2.4.2: Reading and spelling

Children with dyslexia will often show difficulties in reading and/or spelling specific kinds of words, depending on whether they have problems establishing the decoding system or the direct access system, or both. (You may want to revise these concepts from Section 1, part 4.2.)

More specifically, lists of regular, irregular and pseudo-words can be used to find out in a qualitative way if the systems for decoding and direct access are developing and being used correctly.

Broadly speaking, and as we will explain in detail below, irregular words must be read with the direct access system as their reading with the decoding system would lead to errors. Conversely, pseudo-words must be read with the decoding system as they have never been encountered before and therefore have no corresponding representations in the mental lexicons.

Regular words

In these words the correspondences between ‘letters’ and ‘sounds’ are straightforward and unambiguous, for example:

sat, bin, bat, lamp, basic, canal

and so these words can be read correctly by the decoding system.

Beginner readers proceed in three stages – segmentation, conversion, blending.

During the course of normal development words are progressively stored in the orthographic lexicon and will then be read by the direct system which is faster and more efficient than the decoding process.

In children with dyslexia the establishment of decoding is often slower and more laborious compared to their peers without dyslexia. This will be shown by more mistakes in reading, even for regular words. The list will contain one or two perfectly regular but less common words.

Here’s a PDF with a list of regular words.

Irregular words

In these words the correspondences between some ‘letters’ and their corresponding ‘sounds’ are not regular.

These words can only be read correctly by the system for direct access as the decoding system cannot lead to the correct pronunciation. Instead they lead to ‘regularisation’ mistakes, i.e. mistakes consisting of reading the words with the decoding system as if they were regular, as shown in examples below. Such mistakes are quite normal in the first stages of reading but persist in children with dyslexia because they find it difficult to memorize or to access the representations of common irregular units or complete words.


Here are ten irregular words. With your course partner write down exactly which ‘letter’-‘sound’ correspondence(s) is/are irregular, and how they could be said or written through the decoding system. In other words, what ‘regularisation’ mistakes may the child make by reading or spelling simply by decoding?

Make two lists of observations, one for reading and the other for spelling.

< said >

< yacht >

< knight >

< gift >

< white >

< echo >

< roll >

< busy >

< much >

< sure >

Reading Spelling
said < ai > is usually pronounced [ â ] as in < rain >, so this word would be decoded [ sâd ] instead of [ sed ] < sed >
yacht < a > is usually pronounced [ a ], as in < cat >, and < ch > is usually pronounced [ ch ], as in < bench >, and so [ yacht ] or [ yasht ] instead of [ yot ] < yot >
knight < k > is silent and < -ght > has to be stored for direct access, and so [ k ni..gu…heu-t ] < nite >
gift < g > before < i > is usually pronounced [ dj ] as in < gist >, and so [ djift ] < guift > (as in < guilt >)
white < h > is usually pronounced [ h ] and not silent, and so [ whît ] instead of [ wît ] < wite >
echo < ch > is usually pronounced [ ch ] as in < itchy >, or [ tch ] as in < church >, and so [ echo ] or [ etcho ] instead of [ eko ] < eco >, < eko >, < ecco > or < ecko >
roll the final < ll > usually follows a short vowel, as in < doll >, and so [ rol ] instead of [ rôl ] < rowl > or < role >
busy < u > is usually pronounced [ u ] as in < rust >, and so [ buzi ] instead of [ bizi ] < bisy > or < bizy >
much < ch > is usually pronounced [ ch ] as in [ munch ], and so [ much ] instead of [ mutch ] < mutch >
sure < s > is usually pronounced [ s ] as in < sun >, and < u > before < r > is usually pronounced [ er ], as in < surf >, and so [ ser ] instead of [ choor ]. Also, a final silent < e > usually lengthens the preceding vowel (magic ‘e’), as in < pure > ([ piur ]), and so this word could also be pronounced [ siur ] instead of [ choor ] < shoor > or < choor >

As can be seen from this activity, irregular words examine the child’s ability to access whole word representations in both the visual area of the brain (the orthographic lexicon) and the auditory region of the brain (the phonological lexicon).

Persistent regularisation errors show that the child is overusing the decoding system and that he or she is having difficulties memorizing the complete and structured orthographic representations of words, which is the case in some children with dyslexia.

Here is a PDF with a list of irregular words.

Remember to pay attention to the ‘regularisation’ errors as you listen to the child reading.


These are ‘invented’ words which do not exist in the language, for example < labbit >, < porrid >, and < trank >. They can however be read by the decoding system through the ‘letter’-‘sound’ decoding process. In fact they can only be read by this system because they will never have been seen before and therefore cannot have been stored in the orthographic lexicon.

Good performance in reading pseudo-words shows that the decoding system is working well. In children with reading problems this system is usually weak.

In general, two types of error are found:

  • ‘lexicalisation’ mistakes which consist of reading non-words as words. For example: < labbit > becomes [ rabbit ]. This means that the system for direct access is being overused
  • ‘phonic’ errors which consist of adding, omitting, inversing or substituting ‘sounds’ in sequence. This shows that the decoding path is being used but not effectively and so the correct sequence of ‘sounds’ is lost.

Here is a PDF with a list of pseudo-words.

Remember to pay attention to the ‘lexicalisation’ and ‘phonic’ errors.