English Spelling Reform

dudefellow's picture

Here is a spelling reform for English that I propose [See image "EnglishSpellingReformEmblemJPEG"].

dudefellow's picture

The phonetic symbol in the sixth column and fifth row of the table on the right hand side should be the same as in the cell of the fifth column and fifth row.

George Thomas's picture

How about posting a copy of the image that is perhaps 15% of the size you have it now? It's so damned big I can't in any way see it all, even reducing my browser view to the bare minimum.

dudefellow's picture

Here is the smaller image requested [See Copy of EnglishSpellingReformEmblemJPEG15percent]:


There is no reason for the capital version of the symbol in the eighth column and seventh row of the right hand table except that it was converted by automatic capitalization by the word processor software I was using.

John Hudson's picture

Most would-be reformers propose, in a variety of misguided ways, to 'simplify' English spelling. There doesn't seem to be any risk of that here, so this is in its way oddly refreshing.

hrant's picture

Wow.
Indeed more complexity is what we need.
But you'd need a king to make this happen.

hhp

quadibloc's picture

Although I can't really determine, from the image, how it is proposed to reform English spelling, while most reforms do aim at simplifying the spelling by making the rules uniform, as that implies having one symbol for each sound, that usually involves increasing the size of the alphabet to about 45 letters, because that's about the number of phonemes in the English language.

So the fact of a larger alphabet in itself doesn't mean things will be made more complicated.

dudefellow's picture

Yesterday I pointed out the phonetic symbols in the cell of the sixth column and fifth row and the cell of the eighth column and seventh row. In both of these cells, another allophone represented by their different phonemes is the one that is represented in the International Phonetic Alphabet by "double storey a" where it occurs in diphthongs. It might seem strange to some of you if in my phonemic system, the same allophone can be represented by different phonemes; for example the first part of the diphthongs in the words "how" and "high". There are also differences in pronunciation that arise due to dialectal variation, which is a good reason for the benefits of a phonemic orthography. The table of the right hand side is supposed to contain phonemes rather than the more precise phonetic allophones, but I had to represent them by the symbols of the International Phonetic Alphabet, even though this leads to overcrowding in the cells, because there is not a well-known publically disseminated conventional phonemic symbology for English; although even if there were, it would probably be not the same as the system that I use. The acoustic, phonological and articulatory phonetic theory that I developed indigenously for the English language differs from the one that I expect educated phoneticians to be familiar with. I emphasize that my spelling reform is phonemic more than phonetic to the degree of allophonic. To explain the distinction between a phoneme and an allophone by way of a typographical analogy, consider the graphical difference between a letter modified by the context of its surrounding letters in a ligature compared to how it normally appears when not ligatured. We would not say that the "f" in the ligature "ffi" is a different letter of the alphabet than an ordinary "f". Similarly, phonemes are phonetically distorted from their recitation forms to allophones by the context of their adjacent phonemes, such as by articulatory assimilation and anticipation, but we would not in every case say that they become different phonemes by these processes, unless it is something akin to epenthesis or elision, as examples.

In the cell of the eleventh column and fifth row of the table on the right hand side, the phonetic symbol should be "z" rather than "s".

quadibloc 4 Jul 2013 — 9:51pm: "how it is proposed to reform English spelling [...] having one symbol for each sound, that usually involves increasing the size of the alphabet to about 45 letters"

As you can see from the table on the right hand side, the phonemic system that I propose for the English language has far fewer symbols than 45. In fact, I manage quite well with only thirty, two of which are very rare except etymologically and tend to be pronounced correctly only in words imported from foreign languages. The discrepancy in number of symbols to sounds arises for the most part in what other phoneticians call "rhotacized vowels". I do away with such a notion, as I do not consider them to be phonemes. I also divide diphthongs each into two phonemes, which are then treated as merely allophonic variations of monophthongal targets. This gets rid of some more phonetic symbols for vowels.

dudefellow's picture

The phonetic symbols in the cell of the fourth column and eighth row of the right hand side table should be the same as those in the two cells of the fourth column and fifth and sixth rows. I expect that it should have been possible to guess its approximate correct sound from the patterns assigned to the other vowels in the right and middle tables. The letter when it has this pronunciation may in most cases be considered to be a rhotacized vowel because it most usually occurs before a following letter "r". Nevertheless, this rhotacized vowel would not be considered to be a phoneme in itself separately from all non-rhotacized vowel phonemes, since it does not occur contrastively with respect to them in the absence of a following "r".

JamesM's picture

> But you'd need a king to make this happen.

Or convince the general public to start using the new system, which seems extremely unlikely.

agisaak's picture

Since I can't make heads or tails of your charts, it's rather difficult to comment on this other than by saying that some sort of explanation should be included along with this. Why are there so many duplicate symbols in your lower two charts? (/ɣ/ and /x/ occur at least twice despite the fact that the former is not a phoneme of English, and the latter only occurs in a rather small number of dialects. /f/ and /v/ appear at least four times each for reasons which are a complete mystery, as do many other symbols.) Also, if you claim to be creating a phonemic system, why do /ts/, /ks/, /gz/ and /ft/ appear in your table at all? These are sequences, not phonemes.

Most importantly, if you want to propose a spelling reform, you'll need to point out what you think is wrong with the current system before trying to explain why people should bother with your new system.

André

dudefellow's picture

There is another way of portraying the same information but by means of only a single table, with each letter grapheme of the English Roman alphabet appearing only once on one axis and a list of sounds on the perpendicular axis. Some of the sounds are phonemes, while others are diphones, because sometimes a letter can behave as a diphone in English spelling. Then the cells in the table at the intersections of the columns and rows contain the characters that I propose for the spelling reform. I drew such a table before designing the picture figure with three tables that I have shown above, and perhaps I can scan a version for you soon. The reason that I made the diagram with the three squares is that it suggests how the characters in the middle matrix can be given keyboard encoding shortcuts in such a way that each letter can be typed by a permutation of just two keys selected from only eleven, using the co-ordinates of the columns and rows; and also because I thought it would look better by not having empty cells.

agisaak 5 Jul 2013 — 11:56am: "Why are there so many duplicate symbols in your lower two charts?"

There may seem to be "duplicates" in the lower two tables because, firstly, most letters of English spelling are capable of more than one pronunciation. So a letter will appear in the left hand table as many times as the number of possible pronunciations it could have. Secondly, a given sound, or set of sounds*, can be represented in more than one way by different letters, and therefore those sounds appear more than once in the right hand table. For any given cell in the middle table, its corresponding sound occupies the equivalent row and column of the right hand grid, while the letter of English spelling that was used for it before the spelling reform appears in the same row and column of the left hand square.

*The above paragraph should address the question:
"you claim to be creating a phonemic system, why do /ts/, /ks/, /gz/ and /ft/ appear in your table at all? These are sequences, not phonemes."
Such examples are diphones, and often affricates, consisting of two phonemes each, but are represented by single letters in English spelling.
The right hand table in some cases attempts to show their true diphonic nature as would be required for a proper phonemic system. An algorithm could be used to convert from the spelling system of the middle square table to a perfectly phonemic system that can be placed in the right hand table.

"(/ɣ/ and /x/ occur [...] despite the fact that the former is not a phoneme of English, and the latter only occurs in a rather small number of dialects. "

First of all, "number of dialects" and the number of speakers of a dialect are mainly quite irrelevant to optimality of a system of spelling reform, or even invention of new spellings or orthographies or systems of writing for previously unwritten languages. An optimal system of writing for a particular language should be designed in such a way as to be accommodating to all or as many as possible of the spoken dialects. This can be achieved by explicitly including in the spelling letters for sounds that are pronounced by one dialect although they may be silent or absent or pronounced differently in the same word in another dialect of that language. If it is by being pronounced differently in a dialect rather than by being silent or absent that is the cause of the difference between the dialects, and if that different pronunciation or phoneme is not a member of the inventory or set of phonemes in another of the dialects, then the orthography should represent the sound as though it is being pronounced in the dialect to which the sound is unique. Thereby, speakers of dialects for which the phone is not part of their inventory will still be able to read the spelling by simply substituting what would be probably the phonetically nearest sound that corresponds to the letter in their dialect, while the speakers of the dialect in which the sound is a unique variety will not be at a loss to supply the pronunciation that it has for them, and in that manner local regional and minority accents will not be marginalized or oppressed.

Now, the voiced velar fricative /ɣ/ may not be pronounced as such in the English language, but it can occur in words imported from foreign languages, such as the Irish Gaelic language. Consider the word "ogham". If there be no letter to represent it in spelling, then there would be a very high likelihood that the phoneme would lose its original pronunciation. The unvoiced dorsal fricative /x/ does occur in the Gaelic word "loch", and is maintained as such in certain dialects, so why should it not be given a place in the spelling? After all, people who do not know how to pronounce it correctly are well entitled to use the nearest sound that they can find in their own dialect, and should not encounter any difficulty in reading its specially assigned letter, since they could make their compensatory substitution every time and get used to it.

Moreover, characters for these phonemes are useful in retaining etymological information, which dictionary makers and influencers of spelling have been keen to preserve. Hence, my proposal does not damage etymological clues in the spelling. I think this is important, although other people might not agree, but what is there to lose by an extra silent letter in a word if it is helpful to the reader in understanding the meaning, the history and origins of the word, and its constituent morphemes?

agisaak 5 Jul 2013 — 11:56am: "you'll need to point out what you think is wrong with the current system"
JamesM 5 Jul 2013 — 8:59am: "Or convince the general public to start using the new system, which seems extremely unlikely."

Sometimes people want the problem instead of the solution. I have more to say about this, but I think this post is long enough for now. Let me say that I would intend it to be used initially for teaching non-native students for second language acquisition, because they do not already know how to pronounce the words, and the old spelling does not help them. The new reformed system is analogous to providing pointing of vowels in Hebrew or Arabic for learners who do not know the language well enough yet as speakers. Similar aids should be almost as necessary for learners of English, so that they would not need to carry a dictionary of pronunciations around with themselves wherever they go in order to read. Writing is a different matter.

John Hudson's picture

Let me say that I would intend it to be used initially for teaching non-native students for second language acquisition, because they do not already know how to pronounce the words, and the old spelling does not help them.

This is also the case for native speakers learning to read and write, so the sorry history of the Initial Teaching Alphabet should at least give you pause. I know people who learned to read and write using that system, in its brief experimental heyday: the transition to standard orthography was always problematic.

phrostbyte64's picture

This is very interesting. It reminds me of a replacement cypher I designed when I was a teenager. However, all of this is irrelevant because texting is much more effective in it's reformation of spelling in the English language, than anything that could be proposed here. It is becoming mainstream at an alarming rate of increase.

quadibloc's picture

I could mention I have my own scheme, described on this page; basically, it uses an idea proposed by Mont Follick, except that instead of using the apostrophe to mark where two letters don't form a dipthong, I use the period, on the basis that apostrophes already apppear in the middle of words for contractions, but periods don't, so this avoids confusion.

Dipthongs are used to avoid the need for new characters. So people can reform their spelling right away, without spending money on new fonts, and (well, back in the old days) new typewriters, new matrix-cases, new Selectric golfballs, new phototypesetter negatives. (I'm using "dipthong" in an inexact way, to mean any multiple-letter combination that stands for a sound, so I'm including consonant combinations like th and ng, not just the ones for vowel sounds.)

agisaak's picture

I still cannot really figure out what your diagram is supposed to represent. As mentioned in my previous post, you really require some sort of accompanying explanation. Are the rows and columns in your table intended to be meaningful (in which case they should have labels), or is this simply a list arranged in this fashion for considerations of space?

Trying to figure out your intention on my own is hampered by the fact that it is difficult (at least for me) to try to relate the three tables to one another when I am forced to mentally rotate two of them in order to match up cells in the three tables. It would be considerably easier if you simply presented a single (upright) table in which each cell contained three elements, or better yet a simple list of all of your symbols along with their intended phonetic values.

Also, you need to provide examples of actual words written in your proposed system.

Now, the voiced velar fricative /ɣ/ may not be pronounced as such in the English language, but it can occur in words imported from foreign languages, such as the Irish Gaelic language. Consider the word "ogham". If there be no letter to represent it in spelling, then there would be a very high likelihood that the phoneme would lose its original pronunciation.

This seems like a very strange design consideration for me. The word 'ogham' at one time contained a ɣ in Irish, but it doesn't contain one in modern Irish, and it certainly doesn't contain one in English. Since there is no English phoneme /ɣ/ it *can't* lose it's pronunciation. If you want to indicate how the word is pronounced in the source language from which it was borrowed (or in this case how it was originally pronounced in that language) then you are abandoning any pretext of creating a phonemic system.

Moreover, if you plan on applying this principle more generally, you're going to need a hell of a lot more symbols in your alphabet. How, for example, is one going to spell a word like 'ocelot' without some symbol to represent the voiceless lateralized alveolar affricate? (the Nahua word from which this is derived is pronounced /ōˈtˢēlōt͜ɬ/.) Words of Indian origin would require the addition of symbols for retroflex and breathy-voiced consonants, we'd need tonal marks for words of chinese origin, etc. etc. etc. Why single out the particular cases of x and ɣ?

André

agisaak's picture

I just noticed from your profile that you're located in Dublin. Was this a system which you had originally designed with Irish in mind? That might help explain some of the weird correspondences which had me really thrown off in trying to interpret your charts (for instance, I was perplexed by the fact that you have orthographic M and B corresponding to [v] which makes some sense for Irish but certainly not for English).

André

John Hudson's picture

John Q:I'm using "dipthong" in an inexact way, to mean any multiple-letter combination that stands for a sound, so I'm including consonant combinations like th and ng, not just the ones for vowel sounds.

I presume you mean diphthong, not dipthong. But in any case the word you want for combinations such as th and ng is digraph.

quadibloc's picture

A digraph is any arbitrary pair of letters; I want to note that the two letters are being used to indicate a single sound.

agisaak's picture

I've never seen 'digraph' used to represent any arbitrary pair of letters -- only to pairs which are interpreted as a unit. John Hudson's usage is definitely the standard one.

André

hrant's picture

Actually "digraph", "trigraph", etc. are used to mean arbitrary letter sequences in linguistic frequency analysis.

hhp

R.'s picture

The arbitrary letter sequence is ‘digram’ (or ‘bigram’), I think.

John Hudson's picture

Although I've seen 'digram' used as a synonym for digraph, I've never seen the latter defined as anything other than a pair of letters representing a single phoneme. The term bigram is used in cognitive psychology to refer to arbitrary pairs of letters, but digraph is a standard term in orthographic description for exactly what you describe, along with trigraph etc. all the way up to hexagraph (I know of only one, in an Irish dialect).

By the way, even if digraph were not the appropriate term for two letters indicating a single sound, diphthong most definitely is not. A diphthong is a gliding vowel: not one sound but two sounds, made by movement of the tongue during pronunciation of a syllable (the Greek word diphthongos actually means 'two sounds'). Crucially, a diphthong is a linguistic phenomenon, not necessarily an orthographic one, i.e. a spelling system may or may not record diphthongs distinctly from single vowels. In English spelling, diphthongs are most often indicated by following consonants: hIGHwAY, cOWbOY.

jcrippen's picture

That’s right. Bigram = 2 arbitrary written units taken together, usually for statistical reasons. This term is used in computational linguistics, psycholinguistics, and related areas of research. Digraph = 2 written units used together to represent a single phoneme. This term is used in the discussion of orthography and writing systems. There’s also the graph vs. grapheme distinction (cf. phone/sound vs. phoneme) which overlaps this. A graph is any unit of writing, a grapheme is a unit of writing that is meaningful in a particular writing system or orthography. So <ch> is made of up two graphs <c> and <h>, each of which is an independent grapheme in English, but together they are also a distinct grapheme that happens to be a digraph. The mapping from grapheme to phoneme is not one-to-one, so that <ch> can be /tʃ/ as in <church> /tʃɝtʃ/, /ʃ/ as in <sachet> /sæʃei̯/, or even /x/ as in <loch> /lɔx/.

A diphthong most generally is any pair of vowels that occur in a single syllable. So /ai̯/ in ‘bite’ /bai̯t/ is a diphthong, as is /ei̯/ in ‘bait’ /bei̯t/, /ou̯/ in ‘boat’ /bou̯t/, /au̯/ in ‘bout’ /bau̯t/, and so forth. (The inverted breve below means that the vowel is not syllabic.) In ‘sighing’ /sai̯.ɪŋ/ the /ai̯/ is a diphthong but the /ɪ/ is in the second syllable and is a monophthong, a lone vowel. Notice the mismatch here between orthography and phonology, so the <igh> in ‘sigh’ represents the diphthong /ai̯/ even though <gh> looks like it should be a consonant. Likewise in ‘bite’ the <i> represents /ai̯/, but here only with the <e> after the following consonant: <iCe> = /ai̯C/ (where C is any consonant). Without that final <e> you get /ɪ/ instead, as in ‘bit’ /bɪt/.

John Hudson's picture

Thanks for the excellent explanation and examples, James.

dudefellow's picture

John Hudson 5 Jul 2013 — 7:07pm: "the sorry history of the Initial Teaching Alphabet should at least give you pause. I know people who learned to read and write using that system, in its brief experimental heyday: the transition to standard orthography was always problematic."

As far as I recall, I first read about the Initial Teaching Alphabet from a book about the year 2005 or 2006, where the impression given to me was that it had promoted an encouraging enhanced rate of progress for pupils from the early stages of learning to read compared to traditional spelling. This reported outcome as the interpretation of the experimental evidence suggested to me a glimmer of hope for spelling reform. Clearly there is an opportunity to learn from this experience.

On the other hand, I value that you inform us of the observations of the participants of these studies, as a contribution of objective evidence whose emphasis may not have been consulted for the academic or scholarly publications (Researchers overlook the fact that their subjects can have better insights into the data). But I find this intriguing for a number of reasons. For the people you have encountered who were involved in this project, could it be clarified when you say "the transition to standard orthography was always problematic" whether this attitude was from a feeling of unpleasantness when compelled to replace a more predictable orthography by a more irregular one, and whether these opinions were measured by an actual delay in competency of literacy, and whether the end result was more or less favourable. You see, sometimes faster development can be induced by more challenging conditions. For example, the Japanese do not seem to allow themselves to be hampered by their very intricate system of writing. Or could it be that it is the flexibility of their minds that is willing to tolerate it, and hence how it has endured? However, I understand that education prolonged for a number of years is required for such complex writing. In turn, I worry that our education is impeded by too much time being spent trying to teach spelling, where more time could be devoted to other subjects such as mathematics instead. May I ask, is it that you have come into contact with people exposed to this trial because of mutually living near a location in which it was conducted, or is it that those who had been taught to read by the Initial Teaching Alphabet had a higher probability of being drawn to typography in later life?

In any case, extrapolating the final outcome rather than the earlier advances of the Initial Teaching Alphabet to the system that I am proposing is probably not appropriate. According to how I understand it, an aspect of the Initial Teaching Alphabet included alterations to the permutations of letters within words, particularly of the silent vowel letter as an indicator of long or tense vowels. In contrast, in the system of reform for spelling that I propose, there are no changes to the permutations or order of letters in any words, and moreover, no letter is added or removed, such that all words maintain the same number of letters. If it was the transition to a different arrangement of the letters in spelling that caused problems for moving beyond the Initial Teaching Alphabet, then reasonably these problems are likely to be absent from the system that I propose. All the while, we can expect my system to exhibit the advantage of the original more rapid development in literacy that the Initial Teaching Alphabet was found to impart, because presumably the reason for this enhanced progress was owing to its better correspondence of sign to sound.

.quadibloc 6 Jul 2013 — 10:01am: "I have my own scheme, [...] instead of using the apostrophe to mark where two letters don't form a dipthong [sic], I use the period"

Interesting concept. In my system, a centrally placed dot, which can optionally be shaped somewhat like a tiny asterisk, is used as a syllable marker.

.agisaak 6 Jul 2013 — 10:44am: "it is difficult [...] when I am forced to mentally rotate two of them in order to match up cells in the three tables. "

Admittedly, I have been realising that it would have been more sensible for me to have not arranged the tables in such a way that they appear to be almost upside-down. The opposing orientation, achieved by rotating the tables by a half turn, would have been more logical. But the same information is conveyed nonetheless. A certain amount of obscurity can be a useful deterrent, without obfuscating that fact that the knowledge is there in plain view.

"Also, you need to provide examples of actual words written in your proposed system."

I agree that actions would speak louder than words, and am almost regretful that I have resorted to so much explanation. Something on my wish list would be a really nice compact pangram. In principle, one could be devised with the help of an anagram generator, if it could use a specially-made dictionary of the vocabulary of the English language expressed correctly in the characters of the system of reform for spelling that I am proposing.

"Since there is no English phoneme /ɣ/ it *can't* lose it's pronunciation."

English used to contain a sound similar to [ɣ] and had a letter for it called "yogh", but the letter was removed from the Olde English alphabet, and it is difficult to say whether that was due to increased Latinization, or whether the sound had already disappeared, since we have no voice recordings from that time.

"How, for example, is one going to spell a word like 'ocelot' without some symbol to represent the voiceless lateralized alveolar affricate?"

I hadn't heard of this word until you mentioned it. It would not be difficult to display the original pronunciation of the voiceless lateralized alveolar fricative, by using the character of the eighth column and tenth row of the middle table in combination with a lateralization diacritic. But the damage has already been done, and it is now pronounced as the unvoiced alveolar sibilant. So, to spell the word using the Anglicized pronunciation, use the symbol of the tenth column and fifth row of the middle table. But I think that you will find that if it is spelt (spelled) according to how it was originally pronounced, then people will start to pronounce it correctly (though maybe not the lateralization bit, being too unfamiliar), thereby proving my argument that the correct pronunciation would not be lost if a means of representing it is available. For such rare words, people obtain their pronunciations from the written spelling rather than from oral communication. So for them, all you need to change their pronunciation is to change their spelling.

"Words of Indian origin would require the addition of symbols for retroflex and breathy-voiced consonants"

There is another system of writing that I have that has symbols and procedures for representing such things and others.

"we'd need tonal marks for words of chinese origin"

There is nothing in my proposal preventing tonal marks. I do not consider tones to be phonemes. Phonemes are supposed to be proper segments, but tones typically change across a segment. It is from this same reason that diphthongs are not phonemes. Tones have to be considered to be suprasegmental, or occupying another tier of a featural hierarchy. Tones can be indicated by diacritics, just as they are in Pinyin.

"Why single out the particular cases of x and ɣ"

The phoneme /x/ is included as one of the sounds of English according to the Chambers dictionary that I access, which in my opinion is a much better English dictionary than others I have seen. If you are arguing against me, you are also arguing against the reputable Chambers dictionary. That leaves only the case of the single phoneme /ɣ/. Are you going to try to undermine the predominantly phonemic basis of the system in the right hand table just because of a dispute about whether this one is a phoneme? At this point I point out that the sound in the fifth column and second row of the right hand table probably should be unvoiced rather than voiced. Its character in the middle table was the last to be assigned.

"you are abandoning any pretext of creating a phonemic system."

I never said this was a pretext to the exclusion of other considerations such as etymology and dialectal compatibility. I just wrote, by way of a clarifying explanation:

5 Jul 2013 — 3:45am: "I emphasize that my spelling reform is phonemic more than phonetic to the degree of allophonic."

That it is phonemic more than phonetic does not mean that it has to be completely phonemic. Anyway, the middle table, which is the actualization of the spelling reform, is not supposed to be phonemic, and the way I have portrayed the right hand table is not phonemic. I avoided showing any phonemic system that I have for at least these reasons:
1. If I had used a phonemic system, nobody would have understood it, without it receiving an explanation in itself. It would be silly to explain something by Spanish if the audience only understands English.
2. The phonemes can be conveyed by aggregations of symbols from the International Phonetic Alphabet that is known to linguists. Therefore, I used the International Phonetic Alphabet.
3. Some people would attempt to implement the phonemic system instead of the system in the middle table. A case in point:

.quadibloc 6 Jul 2013 — 10:01am: "avoid the need for new characters. So people can reform their spelling right away"

However, to do so would be destructive to our heritage of literature. It is necessary for any proposed reform to be backwards and forwards compatible, as mine is. I think this is what was meant by:

.John Hudson 4 Jul 2013 — 3:55pm: "misguided ways, to 'simplify' English spelling."

Nevertheless, by means of the right hand table, the system in the middle table can be converted to a phonemic system by a programmable procedure.

I have prepared a scan of the other single table that I discussed on 5 Jul 2013 — 4:38pm, but have yet to upload it. I have a general question about file format relevant to future posts as well. I am not sure whether Typophile can accept PDF format. It would seem to make sense to me for efficiency of file size to use PDF file format for documents containing a mixture of images and font text, so that the text can be viewed crisply at high resolution magnification, instead of posting the whole as an image, where much memory would be wasted on white space. I am aware of the guidelines at http://typophile.com/filter/tips and http://typophile.com/readme.

The phonetic symbol in the sixth column and [edit] ninth row of the right hand table should be j instead of i, but this does not matter from the phonemic rather than phonetic point of view, because although I recognize them to be phonemically contrastive, in my phonemic system they can be represented by the same phoneme when syllable boundaries are shown.

John Hudson's picture

whether this attitude was from a feeling of unpleasantness when compelled to replace a more predictable orthography by a more irregular one, and whether these opinions were measured by an actual delay in competency of literacy

The latter. The success of ITA in getting children reading and writing in that system at an accelerated rate (compared to children starting with standard English spelling) was clearly significant, but eventually the children who were part of the ITA experiment needed to be assimilated into regular classrooms and typical language teaching in the standard orthography. The person whom I know best who went through this experience reported difficulty learning standard spelling. It isn't hard to imagine a better managed transition process, one which took the move from ITA to standard orthography as a separate stage to be properly studied and refined. In the event, ITA kids were simply dumped into regular classrooms, usually under teachers who had neither time nor inclination to give them special help. I reckon the progress of a typical ITA kid could be graphed as a rapid initial acceleration, followed by significant slowing, eventually ending up behind the group in standard orthography literacy.

quadibloc's picture

@dudefellow:
In turn, I worry that our education is impeded by too much time being spent trying to teach spelling, where more time could be devoted to other subjects such as mathematics instead.

Knowing how to spell correctly and well is only important if one has to write, in addition to reading. And the class time spent on teaching spelling is not enough to do the job; a person has to actually do a lot of reading to assimilate the proper spelling of words as second nature.

Thus, an irregular spelling system has the benefit that it makes it harder for uneducated people to obtain jobs for which they are not qualified; it makes it harder to pretend you have an education that you don't really have. This is also why China stuck with characters for so long.

agisaak's picture

The opposing orientation, achieved by rotating the tables by a half turn, would have been more logical. But the same information is conveyed nonetheless. A certain amount of obscurity can be a useful deterrent, without obfuscating that fact that the knowledge is there in plain view.

Why rotate them at all? Why not present everything in the same orientation? I don't see how you can view obscurity as something useful when trying to present information.

Something on my wish list would be a really nice compact pangram. In principle, one could be devised with the help of an anagram generator, if it could use a specially-made dictionary of the vocabulary of the English language expressed correctly in the characters of the system of reform for spelling that I am proposing.

This really would fail to be informative. If you want to provide illustrative examples of your system, I'd present multiple examples illustrating your system through the use of *short* words, each illustrating a single usage of each letter found in standard English orthography. For example, for orthographic u, I'd present your revised spellings for a list of short words such as 'cut', 'put', 'cute' etc. along with other words which would contain the same symbols in your revised system, such as 'boot', 'book', etc.

"Since there is no English phoneme /ɣ/ it *can't* lose it's pronunciation."
English used to contain a sound similar to [ɣ] and had a letter for it called "yogh", but the letter was removed from the Olde English alphabet, and it is difficult to say whether that was due to increased Latinization, or whether the sound had already disappeared, since we have no voice recordings from that time.

Middle English (not old English) ȝ was originally simply a graphic (insular) variant of g. It came to be used as a distinct letter primarily to distinguish hard [g] from palatal [j] since the distribution of these two sounds had become largely unpredictable following the merger of the umlauted vowel series with the front vowels. It's extension to [ɣ] and [w] (other 'non-hard' pronunciations of g) was fairly incidental given that these variants remained predictable. [ɣ] was never phonemic. Its loss was triggered by the extension of y to include those instances of [j] which were derived from /g/ as well as those derived from /y/). We don't need recordings to reconstruct a relatively accurate timeline of when these events occurred and any decent introduction to historical English phonology will provide that.

"How, for example, is one going to spell a word like 'ocelot' without some symbol to represent the voiceless lateralized alveolar affricate?"
I hadn't heard of this word until you mentioned it.

Strange. Its a fairly common word, though perhaps I'm displaying a North American bias here since AFAIK they aren't found in europe (apart from in zoos and as pets of Spanish artists).

It would not be difficult to display the original pronunciation of the voiceless lateralized alveolar fricative, by using the character of the eighth column and tenth row of the middle table in combination with a lateralization diacritic. But the damage has already been done, and it is now pronounced as the unvoiced alveolar sibilant.

No damage has been done -- the English pronunciation is and always has been /ɑsəlɑt/. The original aztec pronunciation is really irrelevant to English orthography, just as the Irish pronunciation of 'ogham' or 'lough' is irrelevant to English. That was the point I was actually making.

Your suggestion that you might have ways of representing retroflexion, breathy voice, tone, etc. makes it very unclear what the actual goals of your system are. You start by claiming you want to make a (largely) phonemic system for English, but then seem to shift to wanting a phonetic system which represents sounds found in the sources of borrowed words, which differentiates between various dialectal pronunciations, etc.

For such rare words, people obtain their pronunciations from the written spelling rather than from oral communication. So for them, all you need to change their pronunciation is to change their spelling.

Why would they want to change their English pronunciation? What's wrong with the one they're using now?

I do not consider tones to be phonemes. Phonemes are supposed to be proper segments, but tones typically change across a segment. It is from this same reason that diphthongs are not phonemes.

Diphthongs most certainly are phonemic. In English there are four phonemic diphthongs (/aj/, /aw/, /ɔj/, and /ju/) -- these behave as single units with respect to the phonotactics of English and thus must be treated as such (the treatment of tones is more complex, since tones behave very differently in the various languages which employ them and their representation will be different for (e.g.) Sinitic and African languages).

The phoneme /x/ is included as one of the sounds of English according to the Chambers dictionary that I access, which in my opinion is a much better English dictionary than others I have seen. If you are arguing against me, you are also arguing against the reputable Chambers dictionary.

Different dictionaries are based around different sets of principles. Chambers is very much an etymologically-oriented dictionary which attempts to treat both modern and archaic forms, so it is unsurprising that it includes representations for sounds found in earlier versions of English, but that does not mean that they are phonemes in modern English (and /x/ was never a phoneme distinct from /h/ -- whether one chooses to label this as /h/ or /x/ isn't particularly important, but there's no reason to introduce a separate symbol to treat them separately when dealing with English).

That it is phonemic more than phonetic does not mean that it has to be completely phonemic. Anyway, the middle table, which is the actualization of the spelling reform, is not supposed to be phonemic, and the way I have portrayed the right hand table is not phonemic. I avoided showing any phonemic system that I have for at least these reasons:
1. If I had used a phonemic system, nobody would have understood it, without it receiving an explanation in itself. It would be silly to explain something by Spanish if the audience only understands English.

Unfortunately, your current table is also fairly difficult to understand. It doesn't really matter what type of system you envision; an explanation is still required.

In an earlier post, I suggested you needed to outline what you thought the problems with modern English spelling are before presenting your own system, to which you responded simply: Sometimes people want the problem instead of the solution. Let me clarify that the reason I feel you need to answer this question is that it will help establish exactly what you think the desirable features of a writing system ought to be, and thus what principles your own system is guided by. In your presentation thus far the latter is particular unclear since you seem to vacillate between wanting to represent phonemics, phonetics, etymology, dialect variation and also claim to be motivated by ease of teaching. Without you actually enumerating your goals (and presenting the system itself more clearly) it is difficult to evaluate how well you achieve these goals.

André

quadibloc's picture

@agisaak:
How, for example, is one going to spell a word like 'ocelot' without some symbol to represent the voiceless lateralized alveolar affricate?

Well, in my system, it would be spelled 'oselot'. A simplified spelling system is intended to avoid the problems introduced by the etymological spelling of Ben Johnson, and so sounds not used by ordinary English speakers would not be represented.

dudefellow's picture

Old English is Anglo-Saxon, but by Olde English I meant Middle English.

agisaak 8 Jul 2013 — 11:06pm: "extension to [ɣ] and [w] (other 'non-hard' pronunciations of g) was fairly incidental given that these variants remained predictable. [ɣ] was never phonemic."

It seems to me that [ɣ] and [x] represent the same phoneme in Middle English. In Old English there is the word "þurh" meaning "through", where the letter "h" represents /x/. This word was spelt "þurȝ" in Middle English.

"Diphthongs most certainly are phonemic. In English there are four phonemic diphthongs (/aj/, /aw/, /ɔj/, and /ju/) -- these behave as single units"

I disagree. Phonologically the diphthongs in English do not behave as single units. This is made evident when they become rhotacized, so that the initial part of the diphthong becomes the same phonemically as another unrhotacized monophthong. /ju/ is a combination of a non-syllabic approximant or glide semi-consonantal onset followed by a monophthongal syllabic nucleus. The other diphthongs are syllabic vowel nuclei followed by non-syllabic semi-consonantal codas. This theory that I devised simplifies the phonological rules for English.

agisaak's picture

It seems to me that [ɣ] and [x] represent the same phoneme in Middle English. In Old English there is the word "þurh" meaning "through", where the letter "h" represents /x/. This word was spelt "þurȝ" in Middle English.

Yogh was also used to represent [x], but /h/ and /g/ remained distinct phonemes. Orthography does not always reflect phonology.

I disagree. Phonologically the diphthongs in English do not behave as single units. This is made evident when they become rhotacized, so that the initial part of the diphthong becomes the same phonemically as another unrhotacized monophthong. /ju/ is a combination of a non-syllabic approximant or glide semi-consonantal onset followed by a monophthongal syllabic nucleus. The other diphthongs are syllabic vowel nuclei followed by non-syllabic semi-consonantal codas. This theory that I devised simplifies the phonological rules for English.

You'd have to give an example of the rhoticization which you have in mind.

If you treat the diphthongs as sequences rather than units, the phonotactic description of English becomes unnecessarily complicated. For example, English normally bars /j/ from occuring in onsets following plosives. We don't get words like *kyat or *pyot. We do however, get words like /pjur/ 'pure', /kjut/ 'cute', and (for many dialects) /tjun/ 'tune'. Similar issues arise with respect to syllable codas. If we treat diphthongs as vocalic units, we can simply rule out CjV and VjC combinations, whereas if we treat them as sequences the phonotactic restrictions turn into a rather complex mess in which specific pairings must be treated as exceptional (and where the exceptions themselves make little sense in terms of the sonority scales which normally govern phonotactic restrictions).

/aw/, /aj/ and /oj/ are pretty much universally treated as phonemic diphthongs. There's less consistency with /ju/ since it exhibits some conflicting properties (initial ju behaves as a sequence (it's a eulogy rather than *an euology) whereas medial ju does not).

dudefellow's picture

agisaak 9 Jul 2013 — 12:04pm: "the sonority scales which normally govern phonotactic restrictions"

So I am not the only one who knew this...

"You'd have to give an example of the rhoticization which you have in mind."

Consider these examples and try to explain them:
1. Ireland;
2. Our.

"If we treat diphthongs as vocalic units, we can simply rule out CjV and VjC combinations"

What about:
1. Fjord;
2. "New Orleans" pronounced with the colloquial dialect;
3. There is a certain informal and casual or slang word that I have not seen written but only heard, and which is a neologistic portmanteau of "yeah" and "no" used to mean an apathetic, disaffected or indifferent combination of their meanings, pronounced like [njɛ]. Perhaps we could spell it as "nyeah".

"Yogh was also used to represent [x], but /h/ and /g/ remained distinct phonemes."

The issue was not whether /h/ and /g/ remained distinct phonemes, but whether [x] and [ɣ] were allophones of the same phoneme /x/. Are you accepting that different phonemes can have the same sound sometimes?

.agisaak 8 Jul 2013 — 11:06pm: "Its loss was triggered by the extension of y to include those instances of [j] which were derived from /g/ as well as those derived from /y/). We don't need recordings to reconstruct a relatively accurate timeline of when these events occurred"

Since you know so much about it, perhaps you can tell us then:
1. When the letter yogh was lost;
2. When the sound [ɣ] was lost;
3. When the sound at the end of the word "cough" became an unvoiced labiodental fricative.
Some examples of dated evidence, such as rhymes in Middle English verse, illustrating these phenomena would support your argument. Those for whom the entire corpus is not sufficiently well-known and who do not specialize in historical linguistics have to take your word for it.

"/x/ was never a phoneme distinct from /h/"

How do you explain the contrasts of these words:
High;
Chi (the letter of the Greek alphabet, as in Chi-squared distribution);
Key;
He
?

agisaak's picture

Consider these examples and try to explain them:
1. Ireland;
2. Our.

What exactly am I trying to explain here? What specific pronunciations did you have in mind?

1. Fjord;
3. There is a certain informal and casual or slang word that I have not seen written but only heard, and which is a neologistic portmanteau of "yeah" and "no" used to mean an apathetic, disaffected or indifferent combination of their meanings, pronounced like [njɛ]. Perhaps we could spell it as "nyeah".

The first is a borrowing and the second is an example of expressive vocabulary, neither of which should be used in phonological argumentation since they tend to violate normal English patterns.

2. "New Orleans" pronounced with the colloquial dialect;

Again, I don't know what specific pronunciation you have in mind, but I can't offhand think of any which might be problematic. The only place where [j] might occur in this word that I can think of would be in /nju/ which is perfectly consistent with what I wrote in my previous post.

The issue was not whether /h/ and /g/ remained distinct phonemes, but whether [x] and [ɣ] were allophones of the same phoneme /x/. Are you accepting that different phonemes can have the same sound sometimes?

Sure, phonemes frequently undergo contextual neutralization, but in this particular case it was simply an instance of the same orthographic symbol being used to represent either coda /h/ (pronounced [x]) or coda /g/ (pronounced in a variety of ways depending on the preceeding sound, but not as [x]).

Since you know so much about it, perhaps you can tell us then:
1. When the letter yogh was lost;

Yogh became increasingly less frequent (being replaced by gh in all of its secondary (i.e. non-[j]) uses) throughout the middle English period and is normally considered to have died out by the end of the Middle English period, though one still finds the occasional instance in Early Modern English. There's no specific time when it was lost.

2. When the sound [ɣ] was lost;
3. When the sound at the end of the word "cough" became an unvoiced labiodental fricative.
Some examples of dated evidence, such as rhymes in Middle English verse, illustrating these phenomena would support your argument. Those for whom the entire corpus is not sufficiently well-known and who do not specialize in historical linguistics have to take your word for it.

June 24th, 1509.

Seriously, though, the changes involving coda consonants in English involve a large number of interacting processes which occurred through the Middle English period and into Early Modern English, each of which must be justified using a large number of converging sources of evidence (none of which have any bearing on your proposed spelling reform). If you are interested in this topic, a good introduction to the history of English would be Baugh and Cable's A History of the English Language (afaik the most recent edition is the 6th from 2012). Another good introduction would be Celia Milward's Biography of the English Language (my copy is from 1996 but I believe there are more recent editions). The former contains more examples, whereas the latter is somewhat more readable.

How do you explain the contrasts of these words:
High;
Chi (the letter of the Greek alphabet, as in Chi-squared distribution);
Key;
He

In my dialect, the above words are pronounced [ha:j], [kʰa:j], [kʲʰi:] ~ [cʰi:] (probably intermediate between the two), and [çi:], respectively. Again, I fail to see what requires explanation.

More importantly, though, variant pronunciations triggered by adjacent vowels aren't the sort of thing which should play any role in orthography, so I don't see how these examples are supposed to clarify anything about your proposed system.

dudefellow's picture

agisaak 10 Jul 2013 — 8:48pm: "What exactly am I trying to explain here? What specific pronunciations did you have in mind?"

You can describe a pronunciation of your choosing, and then, if you wish, leave the explaining to me.

"Again, I don't know what specific pronunciation you have in mind"
I was hoping you would recognize that I meant one of the contracted versions by apocope in which "New Orleans" has only two syllables. Perhaps natives, preferably some who have no understanding of our debate, of that place would illuminate us.

"June 24th, 1509."
A date range would have been more credible, but if you provide a date, it should be accompanied by the references to least some of its original and published source texts, which at least can be dated, since I would not want anyone to be too gullible. You discuss a time for the disappearance of the letter symbol "yogh", for which the visual evidence of only source texts is adequate, but you fail to provide dates as answers to the questions involving the sounds, although these are the questions for which audio recordings might have been useful. You direct the interested reader to some textbooks, but neglect to specify whether these contain the answers to the questions, such as in terms of date ranges, rather than merely the methods by which they may be obtained. If the answers are not in the books mentioned, you may as well tell me now so as not to waste my time. If the answers are in the books, you could tell us the page numbers.

"In my dialect, the above words are pronounced [ha:j], [kʰa:j], [kʲʰi:] ~ [cʰi:] (probably intermediate between the two), and [çi:], respectively. Again, I fail to see what requires explanation."

The initial onset consonant of one of the pronunciations of "Chi" given in the dictionary that I consult (and, having heard it, I think I would have known this pronunciation without the dictionary) differs from that in the pronunciations for the words "key" and "he", while they all have the same vowel. They are different words and the differences are not caused by allophonic variation. Therefore, these consonants are all mutually contrastive phonemes. In Greek, the sound of "chi" was aspirated and differentiated by that property from the unaspirated sound indicated by the letter "kappa". In English, aspiration for plosives is only a predictable allophonic variation (as is indeed the palatalization that you have shown).

Etymology is important for the design of stable systems of writing, so that they will endure to the future by not being as susceptible to further reforms, and so that they will preserve the past by allowing a greater literature to be accessible for longer periods of history. My system of writing addresses these concerns by maintaining existing permutations and numbers of letters in words, whilst at the same time allowing the phonemic pronunciation to be shown explicitly, so that this could be tracked in time without loss of literacy. This is more than our current system can achieve, without, as you say "using a large number of converging sources of evidence". There is more than one meaning in this that I wrote:

8 Jul 2013 — 2:21pm: "It is necessary for any proposed reform to be backwards and forwards compatible, as mine is."

John Hudson's picture

"June 24th, 1509."
A date range would have been more credible...

So would a sense of humour.

Seriously, though...

agisaak's picture

You can describe a pronunciation of your choosing, and then, if you wish, leave the explaining to me.

I pronounce 'Ireland' as [ajɹ̩ln̩d] and for me 'our' is homophonous with 'are', both being pronounced as [aːɹ].

I was hoping you would recognize that I meant one of the contracted versions by apocope in which "New Orleans" has only two syllables. Perhaps natives, preferably some who have no understanding of our debate, of that place would illuminate us.

You mean [n(ə)wɔə̯ln̩z]? There's no [j] in that pronunciation.

As for 'chi', if the pronunciation you have in mind is [xiː] or [χiː], then these represent attempts on the part of English speakers to imitate the Greek pronunciation; they cannot be used as evidence for contrasts in English.

If you want to describe a phonological system for English, you need to first specify which variety of English you are concerned with, and only examples from that particular variety of English can be used in your argumentation. One major reason why different dialects of English exhibit differences in pronunciation is because they do not share the same phonological system.

You direct the interested reader to some textbooks, but neglect to specify whether these contain the answers to the questions, such as in terms of date ranges, rather than merely the methods by which they may be obtained. If the answers are not in the books mentioned, you may as well tell me now so as not to waste my time. If the answers are in the books, you could tell us the page numbers.

Yes, they both contain answers to the questions you ask. Otherwise I would not have referred you to them. I give you a reference rather than an answer because the answers are complex, involving multiple interacting sound changes involving coda simplification and vowel changes, all of which occurred over a period of time of around two centuries or so, and not necessarily the same period of time for all geographical locations in England (some are in fact still ongoing). I cannot give you a specific page numbers because the relevant discussion occupies a major portion of the sections on Late Middle English and Early Modern English phonology. However, a very simple answer would be that these changes were well in progress, but not entirely complete, by the beginning of the Great Vowel Shift (which also occurred over time and not simultaneously in all dialects), and would have been complete by around 1650 or 1700 for London English. The relevant changes occurred well after yogh was no longer in common use, either as a representation of /x/, /j/, or any of the various allophones of /g/.

Etymology is important for the design of stable systems of writing, so that they will endure to the future by not being as susceptible to further reforms

Most of the criticisms of English spelling which I've seen from spelling-reform advocates stem from the fact that English spelling does preserve etymological information, so I fail to see how incorporating this will make it resistant to further reforms. Of course, you still haven't explained what you think is wrong with current English spelling despite having been asked.

My system of writing addresses these concerns by maintaining existing permutations and numbers of letters in words

In other words, your simpler, 'easier to learn' system requires that people be well versed in standard English spelling before they can even attempt to learn your system. That is definitely a unique feature of proposed spelling reforms.

André

dudefellow's picture

agisaak
12 Jul 2013 — 8:35am
"I pronounce 'Ireland' as [ajɹ̩ln̩d] and for me 'our' is homophonous with 'are', both being pronounced as [aːɹ]."

For me, "our" and "are" are also homophonous, but their vowels are monophthongal and have the same pronunciation as the first part of "Ireland". The pronunciation in this dialect is different from the typical American pronunciation that I have heard. We know that it is a diphthong contracted to a monophthong because of the following "r" because in the semantically related adjective "Irish", the initial letter remains diphthongal, and this is allowed because the word partitions into two syllables as "I.rish". However, the word "Ireland" with that diphthong cannot be partitioned in the same way, because to attempt to do so would create a phonologically impermissible consonant cluster, as in (unattested) *[ajrl.ænd] or *[ajr.lænd] or *[aj.rlænd]. Instead, the result is forced to be more like [ɑr.land] or [ɑrl.and]. But another strategy, as manifested in the American dialect, is to add extra syllabicity, as you have shown for the sonorants.

"There's no [j] in that pronunciation."
It is just that I was wondering whether the syllables [nwɔ] and [njɔ] are allowed by the phonotactic rules you advocate.

"they cannot be used as evidence for contrasts in English."
If the contrasts are there in English, however they got there, that is enough evidence for me.

"only examples from that particular variety of English can be used in your argumentation"
There is an intention in the reform that I propose that many dialects will be accommodated, not just from the present time. You have a here-and-now outlook to phonology, whereas I would prefer to adopt a time-dependent and geographically encompassing theory of the patterns of sound arrangements. The reason for this is to encourage stability and depth for the spelling.

"I fail to see how incorporating this will make it resistant to further reforms."
I would suspect that spelling reform advocates criticize etymological information not for its own sake, but rather because it is their way of expressing the difficulty in predicting pronunciation from the spelling. But in my proposal, the pronunciation is shown unambiguously.

"your simpler, 'easier to learn' system requires that people be well versed in standard English spelling before they can even attempt to learn your system. "
The system allows pronunciation for reading and the language to be easier to learn, just as vowel pointing allows the reading of Hebrew text to be easier to learn for those who do not know the spoken language sufficiently. Writing, as I already wrote, is a different matter. The reading material for learners is to be provided for them, not by them.

For convenience, I have prepared paraphrases of my understanding of some points of our arguments from earlier.

.agisaak:

1. /ɣ/ is not a phoneme of English [5 Jul 2013 — 11:56am; 6 Jul 2013 — 10:44am; 8 Jul 2013 — 11:06pm].

2. [ɣ] used to exist in Irish [6 Jul 2013 — 10:44am], and in English, where it was an allophone of the phoneme /g/ or a sound of the letter symbol g, which could be represented by the letter ȝ [8 Jul 2013 — 11:06pm].

3. The letter ȝ could also represent [w] and palatal [j], both of which are allophones of the phoneme /g/ or sounds of the letter symbol g [8 Jul 2013 — 11:06pm]; and [x] [9 Jul 2013 — 12:04pm].

4. Phoneme /g/ was never pronounced as [x] (codally) [10 Jul 2013 — 8:48pm].

5. [x] was an allophone of the phoneme /h, x/ [8 Jul 2013 — 11:06pm] and could be represented by the letter ȝ [9 Jul 2013 — 12:04pm].

6. The phoneme /x/ [I think you meant as though it were a distinct phoneme from /h/ and /ɣ/] occurs in only a rather small number of dialects [5 Jul 2013 — 11:56am] [I think you meant that [x] of Modern English only occurs in a small number of dialects].

.dudefellow:

1. /ɣ/ may not be pronounced as such in English [5 Jul 2013 — 4:38pm].

2. /ɣ/ occurs in Irish Gaelic [5 Jul 2013 — 4:38pm]; and [ɣ] used to exist in English, where it could be represented by the letter "yogh" (ȝ) [8 Jul 2013 — 2:33pm].

3. The letter ȝ might also have been used to represent [x] [9 Jul 2013 — 9:48am].

4. "The issue was not whether /h/ and /g/ remained distinct phonemes" [10 Jul 2013 — 8:18am].

5. [x] (and speculatively [ɣ], see also 10 Jul 2013 — 8:18am) is an allophone of the phoneme /x/, which was represented by the letter h in Old English, and possibly by the letter ȝ in Middle English [9 Jul 2013 — 9:48am].

6. The phoneme /x/ occurs in certain dialects [5 Jul 2013; 8 Jul 2013 — 2:33pm].

Nick Shinn's picture

I don’t understand how this “symbol” relates to spelling reform.
Please explain.

dudefellow's picture

Nick Shinn 12 Jul 2013 — 10:19am: "I don’t understand how this “symbol” relates to spelling reform."

There are many symbols. Which do you mean?

agisaak's picture

@Nick: I don't either. I've asked him a few questions about his system but he seems more interested in discussing these points than providing clarifications on what he has proposed.

@dudefellow:

It is just that I was wondering whether the syllables [nwɔ] and [njɔ] are allowed by the phonotactic rules you advocate.

They're not rules which I 'advocate'; they're simply rules which accurately describe standard, north american English (and, to the best of my knowledge, RP English as well). The rules found in Southern US English will be different because those dialects have a different phonological inventory, and different phonotactics, though the handling of /ju/ isn't going to be any different (/aj/, on the other hand...).

The phoneme /x/ [I think you meant as though it were a distinct phoneme from /h/

No, no dialect of English (past or present) has distinguished /x/ from /h/. I apologise for my inconsistency above. When discussing Old English, its conventional to refer to that phoneme as /x/, whereas for modern English /h/ is more usual, but it's the same phoneme involved regardless of notation.

The letter ȝ could also represent [w] and palatal [j], both of which are allophones of the phoneme /g/ or sounds of the letter symbol g

Not quite. (Old) English *g developed into a number of different sounds including [g], [j], [w], and [ɣ]. However, these represent historical sound changes rather than allophonic variations. [ɣ] and [g] were definitely allophones of /g/, but because /j/ already existed as an independent phoneme in English, those instance of *g which developed into [j] would have merged with that phoneme. There was also a phoneme /w/, but *g developed into [w] only in environments where that phoneme didn't normally occur, so that situation would be open to several different analayses.

dudefellow's picture

agisaak 12 Jul 2013 — 10:45am:
"it's the same phoneme involved regardless of notation."
I expect this to be deemed so because /x, h/ pronounced as [x] is thought to occur only finally in words, while [h] does not occur in that position.

"those instance of *g which developed into [j] would have merged with that phoneme."
The situation seems to be similar to "hard" and "soft" g in current English orthography. They have the same grapheme, but their pronunciation can usually be predicted by whether the following vowel is a front vowel, indicated in spelling by the letters i or e, or a back or central vowel, indicated by the letters u, o, or a. The "soft" g or affricate has the same sound as that usual for the letter j.

"though the handling of /ju/ isn't going to be any different (/aj/, on the other hand...)."
I would have similar concerns for /aw/, considering other pronunciations of the word "our" again. The notion of at least three of the four diphthongs you list as single units seems to be disputable, and this calls into question the sole remaining one.

dudefellow's picture

agisaak 12 Jul 2013 — 10:45am: "However, these represent historical sound changes rather than allophonic variations."
Yes, I was not sure whether you meant allophones of a phoneme or just different sounds of the one letter grapheme, which is why I used the logical conjunction "or". But why do you write of *g as though it is a reconstruction? Are you discussing a phoneme or a letter grapheme of allographs?

Nick Shinn's picture

Dude, I meant the whole mystical emblem image.

John Hudson's picture

Yeah, the overall presentation is a bit odd. In the 16th Century, it would be de rigueur to present one's philological explorifications in such manner, but today it all seems a bit Dungeons & Dragons.

Syndicate content Syndicate content