Professional Documents
Culture Documents
Andreychuk, Margarita
Dhesi, Cristina
Moreno, Mario
Acquisition and development of the English language
Group 81
INDEX
INTRODUCTION
Language and music have long been regarded as unique human abilities. Even
though, animal species can communicate and, some of them also sing, their skills are
very limited when compared to that of humans. The linguistic and musical capabilities
are some of the most remarkable characteristics of our species. Language Acquisition,
together with Neuroscience and other areas of Applied Linguistics investigate the
relations of these disciplines and how they affect human behaviour.
Scientists have long speculated about the effects of music instruction on
Language Acquisition. While it seems clear that the brain is favoured by the
development of music activity, such as learning how to play an instrument, or even
listening to music, it is uncertain how this affects language acquisition. Nevertheless,
the most common assertion is that the influence proves beneficial.
For our project, we were interested in observing if musical education could have
positive influence in the oral component of Second Language Students. We
contemplated the possible enhancement of sound reception and production in those who
had engaged in musical instruction. With this hypothesis in mind, we gathered two
groups, one with musical education and another one without. We decided to put our
subjects through a series of tests that would later help us analyse our findings. These
tests included a grammar test, a test for the reception and production of unknown
sounds and another test for the production of familiar sounds. The first with the
intention of controlling that our subjects had similar level of dexterity in their Second
Language (L2), thus eliminating possible variables. The second and the third tests were
the focus of our project. We examined our collection of data, analysed the results, and
compared the performances of the two groups.
Our main goal was to carry out the project and present the results in an academic
form in order to have a better understanding of how language acquisition research
works. We also hoped that our intuitions met with the findings of the investigation.
Nevertheless, we were fully aware of the scope of our research before beginning our
project. While our research was of a preliminary nature, it is important to expand the
existing literature on the field.
THEORETICAL BACKGROUND
The role of music in Language Acquisition has always been an intriguing theme.
Firstly, because of the relevance of songs as a source for word learning by learning their
lyrics and singing them and secondly, because of the aural acquisition that music seems
to facilitate to those people who pay real attention to each of the musical resources on
them. According to D. Schn et al. (2008), the presence of pitch contours may
enhance phonological discrimination, since syllable change is often accompanied by a
change in pitch, if we look at it from a perceptual point of view. In addition, the
consistent mapping of musical and linguistic structure may optimize the operation of
learning mechanisms. At the acoustic level, we clearly appreciate that music and
speech use pitch, timing and timbre notions as to convey information. But at a cognitive
level, music and speech processing coincide in certain requirements such as memory
and attention skills. Being this the case, musicians show an advantage in processing
pitch, timing and timbre of music compared with non-musicians. These cognitive skills
are also crucial for speech processing; then, we can assume that long-term training with
musical pitch patterns can promote the processing of pitch patterns of foreign languages
too.
If we focus on the relationship between music and neuroscience, we can see that
there are many researches made in the line of observing the link among music and
cognitive functions. First of all, it has been demonstrated that the brain structure of
musicians differs quite a lot from non-musicians. One of the most relevant studies in
relation to this area was that of Gaser and Schlaug (2003), where they compared brain
structures of professional musicians with non-musicians. The important discovery was
that gray matter volume differenced in motor, auditory and visual-spatial brain regions,
what gave way to the assumption of musicians having more and better cognitive and
sensorial capacities than non-musicians though there are always exceptions,
obviously-. Once analyzed several brain regions in both subject groups brains and
having already seen those actually existing differences, it is more possible to say that
they are due to the long-term acquisition and repetitive rehearsal of musical skills rather
than innate principles or cognitive functions.
But for the aim of our project, it is quite interesting what a recent article by
researchers from Northwestern University showed up. Some of the results that they
obtained were that musicians are more successful than non-musicians in learning to
incorporate sound patterns for a new language into words. Thus, as we slightly
suggested before, people who are musically trained manifest stronger neural activation
to pitch changes in speech, having a better vocabulary and reading ability than people
who did not did musical training.
However, an important aspect of this existent aural capacity has been seen by
many doctors as very linked to language capacity also by their learning strategies. If we
look at OMalley and Chamots typology of learning strategies (1987), we will see that
most of the strategies that are given could be also applicable to those carried out by
music learners (metacognitive and cognitive strategies, not social in this case). But
certainly there are several steps that are more general to everyone in the process of
learning. Some of them are listening carefully to how a word/note should sound;
starting with a small part of a sentence/piece of music; trying it once and again until it
is as close to perfect as possible; keep on doing the following sections the same way;
and finally, going over it from beginning to end to make sure you remember it well.
Following this general learning structure, Conversational Solfege by John M.
Feierabend is a general music program that enables students become independent
musical thinkers with the help of a rich variety of folk and classical music, but always
with the dynamic of these previously mentioned learning steps. Thus, in Dr. John
Feierabends words (1995:9), One should learn with his/her ears before learning with
his/her eyes. That means that auditory capacities are fundamental when learning a
language since sounds are the first attempt that we have to get in touch with the
communicative world, due to the fact that when we born, we do not know to read yet,
but sounds can be getting familiar to us earlier than any other thing.
In other to test our hypothesis we focused on the evidence collected from three
different tests. We will now describe the subjects we selected and the tests we chose.
Also we will mention how we reduced the variables in our experiment in order to try to
carry the tests on equal conditions for all the subjects.
5
The data collection obtained was the result of three different tests labelled
within the two different forms of collecting data: questionnaires to elicit our subjects
proficiency in English, and oral production of samples in the target languages. Thus,
the three tests that we applied are the following:
-
Grammar proficiency test: to control that our subjects had a similar knowledge of
one of the tested languages (English). This will be useful when co-relating the next
tests applied.
To learn more about the subjects individually see List of subjects in appendix.
the recorded room so that the subjects could feel relaxed. Making them feel nervous
would have had consequences in our outcome. In general, we focused on making the
recordings feel as natural as possible as to achieve reliable results.
The first of the two oral tests was the reception and production of unfamiliar
sounds. This test consisted in making the participants repeat sentences in two languages
unknown to them, namely Russian and Estonian. One of the members of our team, a
native speaker of both these languages, would sit side by side with our subjects making
them repeat first isolated words, a few times, and then say them within a context: a short
sentence of five words - first in Estonian and later in Russian. These sentences had been
carefully chosen by our team in order to include sounds that could result particularly
challenging for our subjects. We did not look for a perfect reproduction but rather for
the general capacity of subjects to receive and produce unknown sounds.
These two languages differ greatly from what the average subject normally is
acquainted with. Russian derives from the group of Slavic languages and thus has
sounds that might be unusual for most of our subjects. Moreover, Estonian comes from
a different family of languages: Uralic. Both Russian and Estonian have phonological
systems different from that of English, which makes the task of transcribing more
difficult. In order to avoid confusion, we decided to adopt the symbols used in English,
with the exception of the phonemes: // (mid back unrounded vowel; e.g. in the word
udsalt in Estonian), // (mid, rounded and front; as in the Estonian word sel), /y/
(high, rounded, front; as in the word le in Estonian), and // (close central unrounded
vowel as in Russian word ).
The last of our three tests consisted in making our subjects reproduce four
English sentences. We selected sentences that could help us identify the subjects skills
at the time of utterance. Given that 7 out of 10 of our subjects scored C2 level in the
grammar test, we knew that they were going to be familiar with all English sounds. In
order to create a more difficult scenario we chose sentences that could raise problems.
For the purpose of clarity, we will number the sentences in order to refer to them as: S1,
S2, S3, and S4. There were two English tongue twisters: Peter Piper picked a peck of
pickled peppers (S2) and Santas short suit Shrunk (S3) and, placed in 1st and 4th
position, the sentences: Where were you born (S1) and Where were you burnt (S4).
Subjects were given only one opportunity to pronounce the sentences, right after
reading it, to assure their pronunciation was natural and not forced.
Given that tongue twisters can result difficult even to native speakers, this test
allowed us to observe the subjects real ability to produce problematic sentences. Also,
we were able to contrast these results with those obtained from the first test. S1 and S4
challenged them to distinguish between born /brn/ and burnt /brnt/, but that is
something that we will develop later on our analysis. We will compare the general
performance of the two groups.
101
102
103
104
105
16
14
12
10
8
6
4
2
0
201
202
203
204
205
Fig. 1.1> G1
The second tests introduced us into the oral part of our experiment, so we started
with the reproduction of Estonian language, making them repeat the underlined words
in the following sentences:
1a. Mulle udsalt meeldib Prnu sdaelu. (I love the nightlife of Prnu)
[mulle utsalt me:ldip prnu syda,:elu]
1b. sel mina knnin le silla tasapisi. (I quietly cross the bridge at night)
[:sel mina knnin yle silla tasapisi]
After being acquainted with the words, we made them pronounce the whole
sentence in two parts: the first three words and then the rest.
We proceeded then with Russian language, following the same method: separated
words, then short clauses.
2a. . (The bull immediately started and run
away)
[bk srazu vzdrognul ubial]
2b. -. (I am going to Saint Petersburg)
[ja ujeaju v sank piterburk]
Our recordings were made in the way that one of the members of our team had
chosen and previously interviewed our subjects. The rest of us, who had to measure
their ability in their production, did not know who did play an instrument or had to do
anything with music in general. Thus, we avoided the prejudice we might have had in
relation to our hypothesis.
Analysing the recordings, we found some general problems in both groups: the
one that did not play a musical instrument (G1), and the one that did (G2). The first
problem was quite expected by our team. Many people failed to understand and to
reproduce Estonian vowels , and , as well as the Russian . However, the some
subjects that had problems with the sound // () got it at the second time, which could
be explained by its similarity to the English // in bag. While only 2 members of G1
were able to finally identify the sound, none of G2 had any particular problem with it.
Those unable to get the sound replaced it for the first sound of the diphthong /ai/ as in
buy, and thus they were able to reproduce the whole word. As for the rest of the sounds,
they seemed to be more problematic for our subjects.
9
The most difficult came to be the Estonian // and the Russian // in the word
such as udsalt and , respectively. The general tendency was to substitute it for /i/
and /e/, sounds usual and thus easier for them. The results obtained seemed to point out
that G2 performed much better: sound // was captured by 4 subjects from G2 and none
from G1, while // was identified by everybody from G1 and only 2 from G1.
We should say that from the set of problems that Estonian language offered the
most important came to be the number of vowels that occur in a sequence within
compound words. This is quite typical in this language - also in Finnish -, and we can
find it in the word sdaelu, a compound consisting of three words: sda, and elu.
The result is a combination of vowels hard to pronounce: -ae-. In addition, we saw
that people had problems with long vowels within these sequences; the tendency was to
shorten them, as in the case of the long /:/ in the word sdaelu. Others inserted
another vowel, creating diphthongs such as /e/ or /i/. But from all the subjects only 3
could finally make do with these vowels; one form G1 and 2 from G2.
Concerning Russian language, it should be pointed out that the most problematic
feature seemed to be the consonant clusters; namely, in the words and .
However, the difficulty of these consonant clusters was caused by different reasons. The
first one, /vzdr/, coordinated too many consonants, which was quite unusual for our
subject, remembering them being rather a challenge for them. The sounds on their own
were not complicated, and are found in Spanish language. The solution we observed
was that some subjects tended to insert a vocal between the first two consonants:
/vizdrognul/, for example.This insertion of a vocal made it easier to reproduce this
consonant blend. Thus, asked to pronounce this word within a sentence, only one
subject from G1 succeeded in doing it correctly. However, there is a variable we should
take into account: this subjects mother tongue was Polish, a Slavic language, and which
has similar to Russian sounds. That is why this particular subject had no major problems
in reproduction of the words and sentences in Russian.
The case of the second cluster was a bit different; it resulted from the
combination of voiceless alveolar fricative /s/ and alveolar trill /r/. People were
generally trying to voice the /s/, getting /zr/ as a result, which obviously would be a
more sensitive way to pronounce these two letters together, but, which is not the case in
Russian in this particular context. Moreover, the vast majority of our subjects from
10
both groups (3 from each) inserted /t/ into this cluster, [strazu], as a link to make the
way from voiceless to voiced consonants more gradual, while trying to reproduce the
word within a context.
A curious thing manifest in the outcome of both groups was that they failed to
perceive some start or end consonants within a phrase: [_e:ldi_] instead of [me:ldip]. It
is important to point out, that the case of dropping out consonants was found only in 1
subject from G1, against 3 from G2 (especially in the word [vzdrognul]). This data
could weaken our hypothesis. Nonetheless, we could explain it by the thing we
observed while recording our subjects: generally, G2 took their time in starting
reproducing the sounds they were asked to. The fact that they omitted some sounds
might have been due to their fear to pronounce them in a wrong way, and they preferred
to leave them unpronounced instead. Furthermore, 4 out of 5 from G2 could not get the
final /l/ for past tense masculine, as in [ubizal].
An occasional aspiration of a
consonant was found in 3 subjects from G2, being rather a mistake made only one.
However, 2 members of G1 showed to be really influences by this pattern, producing
aspirations regularly.
11
On the other hand, as 90 per cent of the tested people were Spanish native
speakers, it was curious that a subject of G1 tended to replace /b/ in [bk] by /v/ in
vino. This phenomenon must be linked to the general confusion over the use of v and
b, which are both pronounces as /v/ or /b/, depending on the context. On the other
hand, 2 subjects from G1 dropped out the last consonants of the word C [sank],
supposedly associating it with the Spanish word san. The influence of the words
similarity to the same proper name in their mother tongue made some subjects try to
insert an s into this word: [pitersburk], since, as we believe, it is found in Spanish
(San Petersburgo). To be exact, 2 subjects from G1 made this mistake. Thus, we could
see how the L1 transferred to the L2s perception; L1 influenced the way people
perceived and reproduced the sound we were dealing with.
To conclude this part of analysis we could note the fact that G1 was less capable
of remembering a whole phrase, while G2 showed more facility and productivity. Some
G1 subjects mixed the sound form different words and they did not perceive the error.
Moreover, it appears that G2 was more independent from the sound they already knew
from the languages they speak; that is to say, they were more natural in their
reproduction of the input. It was seen in that they made less use of schwa and the case
of aspired consonants was far rarer in their reproductions.
The figures below contrast the performances we obtained. We evaluated the
subjects over 10 points with a general perspective. Figure 2.1 corresponds to G1, figure
2.2 to G2.
12
9
8
7
6
5
4
3
2
1
0
101
102
103
104
105
Estonian out of 10
Russian out of 10
10
9
8
7
6
5
4
3
2
1
0
201
202
203
204
205
Estonian out of 10
Russian out of 10
Finally, our third and last test was based in challenging our subjects to produce
some tongue twisters in order to prove their pronunciation in an already known
language.
S1 and S4 were mostly about being able to distinguish the sounds // and //
since the last words of the sentences were born /brn/ and burnt /brnt/, respectively.
Most of our subjects in G1, for example, were unable to reproduce the difference
between the sound /:/ in S1 and the sound // in S4. Even if this does not point out the
inability of subjects in G1 to produce the sound // it does seem to suggest that under a
challenging context G2 was able to perform better. Also it seems important to highlight
that where some subjects of G1 failed to produce the final /t/ in the word burnt of S4,
every subject in G2 produced it. On the one hand, this again may be due to the
similarity in form to S1. If the subject does not do a complete reading of the sentence,
he or she might simply repeat what he or she said in S1. On the other hand, this factor
13
might point to another factor unrelated to sound performance, which may well be the
general ability of subjects to focus in a particular task. Although this distances from the
main goal of our investigation we shall later comment on this as an external factor that
might influence the subjects performance.
Another pattern we observed was the different realisations of the word where in
initial position. The full form of where is transcribed as /wer/, but the vowels are often
reduced in conversation, plus next to the weak form of were, /wr/, the subject might be
driven to realise the two in a similar manner.
S2 is a very popular tongue twister. The general difficulties are the
mispronunciation of diphthongs for vowels. Some of this confusion is due to the fact
that the letter i is found both as a short vowel /i/ and as a diphthong /a/. This
difficulty is increased by the fact that every word begins with the bilabial plosive /p/
which usually drives the speaker to utter short vowel instead of a diphthong.
In S2 some problems emerge in the pronunciation of Piper /papr/. Contrary to
what we thought, this particular word was performed better by G1 than by G2. Subjects
201 and 204 (of G2) failed to produce the diphthong and instead realised it as a short
vowel /i/ resulting in /pipr/. The past tense verb form picked /pkt/ was heavily
mispronounced by subjects 101 and 105 (G1) whereas it did not raise particular
problems among subjects of G2. Also subject 101 mispronounced pickled /pikld/ as
/pikitn/. The short vowel /i/ of the word pickle /pikl/ was only mistakenly realised as a
diphthong by subject 104. And finally the word peppers /peprz/ was accurately
produced by 3 members of each group. Subjects who mispronounced this last word
realised the open short vowel /e/ as /i/ a common problem that arises when reading this
tongue twister.
The tongue twister in S3 presents the problem of realising the sound /s/ followed
by the sound // in a four word sequence. What it does is that it confuses the speaker as
to what sound,/s/ or //, should follow. Other problems are found, as usual, in the
realisation of vowels.
G1 confused more sounds than G2. The word short was correctly pronounced by
both groups in each occasion, except for subject 101 and 201. Subject 101 did a very
weak realisation of the sound // whereas 201 realised it as /s/. It is worth noting that
14
subject 201 realised every word of S2 closer to /s/. Also, the word suit was realised as /
u:t/ instead of /su:t/ by subject 104. The final sound /t/ is met by //, thus creating
further difficulty. Subject 103, for instance, assimilated this sound and produced it
closer to a /d/ thus resulting in a form closer to /u:d/. Subjects were particularly
troubled by the final word, shrunk, since it is arguably the most difficult to utter,
particularly after uttering the previous sequence. The vowel // was mistakenly realised
by some subjects of G1. 101 failed to produce the vowel and 104s pronunciation was
rather weak. The rest of the subjects seemed to achieve a quite acceptable
pronunciation.
Now let us revise our previous analysis and compare the performance of the two
groups. The tables below show the number of correct words uttered by each subject in
each sentence2. Figure 1 presents the results of subjects in G1 and Figure2 those of G2.
It seems important to clarity that S1, S3, and S4 are evaluated over 4, whereas S2 is
evaluated over 8, corresponding to the number of words in each sentence.
8
101
102
103
104
105
0
S1 out of 4
S2 out of 8
S3 out of 4
S4 out of 4
Fig.3.1>G1
10
8
201
202
203
204
205
0
S1 out of 4
S2 out of 8
S3 out of 4
S4 out of 4
Fig.3.2>G2
A quick glance of these numbers seem to point out that G2 performed better than
G1. However, the number of factors that should be taken into account to give an
2
We should note once again that we qualified the subjects pronunciation ourselves and therefore, the
results should be taken as nothing more than an orientation for the reader.
15
approximate conclusion of this type of study exceeds the scope of our research. We
must, then, restrict our project to the data we collected and the variables we took into
account.
One important factor, not reflected by the numbers in the figures above, was that
G2 usually took longer to reproduce the sentences. This is interesting since it seems to
point out that those that took longer to utter the sentences generally performed better.
Whether musical education has anything to do with this is uncertain and unrelated to
our project. However, it seems to suggest that subjects who had received musical
education were less anxious which could help in their performance. On the other hand,
G1 usually reproduced the sentences faster and hesitations during their performances
were more common. This could suggest the opposite, that is, that G1 subjects were
more anxious during their performance. Then again, anxiety is an extremely
complicated variable and it seems impossible to measure it.
Individual and group results may be misleading. The results obtained by subject
101 suggest some particular problems at the task of pronunciation. This corresponds
well with its grammar test which scored lower than most subjects. Therefore it might
point to a personal rather than a group factor. This would, again, reduce the reliability of
the numbers indicated by figure 1 and figure 2 above.
CONCLUSION
All the test being realized, we would like now to conclude our research. After
realization of all the three parts of our tests, the results we have got are quite interesting
and challenging for further and more profound research. Even though the difference
between the data collected within the subjects with musical education and the ones
without was not so striking, if we analyze the figures in the tables, we will see that it
(the difference?) is there. When we were selecting people for our project, we were
absolutely aware of the variables and also similarities our subjects would have. First, all
of them were our classmates, which supposed they would possibly have a similar level
of English. This would imply that the grammar test and the test of English sounds had to
be scrutinized in order to obtain any possible clue for our hypothesis. On the other
hand, there were people who spoke languages similar to Russian and Estonian.
Obviously, they could obtain better results in the test of unknown sounds as compared
to others. All these details taken into account, we intended to be impartial observers of
16
the outcome. Finally, when we analyzed all the data, we got enough evidence in all
three tests - to believe our hypothesis was proved.
In addition, we would like to point out that generally, both groups took their
time in reproducing the words and phrases, although people that got better results in this
test, were mainly slower in starting the repetition. They asked to repeat the word or the
phrase twice before they took the risk of pronouncing it, thus achieving a more accurate
result. An important note was taken on the fact that G1 was less capable of
remembering a whole phrase, while G2 showed more facility. Moreover, in the test of
unknown sounds, it appears that G2 was more independent from the sounds they
already knew; that is to say, they were more natural in their reproduction of the input. It
was seen in that they made less use of schwa and the case of aspired consonants was far
rarer in their reproductions.
To conclude, we should say that, on the contrary to the grammar test, which
consisted in choosing an appropriate answer, and which was easy to correct, the
transcription and the evaluation of the data collected on the basis of recording and was
much more complicated, taking into account that none of us is a specialist in phonetics.
On the other hand, the overall results might have been rather different if the scope of
our investigation was not so limited.
17
REFERENCES
Chamot, A.U. & O'Malley, J.M. (1987). A cognitive academic language learning
approach: A bridge to the mainstream. TESOL Quarterly.
Feierabend, J. M. (1995). First steps in music for nursery and preschool series. GIA
Publications.
Gaser, C.; Schlaug, G. (2003). Brain structures differ between musicians and nonmusicians. The Journal of Neuroscience (23(27):9240924). Available online at
<< http://www.jneurosci.org/content/23/27/9240.full.pdf+html>>
Schn, D.; Boyer, M.; Moreno, S.; Besson, M.; Peretz, I.; Kolinsky, R. (2008). Songs as
an aid for language acquisition. Cognition 106 (975983). Available online
at <<http://www.incm.cnrs-mrs.fr/publication/schoncognition_08.pdf>>
18
APPENDIX
List of subjects
Do not play an instrument
101
102
103
104
105
ESTONIAN
6/10
8/10
5/10
4/10
6/10
RUSSIAN
5/10
6/10
4/10
4/10
6/10
Play an instrument
201 (violin)
202 (guitar)
203 (clarinet)
204 (guitar)
205 (piano)
ESTONIAN
7/10
8/10
8/10
8/10
9/10
RUSSIAN
6/10
7/10
7/10
6/10
7/10
L1
Other Ls
Lived abroad
101
102
103
104
105
Sp
Polish
Sp
Sp
Sp
Studying
English for
15 years
4 years
17 years
15 years
15 years
/
Spanish, German
French, German
French, German
Japanese
No
Berlin (1 year)
No
No
No
Play an instrument
L1
Other Ls
Lived abroad
201 (violin)
202 (guitar)
203 (clarinet)
204 (guitar)
205 (piano)
Sp
Sp
Sp
Sp
Sp
Studying
English for
18 years
14 years
13 years
13 years
12 years
French
No
No
Paris, 1 year
No
No
19
French, Italian
French, Finish.
20
d- Does
10- How ____ have you and your family lived in this flat?
a- Soon
b- Much
c- Often
d- Long
11- Its a great place to live apart from the increasing volume of _____ that passes under my window
every day.
a- Transport
b- Traffic
c- Vehicles
d- Circulation
12- We _____ a lovely three weeks in the south of Spain last year
a- Passed
b- Took
c- Did
d- Spent
13- Fiona is very angry _____ her bosss decision to sack several members of staff.
a- Against
b- About
c- For
d- By
15- This meat is beautifully ____. What recipe did you use?
a- Gentle
b- Tender
c- Mild
d- Soft
Name:
Gender:
Studying English for (years):
Play instrument: yes/no
L1:
other languages:
lived abroad: (where/for how long)
if yes, what instrument:
21
101
[eutsalt]
[parnu]
[suda,j:elu]
[o:sel]
[kinnin]
[yle silla]
[srasu]
[vizdrognul]
[ubzal]
[ubezaju]
[san piterburk ]
[iutsalt]
[prnu]
[sytae:ilu]
[:sel]
[kinnen]
[yle silla]
[srazu]
[vzdrognul]
[ubial]
[ujizaju]
[sant pitersburk ]
[iudsalt]
[parnu]
[syta:elu]
[mulla eutsalt me:ldip parnu syda,:elu]
[:sl]
[kinnin]
[ule silla]
[sraz]
[vzdrognl]
[ubidal]
[jueaju]
[sank piterburk ]
[eutslt]
[prn]
[sydae:lu]
[:sel]
[knnin]
[yle silla]
[zrazu]
[vzdroznl]
[ubidal]
[ujedaju]
[san_ pitersburk ]
[eutsalt]
[prnu]
[sytailu]
[:sel]
[knnin]
[ule silla]
[zrazu]
[vzdrognl]
[uvizal]
[ujezaju]
[sank peterburk ]
23
Play an instrument
201
[eutsalt]
[prnu]
[sydao:elu]
[:sel]
[kjnnin]
[yle silla]
[srazu]
[vzdrodnul]
[ubizal]
[ujezaju]
[sank piterburk ]
[eutsalt]
[prnu]
[syda: ilu]
[:sel]
[knnin]
[yle silla]
[srazu]
[vzdrogn_]
[ubijal]
[ujejaju]
[sank piterburk ]
[utslt]
[prnu]
[syda:elu]
[:sel]
[knnin]
[yle silla]
[zrazu]
[vzdrognu_]
[ubizal]
[ujezaju]
[sank piterburk ]
[utsalt]
[prnu]
[sydae:lu]
[:sel]
[knnin]
[yle silla]
[srazu]
[vzdrognul]
[ubizjal]
[ujezaju]
[sank piterburk ]
[utsalt]
[prnu]
[syda:elu]
[:sel]
[knnin]
[yle silla]
[srasu]
[vzdrognu_]
[ubizal]
[ujedaju]
[sank piterburk ]
25
English transcriptions
Group 1: subjects with no instrumental education
Where
were
you
/wer
wr
j
born?
brn/
101
/wer
wr
brn/
102
/we:r
wr
brn/
103
wr
wr
brn/
104
/wr
wr
brn/
105
/wr
wr
ju
brn/
peck of
pek
v
pickled
pikld
peppers
peprz/
101
pek
pikitn
piprz/
102
pek
pikld
peprz/
103
pek
piklt
piparz/
104
pek
pakl
peprz/
105
pek
pikl
peps/
Santa's
/sn.tz
Short
rt
Suit
sut
Shrunk
rk/
101
/sant
rt
sut
rk/
102
/snt
rt
sut
rk/
103
/snt
rt
sud
rk/
104
/snt
u:t
rk/
105
/snt
rt
sut
rk/
26
Where
/ wer
were
wr
you
j
burnt?
brnt /
101
/wer
wr
brn/
102
/wer
wr
brnt/
103
/wer
wr
brn- j brnt/
104
/wer
wr
brnt/
105
/wer
wr
brnt/
were
wr
you
j
born?
brn/
201
/wer
wr
bn/
202
/wer
wr
brn/
203
/wer
wr
brn/
204
/wer
wr
br/
205
/ wer
wr
brn/
peck of
pek
v
201
pek
piklappiprz
202
pek
pikld p...peprz/
203
pek
pikl pi...peprz/
204
pek
pikel piprz/
205
pek
pikld peprz/
27
pickled peppers
pikld peprz/
Santa's
/sntz
Short
rt
Suit
sut
Shrunk
rk/
201
/sntz
srt
sut
srk/
202
/sntz
rt
sut
rk/
203
/sntz
rt
sut
rk/
204
/sntzrt
sut
rk/
205
/ sntz
rt
sut
rk/
Where
/ wer
were
wr
you
j
burnt?
brnt /
201
/ wer
wr
brnt
202
/we
wr
brnt/
203
/w...wer
wr
brnt/
204
/wr
wr
brnt/
205
/wer
wr
brnt/
28