes from a label that in modern Hebrew takes the form /dεlεt/ (door). Because of the phonological complexity and length of these names, there are few words that include the entire sequence of sounds comprising a letter name. Given that letter names are rarely heard in spoken words in Hebrew, letter name knowledge might not be as helpful in bridging spoken and written words as it is in Indo-European languages. However, the largely acrophonic character of letter names in Hebrew may prove helpful in supporting reading or spelling acquisition. As described previously, acrophonic English letter names were found to provide useful cues for their sounds to a greater extent than were letter names that ended with the sounds or did not include the sounds.

To examine whether Hebrew’s acrophonic letter names support word recognition, Levin and colleagues (2002) showed kindergartners a series of printed words. For each printed word, children were asked to select which of two spoken words matched the print. Two types of word pairs were used: the name pair, where each of the two words started with a sequence of sounds matching its initial letter’s name, for example, tafkid–pεjrot (role–fruits), which begin with the letters Taf and Pεj, respectively; and the sound pair, where each of the two words started with only one or two sounds of its first letter’s name, for example, talmidpizmon (pupil–pop song), which begin with the letters Taf and Pεj, respectively. Word recognition was higher on name pairs than on sound pairs.

The conclusion drawn from studies of word recognition and word reading is that knowledge of letter names facilitates recognition or reading of words from very early on, even when children are still nonreaders with limited knowledge about letters. However, the reasons for this phenomenon remain unclear. Perhaps children are better able to use cues based on letter names than to use cues based on letter sounds because they recognize more letters by name than by sound. Alternatively, letter names may be more helpful because they are phonologically longer than letter sounds.

We asked whether training on letter names or on letter sounds would be more effective in promoting recognition of printed words. The task required children to choose which of two spoken words was written on a presented card. As mentioned previously, only rarely do Hebrew words comprise name words, that is, words beginning with a series of sounds that form a letter name. The vast majority of Hebrew words constitute sound words, that is, words beginning with the initial sound (C) of the letter’s name. For example, limon (lemon) starts with /l/, the sound that the letter Lamεd represents, but continues with a different vowel in the second position /i/ than that appearing in the letter’s name /a/. However, many Hebrew words do begin with more than one of the sounds that comprise a letter name; therefore, we also included partial name words that start with a CV that matches the beginning of the initial letter’s name. For example, lavan (white) includes the /la/ sound depicting the CV at the beginning the letter named Lamεd. Our design allowed a comparison between partial name (CV) words and sound (C) words to assess the leverage given by letter name to word recognition.

Following the initial screening, children underwent training on names and sounds of letters with the aim of comparing how readily they could learn these two skills. After learning the first skill (either names or sounds), they were tested on both names and sounds to assess immediate transfer from name to sound and from sound to name. Next, children underwent a second training phase during which those who had studied names were trained on the sounds of the same letters and vice versa. After learning the second skill (either names or sounds), children were again tested on both names and sounds to compare name-to-sound and sound-to-name facilitation of learning. Children were also tested on word recognition three times—pretest, Posttest 1 (following the first training phase), and Posttest 2 (following the second training phase)—to compare the effect of learning letter names with that of learning letter sounds on word recognition as well as to examine whether the sequence of training on names and sounds had an impact.

We hypothesized that children would learn names more quickly than they learned sounds, that name-to-sound facilitation of learning would be stronger than sound-to-name facilitation of learning, and that letter name knowledge would enhance word recognition more than would letter sound knowledge. Prior findings seemed to support these assumptions, although competing interpretations were offered previously.

Section snippets

Participants

A total of 123 children, with a mean age of 5 years 1 month (range: 4 years 4 months to 6 years 4 months), were screened for knowledge of letter names and letter sounds in a counterbalanced order. They were sampled from eight prekindergartens and kindergartens in middle-socioeconomic status (SES) neighborhoods. Children were included if their mother tongue was Hebrew, they had no special needs as described by the teacher, and their parents agreed to their participation. The screening was

A priori letter knowledge

A total of 123 children underwent screening that included the letter naming and letter sounding tasks for all 22 letters. Children’s mean percentage correct responses (and standard deviations) on names, extended sounds, and phonemic sounds, respectively, were as follows: M = 36.3 (SD = 33.9), M = 20.9 (SD = 24.2), and M = 8.2 (SD = 19.4). One-way ANOVA followed by Bonferroni tests indicated that children performed highest on naming, second highest on extended sounds, and poorest on phonemic sounds. All of

Discussion

This study has provided both expected and unexpected results, raising new questions. As expected, prior to training, Israeli prekindergartners and kindergartners were substantially more advanced at naming letters than at providing their phonemic sounds, similar to their age mates in various countries (e.g., Brazil, Canada, France, The Netherlands, United States). Likewise, children were slightly more advanced at naming letters even when we accepted extended CV sounds as correct (e.g., /l/,/la/,

Acknowledgments

This study was funded by the Chief Scientist Competition in the Israeli Ministry of Education, granted to Iris Levin.

References (57)

  • E. Teubal et al.

    Children’s developing numerical notations: The impact of input display, numerical size and operational complexity

    Learning and Instruction

    (2005)
  • M.J. Adams

    Beginning to read: Thinking and learning about print

    (1990)
  • E.W. Ball et al.

    Does phoneme awareness training in kindergarten make a difference in early word recognition and developmental spelling?

    Reading Research Quarterly

    (1991)
  • L. Bradley et al.

    Categorizing sounds and learning to read: A causal connection

    Nature

    (1983)
  • M. Bruck et al.

    A cross-linguistic study of early literacy acquisition

  • B. Byrne et al.

    Phonemic awareness and letter knowledge in the child’s acquisition of the alphabetic principle

    Journal of Educational Psychology

    (1989)
  • California State Board of Education. (n.d.). Kindergarten English-language arts content standards. Retrieved November...
  • T. Callaghan

    Early understanding and production of graphic symbols

    Child Development

    (1999)
  • C. Cardoso-Martins

    Sensitivity to rhymes, syllables, and phonemes in literacy acquisition in Portuguese

    Reading Research Quarterly

    (1995)
  • C. Cardoso-Martins et al.

    Letter name knowledge and the ability to learn to read by letter–phoneme relations in words: Evidence from Brazilian–Portuguese-speaking children

    Reading and Writing: An Interdisciplinary Journal

    (2002)
  • A. Castle et al.

    Is there a causal link from phonological awareness to success in learning to read?

    Cognition

    (2004)
  • M. de Abreu et al.

    Alphabetic access route in beginning reading acquisition in Portuguese: The role of letter-name knowledge

    Reading and Writing: An Interdisciplinary Journal

    (1998)
  • P.F. de Jong et al.

    Specific contributions of phonological abilities to early reading acquisition: Results from a Dutch latent variable longitudinal study

    Journal of Educational Psychology

    (1999)
  • L.C. Ehri

    A critique of five studies related to letter-name knowledge and learning to read

  • L.C. Ehri

    Sources of difficulty in learning to spell and read

  • L.C. Ehri

    Reconceptualizing the development of sight word reading and its relationship to recoding

  • L.C. Ehri et al.

    Movement into reading: Is the first stage of printed word learning visual or phonetic?

    Reading Research Quarterly

    (1985)
  • Evans, M.A., Shaw, D., Moretti, S., Page, J. (2004). Letter names, letter sounds, and phonological awareness: An...
  • Cited by (105)

    • Using tablets and apps to enhance emergent literacy skills in young children

      2018, Early Childhood Research Quarterly
      Citation Excerpt :

      Development of emergent literacy skills is important in the preschool years as these skills strongly influence future reading and writing ability (Adams, 1990; Snow, Burns, & Griffin, 1998). The emergent literacy skills young children need to acquire in order to become successful readers and writers include print concepts (Clay, 1998; Justice & Ezell, 2001; Lomax & McGee, 1987), alphabet knowledge (letter sounds and names; Bowman & Treiman, 2004; Foulin, 2005; Levin, Shatil-Carmon, & Asif-Rave, 2006), phonological awareness (Byrne & Fielding-Barnsley, 1989; Mann & Foy, 2003), and emergent writing (Aram & Biron, 2004; Welsch, Sullivan, & Justice, 2003). These are foundational skills upon which children build proficient literacy skills with alphabet knowledge being one of the strongest predictors of future word reading ability (Adams, 1990; Ehri & Roberts, 2006; Molfese, Beswick, Molnar, & Jacobi-Vessels, 2006; Storch & Whitehurst, 2002).

    View all citing articles on Scopus
    View full text