Anglish Pronunciation

From The Anglish (Anglisc) Wiki
(Redirected from Anglisc Pronunciation)

French, Latin, and Greek do not seem to have influenced English pronunciation much (unless it turns out that French influence caused the Great Vowel Shift, as some scholars postulate), but this article will discuss what little we have found.



In Middle English, two new diphthongs appeared as a result of French loanwords: /ɔɪ/ and /ʊɪ/. Examples include joy, boil, and destroy. In Early New English, words with /ʊɪ/ came to be pronounced with /aɪ/, which explains why in certain nonstandard dialects, line and loin are pronounced the same. However, this merger was never completed in standard speech and ultimately failed to establish itself, as /ʊɪ/ would end up being replaced with /ɔɪ/. In the end, words with /ɔɪ/ are foreign or ultimately show foreign influence.

How can Anglish nativize a loanword with /ɔɪ/? Probably, oi would be felt to be two separate vowels and would become something like /oʊɪ/ or /oʊi/, much like how speakers may pronounce the vowels in buoy (/buːi/) separately and not as a diphthong.


/ʒ/ (the zh sound) does not appear in Old English at all. /ʒ/ instead arose from French loanwords that historically had /zj/ such as measure and vision. The /zj/ sequence was not present in native words, so no native word in standard speech has /ʒ/. Hence, French is indirectly responsible for the introduction of the phoneme /ʒ/. The phoneme itself arose from native sound changes, but the environment in which it arose was present only in foreign words, which is why it can still be considered foreign. This applies only to medial /ʒ/ resulting from historical /zj/; in recent loanwords such as genre, /ʒ/ is from the original foreign pronunciation.

If a loanword with /ʒ/ entered Anglish, and someone attempted to nativize the pronunciation, /ʒ/ would likely be replaced with /ʃ/, the voiceless equivalent. This accords with how /tʃ/ was occasionally substituted for /dʒ/ as explained above. While one can argue that it would be replaced with /dʒ/, as is sometimes the case with non-medial /ʒ/ in today's English, initial /dʒ/ would be rather unfamiliar without the borrowing of French loanwords to help establish initial /dʒ/, whereas /ʃ/ in initial position would be familiar in Anglish. Of course, it can go either way with final /ʒ/ since /ʃ/ and /dʒ/ are both common in final position.

Initial /v/[edit]

In Old English, [v] was an allophone of the fricative /f/ that appeared between voiced sounds. For example, /f/ in þēof (thief) was [f], but /f/ in þēofas (thieves) was [v]. This explains the consonantal changes in singular-plural pairs such as wolf-wolves as well as noun-verb pairs such as thief-thieve. Loanwords that had initial /v/ were altered to have [f], as shown in OE fers (Latin versus) and fann (Latin vannus), so clearly there was a rule that converted initial [v] to [f].

In Middle English, this rule broke down because of the goodly number of French loanwords beginning with /v/, such as venom, virtue, vice, and violence. As a result, [v] became a phoneme in initial position, and /f/ and /v/ can now contrast, e.g., fine and vine, file and vile.

Note that in Middle English, the southern dialects had undergone a change in which all initial fricatives were voiced, so finger and fox became vinger and vox. In the regional dialects that have inherited these voiced forms, /v/ is of course a native phoneme. But these dialects evidently had little influence on the standard language in this aspect, as very few dialectal forms of this kind became standard; these include vat, vane, vixen (the native set), and vial (a variant form of phial). As a result, it can be safely assumed that the southern dialects played no role in the phonemicization of word-initial /v/ in the standard language. Since French loanwords were responsible for the allowing of initial /v/, it is reasonable to assume that without French influence, we would say fat, fane, and fixen in the standard language instead.

How would Anglish deal with a loanword with initial /v/ such as violin? /v/ might be retained in pronunciations that clearly attempt to emulate the original pronunciation, but a nativized pronunciation would probably change it to /f/, so violin would be pronounced as if spelled fiolin.

Initial /z/[edit]

Like /v/, /z/ was not a phoneme in Old English. Rather, [z] was an allophone of /s/ that appeared between voiced sounds. For example, /s/ in hūs was [s], but /s/ in hūsian was [z], which explains the consonantal difference in the noun house and the verb house. Loanwords with initial /z/ appear to have been altered to have [s] in Old English, though the only evidence is alliteration with foreign names, e.g., zefferus (Zephyrus) was once made to alliterate with swifta (swift).

In Middle English, [z] became a phoneme. Unlike /v/, however, foreign words played a far smaller role in the phonemicization, as foreign words with initial /z/ were rare, and these sometimes were spelled with s (suggesting /s/), e.g., zodiak (also spelled sodiac), zele (also spelled seele). Later on, a few Ancient Greek words with initial /z/ were borrowed, and initial /z/ was firmly kept, e.g., zeugma, zone. These words are generally learned or technical in character, which explains the rarity of initial /z/.

How would Anglish deal with a loanword with initial /z/ such as zodiac? /z/ might be retained in pronunciations that clearly attempt to emulate the original pronunciation, but a nativized pronunciation would probably change it to /s/, so zodiac would be pronounced as if spelled sodiac.

Initial /dʒ/[edit]

In Old English, the sound that would later become /dʒ/ arose from palatal g that was either geminated (e.g., ecg, edge) or after n (e.g., sengan, singe). Thus, we should expect /dʒ/ to never appear in initial position. This, however, changed when many French loanwords with initial /dʒ/ appeared such as judge and general. Thus, /dʒ/ now may appear in initial position.

There is a small number of supposedly native words with initial /dʒ/ that are traced back to older forms with /tʃ/, but all these words have a murky history:

  • Ajar - said to have undergone a sound change in which a stressed syllable after an unstressed one is voiced. In this case, the older form is achar, in which char is an obsolete native word meaning turn. But if this is the case, then ajar not only appears to be the only native example showing this sound change but also is the only example showing /tʃ/ > /dʒ/; other examples of voicing are found in loanwords or foreign names and show /(k)s/ > /(g)z/, e.g., resign, executive, Alexander.
  • Jaw - the etymology is uncertain, but it may be connected to French joue (cheek) or have been influenced by it.
  • Jowl - an extremely perplexing word that may be connected to two obsolete words: chavel (jaw, from OE ceafl) and cholle (throat, possibly connected to OE ceole). Because of the similarity in meaning that these words have to jaw, the consonant may have been altered from influence of jaw and indirectly French joue.

In short, evidence that this is a native sound change is dubious. And in any case, these words are extremely few in comparison to the far greater number of loanwords with /dʒ/.

The one case in which initial /dʒ/ may appear in native words is due to a sound change called yod coalescence, in which certain consonant clusters with /j/ as the second element undergo palatalization. In this case, /dj/ has become /dʒ/. This sound change has occurred in a few dialects (mainly British English, but not American English), so in those dialects, the word dew may be pronounced like Jew. Note, however, that all these cases stem from a word that historically began (and still does for many dialects) with /d/, and only occurs before historical /uː/, e.g., dune, endure. So even with these words with historical /dj/, the distribution of /dʒ/ without foreign influence would still be extremely limited, and for speakers whose dialect this sound change has not affected, initial /dʒ/ would not exist in native words.

A useful rule of thumb is that any word spelled with initial j or soft g is foreign or shows foreign influence, and any word spelled with initial d shows a native sound change. However, not all words fulfilling the latter condition are native. For example, the word duel may be pronounced with /dʒ/, but the word is from Latin.

How would Anglish deal with a loanword with initial /dʒ/ such as gelato? One solution is to simply keep /dʒ/, just as one may keep /ts/ in the Japanese loanword tsunami. Another solution is to nativize the pronunciation by replacing /dʒ/ with another sound, just as one may nativize the pronunciation of tsunami by changing /ts/ to /s/. Evidence from Middle English indicates that some Englishmen may have substituted initial /dʒ/ with /tʃ/, which is phonetically the voiceless equivalent; jealous can be found in Middle English spelled as chelous. Hence, a nativized pronunciation of gelato would probably pronounce it as if it were spelled chelato.


Hard and soft C and G[edit]

It was due to French influence that the letters c and g have respectively been given the values of /s/ and /dʒ/ before the letters e, i, and y, as shown in city and gentle. Though this was originally meant to reflect the French pronunciations, this has had long-lasting effects on pronunciations of many later loanwords from Latin and Greek. For example, the Latin loanword censor is pronounced with /s/, even though it had /k/ in Latin, and the Greek loanword gymnasium is pronounced with /dʒ/, even though it had /g/ in Greek. Thus, even if one accepts certain loanwords from Latin and Greek, one must account for the fact that French has affected the traditional pronunciations of those words.

How can Anglish pronounce foreign words like these? Suppose that one were to try pronouncing the name Cicero. Without French influence to give the letter c the value of /s/, how would Cicero be read? Classical Latin pronunciation was reconstructed relatively recently, so early speakers would not have based the pronunciation on Classical Latin. Rather, they would have based it on their spelling conventions. According to Anglish spelling conventions, the rules for using c are:

  • c before a, o, and u - /k/, e.g., cat, corn, cunning.
  • ce before a, o, and u - /tʃ/, e.g., cearlock (for charlock), ceore (for chore), ceuck (for chuck).
  • k before e and i - /k/, e.g., keep, kin.
  • c before e and i - /tʃ/, e.g., ceek (for cheek), cild (for child).

This is essentially the rule about hard and soft c, except that soft c represents /tʃ/ instead of /s/, and there is no separate letter for /tʃ/ before a, o, and u (unlike /s/ before those letters). With these rules, the traditional Anglish pronunciation of Cicero would have /tʃ/ as if it were spelled Chichero.

What about a loanword such as Germania? According to the rules in Anglish Spelling:

  • g before a, o, and u - /g/, e.g., game, god, gum.
  • ge before a, o, and u - /j/, e.g., geard (for yard), begeond (for beyond), geup (for yup).
  • g before e and i - /g/, e.g., geese, begin.
  • g before e and i - /j/, e.g., ges (for yes), gippee (for yippee).

Unfortunately, Anglish Spelling does not assign a distinct letter to either soft or hard g before e and i. It is not quite clear how Old English writers would have read Latin words with ge or gi. There is, however, an instance of a scribe writing Latin Georgius as Iorius, indicating that he read Latin ge as having soft g and thus used consonantal i to show /j/. Hence, it is assumed that Latin ge and gi would be interpreted as soft, and Germania would have /j/ as if it were spelled Yermania. As for why the Old English attestation was spelled Iorius rather than something like *Ieoriius, this probably was from a desire to avoid three vowel letters in a row; this is also shown in iāces, a rare spelling for gēaces (cuckoo; the word became yeke in Middle English).

There is, however, one exception to how soft g is read: after n. In words such as singe, it seems that Anglish spelling uses ge for /dʒ/ after /n/. This is the only time when soft g should be read as /dʒ/.

In short, because of French influence, soft c and g have been given the values of /s/ and /dʒ/, so for any Latin and Greek borrowings with soft c and g, they should never be pronounced with the French sounds (except for soft g after n). Instead, the traditional Anglish pronunciations of soft c and g would be /tʃ/ and /j/, respectively.


In Old English, the letter f was used for [f] and [v], since the voiced fricative was simply an allophone of /f/. As a result, the sound value of f was generally predictable. For example, in the word drīfan (drive), since f was flanked by two voiced sounds, f represented [v].

It was due to French influence that the letter v (also u if not used word-initially) was made common in scribal use, and the two sounds were distinguished. Not only were they now separate phonemes, but they were distinguished orthographically. Hence, the letter f lost its ability to represent [v]. The only exception is the function word of, which originally had [f] and came to have [v] as part of a later development.

Anyway, as a result of f now being read only as [f], however, revived names such as Alfred have been given a spelling pronunciation. In Old English, Ælfrǣd was certainly pronounced with [v]; this is supported by Middle English spellings such as Alured (remember that u stands for [v]), as well as the modern name Leverich (from OE Lēofrīc). Ælfrǣd was not an ambiguous spelling since /f/ was right between two voiced sounds, so it would be read as [v]. The name apparently declined in use sometime after the Norman Conquest and was revived later on. But the spelling with f was kept, so people thought that it was pronounced with [f]. As a result, the current pronunciation of Alfred is a spelling pronunciation showing indirect French influence removing the ability of f to be read as [v].

As for what Alfred would be like if the name had survived naturally, it would probably be something like Alvered or Alred from loss of /v/ before a consonant, as shown in words like head (OE hēafod) and lord (OE hlāford).


The letter j was not used in Old English. In fact, the letter did not even exist at the time. Instead, to show its consonantal value, the letter i was used, which is the case in Latin. The letter j is a prolonged form of i that was used at first to help distinguish it from i.

Anyway, in Old English, consonantal i represented /j/ and appeared mainly in foreign names and in occasional spellings of a few native words with initial /j/, which was otherwise represented as ge and before /u/ as geo, e.g., iung for geong (young), iūguþ for gēoguþ (youth), ioc for geoc (yoke), for geō (formerly), iesca for gesca (hiccup). However, after the Norman Conquest, from French influence, consonantal i (and its later form j) was given the value of /dʒ/ instead, so names like Jacob and Jesus are not read with /j/. Hence, j having /dʒ/ instead of /j/ is due to French. This use of j is unlike how it is used in the other Germanic languages, in which j normally represents /j/, e.g., German Jahr (year), Dutch jong (young).

It is reasonable to conclude that without the Norman Conquest, consonantal i and its later form j would represent /j/ in English. Hence, Latin words and names spelled with initial j such as janitor and Jacob would be pronounced as if they were spelled yanitor and Yacob. In other words, with a few exceptions such as modern Romance names and borrowings with different values for the letter j (e.g., French Jacques, Spanish jalapeño), j in words and names would normally be read as /j/.


The vowel letter o came to have the value of short u in a few words apparently as a result of a French orthographic practice of writing o for u to prevent misreading. For example, in the words come and monk, o is written instead of u before m and n. Hence, without French influence, it is reasonable to assume that o would generally not have come to be written for short u in certain words.

As a result of this practice, a few uncommon or disused words have gained spelling pronunciations. For example, wort (from OE wyrt) rhymes with hurt, as expected. However, the alternative pronunciation that has it rhyme with sort is unetymological and is based on the spelling. Another example is wont, which is supposed to rhyme with hunt, but because it is now an uncommon word and found mainly in writing, a few unetymological spelling pronunciations have arisen. Hence, the Anglish pronunciation of wort and wont should always have them rhyme with hurt and hunt, respectively.

Note that there are a few words that originally had o, but came to have u through a sporadic sound change, e.g., among, oven. In this case, the o in the spelling is etymologically justified, even though it is no longer aligned with the modern pronunciation.


In Old English, ss represented /ss/, and after degemination, this became /s/, and ss afterwards became generalized as the spelling for /s/ after a short vowel, e.g., kiss, blessing. For native words with [z] in Old English, only one s was used e.g., OE bysig > NE busy (see U for an explanation of the spelling of the vowel). After short vowels, s was generally replaced with zz, e.g., OE dysig > NE dizzy. Hence, after short vowels:

  • /s/ - ss (rarely s), e.g., bless, miss, grassy, us.
  • /z/ - zz (rarely s), e.g., dizzy, buzz, busy, is.

However, in a few foreign words, ss had the value of /z/, e.g., possess, scissors, dessert.


Th was used in early Old English for the dental fricative, but it was later replaced with þ (thorn) and ð (eth). The two letters appear to have been used interchangeably, but there was apparently some tendency to use thorn word-initially, and eth elsewhere. In Middle English, the two letters continued to be used at first, but eth soon fell into disuse, and thorn was increasingly replaced with th. In some cases, early printers replaced thorn with y, so ye appeared as a printed form of the; there is absolutely no historical basis for pronouncing ye (representing the) with /j/.

In any case, th is now used for the voiceless fricative /θ/ (like in thin) and the voiced equivalent, /ð/ (like in other). The use of th is fairly predictable in native words in that medial th shows the voiced fricative, except for compounds such as bathroom, derivatives such as wrathful, and inflections such as deaths (but even then, there are exceptions such as worthy and northern). However, this predictability in the pronunciation of medial th is disrupted by Ancient Greek influence in loanwords such as anthem and cathedral. As a result, medial /θ/ can now appear in simple words, and medial th in simple words can represent /θ/.

Moreover, some words with /t/, whether native or foreign, are spelled with th from Ancient Greek influence; for example, thyme is an Old French loanword, but is pronounced with /t/ from its Old French pronounciation. The word was often spelled with th because the word was ultimately from Ancient Greek thymon. Sometimes, etymological spelling has caused the word to be pronounced with /θ/ again; for example, authentic was borrowed from Old French autentik, but as the word was from Ancient Greek authentikos, it was respelled with th, and the pronunciation with /θ/ is from the spelling.

In some cases, the spelling with th is unetymological. For example, the Thames (from OE Temes) is pronounced /tɛmz/, but it came to be spelled with unetymological th under a mistaken assumption of an Ancient Greek connection. This has even caused unetymological pronunciations. For example, Anthony is from Latin Antōnius, which also entered English as Antony; the unetymological th spelling has caused Anthony to be pronounced with /θ/.

In short, there are four cases with foreign th and /t/:

  • Etymological th with /t/, e.g., thyme.
  • Etymological th with /θ/, e.g., authentic.
  • Unetymological th with /t/, e.g., Thames.
  • Unetymological th with /θ/, e.g., Anthony.

In any case, any word spelled with th but pronounced with /t/ either is foreign or shows foreign influence.

The use of th for the dental fricative has caused a few spelling pronunciations as a result of misreading th as /θ/. For example, the place name Gotham is pronounced with /t/ (h in -ham having become silent), but when this was later used by speakers as a nickname for New York City, they gave it a spelling pronunciation with /θ/.


In Old English, u represented /u/ or /uː/ (length was not consistently marked). Thus, like the other vowel letters, the expected values that it would represent in modern English would be /ʌ/ or /ʊ/ for "short" u and /au/ for "long" u. However, two changes disrupted this.

First, in Middle English, /uː/ gradually came to be represented with ou/ow from French influence. Hence, instead of using silent e like the other vowel letters to mark historical vowel length, a separate digraph is used for this. As a result, OE hūs came to be spelled as house and not *huse.

Moreover, in Middle English, u also represented /y/ in French loanwords such as measure and music. The French sound, however, was unfamiliar to English speakers, so as part of naturalizing French /y/, it was replaced with /u/ in stressed closed syllables (e.g., judge), but was otherwise replaced with /iu/, which ultimately became /juː/. Also, because ue was no longer used for historical /uː/ (for which ou was now used), it now represents /juː/. Hence, OE hīew became hue in spelling, not hew. If French had not affected English spelling in this aspect, due would only represent /daʊ/, never /djuː/.

Both changes have affected the traditional English pronunciation of Latin words, since u does not accord to the traditional rules of "short" and "long" vowels. Instead, the pronunciation of u depends on whether the syllable is open or closed, and the underlying sound for open u is /juː/, not /au/. Hence, the traditional Latin pronunciation of tutor does not have it rhyme with outer.

In a minor case, as part of a French convention, u was used in some Middle English dialects for /y/; other dialects that had already lost /y/ thus did not use u that way. This use was kept when /y/ became a genuine u sound (e.g., ME crucche > NE crutch), but when /y/ was unrounded to /i/, the sound was represented as i again. However, busy (OE bysig) still shows the French use of u for historical /y/; the normal spelling of this word would be bizzy, as shown by how OE dysig became dizzy. Hence, the use of u for i shows foreign influence.


In English, the native value of x is /ks/. However, in a few words, x has two values: /gz/ and /z/. Neither of these is found in native words.

For /gz/, this originated in French words with x, which were often words with the ex- prefix. In French, x generally represents /gz/ if followed by a vowel, whence the pronunciation in words such as example, examine, and exist. This is also reflected in the modern French pronunciation of these words. However, a few words such as exile and exorcise may show /ks/ from Latin influence, since in Latin, ex- always has /ks/.

These two influences have led to a pattern in English in which if the syllable following x is stressed, then x usually has /gz/; otherwise, if the syllable with x is stressed, it usually has /ks/. This is shown by the pronunciation of executive and execute. Because of this pattern, the pronunciation of a few other words with x has been affected as part of an English development. For example, Alexander is pronounced with /gz/ in English, even though it is pronounced with /ks/ in Greek (as well as in French). Likewise, anxiety has /gz/ in English, even though the Latin pronunciation has /ks/. These pronunciations are ultimately due to the French influence on x. Hence, to undo this, one would always pronounce non-initial x as /ks/ in Latin and Greek words, regardless of stress.

For /z/, this arose from a reduction of /gz/, whence the pronunciation of such words as Xerxes and xylophone. This sound value is not from Greek, since initial x in Greek represented /ks/. Since initial /ks/ in not allowed in English, it would be simplified to /s/. Thus, one would pronounce Xerxes as if it were spelled Serxes (which also aligns with initial /z/ not appearing in Anglish).


There is some evidence that z in Old English, though very rarely used, was pronounced as /ts/, e.g., bezt for betst (best). This continued on in Middle English, e.g., milze for miltse (mercy). This use of z was also present in Old French, e.g., voiz (voice). In English and French, /ts/ was later simplified to /s/, and since z became redundant in its use for /s/, the letter was given the new use of representing /z/ in French use. This use of z was then brought over to English, though never used as fully as v. As a result, z used for /z/ came from French.

In Old English, for a few foreign names from Greek, z was used apparently with the value of [s] in initial position, as shown with how zefferus (Zephyrus) was once made to alliterate with swifta (swift).

The use of z for /z/ has led to some unetymological pronunciations. For example, the German loanword zeitgeist is pronounced with /z/, but in German, z is used for /ts/ instead. Since initial /ts/ would be simplified to /s/ in a nativized pronunciation, it would be more correct to pronounce zeitgeist with initial /s/ instead (which also aligns with initial /z/ not appearing in Anglish). Moreover, the normal pronunciation of a few Scottish names such as Mackenzie has /z/, but the spelling with z arose only because z was used for ȝ (which represented /j/) in some early Scottish printers, so the current pronunciation is a spelling pronunciation.



The vocalism in briar is unexpected, as the OE form was brēr, which would have yielded breer. In ME, the vowel was /eː/ and was raised to /iː/, which happened before the Great Vowel Shift began and caused /iː/ to become diphthongized. This sound change happened in a few other words, e.g., ME frere (friar), de (die), quere (choir), and entere (entire). The fact that the raising of /eː/ happens almost entirely in French loanwords highly suggests that the French vowel was the key in this sound change, so it was not a sound change that simply affected all instances of ME /eː/.

It is not quite clear why this change happened. One explanation from the scholar Richard Jordan is that the French /eː/ appears to have been tenser than the native equivalent, whence the raising. Hence, it appears that ME brere became briar from influence of phonetically similar French words such as frere. However, this does not explain why other French words with /eː/ such as chief and brief underwent the usual development of /eː/ > /iː/. Whatever the cause may have been, as French influence in some way has affected the vowel in briar, the Anglish form to use would be breer.

The verb tire is probably not an example of this; though it is attested in OE as tēorian, a variant of tȳrian is also attested.


The traditional pronunciation of brothel has the voiced dental fricative /ð/, which is expected from the etymology. But brothel is now sometimes pronounced with /θ/. There is no reason to think that the fricative was naturally devoiced or that the voiceless fricative was introduced by analogy since brothel is certainly not felt to be a derivative. Instead, the pronunciation with /θ/ must be a spelling pronunciation as a result of Ancient Greek influence allowing for medial /θ/ in simple words such as cathedral (see TH above). The Anglish pronunciation of brothel is thus the traditional one and uses /ð/.


The word came from a shortening of housewife, which was later formed again once hussy gained a pejorative meaning. The word was traditionally pronounced with /z/, and the usual spelling should have been huzzy (which is attested). However, ss was used for hussy, and the use of ss for /z/ is foreign (see SS above). As a result of ss being allowed to represent either /s/ or /z/, and hussy not being a particularly common word, an unetymological spelling pronunciation with /s/ later came to be used for hussy. The Anglish pronunciation of hussy ought to have /z/.

Foreign Influence on Stress[edit]

Old English followed the Germanic Stress Rule (GSR), in which the main stress was placed on the leftmost syllable of lexical roots. Hence, for the word father, the stress is always on the first syllable of the root, even if one adds affixes to it, e.g., fatherlessness.

Some prefixes are stressed, so in a word such as andswaru (answer), the stress falls on the prefix and- because and- is a stressed prefix, with secondary stress assigned to the second syllable. But in the word gewriten (written), the stress falls on the second syllable because ge- is an unstressed prefix. The first pattern is the one used in compound nouns, as shown in words like manslaughter and churchgoer.

Moreover, there is also no shift in stress in derivatives belonging to a different part of speech; for example, the noun andswaru has the stress on the first syllable, and even in the derivative verb andswarian, the stress is still on the first syllable. Likewise, the verb forgiefan (forgive) has the stress fall on the second syllable, and in the derivative noun forgifnes (forgiveness), the stress stays in the same place.

An important part about the GSR is that syllable weight plays no role. To be clear, a syllable can be either light or heavy. In this context, a light syllable consists of a short vowel. All other syllables are heavy. With the GSR, the stress is always on the first syllable of a root, regardless of weight, as shown in written and writing.

The GSR was challenged by French loanwords in Middle English, however, as they had a different pattern of stress.

  • If the final syllable does not have a schwa, then the stress falls on that syllable.
  • Otherwise, the stress falls on the next to last syllable.

For example, in Middle English, the French loanword cite (city) originally had the stress on the second syllable since it had a long vowel. And the French loanword pilgrimage (in which the last e was pronounced) had the stress fall on the third syllable since the last syllable had a schwa and was thus deemed extrametrical.

Of course, the French rule never became dominant, since many words such as city and charity were later altered such that the stress was now on the first syllable. Some loanwords still keep their old stress, as shown in divine and reward. It should be noted, however, that words like condicioun (condition) and adversitee (adversity) were stressed on the second syllable instead, the first syllables seemingly treated as unstressed prefixes. But these prefixes must have been seen as having pretty much no meaning, given that the simplexes of ad- and con- words seldom appeared in ME and would have no clear meaning to native speakers. As a result, the prefixes served as a way to contrast with other prefixed words (e.g., adduce from reduce, duce having no meaning of its own to English speakers).

What changed English stress much more was the later flood of Latin loanwords. The stress rules for Latin were:

  • If the next to last syllable is heavy, then it is stressed.
  • Otherwise, the stress goes to the third to last syllable.

Basically, one-syllable and two-syllable words had the stress fall on the first syllable, which made it identical to what the GSR would yield. But things became more complicated with words of three or more syllables. The word instrument (an early French loanword) is stressed on the first syllable, but with the addition of the Latin suffix -al, the stress is placed on the next to last syllable since ment is heavy.

Because of these Romance stress rules, it can be said that our current stress system is essentially a mix of Germanic and Romance stress rules. Here is a relevant passage from a book: "It is therefore often said that Middle English replaced the [Germanic Stress Rule] with a Romance Stress Rule (e.g. Lass 1992: 83-90), which it imported together with a large number of French and Latin loans in the wake of the Norman Conquest."

Incidentally, English is not unique in having its stress system affected by Romance loanwords. The same thing has happened to Dutch and German, though their systems work a bit differently. However, Icelandic has been unaffected by this.

With this in mind, the pronunciation of a few native words diverges from what we would expect if they had kept their native stress pattern. It is unclear right now whether the foreign influences on the stress system were involved, but these divergences are noted here anyway.

  • In mankind, the stress should be on the first syllable (as shown in other compounds like manslaughter and maneater), but instead, when the word has the meaning of humanity, the stress is on the second syllable.
  • In eleven, the stress is unexpectedly on the second syllable, even though the stress was originally on the first syllable (as shown in the Icelandic cognate ellefu). The stress shift happened in Middle English, as shown by the rare and obsolete variant leven (in which the first syllable of eleven, now unstressed, had become dropped). If the stress had not shifted, the word would have become elleven instead. According to the OED, the current pronunciation may have come from four-syllable inflected forms in Middle English such as elleuene.
  • In -teen words such as thirteen, the word is often stressed on the second syllable, even though the stress originally fell upon the first syllable only. While this change in stress may be due to rhythmic reasons, it is noteworthy that it never happens in -ty words such as thirty as well as words with other native suffixes such as hopeful and happy.

The stress of certain prefixes has also been affected in a few words.

  • In afternoon, the stress is generally on noon, even though the prefix in Old English was stressed when used in nouns, as shown in words like aftermath and afterthought.
  • In outlandish, the stress should fall on the prefix out- (as shown in the pronunciation of outlander), but instead, the stress is on the second syllable.
  • In inside and outside, the stress varies instead of having one fixed stress. The stress on the second syllable can be attributed to analogy with beside(s), as inside and outside are often used as adverbs and prepositions.
  • In misdeed, the stress was on the first syllable in Old English since the nominal prefix mis- was stressed. Instead, the noun deed is stressed. Note that in mistake, the noun was derived from the verb, in which the stress was on the root rather than the prefix, so the stress pattern in mistake is in accordance with the GSR.
  • In ordeal, the prefix was stressed in Old English, and in fact, an older modern pronunciation of this word, as noted by old dictionaries, had the stress fall on the prefix. But the stress has now shifted to deal.