Helen Grech

Institute of Health Care

University of Malta

Abstract

This paper discusses recent findings following an exploratory study of the phonological development of twenty-one normally developing Maltese speaking children. The children were recorded in their natural settings at four different stages between ages 2;0 and 3;6. The data collected was transcribed narrowly and analysed. Each child's phonetic/phonological inventory was identified together with the various developmental phonological processes that were applied throughout the period of study. A developmental profile was collated for the group, indicating trends of stages of phonological development. This profile was compared cross-linguistically. The data fits in with current theories highlighting universal phonological acquisition particularly in the early years. As predicted some language-specific behaviour was also observed. The data is expected to be useful to local speech therapists and allied professionals as an indication of the 'norm'. This information is needed for comparison of children's speech patterns, identification of the characteristics of speech disability and for planning intervention.

Introduction

To date theories explaining phonological acquisition are still inconclusive or incomplete. The findings of cross-linguistic studies have led researchers to an awareness that questions cannot be answered from one single language. Only cross-linguistic data can reveal which aspects of acquisition are universal and which are language-specific. This study is meant to contribute to the data base which already exists in search of the theory of phonological acquisition. It is the first major contribution to develop 'norms' for the Maltese child population regarding the development of phonology. Studies involving pronunciation systems used by normal speakers are a pre-requisite for identifying disordered speech. Data and conclusions from this project are meant to become a help to clinicians. The study deals with:

• describing aspects of Maltese child phonetics and phonology;
• identifying trends of commonalities (construction of developmental profile);
• comparing data cross-linguistically;
• contributing to the understanding of child phonology;
• constructing conventions for transcription of speech of children exposed to a bilingual environment.

Maltese is a derivative of Arabic, spoken by approximately one third of a million people living on the Maltese Islands. The islands are situated in the Mediterranean, 100 km from Sicily and 290 km from North Africa. Maltese shares its status with English as the official language. Borg and Azzopardi-Alexander (1997) report that most of the Maltese has been standardised, though there are still some areas such as intonation and syntax which need to be researched further. The Maltese grammar is Semitic with influences of other languages amalgamated with it. The Maltese lexicon is influenced by English and Italian giving rise to a mixed vocabulary.

Table 1 below highlights the 22 Maltese consonant phonemes as derived by Azzopardi (1981), later updated in Borg and Azzopardi-Alexander.

	bilabial	labiodental	alveolar	postalveolar	palatal	velar	glottal
plosive	p        b		t          d			k     g	?
nasal	m		n
fricative		f          v	s          z				h
affricate			ts	t      d3
approximant/ tap/trill	w		r		j
lateral approximant			l

Method

The 21 normally developing Maltese speaking children were audio- and video-recorded at home, at four periodical intervals starting from 2;0 years of age and finishing off when they were 3;6 years old. The age at onset of recording was selected in view of the existing indications that segmental organisation could be possibly approaching by 1;4-2;0 years of age (Oller and Steffens, 1994). By age 3;6 it was expected that most of the phonemic development would have developed.

The data were then transcribed and organised on data sheets adapted from Grunwell's (1985) Phonological Analysis of Child Speech protocol. Following linguistic analysis the phonetic, phonemic and phonotactic components of the phonology of the Maltese children were drawn out. Elbert (1992) suggests that for children to be able to use the phonological system of a language for communication optimally, they need to be able to master all the above mentioned components. In addition some processes observed among the subjects' repertoires were described using nonlinear phonology frameworks, such as feature geometry. This was done as some authors (e.g., McCarthy, 1979) argue for its relevance in their attempt at explaining child phonology, particularly to languages such as Maltese.

Results

The developmental profile of the children's phonetics and phonology between the ages of 2;0 and 3;6 is given below. This profile is meant to have a triple evaluative role; it evaluates the emergence of the childís system of phones and contrastive phones, it evaluates mastery of phonemes and it evaluates the sequence of disappearance of children's processes. This constitutes the central core of the clinical implications of this project.

AGE	NASAL	PLOSIVE	FRICATIVE	AFFRICATE	APPROXIMANT
2;0	m               n	p            b t             d k                ?	h		w                   l                     j
2;5	m              n	p            b t             d k               ?	f s                     h		w                    l                    r                    j
3;0	m              n	p            b t             d  k            g               ?	f                 v s                       h	t	w                    l                    r                    j
3;6	m               n	p             b t              d k             g                ?	f                 v s                z                   h	t                 d3	w                    l                    r                     j

AGE	NASAL	PLOSIVE	FRICATIVE	AFFRICATE	APPROXIMANT
2;0	m               n	p              b t               d                ?			l
2;5	m              n	p              b t               d k              g                 ?	f                 v s                   h		w                       l                       r                       j
3;0	m               n	p              b t               d k              g                 ?	f                 v s                z                   h	t	l                        r                       j
3;6	m               n	p               b t                d k                g                   ?	f                 v s                z                  h	ts t                 d3	w                       l                       r                       j

/w/ is not represented as a phoneme at age 3;0 but is observed at age 2;5. This is because some children did not have the opportunity for its occurrence and not because it had disappeared as a phoneme since the earlier recording stage. Although this phoneme inventory seems to follow a progressive hierarchy, yet some discrepancy may occur because of not having enough opportunities for realisation.
 
 

/f/, // and /w/ are not represented as a phoneme at age 3;0 but are observed at age 2;5. This is because some children did not have the opportunity for its occurrence and not because it had disappeared as a phoneme since the earlier recording stage. Although this phoneme inventory seems to follow a progressive hierarchy, yet some discrepancy may occur because of not having enough opportunities for realisation.

Phonemes which still show up with substitution at age 3;6 include /n, d, s, z, , l, r, ts, t, d3/.

/f, j/ are not highlighted at 2;5 though they seem to have been mastered at age 2;0.
/m, p/ are not highlighted at 3;0 though they seem to have been mastered at age 2;5.
/d/ is not highlighted at 3;6 though it seems to have been mastered at age 3;0.
This 'apparent regression' is discussed in Grech (1998).

Discussion

Comparison of findings from this study with those from other languages was carried out, as it has various implications, particularly for intervention with the disordered population. Findings indicate:

• universal trends in the Maltese subjects repertoires, for the development and sequence of phones (e.g., nasals and plosives of the adults were used commonly from very early on). Apparent lack of stability of voicing was observed and this is also universal. However, this could not be solely developmental as it may not be easy to distinguish voicing in some contexts. The phonemic inventory expanded progressively in Maltese subjects; variations diminished progressively, as expected universally. Some developmental processes applied by the Maltese subjects, such as fronting, stopping, weak syllable deletion, cluster reduction and gliding are universally common.
• The influence of the specific language was noticeable for phone and phoneme establishment. Although the phonetic and phonemic sequence of development tallied with universal hierarchical models, this was only so, for types of sound classes; within class distinctions associated with place and voice in particular, did not correspond on a one-to-one basis. While commonalities exist in developmental processes, others including gemination, compensatory vowel lengthening, devoicing and lateralisation of /r/ seem to be more accountable by specific adult language phenomena such as Maltese phonological processes/rules. Besides, some processes may be common to Maltese and non-Maltese children in type but not in frequency of occurrence.

Implications of findings for assessment of Maltese child phonology

• The results obtained give clear indications of stages of phonological development which may be applied with a high degree of confidence for assessment and intervention purposes in speech therapy clinics as well as in allied professional settings. The data should lead to the identification of three types of possible developmental abnormalities, that is, delayed, uneven and deviant development. However, this may not be as clear cut as other researchers infer. The Maltese findings (which reflect a complex language structure) have demonstrated that:
• the clinician may find it difficult to decide whether a mismatch is the result of a phonological immaturity or a morpho-syntactic one;
• classification of language disorders by linguistic analysis may still leave some unresolvable possibilities;
• developmental profiles concerning one language could be misleading when applied to different languages.

Implications for intervention

Specific implications for intervention can be drawn out from the Maltese findings. These include:

1. Affricates should not be tackled prior to respective fricatives/plosives constituting the affricate, in articulation therapy.
1. If a clinician is treating Maltese speaking children, she or he should follow profiles specific to that population when devising treatment strategies.
1. Complex structures may give an apparent view of regression of mastery of phonemes. Perhaps if clinicians consider the developmental hierarchy of phonotactics in prioritising strategies for intervention, it may help intelligibility to be achieved earlier.
1. It is now a well established fact that developmental verbal dyspraxia may be characterised by inconsistent patterns. The complexity of Maltese phonotactics and the numerous possible combinations of consonantal clusters makes this statement very difficult to apply in the Maltese context. One would expect inconsistent patterns even in a 'normal' Maltese child, especially if processes are studied from a limited speech sample. Clinicians need to base analysis on a large enough transcript that incorporates as many possible phonotactic possibilities as possible.
1. It has been found that contrastive clusters take longer to be mastered even though they may surface early. Clinicians need to be careful not to stop intervening too early when full mastery may not yet be complete.
1. Intervention for phonological impairment needs to be planned in parallel with the development of other sub-components of the language.

Azzopardi M. The Phonetics of Maltese: some areas relevant to the deaf. Unpublished Ph.D. thesis. University of Edinburgh. 1981.

Borg A.J. and Azzopardi-Alexander M. (1997) Maltese. London & New York. Routledge.

Elbert M. (1992) Clinical forum: Phonological assessment and treatment consideration of error types: A response to Fey. Language, Speech and Hearing Services in Schools 23: 241-246.

Grech H. Phonological Development of Maltese Speaking Children. Unpublished Ph.D. thesis. University of Manchester. 1998.

Grunwell P. (1985) Phonological Assessment of Child Speech. NFER: Nelson.

McCarthy J. (1979) Formal Problems in Semitic Phonology and Morphology. PhD Thesis. Cambridge Massachusetts. MIT. Published by Garland Press, New York, 1985.

Oller D. K. and Steffens M.L. (1994) "Syllables and segments in infant vocalizations and young child speech" in Yavas Oller(ed.) First and Second Language Phonology. San Diego, California. Singular Pub. Group Inc.