Why phonology comes first

The very first essay I wrote in my undergraduate linguistics course was a defence of the English spelling system. My argument – inasmuch as my callow 18-year-old self was able to construct one – was that, given the unsuitability of the Roman alphabet to the English phonological system, not to mention the varied and often overlapping influences on the English language, our ramshackle orthography was not a bad compromise. (Those who like to draw social parallels could point to the trial-and-error accretion of English common law, or the outwardly bizarre ‘imperial’ system of weights, measures and currency.)

By Michael Salter

Download PDF

Morphology and etymology played an important role in the essay. I wrote about sign and signal, and how the morphemic identity outweighed the phonological discrepancies. About debt and debit. Receipt and reception. And I wrote, too, about the need to preserve the orthographic form of a morpheme in the wake of the shifting schwa vowels occasioned by the stress patterns of English (think informant versus information, or photograph versus photography).

Nearly three decades later, my view hasn’t substantially changed. I still think that, all things considered, the English spelling system is not nearly as bad or as capricious as it is sometimes depicted, and I am frankly pretty dismissive of those who advocate a wholesale, Bernard Shaw-esque reform of English orthography.

Given all this, you would expect me to be supportive of approaches to literacy which assign prominent roles to morphology and etymology. And in many ways, I am. It is important for young people learning English to get to know something about them, even if they do not become familiar with the exact terminology. The basic idea behind these currently popular approaches – of dividing a word into its etymological or morphemic units – is sound enough.

But there are good reasons why phonology should still come first. And that means phonics.

First of all, it is not as though familiarity with morphology and etymology will clear up all the mysteries and inconsistencies of English orthography. To go to the opposite extreme of the Bernard Shaws and claim that English orthography is entirely rule-governed and bereft of exceptions is futile; still less is it true that there is strict regularity in the adoption of Greek and Latin roots and morphemes into English. Recede and precede come from compounds of the Latin verb cedere, but so do succeed and proceed. The almost identical Latin verbal adjectives nobilis and mobilis give us noble, but mobile. (Yet, of course, nobility and mobility.) The Greek verbal noun suffix -ma looks to have entered English regularly enough if we consider schematic, idiomatic and dramatic, but not when we consider the base forms scheme, idiom and drama.

English morphology, too, is tricky territory for the uninitiated – and it is unlikely that trainee teachers will be able to negotiate it with confidence. Is the suffix that produces the noun which describes the process of a verb -tion/-sion or -ion? As a matter of fact it is the former, but this is a linguistic minefield. If -ing is a proper English suffix (it is), and if we can change make into making by removing the silent e, why should it not be the same for, say, create and creation?

The problems with this logic are numerous, but they can be quite hard to discern. (Here, for the record, are two of them: (1) we add -ation, not simply -ion, to verbs not ending in -ate such as condemn or flirt; (2) the letter i in the -tion ending has no phonetic value independently of its preceding consonant. Some of the other reasons have to do with Latin verb conjugations, and are rather obscure.)

Secondly, and far more importantly, there is a basic problem (another -ma word there!) with approaches to literacy which suggest a complementary focus on morphology and etymology from the outset. One such approach is known as SWI (Structured Word Inquiry), and in an introductory article about SWI by Professor Jeffrey Bowers, one of its chief advocates, we find the following:

English prioritizes the consistent spelling of morphemes over the consistent spellings of phonemes … A language that prioritizes the consistent spelling of morphemes over phonemes is not “fundamentally alphabetic”. (p. 4)

The problem with this plausible contention is that like is not being compared with like. Morphemes are not unitary in the way that phonemes are: indeed, they are made up of one or (usually) more phonemes, in a specific pattern. And the orthography of the basic morpheme is, of course, determined by the phonology: it is not arbitrary.

The clearest indication of this comes, in fact, with new additions to the language. Foreign words, onomatopoeic words, and borrowings from slang are all initially adopted according to phonology (it could hardly be otherwise, since they will constitute a morpheme that doesn’t exist yet in the language). They may acquire –ed, –s, –ing and others along the line, and morphophonemic changes may occur. But it is, of course, phonology which determines the spelling of the new word. A can hardly be more ‘fundamental’ than B if it depends on B for its component parts.

Furthermore, the number of English roots, rather than affixes, which undergo morphophonemic change is surprisingly small. Yes, we have please and pleasure, with /iz/ becoming /ɛʒ/, and sign and signal, with /aɪ/ becoming /ɪg/. But cast your eye over a random page in a book and you are unlikely to come across more than one word in ten which features such a morphophonemic quirk.

Elsewhere in Prof. Bowers’ article, there are hints (though not outright declarations) that a phonics-based approach ignores, rather than defers until a developmentally appropriate stage, issues of morphology and etymology:

However, unlike phonics, SWI considers grapheme-phonemes within the context of morphology and etymology… (p. 5)

[P]honics instruction … explicitly teaches children grapheme-phoneme correspondences in English without reference to morphology and etymology. (p. 2)

It would have been fair of Prof. Bowers to note that no serious proponent of phonics instruction, not one, decries the value of morphology and etymology at a later stage, or claims that familiarity with grapheme-phoneme correspondences alone is sufficient to become a competent reader and writer of English, with its deep orthography.

Before we leave the article, a tangential but important issue:

There is an overwhelming consensus in the research community that systematic phonics is best practice for early reading instruction in English.

This is undoubtedly true, but it is not the whole story. Those actually involved in proper research into early literacy have indeed consistently confirmed what common sense would already suggest, namely that thoroughly familiarising children with letter-sound correspondences initially is the most effective approach. But it is not in research publications that the battle for influence over the hearts and minds of trainee teachers is really fought. It is in the lecture theatres of initial teacher education courses.

What 20 years of interactions with trainee and first-year-out teachers has shown me is that attitudes to proper phonics teaching among initial teacher education (ITE) lecturers are almost uniformly negative, whatever the accumulated research may suggest. Phonics is simply lumped in with the other ‘traditional’ practices and attitudes, and trainee teachers are implicitly encouraged to react from the gut in such matters, not from the evidence.

Morphology and etymology are fascinating, and very important. But they have their place, and it is not at the very beginning of reading instruction. There is a good reason why, when field linguists produce a grammar of a language, they traditionally deal thoroughly with the phonology before moving to matters of morphology and syntax. It is simply the systematic way to proceed: deal with the building blocks first, then move on to the more exciting stuff. Mutatis mutandis, the same principle holds with initial literacy instruction, and for the same reasons.

This article was published in the June 2020 edition of Nomanis.

Michael Salter has taught classics and modern languages in both government and non-government schools in Sydney for 20 years. He has written two e-books on English etymology, and has written problems for linguistics competitions both in Australia and overseas. He has twice led Australian teams at the International Linguistics Olympiad. He blogs about education issues at

Similar Articles