Spelling principles - in detail

PC2 and Orm accord

A majority of spellings are common to PC2 and Orm. This includes all vowels, except new eME diphthongs, and most consonants. The fact that PC2, like most OE texts, doesn't mark vowels for length, is ignored. eME distinguishes long vowels from their short counterparts, as is the convention with modern editions of OE texts. So, those spellings that emerge from the first step in the normalisation process outlined above are:

a, a__, b, d, e, e__, f, g, h, i, i__, k, l, m, n, o, o__, p, r, s, t, u, u__, x, ae, ng

where PC2 and Orm don't see eye to eye

When there is conflict beween PC2 and Orm, two options are available:

1. other spellings are considered, either in PC1, OE or secondary eME sources, or
2. the contexts in which certain letter (combinations) can represent different sounds, are defined more clearly.

Those conflicting spellings are:

gg|? /dZ`/, wh|w /xw/, cw|qu/cu /kw/, f|v /v/, ch|c /tS`/, sh|sc /S`/, y%h`|ch/g /G``/ /J`/, y%|g/i /j/, ey%|ei/eg /ei/, ay%|aei/aeg /ai/

more on /dZ`/ and /gg/ in OE and ME

You won't find anywhere in the PC2 texts, a word which had the sound /dZ`/, represented by <cg> in OE brycg and <dg> in ModE bridge. But PC1 has brigge. That's the spelling Orm uses in biggen and egge (OE bycgan, ecg). gg is also the spelling used exclusively by Lmn, AW, Owl, SO and Ch for /dZ`/. So, the solution to the conflict between PC2 & Orm over the spelling of /dZ`/ appears at first glance to be gg. However that digraph is also needed for /gg/ (e.g. dogge, frogge, hagge, hogge, pigge, stagge, wigglen) in eME. Hence the use of cy% for /dZ`/ in the default spelling of eME in this site.

It's not entirely clear what sound(s) cg represented in OE. Just /dZ`/, or /gg/ as well? The development of palatalised /gg/ in WGmc and OE is disputed. It began life as either /gg/ or /G``G``/, but has probably been /dZ`/ since the mid to late OE period, although this view is not held unanimously. Wright believed it may have represented /ɟɟ/ (a voiced palatal plosive) until late OE, and Sweet - until early ModE.¹.

What we can be certain of is that OE words spelt with cg emerge in ModE with both sounds, e.g. OE docga frocga (gg) picg- stacga (gg) brycg hrycg ecg² > MED dogge frogge pigge stagge brigge rigge egge > ModE dog frog pig stag (/g/) and bridge ridge edge (/dZ`/). I've assumed that the eME reflexes had a similar dichotomy of sounds and are thus best represented by dogge frogge pigge stagge bricy%e ricy% ecy%e

turning to OE

PC2 uses two different spellings for /cw/ - cu and qu. However, the latter is used only once in quarterne, while the former is used in - cuen, which appears twice in the texts. Orm, on the other hand, has only cw - in i.a. cwarterne cwellen cwemenn cwenkenn. OE resolves the deadlock for cw.

For OE hw (/xw/), PC2 has either w or uu, while Orm uses wh consistently. To resolve the deadlock, we have hw, which presents no potential confusion. It's also used by AW.

PC2 vacillates between u and f for OE f between vowels (/v/), while Orm uses f in all contexts - initial, medial and final. Since context always dictates the pronunciation of OE f, there's no reason we can't use f medially in eME as well. We can be confident that initial and final f are /f/ and medial f is /v/, just as initial s is /s/ and medial s is /z/ in ModE.

Context isn't quite as clear with OE c, sc, g. Potential confusion arises, so we'll shift those on to the next step.

The spellings which emerge from this step are: cw (/kw/), hw (/xw/) and medial f (/v/)

refining (and extending) the context - c, sc, k and g

In Orm, c is always pronounced /k/ and ch is used for /tS`/, while PC2 retains the dual pronunciation of c, which we find in OE. OE c is pronounced either /k/ or /tS`/, depending on the context. Most OE primers will give the readers a general idea of those contexts. Unfortunately, no matter how you state them, there will always be exceptions. This was already an issue in pre-Norman times apparently, since k was introduced in lOE to represent /k/ before i and e where /tS`/ would be expected. The most notable example is king. By the time PC2 was being written, this use of k was much more widespread. In fact, the basis of a clear and consistent system had emerged: c was /tS`/ before [e e__ ae i i__] and finally, when preceded by [e e__ ae i i__] and an optional [l n]. Elsewhere c was /k/. Conversely, k was only used before [e e__ ae i i__] to indicate /k/. The rules pertaining to final c were still a little murky and inconsistent in PC2. But with a minor tweak they could be made clear for modern users. So rather than shafting this scheme and moving on to the secondary eME sources for a solution, I propose the following: that final c always be pronounced /tS`/. Consequently, k must be used for any /k/ at the end of a word (including those in compounds). That gives us the eME spellings ce__sen, caep, hwilc, ic, king, kin, bo__k.

The rules pertaining to c in eME may seem at first glance, more difficult for the speaker of modern English than simply introducing ch, but in fact they mirror closely the ModE rules for c. c in ModE also changes its pronunciation in front of e and i (to /s/), and is never found at the end of a word as /k/. At the end of word in ModE, /k/ is always represented by k or ck.

It should also be pointed out that final k is not so out of kilter with Orm, PC2 and the secondary eME sources, as it might seem at first. The spelling bok for example, is perhaps more numerous in the examples listed by the MED, than boc. That's because it features in the dative and plural forms - boke bokes. And Ch has bo(o)k in the nominative. The use of final k gives greater consistency in both noun and verb paradigms - bo__k bo_kes³ and scaken sco__k sco__ken iscaken than would otherwise be the case: bo__c bo__kes and scaken sco__c sco__ken iscaken.

Orm has sh for /S`/, while PC2 retains the sc of OE (as does Lmn). OE sc was used almost universally for /S`/, so there's no practical reason not to continue that spelling in eME. As long as we use sk in the few cases where we need to indicate an /sk/ pronunciation - asken, Skotland, sko__l, there will be no confusion.

PC2 has one example of <g> representing a fricative (palatal or velar) /J`/ or /G``/, rather than a plosive /g/ - heglice /heJ`li:tS`@`/, and one example of <gh> in that role - sloghen. More commonly, PC2 has <ch> representing a velar fricative /G``/ (or perhaps palatal /J`/). There are four examples - Burch halechede halechen folecheden. Orm has <y%h`> for /G``/ and /J`/. The deadlock is resolved by OE - g. Context decides how it is pronounced - /g/ at the start of a word, /G``/ elsewhere. OE g also mirrors neatly the behaviour of h, which has two quite distinct sounds - one initial and one medial/final. Note that after front vowels, the latter fricative is actually palatal - /J`/, rather than velar.

In the above cases, in which there is conflict beween PC2 and Orm, the eME spelling is sought by returning to OE but at the same defining more clearly the contexts in which certain letters are used to represent different sounds.

So the spellings which emerge from this step are: c, k, sc, sk, g

A note on alternate schemes: If we allow secondary eME sources, rather than OE, to determine spelling in the case of conflicts between Orm and PC2, then we would have a quite different set of spellings here - ch, k, sch, sc, y%(h). Setting aside for the moment the principle that eME should provide a bridge between ModE and OE, and basing spelling on taste alone, I would still opt for the former rather than the latter. There's something appealing about spellings such as "Englisc", "scip", "hwilc" and "ic", while "Englisch", "schip", "hwilch" (or "whilch") and "ich" strike a slightly discordant note by comparison, perhaps because <sc> for /S`/ and <c> for /tS`/, are unique to past English.

Orm and the secondary eME sources step in

new letter combination ch

Context can't fix all. For occasional /tS`/ before [a a__ o o__ u u__], Orm, backed by most secondary eME sources, provides the following: ch

third sound - new letter y%

In the final case, in which there is conflict beween PC2 and Orm, and OE is unable to provide a solution, the eME spelling is taken from Orm, as its spelling is backed by most secondary eME sources, though in the slightly modifed form known as yogh - y`. The spelling that emerges from this step is: y%

It should be pointed out here that y% is not actually a new letter. It is in fact the shape of OE's insular 'g' - y% - used in OE writing for /g/, /j/ and /G``/. g is in fact the newcomer, borrowed from continental Europe for /g/ (generally found initially). In a sense, y% is simply carrying on a long OE tradition.

extending y% to the diphthongs

PC2 has uureide, eie, rachenteges, daeies and saegen. Orm has ey%y% and ay%y% throughout its texts.

There are two ways of broaching the spelling of the diphthongs in eME. Firstly, we could see them through OE eyes as reflexes of ae + g` and e + g` etc, in which case they simply become a + y% and e + y%, by applying both the established OE to ME sound changes and the spelling established thus far. The merging of the long and short vowels in these combinations doesn't affect the choice of the final letter (y%). Thus we arrive at the eME spellings - ey% and ay%.

Or we can assume that the second element in these diphthongs is something like /I`/ found in the corresponding ModE diphthongs. From that perspective, PC2 aei and ei fit the bill, and they match spellings in the secondary eME sources. But a minor tweak is needed to conform with the established eME spelling for /a/, giving the result - ai and ei.

I prefer the former perspective, for a couple of reasons. A clearer path to OE is presented by ey% and ay%. Also, verb paradigms are less confused and relations between cognate words are less blurred with a g/y% alternation, rather than g/i. And last but perhaps not least, the beginner has an easier path. He or she doesn't need to conceive of ey% and ay% as a separate "beast". They don't need to create a separate compartment in their heads for diphthongs. They can simply view a + y% and e + y%, as they would a combination of a or e with any consonant.

The spellings that emerge from this step are: ey% and ay%