The Clearest Issues with Modern English Spelling that could be corrected without replacing it with an entirely new system.Edit
This is first of all an essay about standard English spelling. However, it may also be of interest to those involved with the Anglish project who wish to devise new spellings or new systems of spelling. For convenience, it is written in standard English. It is a counterpart to my other essay, Overlooked Benefits of Modern English Spelling .
A note on possibly confusing conventions used here: when discussing particular letters or digraphs used in spellings, C represents any single consonant letter, while # represents the end of a word. Here's an example: oaC represents the vowel digraph in words like "oak" and "boat", oCe represents the spelling of the vowel in hope and note, and oe# represents the spelling of the vowel in foe and toe. To discuss the actual sound of a vowel when spoken, I will use example words or "lexical sets", which are just words in all-caps that have the vowel in question. So GOOSE represents the "OO" sound found in words like "goose", "loose", "lose", and "prove".
I think the clearest place where English spelling is sub-optimal is the use of spelling variants (mainly redundant vowel digraphs) that never correspond to different diaphonemes (in words that follow regular spelling-pronunciation correspondences). For a discussion of the scale of the problem and the number of alternative spelllings English has for the same sounds, see Masha Bell's analysis . However, some people also think that besides pronunciation, it's also useful to have etymological information encoded in spelling. Interestingly, most of the spelling variants that never correspond regularly to a distinction in speech, also do not correspond regularly to a distinction in etymology. At best, they give hints to the source language of the word; they generally don’t mark distinctions between different sounds in historical or dialectical pronunciations.
The only arguments I can think of that remain in favor of these spelling variants would be:
a) having arbitrary variant spellings for the same sound in different words causes different words to have a more distinct shape. Because of this, it is easier for experienced readers to read (the “bouma” hypothesis). I think this argument is pretty lame, because even if the variants don’t look like each other, one of them might look more like another word (for example, the digraph oa has the same bouma as the digraph ea). And of course, the “bouma” hypothesis was never proven.
b) Having arbitrary variant spellings for the same sound allows us to spell homophones differently, which is a good thing (because it prevents confusion or something like that).
Well, maybe. If this is a real advantage, it would still have to be weighed against the cost of learning and remembering the variants and which words have which spelling.
Neither of these arguments convinces me, so I’m going to make some recommendations for what the optimal respelling of these sounds should be, if their spelling were to be reformed. The recommendations are mainly for a reform that keeps most of the rest of the current English spelling system in place.
List of phonetically and etymologically redundant vowel digraphsEdit
Long O/digraph OE and digraph OAEdit
The spelling patterns oCV, oCe#, oaC, oe# and o# all represent the same vowel sound for all English speakers.
Therefore, the distinction in spelling between words like "bone" and "loan" is in general completely useless. These words rhyme for all speakers of English; and there is no consistent historical difference between "oa" and "o" either. The only slight information the spelling encodes is origin language: words with the “oa” digraph are not Latin or Greek-derived, but generally from Germanic (through Old English or Norse) or French.
Use of o# and oe# is slightly more regular:
- In monosyllables, oe# generally indicates the GOAT lexical set, while o# represents the GOOSE vowel in a subset of frequently used common content words (i.e. to, do, who, two). However, "shoe" and "canoe" are exceptions to this generalization, so it isn't even a regular pattern.
- In multi-syllabic words, generally o# is used to represent the unstressed GROTTO vowel, while oes# is used to represent the plural. Although this rule is somewhat regular, it still causes difficulty for native speakers; consider Dan Quayle’s confusion about the spelling “potatoe”.
Ways to resolve the problem using current sound-spelling correspondences:
Etymology-based: If we wanted to take advantage of the already existent different spellings to repurpose them to mark an etymological distinction, it might make sense to use oCe to represent originally short vowels that became lengthened due to being in an open syllable, and oaC to represent originally long vowels, generally corresponding to OE ā. However, the problem with this is that the reflex of ā was also shortened in closed syllables or due to trisyllabic laxing, as in the first syllable of the word “holiday”. So this method won't be able to indicate the etymology in all cases. And an etymology-based spelling system that indicates a misleading etymology part of the time is worse than useless; under this system, the historical connection between the words "holy" and "holiday" would be even less clear than in the present system.
Simplicity-based: We might choose to use only the digraph “oa,” in order to avoid using the problematic silent e. However, this looks odd in Latinate words, and anyway it is not possible to get rid of silent e without completely breaking the current system of English orthography.
Since we’re going to have to use both long “o” and “silent e” in some cases anyway, we might as well use them in all cases. Therefore, the digraph “oa” should be eliminated, and replaced with “o”, with silent e’s added where they would be necessary. (Masha Bell reaches the same conclusion as I do, coming from a slightly different perspective).
Long E/digraph EA and digraphs EE, IE Edit
For a few dialects, there is a difference between the vowel sounds in "meat" and "meet." But for the great majority of people, they are pronounced the same way, and the vowel in words like "THIEF" is pronounced the same way. In Middle English spelling, such as Chaucer's, it was common to use E and EE for the ancestor of both of these sounds. Furthermore, some words display variation even in Old English, and modern words such as "discreet" and "discrete" show that they have been somewhat interchangeable since then as well.
It would probably be best to standardize the spelling of these words to either long E, the digraph EE, or the digraph IE. The digraph EA is unnecessary.
Digraphs EI/EY and AI/AYEdit
The vein/vain merger is complete for everyone (there is no modern dialect where these two words are pronounced differently). Therefore, the variation between words spelled with ei/ey and ay/ai is in general completely useless. It doesn’t regularly reflect any etymological distinction, as seen by words like “way” and “grey, gray”, and there are signs of this merger even in Middle English, in Chaucer’s time. The merger is even older than the loss of distinct pronunciations of gh, wh and kn. In fact, though, there are very few English words spelled with the digraph EI/EY.
Ways to resolve the problem using current sound-spelling correspondences: Simply choose one or the other.
- Arguments in favor of ei/ey: The sound is phonetically usually more similar to short e than to short a; the spelling ei/ey might also make English spelling more harmonious with other language’s spelling systems. In addition, “ai” is fairly frequently used in English to represent PRICE, although this is mostly in loanwords.
- Arguments in favor of ai/ay: it is vastly the more common representation of this sound in the current spelling system. In some words from French, like “plain”, the entire sound/diphthong developed as a whole from “a”, so there is a relationship. In addition, ei/ey is already used to represent multiple other diaphonemes in English, such as “conceive,” and it represents PRICE in German loanwords. So “ei” could be confusing.
If we want to make as little change as possible from current usage, ai/ay would be best. If we don't care about how much we change the spelling of specific words, but we want to make pronunciation clearest for English speakers, keeping in mind the current pronunciation-spelling correspondences used in English and in the foreign languages that English has loanwords from and that English speakers are likely to be familiar with, using “ey” in all positions would be best. Using “ay” in all positions would also be fairly effective, and would be a slightly smaller change from the status quo.
- using “ei” in all positions: How to pronounce this would probably be unclear to English speakers, who might think of the “ei” in German or the “ei” in English words like “conceive”.
- using “ai” in all positions: This looks foreign to English speakers, and for most generic foreign languages, an English speaker is likely to take “ai” as representing the PRICE vowel.
Note: for speakers of some dialects, either the EIGH sound or the long A sound or both may be distinct from the sound in VEIN-VAIN.
Long U/digraph UE and digraph EU/EWEdit
The spellings ue and ew/eu generally represent the same sound for everyone. Therefore, the systematic difference in spelling between words like "flew" and "clue" is completely useless. It doesn’t consistently reflect any historical or current distinction in pronunciation. The only slight information it encodes is: words from Latin, Greek, or other foreign languages tend not to be spelled with “ew” (however, words from French can be, such as "view" and "curfew"), and with native Germanic words, “ew” is consistently used at the end of past-tense forms of verbs (as with grow-grew or fly-flew), while in other words “ue” is often used, contrary to the historical origin of the sound (as in "hue", "clue" or "true").
Recommendation: this one is tough, actually. The spelling "ew" is in most cases more etymologically accurate for Germanic words, but "ue" could be considered to fit in better with the use of silent e after other vowel sounds. In a spelling system that uses "soft" c and g, using "ue" to replace "ew" would also lead to fewer changes in the spelling of consonants.
Irregular variation in digraphs between W/U or I/YEdit
The general pattern is to use the digraphs Vw and Vy at the end of words and before vowel letters, and the digraphs Vu and Vi before consonant letters.
There are also more or less consistent sub-patterns; for example, Vw tends to be used before word-final n and l, as in "own," "lawn," "strewn" and "owl, "awl". (patterns Vwn# and Vwl#)
(Another subpattern is Vwth, with the suffix -th. Examples: growth,
However, there are many exceptions to the sub-pattern of using w before l: "haul," "maul," "caul."
There are also a few exceptions to the sub-pattern of using w before n. The most common one is "noun"; "faun" is another.
Even considering sub-patterns, some words are just irregularly spelled.
Before d: “crowd," "dowd/dowdy," "howdy," "powder" versus “cloud”, “loud”, “proud," "shroud."
Before voiced s: dowse, browse, drowse/drowsy...
Irregular aw: hawk, gawk, bawd...
Complicated use of digraph IE vs. YEdit
In words like tie, die, pie, lie, that start with one consonant, the digraph -i.e. is used. Other words tend to be spelled with -y, such as cry, dry, fly. The -i.e. words could all be respelled with -y, to give ty, dy, py, ly.
Variations in the indication of vowels affected by a following LEdit
oal, oul and oll (as in roll) seem to be the same for everyone, and the same as olC in general. aul/awl and all seem to be the same for everyone, and generally the same as alC. The use of "oll" and "all" is also undesirably because it leads to an inconsistency when comparing words like "roller" and "dollar," or "taller" and "tallow."
For this topic, it's useful to distinguish between these sequences when word-final and before other consonants.
Word-finally, it seems it would be best to use only awl. So fall, tall, mall > fawl, tawl, mawl. It looks a little odd, but not too bad. For o, the parallel solution "owl" might lead to ambiguity with the "ow" diphthong in words like "howl." Therefore, I'd recommend using "oul", so goal > goul, etc.
Before consonants, it is very rare for short ol to appear (as far as I know, it can only appear before the sounds "f" and "v," in words like "golf" and "solve"). So, the current system of using "ol" might be continued, with exceptions like "shoulder" respelled ("sholder").