As mentioned above, English spelling is considered notoriously complex and difficult to learn due to the inconsistencies of the system.

–  Wrighting in Englesh is harde; An error analysis of Norwegian pupils’ L2 English spelling

Fortune doth address our affairs better …

– The History of Don Quixote, Vol 1

One of the first tests even very young children are familiar with is the spelling test. If spelling was not considered demanding, effort may not have been put into designing spell-check and auto-correct functions in word processors. A researcher in Norway analysed many samples of student work to understand better the type of spelling errors students make.

She discusses these different types, why these errors were made, how spelling is learnt and what some effective ways of correcting (teachers) and doing corrections ( students) for spelling errors are in her paper published this year.

The experience of students in Norway will be useful because both we and them can for the most part be considered as L2 learners of English. English is not our native tongue. Indeed, the types of errors the Norwegian students made are not all that different from those students here make.

Though upper primary students may not have as many or have at all any spelling tests the way their lower primary peers do, they too are penalised for spelling errors across components. There is one component which tests spelling explicitly even at the primary 6 level. This is the editing component of the Language Use paper. Students are given a text with 12 errors. 6 of these will be spelling errors. There is a lot of thought that has gone into how this component has been designed as we will see later.

In compositions, spelling will affect how much is awarded under the Language component. Though it is not the case that each instance of error would result in a commensurate reduction of marks (especially for repeated instances of the same misspelled word), spelling can affect a reader’s understanding and enjoyment of a story and this can affect the overall score.

There are two main ways to learn spelling. Spelling can also be acquired through repeated exposure without deliberate effort.

The first way is what is known as the “lexical route” (Treiman, 2017 as cited in Wedge, 2021). This is when students memorise the correct letters and arrangement of letters of a word. So, when they hear the word, they simply reproduce from memory without paying attention to how the word is made up. This way of course is cognitively demanding and those who have found ways to memorise more effectively find it less tedious.

The second is what is known as the “nonlexical route” (Wedge, 2021) which is when students rely on their experience and understanding of what kinds of letters produce which kinds of sounds and the patterns which exist in how words are spelt. For example, students could have an awareness of the ‘ous’ suffix commonly found in adjectives. So, they could be reasonably certain that the last three letters must be ‘ous’ even if they have not spelled the word in question before.

Though this way saves the effort of memorising, it will not always work for the following reasons according to Wedge (2021). There is more than one letter which is capable of producing the same sound – ‘c’ and ‘k’, ‘e’ and ‘i’ or ‘a’ and ‘e’ for example. Double letters have no distinct pronunciation – ‘little’ sounds not very different from ‘litle’. Also, some words are not spelled the way they sound – rendezvous.

In practice, students use both the non-lexical and lexical routes (for words which do not have a straightforward letter-sound connection) to overcome the challenges of spelling. If a word is a high-frequency one and the letter-sound connection is not straightforward, students would still be able to spell it correctly through the sheer frequency of use. An example of this is ‘have’ – not spelled as ‘haf’ (Wedge, 2021). Repeated spelling leads to automaticity. This is one reason some students do corrections thrice or five times or ten times for misspelled words. Whether this method works depends on if a student is paying attention to which letters were wrong and what the correct arrangement or letter choice should be.

Usually, students should be able to spell a word correctly after getting it wrong a number of times especially if it is a high frequency word and if the error is taken note of.

Wedge (2021) referred to two sets of data. The first was from student answers to their National Writing Tests. The students were aged 10-11 and 15-16. The tests were open-ended in the sense that they had to, “Write a postcard” or “Describe what you see in this picture”. The total number of words analysed in this dataset was 129420 words. She also referred to the answer submissions of students aged 15-16 to their school written assessments. In the school written assessments, students had to give one or two short answer responses and one long answer response. These responses provided 66079 words for analysis. While the responses to the National Writing Tests were handwritten, the responses to school assessments were typed out on laptops.

Wedge (2021) noted that when working with word processors as they did in their school assessments, student work would be more error free because of spell-check and autocorrect. She also suggested that the spelling errors made while typing may be typing errors more than spelling errors. Students working on word processors may want to turn the spell-check and auto-correct functions off while writing to train their spelling ability more.

The error types, in order of frequency of occurrence (try spelling this by hand); are: errors of omission (dropping a letter), errors of substitution (replacing ‘s’ with ‘c’ for example), errors connected with double letters, errors of insertion (including additional letter(s) – especially the silent ‘e’; reasones (Wedge, 2021), errors of transposition – one example would be reversing the order of ‘e’ and ‘i’ in ‘receive’. There was one last category for errors which did not follow any identifiable pattern. This category was labelled, ‘unclassifiable’.

She added that in many instances of misspelled words, the word though spelled wrongly, sounded right. This is why in the Editing component of the Language Use paper, spelling errors will sound right. Students through practising this component should be able to appreciate that the sound-letter connection is not straightforward and engage in self-correction of their mistakes in other components.

She concludes that the vowels, ‘a’, ‘e’ and ‘i’ are problematic for many students. Students in the Norwegian study, often omitted, included and wrongly positioned these vowels. Students tended to have more difficulty with double consonants rather than vowels. This pattern of errors is also observable in students here.

Drawing from the research of fellows, she suggests that there tends to be a pattern unique to every student in errors made. She recommends that teachers conduct error analysis of student work over the long term to discern such patterns. For example, with sufficient instances of student work, a marker would be able to tell if an error is a competence error (awareness must be increased) or a performance error (attention must be increased). This would help in the crafting of customised solutions. She submits that there is currently no consensus on the most effective way for teachers to correct spelling errors or for students to do corrections. The applicable pedagogical principle is as much support as and only as necessary.

When it comes to right and wrong, the eyes have it.

Supercalifragilisticexpialidocious

Supercalifragilisticexpialidosious

Supercalifragilisticexpiallidocious

The Brain Dojo