Not part of the project; just for fun.

Last chance

For this my final recording, I elected not to be listening to the original while I recorded, instead doing my utmost from memory. At the beginning of this project, I drew my inspiration from the list of elements to work on provided by Dr.O in the project description, but I never did follow the suggested order. Instead I tackled the passage in an order that made sense to me based on my awareness of certain difficulties. I ended up spending a lot of time improving the original recording quality via Audacity's multiple track and playback features. I found myself immensely grateful for my knowledge of how to use Audacity, since my cousin used a recording format that no other program on my computer can even play. I think, that when embarking on any kind of project like this, it will be important to instruct students on the tools available. Audacity also provides a great way of commenting. You can record commentary on another track, and hear it as things happen in the original recording, then delete it for subsequent work. The only thing Audacity misses, and should have given its conceptual use is for music editing, is an ability to write on the tracks, a lyric track (like Pratt's annotation tracks). I started with Rhythm rather than voice quality or consonants and vowels because Rhythm and Intonation are the most identifiable features of a language. I took my guidance on this not only from my own previous experience with accent work (I learned both RP and cockney for a single play) but from Child Language which talks about the fact that extremely young children can recognize their mother's language based on Rhythm and Intonation. It made sense to me that since it seemed to be among the first thing children pick up, it would be a good starting point for my project attempts. Rhythm work led naturally to vowels, since for me, the trickiest areas to keep fluid centered around the transitions across multiple discrete vowels in 'matuod' and 'tuhoi'. Repetitions of the same vowel, as in 'paabuton' were far easier. Partially because of the lack of movement, and more because 'Paalam' means 'Goodbye' in Tagalog, the related language I'd actually studied before beginning this project and it was a grouping I was fairly practiced at saying. Analysis of the vowels in Pratt, which I undertook because I couldn't figure out what I was doing wrong by ear alone, led me to the startling and frustrating discovery that most open syllables in Binisaya were cut off with a glottal stop. Between that and the unaspirated consonants, double releases were common and my glottis was spending more time shut than open. In practicing this, a physical technique worked extremely well. I was able to incorporate the glottal stops initially by practicing while lying down, thereby reducing the airflow I needed to cut off. To the end, I used a similar physical technique. None of the good recordings were done with my head held high, but with my chin slightly tucked to make it easier to cut off airflow. Of course the challenge there was not allowing that vocal posture to drive air into my nasal cavity either. There, I was thankful for voice lessons. My singing teacher spent ages teaching me to shape my airflow so that it was drawn across my palate rather than shifting into my nasal cavity. I was able to make use of those old lessons to maintain voice quality while incorporating the glottal stops. If I was doing this with a student however, maybe to work on a students' nasality, I would want to get in touch with her about the exact exercises. I remember a lot of ridiculous analogies to dolphins, but not exactly how she taught me the technique. While I was making use of that voice technique, I was also taking advantage of the fact that my archetype was a song. I may have been using the spoken lyrics for the actual archetype, but that didn't mean I couldn't use the sung version as a practice tool. I used it to get my vowels where they needed to be in my mouth, and also to practice general accent features. Singing is a method I often use to learn accents, as it tends to eliminate accentual features unless the singer makes the effort to keep them, like Herman's Hermits or The Proclaimers. I've always been able to sing an accent before I can speak it. This did create rhythm problems though, I kept trying to say the words to the much, much slower song rhythm than to the spoken rhythm, which is why I created the clap along and dual tracked versions where the recordings included both my cousin's voice and mine. Next, I entered into my battle with aspiration. This battle would continue throughout the remainder of the project with brief breaks to work on other problems such as the stress work needed as I shifted from a stress-timed to syllable-timed language and began working on matching the archetype's prominence. The most important part of the prominence and stress work ended up being mostly keeping the syllables distinct. I kept trying to use US style reductions, obviously a serious mistake. Binisaya not only has different reductions, but the archetype is someone reading the lyrics to a song, and the speaker knew when recording it that I was using it for school. The recording is careful speech. No reductions except those caused by the grammatical forms (Ako to ko when the agent occurs at the end of the sentence instead of the beginning, following the Patient-Agent pattern which makes the language look like VOS). This was of course made especially difficult by the fact that stress can be discriminatory in Binisaya. For the remainder of the project, I mostly focused on battling aspiration. I received some help from Kira's explanation to Mohamed of what happens during the reduction of the present participle. Her example of walking to walkin gave me a link to use in English that I could connect to what I was trying to do in Binisaya- that k in walkin isn't aspirated. This is exactly the same primary technique I used with my tutee, although in that case we were working on making connections from already mastered English rather than from L1 to another L. Beyond the aspiration, I mostly just focused on not losing what I had already fixed while I concentrated on aspiration and on getting the memorization down. I even ended up singing it into a mic, acapella, for an audience at the East Elementary International Food Fair. Due to my mind glitching and continuously trying to switch to the Philippine National Anthem (which is in Tagalog, not Binisaya) due to some superficial similarity in line beginnings and endings, it wasn't until that performance that I truly managed to get the whole thing memorized- just in time for the tricky holistic attempts at the end of the project. All in all this was a major challenge. I think that while this careful speech recitation of song lyrics provided me with a lot of useful ways to approach it, if I was doing this again, I'd prefer to do more natural speech. Of course, as I pointed out before, most Filipinos code-switch through at least three languages in the course of their conversations, so it would have been extremely tricky. I also should have perhaps taken more advantage of my family's status as native speakers, getting them to check on my progress. On the other hand, I found my dad's contribution of immediate and constant interruption of 'that's not how you say it' and 'you sound like an American' during Spring Break distinctly unhelpful and extremely annoying. I might also, with another semester of Japanese available, try to muddle through more of the established linguistic literature on Philippine languages, though it is mostly on Tagalog and much of it is in Japanese. Finally, I would, if such a thing is available, really appreciate an actor's guide to the accent/language, such as the one I used to pick up Cockney or the one I saw in Scotland written by Dr. Chris. Brian's tips were extremely helpful, and there's a lot more crossover than most realize exist. I would really encourage any student working on their pronunciation and presence (like the TAs in Lara's class) to try some basic acting exercises or, if they have time, classes. Nothing teaches volume and pitch control like being on a stage reciting spoken words, not even singing.

Step together now

As the semester moves towards the end, I need to move towards wrapping up. To that end I started working on combining two aspects of concentration for each recording. This one for example: combines my voice quality practice and my aspiration practice. Interestingly, I found that concentrating on both at once solved the distraction problem, albeit with the addition of a rather ridiculous looking mnemonic gesture where I closed my spread fingers into the shape of a talking hand continuously as I spoke. This gesture reminded me to produce less breath, to round my lips a bit more and to draw my vowels out more fluidly rather than kind of spittting or snapping them out as I do in English. Thankfully, in later attempts I was able to drop the ridiculous looking gesture, but it really helped during the earlier combination practices.

Full of hot air.

The most serious difficulty I've had with this language and this passage throughout the semester has been aspiration. Binisaya does not aspirate initial consonants. This was a difficult thing to catch and the single most salient difference between the accent of a Visayan speaker and an English accent. In English, our unaspirated consonants are limited to stops in consonant clusters. When my father came to visit, I even had to point this out to him, a native Visayan speaker. He'd never noticed the addition of breath when speaking English. After working on this for so long, I think I can safely claim that my cousins could rightly say that Americans are full of hot air. To work on aspiration for this recording, I put my hand in front of my mouth, and if I pronounced a stop with aspiration, I repeated it until I couldn't feel anything during the consonants. This was especially difficult when the stop was in front of a breath heavy lax vowel like 'a'.

Double time

I'm not entirely certain how that 'kay' ended up in the wrong line, but as I was consistently saying it in the wrong place, trying to fix it, and spending half the practice saying those lines just behind when I'm supposed to. Timing is everything. Therefore, to check and be certain I had the rhythm down, especially that 'kay', I practiced saying it with the right intonation and pauses in various contexts, at all speeds. Here's a better example of one of my sped up attempts. You can hear that I kept the rhythm of the language constant despite the varying speeds.

Imitation is the sincerest form of flattery.

One of the most interesting things to remember in this project is that I'm not just learning a passage,  I'm doing an impression. I wasn't sure what was different about it, but I could tell that I wasn't quite where I wanted to be. To my own ears, I sounded much more native, but I didn't sound like Barbie. Therefore, I spent this bit working on matching her exact voice quality. To do this, I recorded my best attempts with and without listening to her as I recorded. While my rhythm and voice quality were definitely improved by speaking with the recording, I can definitely note for both attempts that my voice quality improved by the second verse. The first verse always sounded off until I got into my stride.

Solving the Multitask Problem

I am a consummate multitasker. My father often expresses his disbelief; he can't understand how I can read, watch TV, play a game, bike, whatever all at more or less the same time. What does this have to do with learning? The problem is that I'm so used to multitasking I have immense trouble settling my brain down enough to engage in deliberate practice. My body and at least half of my brain scream in protest at being asked to sit quietly while the rest of my brain is engaged in concentration. Further, while I consider myself a visual learner- I remember most of what I read- that only applies to content retention NOT to word for word recitation. Between years of vocal performance, marching band, and drama, I connect memorization to movement. As Dr. Oshita put it, I need to dance when I remember things. Previous encounters with deliberate practice have only reinforced the need for movement: I've used it for martial arts, percussion practice, and plays. Beyond the recitation-movement connection, part of my brain needs to be engaged in some background activity for me to concentrate fully. I study with music or television that I've seen before. Contrary to most study advice about minimizing distractions, if I don't have something going on in the background, my brain distracts itself. As pursuant to this project, I'd find myself halfway through a line and wondering if the word was the same in Javanese or looking up an article on comparative lullaby topics or thinking about something else entirely. Surprisingly, I found an answer. I found myself, without any real decision to so, practicing in my car. I spend a lot of time driving, and for whatever reason, while most of my brain and body were engaged in driving, I found it considerably easier to do my practicing. Also, it was isolated, giving me an environment free of people, unlike my office or home. Of course that led to a different problem. While I can manage to practice while driving, I am not so stupid as to try recording while driving, and my brain usually skipped subjects by the time I got back to somewhere I could open my computer. In the future, when I work on something like this, or have a student experiencing problems with concentration: I can suggest taking the exercise for a walk, or giving a presentation while doing it (pointing out things on the screen, making good use of gesture). After this, I have considerably more sympathy and ideas for kinesthetic learning. After this attempt, I sounded more like a native speaker to my own ears, but not so much like Babs. I sounded more like myself. I need to work on reducing the intensity and audible effort it takes me to produce those sounds at those speeds with the reduced aspiration. I also need to go lighter and a bit higher. Honestly, I feel like I'm back at choir practice with my teacher telling me to stand a foot further back from the microphone.

Come back, don't go home.

So I've spent the last week or so working on stress, prominence, and intonation. Binisaya, like most Philippine languages is syllable-timed, not stress timed, which takes some getting used to. What's more, stress does act to distinguish words in Binisaya, Uli means 'return', but uli means 'go home'. Similar, but noticeably different meanings. 
Most longer words have stress on the second to last syllable, e.g. hikalimtan, but words ending in open syllables with an 'a?' or an 'o?' put stress on the stopped vowel, though it isn't always primary stress.
Here's my analysis of the stress in my archetype below. Primary stress for a word is underlined, secondary is in italics, and the most prominent words in the line are in red. Despite being a function word 'sa', which can translate to 'in', 'on', 'at', 'of', or 'the' often carries the primary stress in a sentence because of the nature of the grammar of Philippine languages. Philippine languages belong in a group called either Philippine or Austronesian type voice system languages. When 'sa' receives prominence it's acting almost as a topic marker, indicating an actor or the equivalent of a direct object.



Buhi sa kanunay
Ug di ko hikalimtan
Sa hilom gibati ko.
Ang kamingaw ning dughan

II
Bisag giyam-iran, bisan pa’g gitamay
Molambo ug molipang 
kay_ang gugmang matuod di mamatay

Refrain:
Tuhoi intawn ako
Salig sa gisaad ko
Gugma kining way pagkalaya
Bisa’g ibanlas sa luha

III
Kon ang kasing-kasing ko
Sa palad sakiton
Didto sa kalangitan
Ako kang paabuton



Note to self: paabuton keeps coming out as paabutaen, complete with an utterly American [ae] in there. Really need to work on it. Also, when I get the stress right, I tend to over aspirate. I really need to work on (still) aspiration, or rather, non-aspiration, but I'm finally starting to get the speed and sound of it right, if over-exaggerated. I'll probably spend this week working on aspiration more, because I really need to work on it.

Konsonan at linking

Consonants

The first thing I had to remember when working on consonants is the at rest position. The 'at rest' position, when you aren't saying anything, is more rounded in Binisaya than in English. The result of this seemingly minor change is that the entire language is slightly more rounded, so instead of saying 'b' like I'm popping my lips after putting on lipstick, it's more like I'm blowing a kiss, though not that exaggerated

The second thing I had to work on was aspiration. Voiceless stops (k in particular) are unaspirated in Visaysan. Because of the short duration and, generally, the way most syllables (especially those centered around the vowel /a/) end in a glottal stop, the /k/ sound tends to disappear between /a?/ and /a?/ or /a?/ and /o?/, especially when it is in that environment in a word initial position. In the line /bu?hi? sa? ka?nunay/ for example, the /k/ is produced almost as a click, with the combination of the glottal stop and what should be a /k/ producing a double release effect which does not become a full /k/ sound. Because my archetype is careful speech, a set of song lyrics, this is the most significant instance of linking as well.

As a native English speaker, keeping my consonants unaspirated is incredibly difficult, and I still haven't managed it with the voiced stops, which are also unaspirated, though in careful speech, some minimal aspiration may occur in /b/ and /g/, but not /d/ in the archetype, which I suspect is actually back transfer from English. The word initial stops are especially hard to eliminate aspiration in, I can just about manage it intervocally and it's not terribly hard in the rarer coda positions, but it usually takes at least three tries before I can eliminate the aspiration in a word initial voiced stop.

Interestingly, the lack of aspiration is what produces one of the most characteristic features of a Visayan speaker's English. To native English speakers, a Visayan saying 'big backpack' generally sounds like they are saying 'pig packback' or 'pig packpag' or even 'pig pagpag'. In order to practice that lack of aspiration, despite my archetype being in Binisaya, I have been trying to apply that accent to English speech. If I can get it when I'm speaking English, it should be easier to do when I'm speaking Binisaya. 

I started this attempt with singing, also incorporating my vowel work. This is an old trick I've been using to learn accents for years- if I can sing in that accent, when it is easier to adjust things like duration, vowel placement, and airflow (or at least when I have been trained to do so and therefore find it easier), it's usually a short step to carry that over to my speech. When the variable in question is aspiration, this is much less simple than normal, but to me it seems as to be even more helpful as well- if it's harder when singing and I can do it, the speech should be no problem.

The song I usually use for this is Seven Wonders by Nickel Creek. Here are the lyrics for the first verse and chorus with the Visayan accented words approximated next to the English (I will fix this to IPA at some point):

When shadows fall/ wen shado?s faw

He'll close his eyes/ hil klos his ays

To hear the clocks unwind/ to hirr da klox unwin

Powerless to leash the hands of time/ pawerliss to lish da hans ?op daym

Seven wonders crowed the man/ seben wonders kro?t da man 

Knowing six are gone/ no?ing siks are gon

And how the great illusion lingers on/ an haw da gred iwlusho?n li?ngers awn

It's not the best choice for working on /b/ or /g/, but it is the song I'm most used to working on and is very useful for some of the other tricks- /v/ is /b/,  /f/ is /p/, /z/ is /s/ and coda position /l/ becomes /w/. Even when it isn't working directly on the problem, it helps my mind keep track of what sounds do and do not exist in Binisaya.

 I also practiced unaspirated consonants in English and Binisaya by the simple expedient of putting my hand in front of my mouth, and if I felt my breath strongly, I had to do it again. Luckily, while I can almost hear it when a native Visayan speaker says it, when I say 'big backpack' with no aspiration it sounds like 'pig bakpag' to me and I know what I'm saying. A lot of the time, I end up with popping sounds instead of the actual consonant, but it sounds closer, so I'm still working on it.

The next thing to work on is stress, because that becomes one of the ways to distinguish between word initial /ba/ and word initial /pa/, stressed is /ba/, unstressed is /pa/, and stress differences do create minimal pairs in Binisaya.

Vo'wel Tro-uble

           The first thing I did was identify the five orthographic monophthong vowels /a/, /e/, /i/, /o/, /u/ and the two orthographic diphthongs,/ay/ and /aw/. The second thing I did was identify environments: cV cVc, Vc, nV, nVn, Vn, VV, V'c, cVV, VVc, VVn, nVV, based solely on the orthographic representation. Unlike in English, orthographic representation in Visayan is a fairly good representation of the language's sounds. I then identified the minimum lines I needed to cover all these environments (that occur in this script) for as many vowels as occur in them.

Buhi sa kanunay
Ug di ko hikalimtan

Bisag giyam-iran, bisan pa'g gitamay
Molambo ug molipang
Kay ang gumang matuod di mamatay
Tuhoi intawn ako

Bisa'g ibanlas sa luha

Didto sa kalangitan
Ako kang paabuton

I also highlighted the trouble spots I anticipated. It's worth noting that  - and ' both represent glottal stops in Visayan orthography.
What I found was that the vowels were all about environment. Listen to the clip above. That is every /a/ in the archetype laid end to end, cut as best I could to just the /a/ sounds. How many consonants can you still hear? Word final /a/ and /o/ all have glottal stops following them. The marked glottal stops in pa'g and bisa'g are only marked because they occur within a syllable instead of at the end of it. Vowels on either side of and especially between nasals all but disappear into said nasals. The /u/ in kanunay is all but indistinguishable.
In English, the sound of a word is dominated by the vowels. Get the vowels right, and most English speakers brains will fill in the consonants, regardless of which ones you actually say. In Visayan, there isn't a lot of bleed over, so you still have to say the right vowel, but the consonant's place of articulation is what dominates the rhythm.

The thing I had to practice most was cutting off the airflow to my vowels. English vowels tend to be prolonged, only cutting off when they hit an obstruent. Visayan vowels cut off, except for the diphthongs and coda /i/. English also tends to blend vowels. In Visayan, each vowel is distinct, no matter how many of them are in a row. I found that practicing this laying down helped me cut off the airflow, then I moved to doing it sitting up while I still had the correct sound and sensation in my head.
I practiced saying o-i-in, ho-i-in, ho-i-in-tawn, tu-ho-i-in-tawn for an hour before I could get it out at speed with each vowel distinct.
pa'g is still difficult, but that's more a matter of the consonants. I can get the vowel right; it's getting the /g/ in before the following gitamay.
Overall, the vowels in Visayan are fairly low energy as well. Highest energy goes to /s/ followed by the nasals, then either the vowels or initial voiced consonants. The vowels are strongest as diphthongs, followed by the multi-vowel sequences.
Another thing that helped me was using Praat to take a look at exactly what it was I was trying to say.

The final vowel trick I worked with was practicing the sung version. This is a set of song lyrics. Singing tends to disguise or erase accent markers, and I have a lot more control over what I'm doing, especially with the vowels. Also, the vowels are generally elongated and easier to work with individually. 

The next thing I need to work on is definitely consonants and linking.

Rhythm is gonna getcha.

This past week I've been working on rhythm and speech speed.
The first thing I did was create a beat-count, like a metronome for the archetype. It helps me to match the reading's speed and not instinctively go for the song's rhythm. I'm a percussionist, so I do tend to think in rhythms and beats. Once I had a tempo set, it really helped me stay on track.
The second thing I worked on was practicing reading along with the recording. The first dozen run-throughs showed me the trouble spots, the places where I always, always got off rhythm and the places where I had no trouble at all. Green is good to go, yellow is some trouble, red is where I keep messing up.

I
Buhi sa kanunay
Ug di ko hikalimtan
Sa hilom gibati ko
Ang kamingaw ning dughan

II
Bisag giyam-iran, bisan pa’g gitamay
Molambo ug molipang 
kay ang gugmang matuod di mamatay

Refrain:
Tuhoi intawn ako
Salig sa gisaad ko
Gugma kining way pagkalaya
Bisa’g ibanlas sa luha

III
Kon ang kasing-kasing ko
Sa palad sakiton
Didto sa kalangitan
Ako kang paabuton




https://soundcloud.com/murikakari/rhythmakp

Once I'd identified the trouble spots, I practiced listening to the recording and just chiming in on those spots. I spent a full session on the yellow. The next session worked on saying the yellow and green. Once I had the yellow mostly down and incorporated into the full text, I practiced on just the red. So far, I've managed to usually get the first and last syllables correct for the red lines, but find myself rushing or slowing mid-line time after time. Part of it is what I intend to spend the week after next working on, the change over between coda and initial consonants g-g folowed by g-m followed by ng-m then d-d. 
There's also this week's work- vowels. two or three sequential separately articulated vowels all of which are not quite where my mouth wants to place them. o-u o-i-in, u-o. It's easy enough when they are the same/similar vowel, as in gisaad in Line 9 or paabuton in Line 15, but the quick change between low and high causing problems. I'll be spending some time with those vowels in Praat this week, trying to analyze, exactly what's different about my vowels and Barbie's.

How Many Times Must I Recite.

This is my first attempt at replication.

Things I need to work on:
Voice Quality- Less Tense, Key is about right, but maybe a start a note lower

Rhythm- First attempt was focused on sound not on speed. Increase speech speed and watch stresses.

Consonants- Need to be sharper, better enunciated. I need to watch that I'm not dropping or softening coda-consonants, especially [g]. Also watch out for glottal stops in the codas. Should use Pratt to check for degree of aspiration, especially on [p] and [b], as /p/ /b/ /f/ /v/ can all occur as allophones. Previous experimentation has shown that my personal phonetic system is capable of substituting any of the four for variations not heard in English. (Unaspirated [p] can sound like any of them, though not usually [v] to me.)

Vowels- Rounder, laxer. Need to pay special attention to vowel length and pitch. Textbooks say Binisaya (native name for the language family which includes Boholano) has five vowels [ɑ] [e] [i] [o] [u]. [ɪ] is also in there, as well as others.

Connected Speech- Tendency to hesitate moving from coda-[g] or [d] to inital-[k]. I keep trying to hold coda nasals longer than they should be held as well.

Stress- need to identify phrase level stresses and map. Should probably create a meter chart. Also need to determine just how much stress is given to stressed syllables and how stress is primarily expressed. 

Basic Audio Quality- Forgot to unplug my laptop when I recorded.

NS- Much, much harder to relax into what I already know of the language accent when it's this cold outside.

First thing to work on is rhythm. I'm going to try a sing along technique, namely speaking while listening to the recording. I will also create a count. One rhythm difficulty I've encountered so far is caused by the fact that these are song lyrics. I'm defaulting to the rhythm of the song instead of the rhythm of the spoken word version.

Learning to speak like I'm from Boholan.

O.K. Lang. The dialect I've chosen to attempt for this project is Boholano. Boholano is a subset of a language that is variably called Binisaya, Visayan, and Cebuano. The differences between Boholano and Cebuano are fairly minimal, such as pronouncing /nj/ as [nj] or as [nʤ], but to minimize complications, I used material written by a native Boholano recorded by a native Boholano. I don't want to go to all the trouble of mastering the accent just for my cousins to say I sound like I'm from Cebu.

Finally got my archetype delivered and converted to something I could upload. I was having some trouble with that. It's actually just audio, but that's fine with me.

Below are the lyrics that my helpful cousin recorded for me.
Ako kang Paabuton (Buhi sa Kanunay)
Music and lyrics by Justino 'Ning' Romea

I
Buhi sa kanunay
Ug di ko hikalimtan
Sa hilom gibati ko
Ang kamingaw ning dughan

II
Bisag giyam-iran, bisan pa’g gitamay
Molambo ug molipang kay
Ang gugmang matuod di mamatay

Refrain:
Tuhoi intawn ako
Salig sa gisaad ko
Gugma kining way pagkalaya
Bisa’g ibanlas sa luha

III
Kon ang kasing-kasing ko
Sa palad sakiton
Didto sa kalangitan
Ako kang paabuton
(Repeat Refrain and III)
This is a love song about a lost love.