951585

research-article2020
CDPXXX10.1177/0963721420951585TreimanLearning to Write Words

ASSOCIATION FOR
PSYCHOLOGICAL SCIENCE
Current Directions in Psychological
Science
Learning to Write Words 2020, Vol. 29(5) 521–526
© The Author(s) 2020
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/0963721420951585
https://doi.org/10.1177/0963721420951585
www.psychologicalscience.org/CDPS
Rebecca Treiman
Department of Psychological and Brain Sciences, Washington University in St. Louis

Abstract
Learning to produce the written forms of individual words is an important part of writing. In this article, I review
research on how children acquire this skill. I begin by discussing young children’s knowledge about the visual
appearance of writing and then consider how learners of alphabetic writing systems begin to use letters to symbolize
the sounds they hear in words. The English writing system, the focus of this review, is complex. In the final section of
the article, I discuss how older children learn about its subtler patterns. Implications of the research for how children
learn and for how spelling should be taught are considered.

Keywords
spelling, writing, child development

Learning to write is important for success in today’s For example, the writings were, on average, smaller than
world, and spelling is an important part of writing. the drawings. In addition, the children were more likely
Parents and teachers have many questions about how to choose a pen or a pencil, as opposed to a colored
children acquire this skill. For example, the toddler who crayon, for writing. Before children’s writing includes
produced the scribble shown in Figure 1 identified the identifiable letters, these results suggest, it may have
part with small squiggles near the bottom left as writ- some characteristics that are typical of writing.
ing. Could a child so young have learned that writing In another study, Chinese 2- to 5-year-olds were
is usually small and dense? As another example, a asked to write the same words as in the above-described
5-year-old might draw some shapes and label them as study with U.S. children, and adults who knew both
shown in Figure 2. Why would a child produce spelling English and Chinese were asked to judge whether each
errors such as these? Does the reversal of the ‹d› in the production was made by a U.S. or a Chinese child
spelling of “diamond” mean that the child may have (Otake, Treiman, & Yin, 2018). The bottom row of Fig-
serious difficulties in learning to read and spell— ure 3 shows some of the Chinese children’s writings.
dyslexia? My goal in this article is to review the research The adults performed above the level of random guess-
on how children learn to write words and consider its ing in determining the language of the writer, even with
implications. the productions of the 2- and 3-year-olds. This result
suggests that the productions contained some cues to
Children’s Early Knowledge About the writing system to which the child had been exposed.
A Chinese child’s production might not be identifiable
Writing and Spelling
as a specific character, for example, but it might have
Modern children see many examples of writing, and the squarish shape that is typical of Chinese characters.
they start to learn about its appearance from an early Interestingly, the adults found it easier to identify the
age. In one study, Otake, Treiman, and Yin (2017) asked writer’s language when children wrote their own name
U.S. children between the ages of 2 years 0 months and than when they wrote another word. This difference
5 years 5 months to write four words and draw pictures suggests that youngsters write their names in a more
of the corresponding objects. The top row of Figure 3
shows several U.S. children’s drawings of the sun, and
the middle row of Figure 3 shows some writings of the Corresponding Author:
Rebecca Treiman, Washington University in St. Louis, Department of
word “sun.” Although one might label all of these Psychological and Brain Sciences, Campus Box 1125, 1 Brookings Dr.,
productions as scribbles, there were some differences St. Louis, MO 63130
between the writing scribbles and the drawing scribbles. E-mail: [email protected]
522 Treiman

some systems and vertically in others. Evidence for

children’s learning about spatial arrangement comes
from a study in which 3- to 5-year-old U.S. nonreaders
were shown pairs of displays with different spatial
arrangements (Treiman, Mulqueeny, & Kessler, 2015).
Some pairs contrasted horizontally arranged letters, as
in Figure 4a, with randomly scattered letters, as in Figure
4d. Other pairs had other contrasts, such as vertical (Fig.
4b) and diagonal (Fig. 4c). When horizontally arranged
letters were paired with other types of arrangements,
the children picked the horizontal arrangement signifi-
cantly more often than expected by chance. Thus, the
children seemed to have some knowledge about a lan-
guage-specific characteristic of their writing system.

Representing the Sounds in Words

A U.S. child of 4 or 5 years may produce a roughly
Fig. 1. Production made by a 2.5-year-old U.S. child.
horizontal string of recognizable letters when asked to
write a word. However, the letters may not make sense
advanced way than other words (Levin, Both-de Vries, given the phonemes (sounds) in the word: the phonol-
Aram, & Bus, 2005). ogy. For example, a child may write ‹OAB› for “dirt.”
Young children learn not only about the general Such a child may perform better when asked to write
characteristics of their writing system’s symbols but also “deal,” producing a spelling such as ‹DIT› that at least
about their typical spatial arrangement. This differs begins with a phonologically plausible letter. Children’s
across languages. Symbols are arranged horizontally in better performance on “deal” reflects the fact that its
first two sounds form the name of the letter “d.” Chil-
dren who are familiar with letter names, as many U.S.
preschoolers are, sometimes use a letter to symbolize
all the sounds in its name (Treiman & Wolter, 2020).
Knowledge of letter names, therefore, can help children
produce spellings that are at least partly phonologically
plausible.
Children who know the spelling of their own name,
as many preschoolers do, can use this knowledge to
move toward plausible spellings of other words. For
example, Ben is likely to be good at using ‹b›, the first
letter of his name, in a phonologically appropriate man-
ner. This allows him to produce better quality spellings
of words such as “button” and “bear” than of words that
do not contain the name of the letter “b” (Zhang &
Treiman, 2020).
Results such as those just described indicate that
children do not move neatly from a stage or phase of
literacy development during which they use random
letters to spell words (see Ehri, 2015; Gentry, 1982) to
a period during which they represent some sounds in
words with phonologically appropriate letters. A child
may produce spellings that vary in phonological plau-
sibility at a given time, depending on the properties of
the words.
Spellings such as ‹DIT› for “deal” and ‹b› for “button”
Fig. 2. Typical spelling errors of beginning writers when asked to represent the first phoneme of the word with a plausible
label the names of shapes they have drawn. letter, but a parent or teacher who saw the spellings out
Learning to Write Words 523

“Draw the sun”

U.S. Child U.S. Child U.S. Child

(2 years 0 months) (2 years 5 months) (3 years 4 months)

“Write sun”

U.S. Child U.S. Child U.S. Child

(2 years 0 months) (2 years 8 months) (3 years 2 months)

“Write ri ”

Chinese Child Chinese Child Chinese Child

(2 years 7 months) (2 years 8 months) (3 years 3 months)
Fig. 3. Sample figures produced by preschool children when asked to draw the sun, write
the word “sun” (English), and write the word [rì] (Chinese).

of context would not know what word was meant. The such as “triangle” and “star,” producing spellings like
spellings in Figure 2 communicate better in that they ‹chiego› and ‹SR›. These spellings reflect children’s dif-
represent more of the sounds in the words. The child’s ficulty in breaking spoken words into phonemes and
use of ‹chR› at the beginning of “triangle” may seem odd, thinking about and manipulating the phonemes: pho-
but the first sound of “triangle” is similar to the sound nemic awareness. Consonant clusters such as those at
that is conventionally spelled with ‹ch›, as in “chin.” The the beginning of “triangle” and “star” are especially
child’s spelling shows a recognition of this similarity likely to be treated as units (Treiman, 1991). The omis-
(Read, 1975) as well as some confusion about the proper sion of the vowel in spellings such as ‹SDR› and ‹SR› for
use of uppercase and lowercase letters. Another letter “star” reflects the use of ‹R› to represent both the vowel
choice in Figure 2 that may seem odd to adults is the sound and the [r] in the letter’s name. These spellings
‹d› in the child’s spelling of “star.” But this word’s second are another example of how children use knowledge of
sound is similar to [d], and the use of ‹d› reflects this letter names, if they possess it, to construct spellings.
(Hannam, Fraser, & Byrne, 2007). Other beginning spell- Instruction that systematically promotes phonemic
ers omit the second consonants when spelling words awareness and knowledge of correspondences between
524 Treiman

a b
Horizontal Vertical

c Diagonal
d Scattered

Fig. 4. Displays with letters presented (a) horizontally, (b) vertically, (c) diagonally,
and (d) scattered, as shown to U.S. 3- to 5-year-olds in a study by Treiman, Mulqueeny,
and Kessler (2015).

sounds and letters, phonics, helps children learn to spell spelling of “diamond” shown in Figure 2, may be dys-
and read (Ehri, Nunes, Stahl, & Willows, 2001). Such lexic. But letter reversals are not uncommon among
instruction is most effective when teachers understand typically developing young children. Indeed, the left-
the reasons behind children’s errors and respond right reversal of ‹d› in Figure 2 suggests that the child
accordingly. For example, a teacher can tell a 6-year-old has learned that letters of the Latin alphabet are more
who wrote “trick” as ‹chrik› that he did a good job of likely to have a vertical stem and an appendage that
listening to the word and writing the sounds he heard. faces right, going in the direction of writing, than a
The teacher can explain, though, that the first sound of vertical stem and a left-facing appendage. A child who
“trick” also sounds similar to [t], as in “tick,” and that writes ‹d› as ‹b› is making the shape conform to the pat-
we use ‹t› to write it. A teacher might not tell a child of tern shown by letter shapes such as ‹h› and ‹E› (Treiman,
this age that [k] is normally spelled as ‹ck› when it Gordon, Boada, Peterson, & Pennington, 2014).
occurs after a single vowel letter in a one-syllable word,
although this might be covered in a lesson for older
students. Unfortunately, many U.S. teachers do not have
Beyond Simple Sound–Letter Mappings
sufficient opportunities during their training to learn Learners of some writing systems, such as Spanish, will
about the structure of language and how to use it to spell most words correctly if they can analyze spoken
inform spelling instruction (Carreker, Joshi, & Boulware- words into phonemes and if they know the letter that
Gooden, 2010). is most often used to spell each phoneme. The English
Parents and teachers sometimes worry that children writing system, however, is complex. Many sounds have
who reverse letters, as with the use of ‹b› for ‹d› in the more than one possible spelling. Some sound-to-letter
Learning to Write Words 525

correspondences are less common than others, but However, even adults have not fully internalized the
even a less common correspondence may be systemati- morphological patterns in the English writing system
cally used in certain contexts. For example, [k] is almost (Treiman, Wolter, & Kessler, 2020).
always spelled as ‹ck› when it follows a single vowel Is it worthwhile to teach older students about con-
letter at the end of a one-syllable word (e.g., “trick,” text-dependent spelling patterns and about how spell-
“back”). This same sound is almost always spelled as ing is influenced by morphology and etymology?
‹k› when it occurs at the beginning of a word before ‹e› Students learn some of these things on their own
or ‹i› (e.g., “key,” “kitten”). In such cases, the context that through reading. This process is slow, however, and
is relevant to the spelling of a phoneme is its position in adults do not take full advantage of all of the regulari-
the word’s spelling and the spelling of an adjacent letter. ties that could help their spelling. Explicit spelling
In other cases, the identity of the surrounding sounds is instruction can help, and research shows that it leads
more decisive. For example, [ɑ] is spelled as ‹a› in “wad” to larger spelling gains than informal or indirect
and “squash” rather than with the ‹o› that appears in approaches (Graham & Santangelo, 2014). Formal spell-
“pod,” “gosh,” and many other words. The presence of a ing instruction is typically limited to the primary grades
preceding [w] influences the spelling of [ɑ]. in U.S. schools, but it is effective with older students
Children who receive phonics instruction are taught as well.
common sound–letter correspondences, such as the Improving students’ knowledge of spelling is benefi-
correspondence between ‹f› and [f]. They may be taught cial in several ways. For one, it allows those who read
that some sounds have more than one possible spelling, their writing to avoid the confusion that might occur
and they may be taught that some options (e.g., ‹f› for when “prophet,” for example, has been misspelled as
[f]) are more common than others (e.g., ‹ph› for [f]). “profit.” (Most spell-checkers would not catch this
Often, however, children are not taught how context error.) But learning the conventional spellings of words
can help in selecting the correct spelling. Children pick is not just a courtesy for the reader. Learning to spell
up some context-dependent patterns through exposure fosters students’ phonemic awareness and improves
to words in their reading materials, but this process can their reading, underlining the importance of spelling
take many years (Hayes, Treiman, & Kessler, 2006; for literacy in general (Graham & Santangelo, 2014;
Treiman & Kessler, 2006). Ouellette, Martin-Chang, & Rossi, 2017).
The discussion of context effects so far has empha-
sized surrounding letters and sounds as influences on
the spelling of a sound. In other cases, what is impor- Conclusions
tant for spelling is whether a sound or sound sequence Learning to write words is a long process. Even adult
is a morpheme: a unit of meaning. For example, the users of English have difficulty spelling some words
final [t] in “rapped” is a separate morpheme from “rap.” that they know how to read. The research on spelling
English generally spells the past-tense morpheme as reviewed in this article sheds light on learning, showing
‹ed› regardless of its pronunciation. Thus, “rapped” is how children use what they know—including the
not spelled with the final ‹t› that is used for “rapt.” This names of letters, the spelling of their name, and the
is one of many examples of how English spelling is similarities among sounds—to learn new things. The
influenced by morphology (the structure of words and research findings also show that, although implicit
word parts) or etymology (the history of words). Becom- learning plays a role in spelling development, explicit
ing a good speller requires learning about these influ- teaching is also important.
ences. Younger students will need to write some words
that contain more than one morpheme, such as
“jumped,” and words with multiple morphemes are Recommended Reading
increasingly important for older students. Many of the Kemp, N. (2016). Children’s first language acquisition of the
morphologically complex words that older students English writing system. In V. Cook & D. Ryan (Eds.),
encounter come from Latin or classical Greek. These Routledge handbook of the English writing system (pp.
words have some different spelling patterns than words 193–206). Abingdon, England: Routledge. An accessible
in the basic Germanic vocabulary of English. For exam- chapter about spelling development in English.
Read, C., & Treiman, R. (2013). Children’s invented spell-
ple, [f] is more likely to be spelled as ‹ph› in words that
ing: What we have learned in forty years. In M. Piattelli-
derive from Greek (e.g., “phase”) than in basic words Palmarini & R. C. Berwick (Eds.), Rich languages from
(e.g., “foot”). University students are sensitive to some poor inputs (pp. 197–211). Oxford, England: Oxford
of the spelling differences between words that derive University Press. An overview of research on children’s
from Latin and Greek and words that do not, even early spelling.
though these matters are not usually covered systemati- Tolchinsky, L. (2003). The cradle of culture and what children
cally in schools (Treiman, Decker, & Kessler, 2019). know about writing and numbers before being taught.
526 Treiman

Mahwah, NJ: Erlbaum. An interesting discussion of young Hayes, H., Treiman, R., & Kessler, B. (2006). Children
children’s knowledge about writing and other symbolic use vowels to help them spell consonants. Journal of
systems. Experimental Child Psychology, 94, 27–42. doi:10.1016/j
Treiman, R., & Kessler, B. (2014). How children learn to write .jecp.2005.11.001
words. Oxford, England: Oxford University Press. A com- Levin, I., Both-de Vries, A., Aram, D., & Bus, A. (2005).
prehensive work on writing systems and how children Writing starts with own name writing: From scribbling
learn to use them. to conventional spelling in Israeli and Dutch children.
Applied Psycholinguistics, 26, 463–477. doi:10.1017/
Transparency S0142716405050253
Otake, S., Treiman, R., & Yin, L. (2017). Differentiation
Action Editor: Robert L. Goldstone
of writing and drawing by U.S. two- to five-year-olds.
Editor: Robert L. Goldstone
Cognitive Development, 43, 119–128. doi:10.1016/j.cog
Declaration of Conflicting Interests
dev.2017.03.004
The author(s) declared that there were no conflicts of
Otake, S., Treiman, R., & Yin, L. (2018). Preschoolers’ knowl-
interest with respect to the authorship or the publication
edge about language-specific properties of writing.
of this article.
British Journal of Developmental Psychology, 36, 667–672.
doi:10.1111/bjdp.12240
ORCID iD Ouellette, G., Martin-Chang, S., & Rossi, M. (2017). Learning
Rebecca Treiman https://orcid.org/0000-0002-6819-8338 from our mistakes: Improvements in spelling lead to
gains in reading speed. Scientific Studies of Reading, 21,
Acknowledgments 350–357. doi:10.1080/10888438.2017.1306064
Read, C. (1975). Childrens’ categorization of speech sounds in
I thank the members of the Reading and Language Lab for
English. Urbana, IL: National Council of Teachers of English.
help with the figures and for comments on a draft of the
Treiman, R. (1991). Children’s spelling errors on syllable-ini-
manuscript.
tial consonant clusters. Journal of Educational Psychology,
83, 346–360. doi:10.1037/0022-0663.83.3.346
References Treiman, R., Decker, K., & Kessler, B. (2019). Adults’ sensitiv-
Carreker, S., Joshi, R. M., & Boulware-Gooden, R. (2010). ity to graphotactic differences within the English vocabu-
Spelling-related teacher knowledge: The impact of lary. Applied Psycholinguistics, 40, 167–182. doi:10.1017/
professional development on identifying appropriate S0142716418000516
instructional activities. Learning Disability Quarterly, 33, Treiman, R., Gordon, J., Boada, R., Peterson, R. L., &
148–158. doi:10.1177/073194871003300304 Pennington, B. F. (2014). Statistical learning, letter rever-
Ehri, L. C. (2015). How children learn to read words. In sals, and reading. Scientific Studies of Reading, 18, 383–
A. Pollatsek & R. Treiman (Eds.), The Oxford handbook 394. doi:10.1080/10888438.2013.873937
of reading (pp. 293–310). Oxford, England: Oxford Treiman, R., & Kessler, B. (2006). Spelling as statistical learn-
University Press. ing: Using consonantal context to spell vowels. Journal of
Ehri, L. C., Nunes, S. R., Stahl, S. A., & Willows, D. M. (2001). Educational Psychology, 98, 642–652. doi:10.1037/0022-
Systematic phonics instruction helps students learn to 0663.98.3.642
read: Evidence from the National Reading Panel’s meta- Treiman, R., Mulqueeny, K., & Kessler, B. (2015). Young
analysis. Review of Educational Research, 71, 393–447. children’s knowledge about the spatial layout of writing.
doi:10.3102/00346543071003393 Writing Systems Research, 7, 235–244. doi:10.1080/1758
Gentry, J. R. (1982). An analysis of developmental spelling 6801.2014.924386
in “GNYS AT WRK.” The Reading Teacher, 36, 192– Treiman, R., & Wolter, S. (2020). Use of letter names ben-
200. efits young children’s spelling. Psychological Science, 31,
Graham, S., & Santangelo, T. (2014). Does spelling instruc- 43–50. doi:10.1177/0956797619888837
tion make students better spellers, readers, and writ- Treiman, R., Wolter, S., & Kessler, B. (2020). How sensi-
ers? A meta-analytic review. Reading and Writing: An tive are adults to the role of morphology in spelling?
Interdisciplinary Journal, 27, 1703–1743. doi:10.1007/ Morphology. Advance online publication. doi:10.1007/
s11145-014-9517-0 s11525-020-09356-4
Hannam, R., Fraser, H., & Byrne, B. (2007). The sbelling of Zhang, L., & Treiman, R. (2020). Learning to spell phono-
sdops: Preliterate children’s spelling of stops after /s/. logically: Influences of children’s own names. Scientific
Reading and Writing: An Interdisciplinary Journal, 20, Studies of Reading, 24, 229–240. doi:10.1080/10888438.
399–412. doi:10.1007/s11145-006-9038-6 2019.1662027