IPA Palette now available for 64-bit Snow Leopard

About a year ago I asked if anyone could help me get IPA Palette working on Snow Leopard. I didn't get much of a response. I found some workarounds.

A couple days ago, Brian 'Moses' Hall, the author of IPA Palette, found the post and responded.

Snow Leopard "broke" some aspects of Input Methods (because it suddenly went all 64-bit crazy) so people like me (and those write screen saver plugins and such) suddenly had to scramble. IPA Palette 2.0 addresses all these changes and works great on SL. Cheers, Moses Hall.

So IPA Palette 2.0b4 is available for download.

A couple images:



It's a nice utility. I've set up some text replacement preferences for most of my IPA input, but I'll definitely use IPA palette for more of the fine tuning and narrow transcriptions.

A little phthalo schwa & some alizarin fricative

One of my favorite tools on my Mac is the IPA Palette (found here) that allows me to click on each symbol to choose it when typing. Download the package. Put the component in the /Library/Components folder. Then it's available from the menu bar.

Choosing from here ↓


brings up this ↓


I've recently upgraded to Snow Leopard. The menu option is still there, but clicking on it now brings up nothing. "Show Character Viewer" works. "Show Keyboard Viewer" works. "Show IPA Palette" doesn't work.

There's some sort of workaround that I don't fully understand, and so it isn't reliable. Nor is it practical even if it were to become reliable. I select the palette from the menu then I click on some application in the dock, (seems that only 3rd party applications do the trick) and the palette shows up. And even tho it shows up, it doesn't function. Click click click. No output.

I know most of you don't care. But if two of you do, and one of you knows something about this and has a little advice... Yeaaaah, that'd be greaaat.

Broadcast standard.

We just drove through Iowa and spent an evening in Des Moines with friends.

It's true what they say. I didn't hear a single accent the whole time I was there.

Except for a young boy about a year old who was practicing his linguolabial stops. I think they were voiced because he wasn't aspirating them. They sounded a lot like [b(ə)].

A linguolabial is marked by a subscript 'birdie' (or 'seagull') diacritic. The coronal articulation is noted by the [d] and its labial placement gets the diacritic [   ̼]? The IPA doesn't recognize a [b̼] because the [b] would be a bilabial and the linguolabial doesn't use the lower lip. The 't' and 'd' are the most reasonable coronals to use because symbols like 's' 'n' 'θ' 'ð' 'ʃ' 'ʒ' etc are fricatives not stops. The other coronal stops 'ʈ' and 'ɖ' would make little sense because they are retroflex symbols. Why would a symbol used for a retroflex be used in a symbol for the opposite of retroflex?

Schtupp in the name of love

I was visiting with a former professor this weekend. She's not a linguist but she is very well acquainted with IPA transcription. She frequently teaches communications and speech and she often incorporates units on transcribing speech.

We were speaking of aural "blind spots" and common mistakes of novice scribes. Some vowels are very difficult to differentiate such as the [a]/[ɔ] distinction (especially for the cot/caught mergers that have overrun her campus) and some spelling conventions are difficult to put aside (such as the 'g' that some students try to put after every [ŋ]).

She also noted an emerging allophone in the speech of some students. She has noticed a trend towards a post-alveolar voiceless fricative [ʃ] before [t]. Right after mentioning this the conversation was interrupted and I was left to sit in the corner muttering [ʃt ʃt ʃt...] to myself. Such moments are common.

Buffy will say to me "you never listen to what I'm saying" and I fade into a trance. My mantra [vɻlɪsn̩ vɻlɨsɨn vɻləsən...]

Several minutes later (perhaps about an hour) I approached my former professor (now friend) and asked her if there was any specific environment in which she noticed this post-alveolar. My guess was that it would not occur in the words steep stick or stop but it would show up in strength string and strange because of the following retroflex.

An interesting co-occurrence of features is the labialization (lip rounding) of the initial approximant [ɹ] even before front vowels and the similar labialisation of initial [ʃ]. Is it possible that the concurrent labial feature as could influence a non-labialized phoneme [s] to resemble a very similar phoneme [ʃ] that does take lip-rounding.

She wasn't sure of a more specific environment but she did promise to listen for it. I don't remember seeing this alternation mentioned but I'll be listening for it very intently now. I really doubt that it's at all related to the pre-consonantal palatalization that we sometimes see in strudel and which is increasingly present in smorgasboard.

The owl vowels

Some days I step into the cyber office and the flurry of activity is obvious. Footprints from all over the world are leading up to my doorstep.

One of the nice things about putting up a blog is the freedom to be responsive to readers. One of the nice thing about good statistics software is the insight it gives us into what readers/visitors are wanting to know. Several months ago I put up a list of the ADS Word of the Year finalists. Web searches for "plutoed" and "Cambodian accessory" had all sorts of strangers standing in my foyer.

Another spike was in response to the mousetrap stunt that still baffles me.

The latest spike has to do with the IPA transcription of the owl diphthong.

Here's an answer: I transcribe owl as either [aʊl] or [awl].

Here's a question: Why is this such a popular search term right now? Are students around the world all cheating on the same final exam? Did the Democratic debate pose this as one of its questions? Did Alex Trebek ask this during Teen Week?

Here's some more information: the one we should ask about diphthongs is John Wells. He has written about them before and even in response to some of my questions. A while back I asked him why he prefers the two vowel transcription rather than /ow/ or /oʷ/. He wrote on his blog

I use the two-vowel-symbol notation for English diphthongs because the diphthongs behave as single indivisible units. A vowel-plus-glide notation would imply the identification of the first part of the diphthong with one of the simple (non-diphthong) vowels. If the nucleus of English goat is taken as /Vw/, what is the /V/? It could be (BrE) /ɜː/, the vowel of bird, or /ɔː/ the vowel of thought, or /ɒ/ the vowel of lot, or /ʌ/ the vowel of strut, or /ə/ a schwa. In this context there is no phonemic contrast between these vowels, and no strong reason to choose one solution over the other. By treating the diphthong as indivisible we avoid facing this false choice. (This concerns its phonological analysis. We still have to choose a notation for it in transcription.)

He further explains that the superscript W is not a possible notation of the /o/ because it indicates labialization and the vowel is already labialized.

Although I saw the superscript notation used by my phonology II professor, I like Mr Wells' correction enough to discard the convention.

It's also worth noting that the symbol is a tool. Symbols are not facts and in phonology sometimes they don't even represent facts. They represent analyses of the facts. It's a fine line but an important one.

A whole new level of nerdiness

Over at Speculative Grammarian^TM I found this simple little game that they chose to call LingDoku ("Like SuDoku, But for Linguists"). I know it's a satirical journal so I'm embarrassed to say how much I could enjoy puzzles like this. The one they provided is very simple but imagine how useful a game like this would be to get practice with distinctive features in a phonology class. It's going in my lesson plans.

They provide nine IPA symbols--ɤ n β g b m z ŋ--representing a matrix of 3 places of articulation--bilabial alveolar and velar--and 3 manners of articulation--plosive fricative and nasal. No row or column can contain any symbols of common manner or place. They start with

d	-	ŋ
-	-	-
-	n	-

The completed grid of course would be

d	β	ŋ
m	g	z
ɤ	n	b

That's almost as basic as a grid can be. So how about a larger puzzle. Or a grid in which the symbols are filled in and the job is to identify what features are being used to organize the rows and columns. coronal dorsal labial bilabial lateral voicing +/-ATR retroflex stridency sonorance...

We could add the variable of binary and absolute features or even throw some feature geometry in there.

It'd be as fun as doing compass proofs in geometry.

Truespel concl.

Zurinskas writes

The alphabetical principle holds that letters stand for sounds. We find now that even Egyption [sic] hieroglyphic symbols stand for sounds, and we can speak the writings of 5,000 years ago because of this.

The problem is that the alphabetical principle is not a universal principle. Our alphabet is a Romanized symbol set that corresponds very roughly (or shall I say flexibly) to our phonemes. Acknowledging this is not the same as believing that it 'must/should/better' be this way.

Were I to accept Mr Zurinskas' claim about Egyptian hieroglyphs (which I understand is based on a simplistic summary of a documentary he saw) I would still reference the differences between Chinese dialects that are based on the same symbols and use different phonology. Is one of these dialects incorrect or less "true" to the symbol set?

Arbitrary dialects destroy this relationship and should they take hold lessen the consistency of correspondence between letters and sounds and make English all the harder to read and learn. Not good.

Okay so it's about consistency. But--what? Arbitrary dialects? Who are these people who have done away with phonology and have begun speaking with indecipherable idiolects based on nothing at all but whims and fancies? And on the flipside, who are these people who are choosing to represent the phonemes of English with fabricated symbol patterns? Like "truespel" for instance.

Let's not be artificial dialectizers by misspeaking words, but rather retainers of what semplence [sic] of alphabetic principle we have for English.

Mr Zurinskas wrote this in response to a question about the perfectly natural and reasonable dialectal pronunciation of milk as 'melk'. Apparently something wecked thes way comes. I have found no basis for his contempt of change beyond a premise that change is evil. He wants the English alphabet to correlate in all English dialects to the same sounds his dialect (or if you look at his phonemic transcriptions, his idiolect) recognizes. Why his dialect? Why the current correlations? Is the language at its peak? Have we been waiting ever since the Norman invasion to see English reach its current ideal state? Freeze things now because they'll only get worse?

Zurinskas has no expertise in phonology or phonetics but claims to have training in psychological experimentation. Does he wish he had training in phonology/phonetics? A few weeks ago he made a claim that the first vowel in "English" is a high front [i] as in "bee", and he was challenged by several trained linguists who observed that it is more commonly lax [ɪ] as in 'bit' (and rarely the tense [i]). Some suggested that it might be a high center vowel [ɨ] (which is not a phoneme in English though many speakers pronounce the [i] or [ɪ] phoneme this way in some environments). Zurinskas' reply: "Right. Call [and ask] a friend that is not a linguist, a normal person." His response to the challenge of the rationale behind that suggestion: "Basically, I'm thinking we need unbiased opinion. Linguists are exposed to phonetic notation that could affect judgment." I take it he refers to those manipulative IPA symbols.

When a body of knowledge encourages more accurate and therefore predictable observation I believe it's okay to affect judgment with that information.


_{All quotes and examples are taken from either American Dialect Society LISTSERV postings or the truespel.com website.}

Truespel "Phonemes"

Let us look at a few of truespel's transcriptions. These are taken from the website.

~a
mad = ~mad
hat = ~hat
apt = ~apt
This one seems reasonable.


~aa
father = ~faather,
hop = ~haap,
aardvark = ~aardvaark,
knock = ~naak
I'm not certain why Mr Zurinskas decided to use "aa" as a low center/back vowel--but I'm guessing his inclusion of aardvark in the set is an attempt to justify the symbol as relating to orthography. I can think of a few other examples of that sound spelled with "aa": Paas easter egg colouring kits; Golfer Jay Haas. And when cartoon characters scream it's usually spelled "aaahh!" which we'll assume represents the same vowel.

Even with this mountain of evidence linking it to a pattern in English orthography (usually when spelling Dutch borrowings) it illustrates the inability of this system to avoid the main liability of any system based on orthography--the same symbol (in this case 'a') represents two different sounds. Doubling makes sense if the transcription intends to double the quantitative length of the vowel--but to represent the vowel change from 'hat' to 'hot' doubling isn't any more intuitive than introducing a new symbol.

~ae
sundae = ~sundae
make = ~maek
paid = ~paed

Here he uses 'sundae' to establish the pattern. Why this digraph I'm not sure. My guess is that there is an orthographic clue to the [ej] diphthong when an 'e' follows in the next syllable thus making the 'a' "long." We remember this from the phonics lessons that told us how to distinguish between 'mad' and 'made'; 'fad' and 'fade'; 'tam' and 'tame'; 'bad' and 'bade'. But this clue to pronunciation is best understood as an indicator of the historical two-syllable forms that lengthened a vowel when in an open syllable. Wouldn't it make more sense within this system to combine the two units that represent the two vowels ("e" & "ee") or the vowel and glide ("e" & "y") in this phoneme

But there is no way to distinguish between the tense and lax mid front vowels in "truespel." The closest the system comes to recognizing the two is the introduction of "~air" used to represent what is apparently analyzed as a vocal-rhotic (my term) phoneme, such as we hear in "fair" "bare" and "wear".

Why not spell it "er"? Apparently because that's already used to represent the vocal-rhotic phoneme" in "fur" "sir" and "docter". Again we have a vowel ('e') representing different sounds. Since Zurinskas uses "~u" to represent several lax middle vowels (or schwas) it's not clear to me why he avoids it here.

Here is where I can say "etc" and avoid picking on every problematic phoneme representation in "truespel". I have to save space for an observation or two regarding dialects and another about syllable stress.

"Truespel" has some regional biases. It recognizes the phonemic vowels that are differentiated in some areas (such as the 'cot'/'caught' and 'rot'/'wrought' vowels) and even recognizes some dialectal allophonic alternations as separate phonemes. For some reason he decides to represent the first vowel of "anchor" as the diphthong [ej] ("~ae" in his system). This occurs in some areas of Wisconsin and Minnesota. (But I'll grant that this is a matter of transcription. It doesn't reflect his system--just his analysis and ear.)

Zurinskas does apparently believe in representing certain non-phonemic phones in his transcriptions. Probably because he believes they are pronounced by everyone (like the vowel in "anchor"). In 'entrance' is "~entrints", 'pounce' is "~pounts", and 'insistence', is "~inssistints". Transcribed this way it is impossible to differentiate between 'sense' and 'cents' phonemically.

And transcribing 'imbibe' as "~inbbieb" is just incorrect.

He uses a system of doubled consonants to indicate stress on a following syllable. I'm not sure how he justifies the doubled 'c' & 'h' in "~reecchhaarj", but decides to double only the 's' in "~inssher" ('insure'). I would guess that he only doubles the 't' in 'without'-"~witthout" because he hears it as a voiced fricative and in his system a doubled 'h' would indicate voiceless as in 'rethink'-"~reetthheenk".

So how would this stress indicator deal with a geminate consonant as the [t] in "cattail"? If doubling the vowel indicates a vowel stress he can't use "~tt" or the stress switches to the second syllable. "~cattael" would sound like "caTAIL" wouldn't it?

Though some of his transcriptions look very close to typical English orthography I can't imagine that "~Ummairiku" is the best transcription to teach a child or adult to read "America."

I have one more short post on this topic. Next time: Mr Zurinskas' attempt at a "socio-moral" (my term) rationale.


_{All quotes and examples are taken from either American Dialect Society LISTSERV postings or the truespel.com website.}

Truespel vs IPA part II

Mr Zurinskas says that he wants

a simple notation of the 40 USA English sounds that kids and adults can use to achieve "phonemic awareness".... They can read and write stories in first grade, and transition to tradspel is no problem as demonstrated by IBM's Writing to Read system as tested by ETS. But that system uses special symbols, not as simple as truespel.

If simplicity is what he's going for it must be noted that "truespel"'s allegiance to the alphabet makes it more complex than the IPA where representations are consistent and some dialectal variation is easily accounted for by representing phonemes instead of allophones. I take it he wants something to mimic American orthography. But he wants it to represent pronunciation consistently. I don't know much about IBM's "Writing to Read" program. It involves listening and reading stations and computer programs, using sound and feedback to direct and provide feedback on a learner's reading and writing. I don't know about the "special symbols" that it uses. I will make no claims comparing this system to "truespel."

But Zurinskas apparently has intentions beyond his attempt to debunk the current phonics curriculum. He has put together 4 books on his system and calls it his "life's work."

It integrates the dictionary...with initial teaching of reading..., and eventually translation guides to other languages. No other notation can do this.

Based on his last claim I must assume he is urging his campaign against the IPA. Of course the IPA can transcribe any language only but at the dear expense of integration with the dictionary. But I wonder how "truespel" is supposed to translate into other languages. Because the system is limited by its reliance on standard AmE orthography it can only capture AmE phonemes. How will he capture the difference between Spanish pero - but, and perro - dog? How will he transcribe a front round vowel in German flügel? How will he capture the phonemic difference between an aspirated or unaspirated stop? English does not have these.

"Truespel" cannot capture non-English phonemes unless it provides system of redundant vowels or symbols as diacritics. Will double and triple letters start creeping in? Will the dollar sign start to represent a sound? This system already uses doubled consonants to indicate phonemic stress in a word so that "attack" is transcribed "~uttak" and "implore" is transcribed "~impllor." (Mr Zurinskas has decided to use the tilde as the marker of his transcription system.) So "truespel" would represent that tasty drink, the hot toddy "~haatttaadee"? This would become the equivalent of Ptolemaic deferents and epicycles attempting to take back the astronomical model of the universe.

cont.

_{All quotes and examples are taken from either American Dialect Society LISTSERV postings or the truespel.com website.}

Truespel vs IPA part I

An ambitious gentleman named Tom Zurinksas has formulated and proposed a form of phonetic notation that he believes will better serve English phonetics than will the IPA. He describes his system over at tuespel.com, also providing a table full of examples of his transcriptions.

On the American Dialect Society LISTSERV forum he has provided a snippet of his manifesto sales-pitch. Here I share a few notable quotes.

English is the lingua franca of the world....Because English is the most important language, phonetic notation should be based on English. It should use regular letters so that it's easy to write. That's the big impetus behind truespel notation.

He doesn't say what body of phonetic notation should be done with "truespel." By calling English "the most important language" I must assume that he is ranking this phonetic system against systems that favor other languages. I can't think of any standard phonetic notation system that isn't heavily based on conventional English characters. But I'm not really sure if he means that "truespel" is valuable because it's based on the orthography of English or just the characters. Either way he is mounting an attack on the Internation Phonetic Alphabet for some odd reason that makes sense to him. Apparently he doesn't like using character maps.

But there's a bigger reason; our kids. They are not exposed to phonetic reading and writing because of unusable phonetic notation in our dictionaries.

Oh is that why? I thought we kept phonetics from them because... wait... My nieces and nephew are learning phonics. Isn't that a system that attempts to show the relationships between spelling and pronunciation? And just how unusable is the phonetic notation in dictionaries? There are several systems yes. Confusing? Sometimes. I remember my 5th grade teacher who thought the pronunciation gloss of John Muir's name (MYOOR) meant it was pronounced like flier and buyer. I do wish dictionaries would provide a pronunciation key or som- _{(...What's that? ...which ones?...really?...)}
Never mind.

So I'm not grasping Mr Zurinskas' reason for this new standard he has organized. To his demand that the system base itself more on English (he apparently wants to favour American English) I suggest that is a short-sighted goal given the importance of a system that can be applied to all languages. I suggest also that IPA is already quite close to English orthography. The simple transcription for "bed" is [bɛd]; "forever" is [forɛvər] or the indecipherable [fərɛvər]. Not quite "unusable."

cont.

_{All quotes and examples are taken from either American Dialect Society LISTSERV postings or the truespel.com website.}