Learning to read

This unit looks at how children learn to read and examines the problems experienced by pupils with specific learning difficulties such as dyslexia or troubles with comprehension.

Among the areas covered by the unit are:

  • The simple view of reading
  • The alphabetic principle
  • The dual route cascaded model (DRCM)
  • Phonics and the role phonics teaching plays in learning to read
  • The relationship between memory and reading

Select any of the information circles to find out more about how different learners fit into the simple view of reading.


The simple view of reading

The simple view of reading - supporting grid

Language comprehension


  • Good language comprehension, poor decoding
  • People with good comprehension skills but who struggle with decoding would classically be described as dyslexic.

  • Poor language comprehension, poor decoding
  • In older pupils with dyslexia, inexperience with texts and an inability to use reading as a learning tool can result in poorer comprehension skills than those of their peers. Some pupils may also have dyslexia alongside other language difficulties that hinder reading comprehension.

  • Poor language comprehension, good decoding
  • The model predicts that some learners will have better decoding than comprehension skills. These pupils are described as having specific comprehension difficulties with their reading, which can be strongly linked to problems with language comprehension. For example, pupils who speak English as an additional language or have speech and language difficulties may have more problems with comprehension than decoding.


Phonological awareness

Phonological awareness involves the detection and manipulation of sounds and can be an important indicator in the development of a child’s reading ability.

Its three levels of sound structure are:

  • Syllables
  • Onsets and rhymes
  • Phonemes

In her 1990 book, Beginning to Read: Thinking and Learning About Print, cognitive psychologist Dr Marilyn Jager Adams described four main types of task, as cited in Valerie Muter’s Early Reading Development and Dyslexia (2003).

Select any circle to find out more about that type of task.

  1. Rhyming
  2. Segmentation
  3. Sound blending
  4. Manipulation

Rhyming tasks

Rhyming exercises may involve replacing a word in a nursery rhyme, song, or rhyming string of words.

Other examples of tasks requiring rhyming skills include approaches such as these, used in research cited by Hume and Snowling (2008):

Source: Developmental Disorders of Language Learning and Cognition, Hume and Snowling (2008)

  • Rhyme detection, where a child is shown a picture of something, told what it is, and asked to identify the rhyming word from a selection of three choices.
  • Rhyme production, in which a child names as many rhyming words as possible from a given starting word.
  • Rhyme oddity, where a child identifies a word that does not rhyme with the others, for example ‘rate, mate, bake’.
« Return

Segmentation tasks

These tasks require pupils to break given words up into syllables or phonemes.

So for syllables:

  • al-pha-bet
  • teach-er
  • el-e-phant

And for phonemes:

  • /m/ /a/ /t/
  • /s/ /ea/ /t/
  • /g/ /ir/ /l/

One such approach is the tapping task, which requires pupils to tap along with each phoneme or syllable they say within a word. Other tasks can include asking pupils to:

  • Match common phonemes or syllables within different words (for example compare and compute)
  • Identify the syllable or phoneme to complete a word following a visual or verbal prompt
  • Isolate phonemes within a word (for example, Q: ‘What’s the first sound at the beginning of the word shoe?’ A: /sh/ /oo/)
  • Remove a syllable or phoneme from within a word (for example, ‘Say ‘plane’. Now say it without the /p/’)
« Return

Sound blending

Blending involves ‘sounding out’ words by combining phonemes or word sounds, which are provided to the pupil by the teacher.

Although blending tasks initially seem very similar to those of segmentation, there are several differences:

  • The tester provides the pupil with the segments of the word, and asks them to blend these together, rather than breaking down the word and identifying its composite parts.
  • The pupil only needs to understand that the phonemes or sounds with which they are provided can be pushed together to make words; segmentation requires greater linguistic awareness of the length and purpose of these composite sounds.
  • As it does not begin with an already familiar word, blending offers greater insight into a child’s knowledge of, or capacity to remember, phonemes or word sounds.

So in a blending task, a child sounds out words, perhaps in response to visual prompts:

  • /c/ /a/ /t/ = cat
  • /d/ /u/ /k/ = duck
  • /h/ /ear/ /t/ = heart

This approach also works for non-words, which can be sounded out despite having no meaning; so for example, the made up word, ‘mayfab’, would be /m/ /ay/ /f/ /a/ /b/.

« Return

Phoneme manipulation

These tasks are considered to be by far the most difficult of those featured here.

They involve asking pupils to manipulate the phonemes within each test word, using wordplay such as spoonerisms and phoneme deletion.

Examples of phoneme deletion:

  • Say hand without the first sound = and (with /h/ removed)
  • Say earth without the last sound = ear (with /th/ removed)

Examples of spoonerisms:

  • Q: Switch the first sounds of the words ‘car’ (/k/ /ar/) and ‘park’ (/p/ /ar/ /k/) A: par (/p/ /ar/) cark (/k/ /ar/ /k/)
  • Q: What starts with /l/ and rhymes with hand? A: land
  • Q: What other word can you make from ‘pen’ by adding an /o/ sound to it? A: open

Jager Adams notes that pupils’ performance in manipulation exercises have provided strong predications of, or correlations with, reading ability.

« Return

The dual route cascaded model (DRCM) assumes all readers, even those in the early stages of learning, use the same approaches. These bullet lists show the salient points of, and differences between, the model’s two routes to decoding.


The two routes of DCRM

Non-lexical route

  • A grapheme-to-phoneme conversion (GPC) approach to pronouncing a word, or non-word.
  • Frequently referred to as ‘sounding out’ by teachers, the reader identifies phonemes in a serial process, reading left to right, and blends these into a word/pronunciation.
  • Required to decode any non-words or words a reader does not already know.
  • Only works if the word is sufficiently regular and/or made up of phonemes with which the reader is familiar.
  • Slow and superfluous for words that the reader already knows by sight.

Lexical route

  • The pupil instantly reads a word they already know by sight, recalling its pronunciation and meaning from their memory.
  • Increasingly becomes the main route used by a reader as they learn more words.
  • Must be used when reading irregular words such as ‘was’ and ‘the’, which will result in error when using the non-lexical route due to their phonological structure.
  • Cannot be used to read non-words or unknown words, which need to be learned or ‘sounded out’ using the non-lexical route.

In the 1988, Oakhill and Garnham estimated that 1 in 10 of 7- to 8-year-olds have significant comprehension difficulties. A study by Nation and Snowling, published in 1997, suggested general comprehension difficulties across a wider age range of 7- to 10-year olds – a frequently repeated approximation.



An estimated 10% of primary-aged
               children experience difficulties with comprehension

An estimated 10% of primary-aged children experience difficulties with comprehension

This resource provides brief descriptions of the three comprehension skills that readers should develop, as cited in your PDF. More expansive definitions can be found in the source material – Oakhill and Yuill’s chapter, Learning to understand written language, in the 2002 book, Addressing difficulties in literacy development. See your PDF’s resources list for further details.


Comprehension skills

Inference skills

Authors cannot always explain everything explicitly. As such, readers will often need to make inferences from text, something that causes difficulties for less-skilled comprehenders.

A way of testing inference skills is to ask pupils two types of questions about a given sample of text:

  • Literal questions, based purely on what is explicitly explained.
  • Inference questions, based on what is only implicitly communicated.

An experiment conducted by Oakhill found that less-skilled comprehenders record a high error rate on inference questions, even when given a chance to re-read and check against the text (as opposed to answering questions from memory, as in the majority of comprehension measures).

As well as seemingly not being linked to issues with memory, other studies have also suggested that an inability to answer inference questions is not directly linked to a lack of general knowledge.


Barnes, M. A., and Dennis, M. (1998), cited in Cain K and Oakhill J (2007), Children's Comprehension Problems in Oral and Written Language: A Cognitive Perspective


Other possible explanations for inference difficulties include:

  • Pupils concentrating too hard on ‘getting words right’, thus processing text at a superficial level.
  • Children being unable to integrate inferences with their understanding of the text due to processing difficulties.

Memory and reading

The three essential memory processes...

Visual aids for Encoding (a computer keyboard), Storage (a computer hard drive) and Retrieval (a
               computer monitor)
  • Encoding

    Upon receiving information, the brain processes it, ready to store so it can later be recalled.

  • Storage

    The brain creates a record of the information to be recalled. How this is stored varies between the functions of short-term, long-term and working memory.

  • Retrieval

    When required, the brain recalls the information in reaction to a cue, requirement or activity. Again, this process varies between the different forms of memory.