Reading by analogy/onset and rime

Please contribute past postings from the message board that you have have found particularly inspiring.

Moderators: maizie, Lesley Drake

Post Reply
Posts: 1859
Joined: Wed Nov 05, 2003 7:41 pm

Reading by analogy/onset and rime

Post by g.carter » Mon Aug 17, 2009 7:57 pm

Debbie Hepplewhite

Joined: 29 Aug 2005
Posts: 1829

Posted: Mon Aug 17, 2009 5:29 pm Post subject:


There are many reasons why onset and rime mixed with synthetic phonics is not advisable in my opinion. Here are some which may not have occurred to some people:

1) Onset and rime is based on what children appear to do 'naturally' when asked to break down spoken words. They generally break words into various chunks, some of which have been identified as 'onset' leaving the remaining word chunk as 'rime'.

However, what children may do 'naturally' doesn't necessarily reflect what is best to 'teach' and 'learn' and research is indicating that it is both very possible to TRAIN young learners to recognise units of sound at phoneme level (smallest single sound level) in terms of the printed word (so recognising letters and letter groups that have been specifically taught as units of sound) - for both decoding (reading) and encoding (spelling purposes).

If an alphabetic code is therefore generally taught at 'phoneme-level', then to ALSO teach at onset and rime level is to massively increase the numbers of 'units of sound' that need to be taught.

Not only is this silly mathematically, it also makes no sense on other fronts. 1) That the teaching is inconsistent. You are, in effect, teaching c+a+t = cat as well as teaching that c+at = cat. You could also argue a point to teach that ca+t = cat. Three ways just to teach how to read one simple word!

2) When children are taught through a synthetic phonics route, their skills at blending all-through-the-word at phoneme level and segmenting all-through-the-spoken-word at phoneme level reach very high levels of success. This means that we are really teaching a set of core SKILLS and not just a type of written code.

When onset and rime units are taught, the skill is closer to a snapping together like two pieces of a jigsaw puzzle rather than a sounding out multiple sounds in a row which are 'blended'. This may make no difference to some learners - but it could to others.

Arguably, we should teach at what I call the lowest levels - that is, to account for the weakest children in the class (or slowest-to-learn). This means that we should aim for MAXIMUM CONSISTENCY at all times.

2) I've recently been doing some phonics testing including an oral segmenting test with 40 words - 20 'real' of various configurations and 20 'nonsense' words of various configurations - right up to five sound words.

What I have discovered is that the older children who have not received synthetic phonics type teaching (and therefore have no experience of orally segmenting words at phoneme level) were a real mixed bag in terms of how they orally segment words.

There was evidence of phoneme level segmenting only at the three sound level for words such as 'cat' and 'dog'.

When three sound words included long vowel sounds, it was very clear that many of these children were not orally segmenting per se, but were splitting words up into the spellings that they knew, followed by changing those spellings into the sounds of the letters - or mixing up letter names with some letter sounds.

For example, 'green' might be /g/ /r/ /e/ /e/ /n/ NOT /g/ /r/ /ee/ /n/.

'cake' might be /k/ /a/ /k/ /e/ (no /ai/ in the middle and a sound given for the end 'e').

Furthermore, I began to see some strange patterns of segmenting which, at first glance might seem like splitting words up into onset and rime - but actually were more closely associated with splitting words up into THREE BEATS no matter which 'units of sound' those beats were. So the THREE BEAT issue was greater than the onset and rime.

Interestingly, there was a huge lack of consistency in the results of the older children showing that their oral segmenting was little or nothing to do with any 'teaching' they had - but sheer hit and miss according to the individual child.

Certainly, my findings thus far do not indicate any sense of 'natural' break-down of spoken words.

On repeating the same quite substantial test on two sets of Reception children from different schools but both sets apparently taught by a synthetic phonics route, there was a huge difference also.

Many children in one school (five year olds) achieved almost full scores for oral segmenting at phoneme level because that is the way they have been taught. The younger children in this group resorted to more three beat words but I realised that these younger children had not been taught much segmenting of words beyond three sounds. They reflected their immaturity perhaps - or perhaps they reflected the limits placed upon them by the teaching itself.

This raises the question about limiting children's learning by the teaching - versus - children not being 'ready' yet. Which is it? I suspect a bit of both!

In general terms, however, both Reception classes reflected the teaching they had received (and probably the 'programmes' being used) and they reflected what is possible when taught by certain methods with various degrees of rigour.


Another reason to avoid onset and rime teaching mixed with synthetic phonics teaching is because it may reflect teachers hanging onto old resources.

There is a lot to be taught in terms of a reasonably comprehensive alphabetic code even at phoneme level - and if schools want to ensure good spelling and handwriting, time must be allocated to teaching and rehearsing these basic skills well.

When teachers hang on to their favourite onset and rime worksheets, consumable books and so on, this eats away into 'basic skills' time very drastically.

Teachers need to be VERY discerning about the activities they give children to do.

Also, much 'busy work' is about filling in bits of words such as consonant clusters and rimes etc. but this type of chunking and 'selecting' work is not really the same skill as real reading and real spelling.

For example, a book shows words partially written (perhaps the 'rimes') and dashes are left for children to search out the correct 'consonant cluster' or 'onset'.

For example, the child sees: _ ap and has to choose from scr, tr, bl, sc and so on to make the correct word (the word may be next to a picture clue).

This word is neither all print for reading purposes, nor anything to do with segmenting a spoken word. It is neither 'reading' nor 'spelling'. It's simply a busy work type of game.

Now, this may not hurt - but what about the 'dash' system. Is this a 'sound dash' (which could be for any number of letters) or is this a 'letter dash'? Is there consistency with the general teaching going on in the class for phonics?

And, whilesoever the children are doing this kind of 'game' or 'activity', they are NOT making progress with being taught the next bit of alphabetic code nor rehearsing any of the core skills of blending, segmenting or handwriting.

4) Finally, those of us who have worked with many 'special needs' learners will all be able to describe how children are mixed up with their consonant clusters. They might spell with 'st' when they want a 's' or they might say 'st' when they need to say 's' and so on. They are often mightily confused one way or another.

I would say that the best case scenario is the greatest consistency of teaching and rehearsal - even if there are learners who can dip in and out of work which is with different sized chunks of code.

Beware of games and activities which are time-fillers and not teaching and learning opportunities!!!!

Back to top


Joined: 20 Mar 2005
Posts: 296

Posted: Mon Aug 17, 2009 7:00 pm Post subject:


Just a point of clarification: beginning readers do not decode by analogy, but more experienced readers do.

One finding of Johnston and Watson (see their article published in Reading and Writing: an Interdisciplinary Journal) was that the only readers in their study who could decode by analogy were those who had mastered phonemic decoding, i.e. all-through-the-word synthetic phonics decoding. Children trained with "analytic" phonics were unable to transfer such skills as they had to read new words by "analogy."

My recollection of the research in the 80's and 90's was that onset-rime was favoured initially because it was believed young children could not discriminate phoneme-level units, and were more successful with syllables and rimes (evidence for this in subsequent research has been mixed, at best, but it has been shown that children can be taught to discriminate individual phonemes even if they do not "naturally" do so).

The current "word family" practice, which is mostly one of initial consonant substitution, is deleterious at the outset because it gives the beginning reader some wrong ideas about what reading is all about -- it does not focus attention on variable parts other than the initial letter or consonant cluster, does not pay attention to vowels, does not develop a l-r tracking orientation, and may foster some learning of misrules, such as the idea that all patterns that look the same sound the same -- which is true for phonograms like "ip" but not for "ough" or "oul," or that it is unnecessary to look past a familiar "part" to the rest of the word (this leads to the common practice of dropping or ignoring endings).

Thus we can say with some assurance that teaching a word family, initial-consonant-substitution, onset-rime approach to beginners is counterproductive to many children and inconsistent with SP principles.

However, at a later stage of reading, most fluent readers do decode by analogy at times, and those who struggle to gain fluency benefit from being taught explicitly how to do so. There is plenty of data from successful programs for post-beginners of the effectiveness of teaching decoding by analogy, and doing so in an SP-consistent way. The children have already learned to sound out and blend, and to write short words by segmenting all through the word and writing a grapheme for each. At that point they benefit from being able to use larger orthographic patterns, both for reading and spelling, and in the programs I am familiar with that make significant use of the analogy strategy, this is taught as both a spelling and reading strategy (emphasizing the reversibility of the code).

The linguistic phonics program of the 70's, such as Lippincott, Language Patterns, the Basic Reading Series and others of a similar vein (including the original SWRL Beginning Reading Program) introduced "word classes" or patterns of similar words for reading and spelling, but unlike today's onset-rime approach did not teach children the initial.-consonant-substitution strategy or encourage them to memorize the rime units. Some effort was made in teaching the spelling/writing skills to have the children generalize what they had observed about orthographic patterns to untaught words, both for reading and spelling; however, the instruction was not very explicit and depended on inductive understanding for the most part.

The Spalding program, which is sui generis in many ways, also emphasized the learning and application of a number of phonograms (orthographic units ) in both word reading and spelling at the initial stages. Spalding has very strong evidence of effectiveness, but is such a comprehensive program it does not lend itself to individual classroom use, but needs to be a whole-school (or whole-division) approach. Abecedarian uses keywords for each sound-spelling taught, which is an analogy approach to spelling that overlaps for reading purposes, as children sort words by spelling pattern and group under the relevant keyword.

At post-beginner stages, readers profit by being taught specific decoding-by-analogy strategies, if they have failed to develop them on their own. Maureen Lovett's extensively researched and documented work with young readers employed an analogy strategy along with specific self-talk problem solving strategies, morpheme analysis and comprehension monitoring with very successful outcomes -- AFTER the children had learned to decode all-through-the-word in SP fashion and had a solid decoding reflex. In the last several years, she has expanded this work to secondary students (not exceptional learners) who need to improve their literacy skills to get their high school credits. There are additional strategies used for adolescents, but the analogy strategy is taught explicitly, especially in the context of multi-syllable word decoding, which is a stumbling block for many students, even ones who have a good grounding in SP decoding skills.

For those interested, I gave some references and info on Lovett's program in this thread:

The Clackmannanshire report does not mention the specifics of the "Phonics Revisited" module developed for students in Primary 4, except to say that it reviewed advanced code skills and taught multi-syllable word decoding skills; it would be interesting to know what use, if any, the Clackmannanshire researchers made of the analogy strategy. They may have felt it unnecessary, based on their conclusions in the published article that children who decoded well at the phoneme level could decode by analogy, but not the other way around.

Another well-researched and validated effective program that makes extensive use of analogy is the Benchmark School program, developed by Linnea Ehri and Irene Gaskins. They have a very comprehensive literacy program, of which the analogy strategy is only one (but a significant) component. Their book : ... 1593851693
is an excellent resource for teaching middle grade students a variety of reading skills. The program is so comprehensive I can't envision it being successfully replicated in a regular public school, because of the limits on instructional time and the myriad interruptions, but I found their book useful for fleshing out reading instruction beyond the decoding level for older children.

To summarize, the onset-rime, "word family" approach is not consistent with SP teaching for beginners, but there is a place farther along the continuum (ages 7-8 and beyond) for young readers to learn to read and spell larger orthographic patterns and to use these as part of an analogy strategy for reading and spelling new words. Walpole and McKenna suggest analogy strategies be taught specifically to students in second grade and beyond who have proficient decoding skills and know their advanced code vowel patterns but struggle to apply them to new words; to older students with reading delays, and to students having difficulty with multi-syllable words. Because analogy is cognitively demanding, it is not an appropriate strategy for beginners. Proficient readers typically use analogy strategies without being taught or even aware of it. An example is the discussion on the pronunciation of the name "Carraclough" from Lassie Come Home that I started on the Purely Practical Posts board.. In determining the most likely pronunciation, we all used analogies to names and words we already knew.

One of the many strengths of the SP approach is the recognition that the instructional needs and reading behaviours of children change over time; that what is needed and appropriate for beginners is different from what is needed or appropriate for more mature readers. The WHAT to be taught and the WHEN and the HOW to teach it are all very important considerations which vary with the child's age and stage of reading acquisition and need careful consideration.

Susan S.

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest