This is my final post for this blog. This week I continued working on some of those problem areas listed in the last blog post, especially the sections that need to be longer and shorter and the quality of my vowels. Getting the velarization and nasalization right also continues to be a problem.
Here's the link to my last recording for the semester:
Final Recording!
When I compare this with the original, and then with some of the intermediate steps, here are some things that I notice:
*Whatever I am specifically working on that week tends to be better than other aspects. When I was working on vowels, my vowels were better. When I was working on rhythm, my rhythm was better, and so on. That's encouraging, because it shows that deliberate practice does work but it also shows how difficult it is to put all of the pieces together in a way that lasts, and it shows how the lack of deliberate practice in a particular area may cause the previous successes to atrophy.
*After a while, I developed my own expectations of what the archetype should sound like that were different from the actual archetype. I created this pattern where I was listening to myself and comparing with previous versions of my own speech rather than with the archetype first. This was especially apparent to me in the last third of the archetype. I've tried to address that problem in these last two weeks, but it still needs work. This helps me to understand how accents fossilize . Students of language may often fall into the pattern of trying to sound more like the mental image they have of a given target dialect rather than the genuine copy of that dialect. This is probably more pronounced when native speakers are not around to help out. I am sure that my Australian accent would have developed much more accurately if I'd had a real Australian giving me feedback. A recording is helpful, but it simply isn't the same thing.
*When I perform in front of the mirror, I notice that my mouth shape is not the same as my archetype's. I'm approximating the sounds in ways that are different from the natural speech of the image that I am trying to imitate. Again, without genuine feedback from a real person, I'm not sure that I can fix that.
*Another area of serious phonological difference is the way I make my /r/'s. I know from studying a few other languages that the /r/ sound is very difficult to change, and I can sense that in myself with this project. Especially in those situations where /r/ is intervocalic or when it is word final but spoken very quickly, it is VERY hard to do it accurately.
*Lastly, although I am a perfectionist and can sense all of these issues in my accent, overall I think that I am happy with the way that it turned out. I may not be able to say anything ELSE in a convincing Australian accent, but I think with this specific archetype I was successful at making it sound neither American nor British. That's an accomplishment. I'm particularly proud of the work that I put into accurate word stress, sentence final rising intonation, and the development of diphthongs. I did my best, and so I guess that's the best that I can do.
If I were ever to do something like this again, I would definitely want someone around who speaks the dialect natively who could give me more genuine feedback. Without that, the effort becomes wearying and loses its value quickly. If I had to do it over again, I would spend more time focusing on the vowels and find other archetypes to work with to expand my practice material. Instrumentalists never work on only one piece at a time, because they might die of boredom. They play their scales and exercises daily, yes, but they also work on multiple pieces simultaneously. I think that would have helped my motivation and perhaps increased my success in this project as well.
I would like to thank you for reading to the end of my blog. I hope that it was enjoyable.
Sunday, April 21, 2013
Wednesday, April 17, 2013
Weeks 13-14: From Back to Front
This week, I started by analyzing some of my previous recordings. I noticed that over the course of the semester my stress, prominence and intonation have consistently improved. My overal quality (nasality and velarization) also sounds better. However, my vowels on the other hand are a different story. They improved when I was specifically working on them but have since dropped back toward American the comfortable American production, particularly with the /eu/, /i/ and /ae/ sounds. I also am having a lot of trouble with inserting American r-fulness into places where it shouldn't be. Finally, I noticed that my specific words and phonemes are much worse at the end of the monologue than at the beginning. This probably reflects the amount of time I've spent working on each section. So, this week my goals are:
1) Work more on the last 1/3 of the archetype, some on the middle third, and little on the first third
2) Drill difficult vowels, with the goal of producing them both in an appropriate location (depending on stressed or reduced) and the appropriate length (I'm over-lengthening and over-shortening some of the vowels when I compare my speech to my archetype's speech)
3) Drill words with problem /r/ sounds, either present or absent, so that my speech will be smoother.
I listened to the archetype and then my own speech. On my vowel transcription pages that I used earlier in the semester, I circled those words that had some serious problems in pausing, consonant changes, vowel rounding and phrase/vowel length. Here is a list of the segments in my archetype that I chose to drill. The majority (but not all) are from the last half of the archetype. Additionally, I numbered lines so that I could refer to each problem's location in the transcript:
1) Work more on the last 1/3 of the archetype, some on the middle third, and little on the first third
2) Drill difficult vowels, with the goal of producing them both in an appropriate location (depending on stressed or reduced) and the appropriate length (I'm over-lengthening and over-shortening some of the vowels when I compare my speech to my archetype's speech)
3) Drill words with problem /r/ sounds, either present or absent, so that my speech will be smoother.
I listened to the archetype and then my own speech. On my vowel transcription pages that I used earlier in the semester, I circled those words that had some serious problems in pausing, consonant changes, vowel rounding and phrase/vowel length. Here is a list of the segments in my archetype that I chose to drill. The majority (but not all) are from the last half of the archetype. Additionally, I numbered lines so that I could refer to each problem's location in the transcript:
Here's the link to this week's audio recording: Week 13 attempt
Listening to this, although I did trip myself up a few times, the last third of the speech was MUCH better than before. I sound much more natural. I can see how practice really helps. However, I wonder how much this kind of focused practice transfers over into using the accent in another way? In my last blog post I will continue refining my archetype one more time and reflect on what I have learned from the process of doing this project.
Sunday, April 7, 2013
Week 12: Nasality and Velarization
Over the past few weeks while I was working on the project described in the last post, I began to notice some differences between the speaker of the passages that I was analyzing and the speaker that I was trying to imitate. One of the big differences that was brought to the forefront was the issue of sound quality; I began to sense that my archetype has a nasal and velar quality to some of her words that is hard to hear, hard to imitate, and different from the speaker in the passages from the Macquarie website. Furthermore, I watched another Australian movie this week, and I noticed that this nasalization pattern is REALLY strong in the brogue dialect of male speakers. So, it must be one of those aspects of Australian English that varies based on dialect. My archetype speaker is, I think, a speaker of "general Australian", which is the one that is in the middle. Some of her phonemes sound like the brogue phonemes, but there are times when her English sounds very much like the "cultivated" Australian dialect.
Here are some of those places in my archetype's speech where the nasal and velar quality is most apparent. I will highlight the nasalized words in yellow and the velarized l's and far back (almost swallowed) vowels in red. The one instance of orange is when I thought I heard both.
"we are trying to change telephone companies...... we rang a whole bunch, most of them were..."
"It certainly sounded like..."
"um awful anyway..."
"and then we finally rang this other place..."
"needed a one-word answer..."
"a whole sentence or a whole paragraph or a whole spiel you know in answer"
"I've never been able to notice before..."
From these examples, you can see that nasalization usually occurs when a string of nearby words all have an /n/, /m/ or /ng/ sound combination. In these instances, the speaker just nasalizes the entire phrase. The velarization of super dark /l/ /r/ and back vowel "o" tends to happen on stressed words or words that show a pause in the sentence, like "like".
To practice, I've practiced listening to the speaker and repeating. I've done this in the short context of the sentence and in the partial and full monologue. I've also spent more time looking in the mirror while I practice. I am noticing that my overall mouth movements seem to have become smaller. I don't know if this means that my accent is getting more accurate and therefore less exaggerated or if it means my ears and lips have become lazy and I'm not doing as well as I did originally. However, I can say that when I listen to my archetype, I "hear" her differently than I did at the beginning of this project. I would like to think that I sound more like her now than I did in the beginning, also.
Here's my recording for this week:
Week 12: nasalized and velarized syllables
I think that my nasals and velar syllables were ok in this recording, but they could still be better. I'll keep working on these. I also noticed that I've been doing the end of the dialogue quite differently from my archetype. I usually start at the beginning of the dialogue and work through it, so I think in the next week I'll practice working from the back to the front. I'll also re-visit errors that I've been making with /r/ sounds and vowels that still need assistance.
Here are some of those places in my archetype's speech where the nasal and velar quality is most apparent. I will highlight the nasalized words in yellow and the velarized l's and far back (almost swallowed) vowels in red. The one instance of orange is when I thought I heard both.
"we are trying to change telephone companies...... we rang a whole bunch, most of them were..."
"It certainly sounded like..."
"um awful anyway..."
"and then we finally rang this other place..."
"needed a one-word answer..."
"a whole sentence or a whole paragraph or a whole spiel you know in answer"
"I've never been able to notice before..."
From these examples, you can see that nasalization usually occurs when a string of nearby words all have an /n/, /m/ or /ng/ sound combination. In these instances, the speaker just nasalizes the entire phrase. The velarization of super dark /l/ /r/ and back vowel "o" tends to happen on stressed words or words that show a pause in the sentence, like "like".
To practice, I've practiced listening to the speaker and repeating. I've done this in the short context of the sentence and in the partial and full monologue. I've also spent more time looking in the mirror while I practice. I am noticing that my overall mouth movements seem to have become smaller. I don't know if this means that my accent is getting more accurate and therefore less exaggerated or if it means my ears and lips have become lazy and I'm not doing as well as I did originally. However, I can say that when I listen to my archetype, I "hear" her differently than I did at the beginning of this project. I would like to think that I sound more like her now than I did in the beginning, also.
Here's my recording for this week:
Week 12: nasalized and velarized syllables
I think that my nasals and velar syllables were ok in this recording, but they could still be better. I'll keep working on these. I also noticed that I've been doing the end of the dialogue quite differently from my archetype. I usually start at the beginning of the dialogue and work through it, so I think in the next week I'll practice working from the back to the front. I'll also re-visit errors that I've been making with /r/ sounds and vowels that still need assistance.
Friday, April 5, 2013
Weeks 10/11: Stress and Prominence
Subtitle: The post in which you learn that Jessica is a crazy person who invents informal research projects and performs them on herself. Just because.
In the past few weeks I have been periodically practicing difficult sections of my archetype, especially in vowel sounds and stress. I have also "performed" for several non-linguistic friends and asked their opinions of areas where I sound Australian and areas where I still sound American. This has been helpful in determining what I still need to work on.
From a more academic perspective, I wanted to get a little variety and move away form my specific archetype to listen for and practice stress, prosody and vowel sounds (both stressed and reduced) in other passages that I found on the Macquaire website. Since I have memorized my archetype now, it has become very difficult for me to hear rightly the sounds that I am making when I speak and their relative distance from the target. I thought that if I washed my brain out with deliberate practice of some different target paragraphs, that might help me to get closer to where I want to be.
To compare the prominence patterns, I first recorded myself reading the following four passages in English. I then listened carefully both to my American recordings and the Australian recordings. I marked stressed words with square boxes and highlighted the words in each clause or phrase that I believed was most prominent. In several cases, two or more syllables in a phrase seemed equally prominent to my ears, so I tended to err on the side of over-highlighting prominence. I then analyzed the similarities and differences in prosidy in the four paragraphs, particularly paragraphs 1 and 4.
I must put in a disclaimer here that even individuals with the same cultural background would likely read these passages with slight alterations in prominence. However, I expect that members of a similar speech community and reading fluency level would be far more alike in their determinations of prominence than only 50-ish% of the time. Reading skill level, fluency, and familiarity with a particular genre doubtless play a role. Also, I know from experience that we read aloud according to patterns that we have learned from other native speakers. I read in a similar way to that which was read to me by my parents, and while this likely varies somewhat through families, it must also be generalizable across speech communities. (Otherwise, how would we know what to expect from audiobooks?)
Passage 3 and 4 are excerpts from the Hobbit by J.R.R. Tolkein, a fantasy story written in Britain by a British man in the mid 20th century. I expected that since the American and Australian narrative literary tradition comes from the same roots, and this passage comes from a different dialect community altogether, the prominence markers in these passages would be more alike. However, I didn't analyze these at all scientifically. My basic premise was "mark all of the stress, then go find the one stressed part in each phrase that is the strongest". Additionally, I didn't mark intonation and how that affected my judgements of prominence. There were several instances where I marked two words very close to each other because the intonation went up on one and down on the other, indicating that this idea, rather than a specific word, is what was really prominent. This occurred in both Australian and my personal sample. I also went back and re-checked the American and Australian samples for passage 3 and 4 a second time, because a significant amount of time passed between when I "measured" passages 1 and 2 and when I did the same to passages 3 and 4.
Here are the pictures of Passages 3 and 4 before modification, but after modification I determined that the prominent point count stands thus: #3: Australia 29, Me 25. Passage #4: Australia 44, Me 41.
So, for passage 3 the prominent points lined up 13 times, and for passage #4 it lined up 30 times. If I take the higher number from the Australian passage totals as my point of comparison, that means in passage #3 the prominent points correlated 49% of the time and for passage 4 they lined up 68% of the time. This last one is a lot higher!
One of the things I found most interesting about passage 4 are some of the sections that did line up! For example, both my version (which I recorded and coded BEFORE listening to the Australian passage, by the way) and the Australian speaker highly emphasized the phrases "two big round pale eyes", "quite quietly" and "He paddled it with large feet dangling over the side, but never a ripple did he make. Not he." in exactly the same way. This is pretty cool. It demonstrates to me that there are shared conventions for storytelling and narration that have been and continue to be maintained across the globe for our English dialect group, at least among native speakers and for certain types of stories. We have some mutual expectations for how language is supposed to sound, even though we often, as much as half the time in this informal and ridiculously unscientific case study, emphasize different words and phrases in different ways.
So.....how does this relate specifically to making me sound more Australian? Well....I'm not sure it does. But it does help me to understand some of the features that may be playing a role in narration and conversation in different dialects. Additionally, this side project helped me to hear better the differences between cultivated Australian dialect, which is what the narrator of these passages was speaking, and my Archetype, which has a thicker more "brogue-like" accent. I mentioned at the beginning of the post that hearing these sound distinctions was one of the issues I wanted to address at this point in the project. Therefore, I consider that issue addressed.
Because I have realized that these accents are really not the same, I chose not to try and imitate and record these passages, because I think it might actually be detrimental to making me sound like the archetype in terms of phonology. The narrator of these passages doesn't use as much of a nasal sound or the same quality of back vowels that my archetype does. My post next week will return to things that I have been working on related directly to my archetype, particularly nasality and the velarization of back vowels just mentioned.
In the past few weeks I have been periodically practicing difficult sections of my archetype, especially in vowel sounds and stress. I have also "performed" for several non-linguistic friends and asked their opinions of areas where I sound Australian and areas where I still sound American. This has been helpful in determining what I still need to work on.
From a more academic perspective, I wanted to get a little variety and move away form my specific archetype to listen for and practice stress, prosody and vowel sounds (both stressed and reduced) in other passages that I found on the Macquaire website. Since I have memorized my archetype now, it has become very difficult for me to hear rightly the sounds that I am making when I speak and their relative distance from the target. I thought that if I washed my brain out with deliberate practice of some different target paragraphs, that might help me to get closer to where I want to be.
To compare the prominence patterns, I first recorded myself reading the following four passages in English. I then listened carefully both to my American recordings and the Australian recordings. I marked stressed words with square boxes and highlighted the words in each clause or phrase that I believed was most prominent. In several cases, two or more syllables in a phrase seemed equally prominent to my ears, so I tended to err on the side of over-highlighting prominence. I then analyzed the similarities and differences in prosidy in the four paragraphs, particularly paragraphs 1 and 4.
Me (American) Passage 1
Australian Passage 1
When comparing these first two, I was impressed by how I found the same exact number of prominent words, but that the points of prominence landed on the same syllables only 10/17 times. That's only 58% of the time! Also, these differences weren't specific to one part of speech. We chose to put adverbs, adjectives, nouns and verbs in the points of prominence with great variation. Sometimes we made similar choices, and sometimes we didn't.
For passage 2, I noted 21 prominent words in my own speech and 20 in the Australian sample. This paragraph was the most "conversational" of the four passages, and I think that shows in the prominence disparities. In these passages, only 9 points of prominence were in the same locations, or 45% of the time. My hypothesis is that American and Australian English probably has more differences in prosody when we are simply speaking than when we are reading from passages which share more unified cultural narrative patterns.
My Passage 2 Australian Passage 2
I must put in a disclaimer here that even individuals with the same cultural background would likely read these passages with slight alterations in prominence. However, I expect that members of a similar speech community and reading fluency level would be far more alike in their determinations of prominence than only 50-ish% of the time. Reading skill level, fluency, and familiarity with a particular genre doubtless play a role. Also, I know from experience that we read aloud according to patterns that we have learned from other native speakers. I read in a similar way to that which was read to me by my parents, and while this likely varies somewhat through families, it must also be generalizable across speech communities. (Otherwise, how would we know what to expect from audiobooks?)
Passage 3 and 4 are excerpts from the Hobbit by J.R.R. Tolkein, a fantasy story written in Britain by a British man in the mid 20th century. I expected that since the American and Australian narrative literary tradition comes from the same roots, and this passage comes from a different dialect community altogether, the prominence markers in these passages would be more alike. However, I didn't analyze these at all scientifically. My basic premise was "mark all of the stress, then go find the one stressed part in each phrase that is the strongest". Additionally, I didn't mark intonation and how that affected my judgements of prominence. There were several instances where I marked two words very close to each other because the intonation went up on one and down on the other, indicating that this idea, rather than a specific word, is what was really prominent. This occurred in both Australian and my personal sample. I also went back and re-checked the American and Australian samples for passage 3 and 4 a second time, because a significant amount of time passed between when I "measured" passages 1 and 2 and when I did the same to passages 3 and 4.
Here are the pictures of Passages 3 and 4 before modification, but after modification I determined that the prominent point count stands thus: #3: Australia 29, Me 25. Passage #4: Australia 44, Me 41.
So, for passage 3 the prominent points lined up 13 times, and for passage #4 it lined up 30 times. If I take the higher number from the Australian passage totals as my point of comparison, that means in passage #3 the prominent points correlated 49% of the time and for passage 4 they lined up 68% of the time. This last one is a lot higher!
Australian passage #4
One of the things I found most interesting about passage 4 are some of the sections that did line up! For example, both my version (which I recorded and coded BEFORE listening to the Australian passage, by the way) and the Australian speaker highly emphasized the phrases "two big round pale eyes", "quite quietly" and "He paddled it with large feet dangling over the side, but never a ripple did he make. Not he." in exactly the same way. This is pretty cool. It demonstrates to me that there are shared conventions for storytelling and narration that have been and continue to be maintained across the globe for our English dialect group, at least among native speakers and for certain types of stories. We have some mutual expectations for how language is supposed to sound, even though we often, as much as half the time in this informal and ridiculously unscientific case study, emphasize different words and phrases in different ways.
So.....how does this relate specifically to making me sound more Australian? Well....I'm not sure it does. But it does help me to understand some of the features that may be playing a role in narration and conversation in different dialects. Additionally, this side project helped me to hear better the differences between cultivated Australian dialect, which is what the narrator of these passages was speaking, and my Archetype, which has a thicker more "brogue-like" accent. I mentioned at the beginning of the post that hearing these sound distinctions was one of the issues I wanted to address at this point in the project. Therefore, I consider that issue addressed.
Because I have realized that these accents are really not the same, I chose not to try and imitate and record these passages, because I think it might actually be detrimental to making me sound like the archetype in terms of phonology. The narrator of these passages doesn't use as much of a nasal sound or the same quality of back vowels that my archetype does. My post next week will return to things that I have been working on related directly to my archetype, particularly nasality and the velarization of back vowels just mentioned.
Subscribe to:
Posts (Atom)