When spelling gains fail to generalise to words learners haven’t been taught, the obvious response is to search the research for a better intervention. A better phonics programme. More synthetic phonics. Explicit morphology. Variant vowel spellings. A different sequence of explicit spelling instruction.
 

But perhaps looking for another programme is the problem.

A programme necessarily decides in advance what will be taught. It selects the content, organises the sequence and explicitly teaches it to the learner.

Then we measure whether that learning transfers to words the programme didn’t teach.

What if we turn that around?
 

Don’t teach another body of spelling content. Get them self-teaching.

Learners already show us where they need support.

They struggle to read a word.

They misspell a word when writing.

They aren’t sure how the word they can say is represented in print.

Use that word.

Show the learner exactly how it is mapped according to their jurisdictional standard and accent: which letters are the graphemes and their sound values.

Then use the 60-Second Spelling Routine to rapidly bond speech, spelling and meaning in the orthographic lexicon.

Then move on.

There is no predetermined spelling list and no requirement that the learner completes another sequence of content before accessing the words they actually need.

It’s not a programme.
 

The individual word isn’t the end goal

We aren’t doing this because we want the learner to memorise that one word.

Every accurately mapped word gives the learner another opportunity to develop knowledge of the alphabetic code bidirectionally.

Print → Speech

Speech → Print

The learner is accumulating orthographic knowledge from the words they actually encounter.

The goal is that they need us less and less.

That is the critical difference.

If we teach a programme and then ask whether learners can spell words the programme didn’t teach, we’re testing for generalisation after explicit instruction.

What if we design for self-teaching from the outset?

There will always be another word

No programme can explicitly teach every word a learner will eventually want to read or write.

That is why generalisation matters.

The question shouldn’t only be:

Which intervention produces the biggest gains on the content we teach?

It should also be:

How quickly does the learner become able to learn from words without us explicitly teaching them?

That changes the role of the adult too.

We EXPECT adults to struggle with aspects of bidirectional word mapping. We don’t need to spend months making every adult an expert word mapper before they can help.

The technology shows the code.

The adult facilitates the technology and the 60-Second Spelling Routine.

The learner gets accurate information about the word at the point of need, bonds speech, spelling and meaning, and gets back to reading or writing.
 

As much explicit instruction as they need, and no more.

This isn’t a choice between phonics and morphology.

It’s a different question altogether.

Instead of searching for the next programme to explicitly teach struggling spellers, what happens if we give them the tools to become self-teaching?

Support the word they struggle with.

Show the code.

Bond speech, spelling and meaning.

Get in. Get out.

And measure what really matters:

What can they do with the words nobody taught them?