Transcription of Children’s Speech

Transcription is a tool which is integral to our practice as speech-language pathologists (SLPs) (also known as speech and language therapists). Being able to hear what individuals say is useful, but it is only by documenting it that we can determine whether change has occurred over time or in response to intervention. It is through close observation of the transcription that we are able to identify patterns of ability and error and to use this information to plan clinical management. Indeed, its importance in our work and practice and the need to maintain skills was the driver behind the development of the Good Practice Guidelines for Transcription of Children’s Speech Samples in Clinical Practice and Research by the UK and Ireland’s Child Speech Disorder Research Network [1].

Transcription enables us to break down language-specific constraints and focus on speech using a common, internationally intelligible alphabet. From here commonalities and diversity in speech production both within and across languages can be identified and investigated, supporting our understanding of speech-language pathology practice in an increasingly globalised world.

There are a variety of different types of transcription, but the focus in this special issue has been on transcription of children’s speech and therefore on the use of phonetic transcription in particular. This special issue opens with a review by Stemberger and Bernhardt [2] on the history of transcription of children’s speech. They bring their experience, knowledge and skill, particularly in the field of cross-linguistic speech development and disorder, to describe the landscape for transcription of children’s speech today.

The emphasis on cross-linguistic transcription is discussed further in the reports by Masso et al. [3] and Pascoe et al. [4]. All too often, papers on transcription of children’s speech have used samples from English-speaking populations as their source. Masso et al. [3] bring a refreshing perspective in their investigation of Vietnamese-speaking adults and children and, in particular, the ability of English-speaking SLPs to transcribe consonants in words which are unknown to them. Understanding the level of skill and proficiency that exists in transcribing speech from languages other than the SLPs’ home language(s) helps determine the degree of error – and, conversely, the degree of ability – in management of children who speak other languages.

Multilingualism is the norm, and as a consequence, our ability to transcribe in languages other than our home language(s) is becoming imperative. Pascoe et al. [4] take this further as they consider not one, but four languages in their investigations of learning transcription skills and teaching them to South African speech-language pathology students. Their report provides useful information which can be adopted by other training institutions across the world.

As well as by the need for proficiency in multilingual transcription, the field is being challenged by new tools and approaches, none more so than the technique presented in the paper by Cleland et al. [5], who report on the use of ultrasound. They found that ultrasound-aided transcription and traditional transcription from an audio recording identified a similar number of errors. The number of errors identified in both the ultrasound and audio recording conditions was higher than the number of errors identified during the live recording, suggesting that additional technology is important for accurate diagnosis and identification of errors.

Another approach to determining reliability of transcription is proposed by Seifert et al. [6], who describe an alternative to the oft-reported technique of point-to-point transcription through acknowledging that differences between one transcription and another vary in their significance. Whilst some differences represent a difference in perception between the two listeners, others reflect that there is sometimes more than one way of representing a production using the standard features of the IPA. They propose use of the term “near functional equivalence” to reflect these differences which should be considered the same in measures of reliability.

Practice in transcription varies both across and within nations. The study by Nelson et al. [7] considers how SLPs in Australia use transcription through a survey of practice. They report on the range of tools which are available and used by the profession and what notation systems are in common use. The use of detailed or narrow transcription, whilst less commonly used, is nevertheless vital for those practitioners working with children with childhood apraxia of speech (CAS) or cleft palate. Their findings support the importance of continued professional development for SLPs to maintain these skills – particularly given their findings that few respondents had attended continued professional development in this area since qualification.

Two populations of children who may require use of detailed or narrow transcription, as outlined by Nelson et al. [7], are children born with cleft palate and those with CAS. Two papers in this special issue have considered these specific populations with regard to practices for transcription. Sell and Sweeney [8] report on the use of percentage of consonants correct in an analysis of outcomes following intervention for children born with cleft palate. They highlight the importance and need for additional training for those working with children with potentially complex speech problems, but also that with training, skill in transcription can be improved, leading to greater reliability across professionals.

Barrett et al. [9] report on findings from children diagnosed with CAS. Their report on the development of their Connected Speech Transcription Protocol (Co-ST-P) has the potential for significant impact on enabling SLPs to collect, transcribe and analyse a speech sample which is clinically useful but also more closely represents a child’s naturalistic speech pattern than is observed in single-word samples.

In summary, the work reported in this special issue of Folia Phoniatrica et Logopaedica, supported by the Child Speech Committee of the International Association of Logopedics and Phoniatrics (IALP), shows how transcription has changed over the last few decades and brings readers up to date with the state of the art. Time, training and, therefore, money remain important in facilitating transcription in the clinical setting, but its importance in clinical management is undiminished.