MEASURING PHONOLOGICAL SKILLS IN ADOLESCENCE
Ann French
Introduction
The
focus
An
examination of phonological skills in adolescence might include a range of
measures, including:
·
Perception at both phonetic (acoustic)
and phonological levels;
·
Production at both phonetic
(articulatory) and phonological levels;
·
Phonological awareness,
including awareness of syllables, onsets and rimes, and phonemes;
·
Phonological working memory.
The rationale
Language
acquisition texts may suggest that phonological development is virtually
complete by the end of the primary school years, that is at around 10 to 11
years. Both Gleason (2005) and Owens (2005) for example imply that phonological development
in adolescence is limited to features such as the acquisition of adult-like
stress patterns (for example, the difference between ‘hot dog’ – a
canine that is hot - and ‘hot dog’
- a food consisting of a bread roll and a sausage) and the completion of
morphophonemic rules (the consonant
and/or vowel changes that occur when adding derivational affixes to
stems, for example adding –ion’ to ‘create’ /krieɪt/ to get ‘creation’ /krieɪʃən/.)
Recent
research however suggests that some aspects of phonology may still be
developing during at least early adolescence. These include the development of phonemic
perception (Hazan and Barrett (2002), articulatory agility (Walsh and Smith,
2002) and of metaphonological awareness (Wagner, Torgensen and Rashotte, 1999).
Additionally, vocabulary development continues throughout life, and a key
component of this is the acquisition of new phonological forms, both in input
and output. Learning a second language requires the acquisition not just of new
phonological forms, but often new phonetic forms as well.
Children
with impaired language development frequently have underdeveloped phonological
skills, which are thought to impact on the development of literacy, as well as
on vocabulary learning and other spoken language skills. Similarly, children
diagnosed with dyslexia or developmental reading impairment may have, or have
had, phonological deficits. In each secondary classroom (that is, from 11 years
to 16 or 18 years) there may well be one or two students with a history of identified
phonological deficits, whose difficulties may not have fully resolved. There
may also be other students whose phonological skills are less mature than their
peers, yet who have not been identified as such.
Nittrouer
and
The
productive phonological system is more or less complete for most children by
around 8 years (at least for English first-language children) so that
adolescents are generally intelligible to others. This apparent maturity may
mask more subtle immaturities.
The
role of phonological skills
As well
as intelligibility, phonological skills have a developmental role to play in
many components of the secondary school curriculum. These include:
·
Achieving the alphabetic phases
of reading and spelling, usually acquired in the early junior school years,
that is, at around ages
·
The use of analogy and
morphophonemic rules to read and spell unfamiliar words during the orthographic
phase of reading and spelling development. Once basic reading skills are
acquired the student moves from ‘learning to read’ to ‘reading to learn’ (Paul
1995) and begins to acquire the majority of new words through reading rather
than through listening to spoken language. Again the complexity of English
spelling creates a challenge for some learners, who have to cope with such irregularities as
‘ether’ (/iθə/)
v. ‘whether’ (/weðə/) v.
weather (/weðə/)
· New word learning. This is an important element of the secondary curriculum, with many highly specialised words and their variants that the student may not have encountered before. Some of these are phonologically complex, such as ‘taxonomic’ ‘magnitude’ and ‘chromatography’ – all taken from the Year 7 Science curriculum for 11 to 12 year olds (The Standards Site, 2006).
· Appreciating literary concepts such as rhyme and alliteration in poetry and prose. Rhyme is a much less clear-cut concept than that of the ‘rime’ of a single syllable. Rimes share a vowel nucleus and consonant coda, for example ‘coat’, ‘goat’, ‘moat’. Longer ‘rhyming’ words however may be matched over greater or lesser portions of their length, for example ‘magnesium’/’trapezium’ (close match over 3 out of 4 syllables), compared to ‘neutrality’/’psychiatry’ (close match over 1 out of 4 syllables). (All rhyme examples taken from Fergusson, 1985.)
· Appreciating puns and other phonological jokes in texts and in playground talk.
· Reflecting on accent differences and their role in communication.
· The ability to detect word boundaries and identify potentially unknown lexemes in spoken language, and to assist self-monitoring when reading. (For example, knowing when the word decoded cannot be a word of your language, thus prompting a review of the item.)
· The leaning of new media writing styles such as ‘pop’ spellings, for example <Krazy> (‘crazy’) and <Bitz> (‘bits’) and text messaging, for example <b4> (‘before’) and <tmz> (‘tomorrow’: involving a phonological change from ‘tomorrow’ to /təmɒz/ and then deletion of the vowels.) Students also need to be able to recall which spellings are appropriate to given contexts.
· Learning the phonetics, phonology, phonotactics, and spellings of foreign languages.
Additionally, there is the complex interaction
between what
FWM has two components: processing (the generation of lexical and grammatical linguistic representations from the input) and storage (the temporary retention of the processed material.) These share the same limited resources during complex language tasks, and there may need to be a trade-off when demand exceeds capacity. Both PWM and FWM efficiency have been linked to vocabulary learning and to sentence comprehension. The relationship between them is further subject to a complex interaction of linguistic units at different levels (phonological, semantic, and grammatical and textual) and in both LT and ST storage. For example, the ST storage of new phonological words is facilitated by the presence in long term memory of words with similar phonemic and morphemic structures, while retention of the phonological form of a sentence will be enhanced if the words in it are familiar.
Project aim
The aims of my research are:
In the longer term it may be possible to provide advice to facilitate phonological skills development in any students found to be less mature than their peers and less able to access the school curriculum.
Project Methods
Development of the task battery
In the absence of suitable published tests, the following tasks have been developed:
· A word knowledge test. This has been designed to include both phonological and semantic distractors. Since
existing vocabulary tests (such as the British
Picture Vocabulary Test, Dunn et al 1997, widely used in the
· Rhyme detection and Spoonerisms. These two tasks of phonological awareness each have two
parts, one with a relatively low FWM demand and one with a much higher demand. The aim is to identify whether any problems in performance are due to lack of phonological ability/knowledge or to the demands of FWM. Existing tools were felt to be inappropriate. For example, the Rhyme and Spoonerism subtests in the Phonological Assessment Battery (Frederickson et al, 1997) use predominantly single-syllable target words, most of which would typically acquired early in life. Correct performance on such tests, while indicating solid phonological skills with short familiar words, might give a misleading picture of a student’s response when faced with the longer and less familiar vocabulary of the secondary curriculum. The new rhyme and spoonerism tasks therefore were designed to use words of different syllable lengths (from 1 to 5 syllables) and of different predicted ages of acquisition (AoA) (based on a set of ratings collected by Gilhooly and Logie, 1980.)
· Real word repetition and nonword repetition. These tasks looks at phonological production, with
and without semantic knowledge to support production. Again words of varying syllable lengths and AoA ratings have been used, to measure the effects of these variables. The nonwords were derived from real words by changing the initial consonant of each syllable, but ensuring that none of the resulting syllables are existing English morphemes. This is a problem for existing tests (for example the Children’s Test of Nonword Repetition, Gathercole and Baddeley, 1996) as it changes the task from a purely phonological one to one where morpho-semantic knowledge supports phonology.
· Tongue Twisters. Items were selected from Wiltshire (1999). Wiltshire argues that such a task assesses
phonological rather than articulatory skill if participants are asked to repeat at a steady speed of around 100 wpm rather than ‘as fast a possible’.
During construction of the above tasks literacy demands were kept minimal or nonexistent. The aim was that phonological knowledge would not obscured by presence/absence of orthographic knowledge. However in practise this aim proved impossible to achieve. (See ‘Results’ below)
In addition to these new tasks, the test battery includes:
· Digit Recall (Pickering and Gathercole 2001), regarded as a test of PWM (particularly the articulatory loop.)
As the 9 digits used can be expected to be highly overlearned by 11 year olds, this task should demonstrate the functioning of PWM with minimal demands on phonological perception, knowledge or production.
· Digits Backwards Recall (Pickering and Gathercole 2001) Again a task with minimal phonological
perception, knowledge and production demands, but one thought to assess FWM since the student is required to repeat the digit string in reverse order.
· 3 tasks from The Test of Everyday Attention for Children (TEA-CH, Manly et al, 1999): Skysearch, Score
and Opposite Worlds. These were selected to assess the three types of attention described by Manly et al (selective/focussed attention, sustained attention, and attentional control/switching) but to have minimal verbal content. The aim was to assess the contribution to the phonological tasks of the Central Executive, the working memory component responsible for attention control and allocation. (Baddeley 1986.)
It was considered to be important to separate the effects of attention, PWM, FWM and phonological knowledge as far as possible, since weaknesses in these different components might lead to differing strategies for the classroom.
Participants
A large mainstream secondary school was selected to have a Year 7 student intake:
· Large enough to allow random selection of two successive cohorts of 50 students.
· From a broad socio-economic range.
· Who are predominantly monolingual users of English as a first language. Children growing up in bilingual homes may qualitatively different phonological development to monolingual children as a result of early exposure to two or more distinct phonological/semantic systems, and the metaphonological awareness that this may prompt. At this stage of the project the aim was to eliminate this additional factor, but at a later stage it would be important to include this student group.
56 students have been recruited to date; 28 boys and 28 girls. The first 11 students recruited were used to pilot the new phonological tasks developed. Adjustments were made before administering the tasks to the first main cohort of 45 students.
In order to carry out the correlational analyses, two further sets of data were collected for each student. The parent/carer of each student was asked to fill in a questionnaire addressing aspects of the student’s early years and family characteristics, and the school was asked to supply academic information.
Results
At the time of writing the first phase of data collection has almost been completed, but the results have not yet been fully analysed, and academic data are not yet available. However analysis of the pilot data, supported by informal observations during the rest of the data collection, has lead to the following observations.
1. There are marked differences in students’ responding styles, which impact on the scores achieved, but which are beyond the examiner’s control. Some proceed quite quickly through tasks, preferring to skip items and return to them if unsure; others proceed more slowly, musing aloud over the tasks. Some guess when unsure, others are reluctant. Some are methodical in approach, others haphazard. Some use sub-audible rehearsal of stimuli, others do not appear to. Some need more processing time and/or need more repetitions than others. Some review perceived errors spontaneously, others either do not notice them or see no need to review unless prompted. Some always ask for repetitions if unsure, others rarely do so, even when responses suggest that they are struggling. Some accept task instructions as given, others paraphrase them and check anything they find unclear before starting the items. Some find the tasks more tiring than others, and some take longer than others to complete. Some find task instructions hard to grasp and need slow presentation, others barely needed practice items, seeming to understand the task almost immediately. Such variations are not unexpected in test situations, and some of the students assessed may have special educational needs. However all are placed in classes of 30+ students, and such differences in processing speed and skill may mean that some students miss information while still focussed on earlier material they found difficult, and others miss opportunities to contribute by needing longer than average to do so, or by being reluctant to respond unless sure of the answer.
2. There are differences in response to the word length/familiarity factors. During the rhyme tasks for example some feel that increased word length is the factor that increases difficulty, and that familiarity is not relevant as they are listening to the sound rather then the meaning. Others feel that unfamiliar words make the task harder, especially in conditions of increased memory. One student stated that familiar words made it harder as the meanings distracted her from the sounds. Others are not able to reflect on what makes some items harder than others.
3. The motor programming demands of the rhyme, spoonerism and nonword tasks produced significant differences between students in the pilot group, and are remarked by many. Students comment on the strangeness of producing nonwords, find it makes them giggle, or find ‘creating’ nonwords in difficult, although repeating them was usually easier. One or two said they sometimes found it hard to pronounce even familiar words. Errors in these tasks usually occur on longer words, and often mirror early phonological processes such as reduplication, assimilation, and coalescence.
4. Working memory demands impact on rhyme and spoonerism performance, as expected. Thus some students demonstrate phonological knowledge which then becomes less accessible as memory demands increase.
5. During the rhyme and spoonerism tasks it was apparent that all or nearly all students were using word spellings to underpin their responses. This was sometimes apparent in responses, for example on a spoonerism item requiring the movement of an initial /ʤ/ phoneme, some responses utilised instead the phoneme /g/, which does not appear in the stimulus words but in English shares a grapheme with /ʤ/. More often it was apparent in students’ comments: on spoonerisms for example, despite being instructed to take off and add sounds, they talked of taking off/adding letters, or even words. In the rhyme task they commented on trying to decide how far 2 words’ spellings had to match for them to be regarded as rhyming.
Challenges
Methodological difficulties that have arisen include:
· Lexical selection. Tests words were chosen from the Gilhooly and Logie corpus on the basis of word
familiarity, indexed by word frequency or age of acquisition; word length; and phonological similarity. However it was felt important to include verb target stimuli as well as nouns, since these are important to comprehension and are known to be problematic for some children (Conti-Ramsden and Jones 1997). The Gilhooly and Logie corpus does not provide separate data for noun/verb uses of the same lexical item (for example, ‘despair’ could be either) so word frequency data had to be collected from another source to select verbs. Additionally, the AoA ratings were in some cases inaccurate. This may be because word usage has changed since the corpus was published, for example ‘strop’ is listed as being acquired from 13 years onwards, presumably in its earlier meaning of ‘to sharpen a knife’, but is now known by much younger children in its meaning ‘to show temper’. Other words predicted to be acquired at 13 years+ are know to almost all the students assessed as they occur in the Year 7 curriculum, for example ‘magnesium’ and ‘carbohydrate’.
· The difficulty of creating true phonological distractors for later acquired words. ‘Phonological neighbours’
(words that share the majority of their phonemic structure) are common amongst the earliest acquired/most frequent words of the language, which tend to be short, for example ‘pig/ big/dig’, and ‘carrot/parrot’ but are hard to find amongst later-acquired and phonemically more complex words such as ‘philanthropy’ and ‘insincere’. Thus it is difficult to create a lexical comprehension task which effectively incorporates both semantic and phonemic distractors.
References
Baddeley, A. (1986). Working Memory.
Conti-Ramsden, G. and Jones, M. (1997). Verb use in specific language impairment. Journal of Speech, Language and Hearing Research, 40, 298-313.
Dunn,
L.M., Dunn, L.M., Whetton, C. and Burley, J. (1997). The British Picture Vocabulary Scale. 2nd Ed. Windsor:
nferNelson.
Fergusson,
R. (1985). The Penguin Rhyming Dictionary.
Frederickson,
N., Frith, U. and Reason, R. (1997). Phonological
Assessment
Gathercole,
S.E. and Baddeley, A.D. (1996). The
Children’s Test of Nonword Repetition.
Gilhooly,
K.J. and Logie, R.H. (1980). Age of acquisition, imagery, concreteness,
familiarity and ambiguity measures fro 1944 words. Behaviour Research Methods and Instrumentation, 12, 395-427.
Gleason,
J.B. (2005). The Development of Language.
6th Ed. London: Allyn and Bacon
Hazan, V.
and Barrett, S. (2002). The development of phonemic categorisation in children
aged 6-12. Journal of Phonetics, 28, 377-396.
Just, M.A. and Carpenter, P.A. (1992). A capacity theory of comprehension: Individual differences in working memory. Psychological Review, 99,1, 122-149.
Locke, A.
and Ginsborg, J. (2003). Spoken language in the early years: the cognitive and
linguistic development of three- to five-year-old children from
socio-economically deprived backgrounds. Educational and Child Psychology,
20, 4, 68-79.
Manly, T., Robertson, H., Anderson, V. and
Nimmo-Smith,
Nittrouer,
S.and Burton, L.T. (2005). The role of early phonological experience in the
development of speech perception and phonological processing abilities:
Evidence from 5-year olds with histories of otitis media with effusion and low
socio-economic status. Journal of Communication Disorders, 38, 29-63.
Owens, R.E.
Jr. (2005). Language Development: An Introduction. 6th Ed. London: Allyn and Bacon.
Paul, R. (1995). Language
Disorders from Infancy through Adolescence: Assessment and Intervention.
Pickering, S. and Gathercole, S. (2001). Working Memory Test
The
Standards Site. http:// www.standards.dfes.gov.uk/keystage3/respub/scienceframework/appendix3/year7.
Accessed
Wagner,
R.K., Torgensen, J.K. and
Walsh, B.
and Smith,A. (2002). Articulatory movement in adolescents: evidence for
protracted development of speech motor control processes. Journal of Speech,
Language and Hearing Research, 45,
1119-1133.
Wiltshire,
C.E. The “tongue twister” paradigm as a technique for studying phonological
encoding. Language and Speech, 42,
1, 57-82.