Read, Miscue, and Progress
A Preliminary Study in Characterizing Reading Development in Shallow
Indonesian Orthography
Harwintha Yuhria Anjarningsih
Faculty of Humanities, Universitas Indonesia, Depok, West Java, Indonesia
Keywords: Reading Development, Indonesian Orthography, Syllabic Complexity, Children’s Literacy.
Abstract: Understanding what happens when children learn to read Indonesian is very important, in terms of both
advancing psycholinguistics and improving practices that are done in educational institutions throughout the
country. The current study aimed to characterize the normal development of reading in the under-researched,
shallow Indonesian orthography. A total of eighty-two children aged 7-9 years old participated by reading
aloud 100 words that are of high frequency, monomorphemic, disyllabic, and controlled for syllable structure
(simple, diphthongs, digraphs, and consonant clusters). Reading miscues that were committed by the children
showed that simple disyllabic words were mastered at the end of grade one, and diphthongs, digraphs, and
consonant clusters were mastered later. Results are interpreted based on the predictability of the mapping
between graphemes and phonemes in the Indonesian orthography.
1 INTRODUCTION
Orthographic depth, syllabic complexity, word
length, and use of sub-lexical clusters (e.g., (the st in
the word stop and) and digraphs (e.g., the oe in the
word bloem) are explored in the current investigation
with the general aim to assess how syllabic
complexity influences normal reading development
(see Seymour, Aro, and Erskine, 2003; Zoccolotti, De
Luca, Di Pace, Gasperini, Judica, and Spinelli, 2005;
Marinus and de Jong, 2008). Syllabically, the
Indonesian orthography is very transparent,
predominantly CV with the C being simple
consonants such as that found in the disyllabic word
<guru> (‘teacher’). There are just few exceptions in
the mapping between graphemes (i.e., letters) and
phonemes. What can be considered as exceptions are
diphthongs and digraphs, two letters that are
pronounced as one sound. Diphthongs are <ai>, <oi>,
and <au>. Some examples of digraphs are <ng> /ŋ/,
<ny> /ɳ/, and <sy> /ʃ/. Interestingly, although the
words have the same syllabic structure (e.g., CV-CV),
when the consonants are digraphs or the vowels are
diphthongs, readers see more letters written and thus
words containing digraphs and diphthongs are longer
in length than words that contain simple graphemes.
Not only length, mapping two letters into one sound
may also be challenging for the novice readers.
Therefore, there is possibility that words containing
diphthongs and digraphs present some difficulty for
beginning readers.
From phonetics literature, we know that what
diphthongs and digraphs are. Diphthongs are vowel
sounds that contain a glide from one vowel sound to
another (Roach, 2009). The position of the tongue
moves from one position when producing the first
sound to another position when producing the second
sound. For instance, in the diphthong /ai/ which is
represented by <ai>, the diphthong starts with an
open vowel which is between front and back, and
glides to a closed high vowel. Furthermore, digraphs
are two letters that represent one sound (Robbins,
Kenny, and Robbins 2007). For example, in
Indonesian, the letter <n> maps to an alveolar nasal
consonant, and the letter <g> maps to a velar plosive
consonant. However, when <n> and <g> are in a
digraph, they map to /ŋ/ which is a velar nasal
consonant. Another example, the alveolar nasal
consonant which is represented by the letter <n> and
the palatal approximant represented by the letter <y>
are combined in the digraph <ny> which maps to the
palatal nasal consonant /ɲ/. In practice, when
beginning readers are not reading letter per letter and
are instead mapping diphthongs and digraphs to their
corresponding sound, they will be more successful in
Anjarningsih, H.
Read, Miscue, and Progress - A Preliminary Study in Characterizing Reading Development in Shallow Indonesian Orthography.
DOI: 10.5220/0007176208390843
In Proceedings of the Tenth Conference on Applied Linguistics and the Second English Language Teaching and Technology Conference in collaboration with the First International Conference
on Language, Literature, Culture, and Education (CONAPLIN and ICOLLITE 2017) - Literacy, Culture, and Technology in Language Pedagogy and Use, pages 839-843
ISBN: 978-989-758-332-2
Copyright © 2018 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
839

reading the diphthongs and digraphs. When children
think that the digraph <ng> are individual letters, for
instance, they may read the word bunga as bun-ga,
still preserving the number of syllables but changing
their structure. When children realize that bun-ga
does not map into any word in their lexicon, they start
to grasp that in order to link the concept of FLOWER
in Indonesian to its written rendering, they need to
pronounce the <ng> as /ŋ/. That is, using the lexical
strategy to read.
Syllable onsets in Indonesian can also be
consonant clusters. In the current investigation, only
clusters comprised of two consonants were included
in the materials, such as the <dr> in the word drama.
This phonotactic pattern is only one of the eight
syllabic patterns of consonant clusters in Indonesian
(Hasibuan, 1996).
The research builds on previous investigations of
miscues made by children (e.g., Goodman, 1969;
Goodman and Burke, 1973) that seek to compare
what the young readers say (observed responses) and
what they read (expected response). By doing miscue
analysis, processes that happen during reading can be
mapped and used to determine factors that influence
reading development in Indonesian. More
specifically, the current work builds on and expands
an earlier work (Anjarningsih 2016) that studied 17
preschoolers and first graders. The most relevant
findings for the present investigation were that at the
grade one level, children’s miscues were
predominantly visual and that the consonant clusters
were the most challenging words to read. Therefore,
this study aims to find out what kind of miscues
happen when children read highly-frequent disyllabic
words containing diphthongs, digraphs and consonant
clusters; to seek what the reading miscues can inform
us about the effects of the spelling of diphthongs,
digraphs, and consonant clusters on children’s
reading development; to propose a possible
underlying reason for the miscues that are committed;
and to propose how sub-lexical factors influence
reading development of highly frequent disyllabic
words.
2 METHOD
This study was a qualitative study involving eighty-
two normally developing children living in Depok,
West Java, Indonesia, of which 46 are boys, and 36
are girls. Miscued words produced by the children
were tabulated and subjected to a qualitative analysis.
3 RESULTS
Based on the miscues that were committed by the
children, three kinds of miscues were identified,
including 1) visual: miscues showing change,
substitution, deletion, and transposition of sounds in
the words. The resulting words still share 50% of its
graphemes (letters) with the tested words; 2)
regularisation: miscues which results in the division
of digraphs and diphthongs into one or two simple
graphemes or sounds; and 3) substitution : miscues
resulting in totally different words or pseudo- or non-
words.
Figures 1, 2 and 3 show the total number of each
kind or miscues and its proportion relative to the total
number of words read per grade.
Figure 1: The total number of each kind or miscues and its
proportion relative to the total number of words read by
grade 1 students (n=19 children).
Figure 2: The total number of each kind or miscues and its
proportion relative to the total number of words read by
grade 2 students (n=43 children).
CONAPLIN and ICOLLITE 2017 - Tenth Conference on Applied Linguistics and the Second English Language Teaching and Technology
Conference in collaboration with the First International Conference on Language, Literature, Culture, and Education
840

Figure 3: The total number of each kind or miscues and its
proportion relative to the total number of words read by
grade 3 students (n=20 children).
Within each grade and each kind of miscues, a
paired t-test was run, comparing the number of
miscues committed to the shorter words and that
committed to the longer words. In total, nine paired t-
tests were run, and none of the ps was 0.05 or lower.
This means that there was no statistically significant
difference between the reading performance of the
children when they read the shorter words in all word
groups and the reading performance of the children
when they read the longer words in all word groups.
Comparing the three grades, the youngest readers
made proportionately more miscues than the older
ones who had practiced reading longer. This was true
for all kinds of miscues.
Simple words that simulated the transparency of
the Indonesian orthography were read very
successfully early on. Nobody substituted the simple
words with other words that do not share at least 50%
similarity in spelling. Diphthongs were also quite
successfully read from first grade on, as reflected by
the less than 5% miscue rate per kind of miscue. In
first grade, children made more substitution miscues
(e.g., reading tunai as bumi), but in second grade, the
most committed miscue was regularization (e.g.,
reading da-mai as da-ma), followed very closely by
visual miscues (e.g., reading wahai as wati). In the
third grade, accuracy in reading words with
diphthongs reached ceiling.
Comparing diphthongs and digraphs, in first and
second grade, overall rate of miscues of diphthongs
and digraphs are comparable. However, in third
grade, diphthongs were read with virtually almost no
miscues (i.e., 0.20%), whereas digraphs were
miscued 1.80%. It may very well be that digraphs
were mastered later than diphthongs and that the
second grade is a kind of cut-off grade: after learning
to read for two years, children understand that the
double letter of diphthongs map to single vowel
phonemes, but at the same time, it is still more
challenging to map the two letters of digraphs to
single consonant phonemes.
Of the four groups of words, the consonant
clusters group proved to present the most challenge
for the young beginning readers. Although the
percentage of miscues of this group of words declined
as the grades advanced, in third grade children made
comparably more miscues to words in this group than
to words in the other three groups. In general, children
made at least twice as many miscued consonant
clusters as miscued digraphs.
On closer inspection, in grades one and two, for
digraphs and consonant clusters, the kind of miscue
that was produced the most frequently was
regularization. Recall that this is when readers
separated the letters in the digraphs and when the two
letters in the consonant clusters are read as if they
were single letters (e.g., reading krisis as kirisis.). In
grade three, digraphs were still predominantly
regularized, but consonant clusters were
predominantly read as other words that differ from
the intended words in as much as at least 50% of their
letters.
4 DISCUSSION
The study’s main goal was to characterize the reading
development of normal Indonesian children. In doing
so, the materials were designed to capture both the
regularity of the transparent Indonesian orthography
and some irregularities that exist in the orthography
in order to see how such irregularities influence
reading development. To answer the first research
question, from the sample of young readers, miscues
were identified when they read disyllabic words
containing diphthongs, digraphs and consonant
clusters and three kinds of miscues were found:
visual, regularization, and substitution. Up to the
second grade, diphthongs and digraphs are mostly
regularized and in the third grade, only digraphs are
regularized. A higher proportion of miscues still
happen to consonant clusters, compared to those to
digraphs, even in the third grade. Therefore, in the
children’s reading development, it seemed that they
went through a process, begun by successfully
mastering words with simple spelling, followed by
mastering words with diphthongs, digraphs and
consonant clusters consecutively.
As for the answer to the second research question,
the manipulation to the syllabic structure showed that
children went through a process in mastering
common words. They were not at the same time able
to read all disyllabic words that were given. The same
number of letters that constitute diphthongs and
Read, Miscue, and Progress - A Preliminary Study in Characterizing Reading Development in Shallow Indonesian Orthography
841

digraphs did not seem to have the same effect on the
order when they are mastered. Diphthongs seemed to
be mastered at the end of the second grade, while
digraphs were still a little difficult at the end of the
third grade. Therefore, having the same number of
letter does not seem to be the only factor at play.
Another factor seems to be the ease at which children
grasp the mapping between the diphthongs and
digraphs on the one hand, and the phonemes on the
other hand. It is proposed that the earlier mastery of
diphthongs is because the glide in diphthongs may
have been easier to understand by children due to the
transparent mapping between the two glided sounds
and the two letters in the diphthongs.
The digraphs, in turn, may have been more
difficult because the sounds that the digraphs map to
are not directly evident. There are two possible routes
to memorizing this mapping: analysis and
memorization. The children may have analyzed, in
the case of the digraph <ng>, the letter <n> maps to
an alveolar nasal consonant, the letter <g> maps to a
velar plosive consonant, and <ng> map to /ŋ/ which
is a velar nasal consonant, a consonant with the same
manner of articulation as /n/ and the same place of
articulation as /g/. There are features that are shared
by /n/ and /g/ on the one hand, and /ŋ/ on the other
hand. Once children have done the analysis and
arrived at the correct conclusion based on a good
match with an entry in their lexicon, they may have
started to remember, for instance, the pairing between
the written word <bunga> with the lexical entry
bunga in their lexicon. This memorization may then
lead to them reading not the single letters of the
diphthongs, but the phoneme that each of the
diphthongs maps to.
Consonant clusters seemed to present the ultimate
difficulty as shown by the findings that at the end of
third grade, children still made considerable
proportion of miscues to words containing consonant
clusters. A possible explanation is because there are
many more consonant clusters than there are
diphthongs and digraphs, children took longer to
establish the mapping between consonant clusters and
the phonemes that they represent. In other words, the
greater number of kinds of consonant clusters adds
more challenge, in addition to the clusters having two
graphemes which render the clusters having more
sounds than diphthongs and digraphs.
The above findings go along the lines of those of
Marinus and de Jong (2008). Although using a
different method, it was demonstrated that digraphs
are utilized by beginning readers when they read and
they do influence the details of the syllable structures
that are mastered by children as they progress from
grade one to grade three of primary school.
Furthermore, findings about word length effect that
has been observed previously was also observed in
the current investigation. In line with Zoccolotti et al.
(2005) end of grade three seemed to be the point at
which children are not very much influenced by the
longer diphthongs and digraphs. In addition, the
current results expand the work of Anjarningsih
(2016) in that the consonant clusters continue to
present challenges even up to the end of grade three.
However, the findings about consonant clusters
may raise some questions before they can be
confidently interpreted. There is of course a
difference between digraphs and diphthongs on the
one hand, and consonant clusters on the other hand:
while diphthongs and digraphs map to one phoneme,
consonant clusters map to two or more phonemes, one
for each of the graphemes or letters in the clusters.
While this mapping, intuitively, may seem to predict
that consonant clusters should be easier to master than
digraphs because once children can map single
consonant graphemes to their corresponding
phonemes, they should get the pronunciation of
clusters correctly, our current findings go against this
prediction. Whether the visual and regularization
miscues were caused by children expecting to find
simple consonants followed by simple vowels in
syllables remains to be tested further.
With the above findings at hand, it interesting to
see that children did not master all the tested
disyllabic words at about the same time. They needed
to practice reading for about three years before their
reading performance became accurate. The sub-
lexical details or in this case, the syllabic make-up of
the words did influence how accurate children read
the words. Therefore, in a transparent orthography
such as the Indonesian orthography investigated here,
the mapping between graphemes and phonemes in
“irregular” diphthongs, digraphs, and consonant
clusters influences how early children can read
highly-frequent disyllabic words. Syllabic
complexity and orthographic depth, just like the
findings of Seymour et al. (2003), also influence the
rate of reading development in Indonesian.
It should be noted that the syllabic structures
tested in the current investigation do not comprise all
syllabic structures in Indonesian. In addition to CV or
CVC, Indonesian also has VC which can be written
by vowel and consonant graphemes or just vowel
graphemes (e.g., in the word <dua> meaning “two,”
the grapheme <a> maps to /wa/). Furthermore,
trisyllabic and words with more syllables also exist in
Indonesian, both monomorphemic (e.g., udara “air”)
and polymorphemic (e.g., penggorengan “frying
CONAPLIN and ICOLLITE 2017 - Tenth Conference on Applied Linguistics and the Second English Language Teaching and Technology
Conference in collaboration with the First International Conference on Language, Literature, Culture, and Education
842

pan”). Future research should incorporate other
syllabic structures and longer words in order to map
the reading development of Indonesian children more
thoroughly.
5 CONCLUSIONS
In this research that aims to characterize the
development of reading in normally-developing
Indonesian children, three kinds of reading miscues
were committed by the participants: visual,
regularization, and substitution. These miscues
helped to highlight that digraphs were mastered some
time after the children could read diphthongs
accurately (end of grade 3 vs. end of grade 2), and
consonant clusters were also mastered later that
digraphs. The different paths of acqusition between
diphthongs and digraphs were explained by how
predictable the mapping was between the diphthongs
and digraphs and their sounds. For consonant clusters,
the more time it took the children to master them was
attributed to the existence of much more consonant
clusters than digraphs that could have extended the
time children needed to map the clusters and their
sounds accurately. It is shown by the current results
that sub-syllabic complexity influenced the paths of
reading acqusition of frequent disyllabic Indonesian
words.
REFERENCES
Anjarningsih, H.Y. 2016. Characterizing the Reading
Development of Indonesian Children. In Yanti (ed.).
Kolita 14: Konferensi Linguistik Tahunan Atma Jaya
Keempat Belas. Paper presented at the Konferensi
Linguistik Tahunan Atma Jaya Keempat Belas, Unika
Atma Jaya, Jakarta, 6-8 April, 2016 (pp. 156-159).
Goodman, K.S. 1969. Analysis of Oral Reading Miscues:
Applied Psycholinguistics. Reading Research
Quarterly, 5, 9-30.
Goodman, K.S., Burke, C.L. Theoretically Based Studies of
Patterns of Miscues in Oral Reading Performance.
Wayne State Univ.: Detroit, Mich.
Hasibuan, N.H. 1996. Fonotaktik dalam suku kata bahasa
Indonesia. Depok: Fakultas Ilmu Pengetahuan Budaya
Universitas Indonesia (unpublished master’s thesis).
Marinus, E., de Jong, P.F. 2008. The Use of Sublexical
Clusters in Normal and Dyslexic Readers, Scientific
Studies of Reading, 12, 253-280.
Roach, P. 2009. English Phonetics and Phonology. 4
th
edition. Cambridge: CUP.
Robbins, L. Kenny, H.A., Robbins, L. A. 2007. A Sound
Approach: Using Phonemic Awareness to Teach
Reading and Spelling. Portage & Main Press.
Seymour, P.H.K., Aro, M., Erskine, J.M. 2003. Foundation
literacy acquisition in European orthographies. British
Journal of Psychology, 94, 143–174.
Zoccolotti, P., De Luca, M., Di Pace, E., Gasperini, F.,
Judica, A., Spinellli, D. 2005. Word length effect in
early reading and in developmental dyslexia. Brain and
Language, 93, 369-373.
Read, Miscue, and Progress - A Preliminary Study in Characterizing Reading Development in Shallow Indonesian Orthography
843