There are many reasons that understanding a child’s language development, when the child has been exposed to one or more languages other than English, can be difficult - the language barrier itself is only the start!
Building the cultural and linguistic diversity of the speech pathology workforce is essential - both to support families with similar backgrounds, as well as to help build the field’s understanding of cultural and linguistic diversity more broadly. But in a culturally diverse city, there will always be reasons why finding a cultural and linguistic ‘match’ between family and clinician is not possible, or not realistic, in every case.
Standardised assessments
When standardised tests are available for a given language, these may be able to provide one piece of evidence in the assessment process. But using test results to help determine whether a child experiences a speech or language disorder relies on a series of assumptions.
A normed task aims to rank children’s performance reliably (i.e., consistently). For this to differentiate between children experiencing difficulty in learning language from those developing typically, it makes the assumption that all children have equal experience with the language tested.
I.e., assuming that all children have the same opportunities to learn particular language skills, children with a speech or language disorder will have learned less than other children.
However, a child exposed to more than one language won’t have the same level of experience with the language of testing as a monolingual child. So if the child attains a low score on a normed task (i.e., knows less of the language than other children), it is not clear:
- whether this is because the child hasn’t had enough experience to learn more of the language,
- or whether there is some kind of cognitive, socio-cognitive, sensory, or physical reason that is causing the child to have difficulty learning the language.
This is true for both the language of education (e.g., English), and the home language. For example, a 4-year-old child growing up in a Mandarin-speaking family in Canberra (Australia) won’t have the same amount or type of experience with Mandarin as a child growing up in a Mandarin-speaking family in northern China, even if Mandarin is the child’s primary or dominant language. This effect becomes more significant with age, as children are exposed to different types of language, for example through reading and academic experiences, versus social experiences and cultural events.
For further discussion about assumptions and uses of standardised assessment tasks, see Denman et al. (2023) (open access), as well as this podcast interview with Elena Plante.
Other types of evidence about language skills
My approach here is somewhat similar to Castilla-Earls et al. (2020)’s converging evidence approach (full text available from PubMed).However, while these authors’ tutorial is relevant to clinicians working with preschool- and early school-aged English-Spanish bilinguals, families I work with use a much broader range of languages. I have attempted to include other useful sources of evidence that are relevant to the languages and populations I work with.
I am hoping to flesh out these strategies and tasks more fully when I get the chance (ha!).
- Case history will always be an essential source of information about a child’s language experience and abilities. Without case history, other pieces of information we gather through our assessments (e.g. test scores) are just data points, and could be consistent with many different interpretations. For more information about using case history to identify language disorders in multilingual children, see Paradis et al. (2010), Boerma & Blom (2017), and Li’el et al. (2019). These studies use the Alberta Language and Development Questionnaire (ALDeQ), available from Paradis (2024)’s website, and derived measures.
- Communicative development inventories (CDIs) are parent report tools that are useful for capturing the early vocabulary skills of toddlers between around 12 and 30 months old. At the time of writing (January 2025), authorised adaptations have been developed for around 100 languages, with details listed on the MacArthur-Bates CDI website by CDI Advisory Board (2015). Some are available to download from the author or institution’s website, some for purchase, and some are not currently publicly available. A number of scoring methods have been tested for identifying language difficulties in young multilingual children, including conceptual scoring (Bedore et al., 2005), and total vocabulary (Core et al., 2013).
- Assessing speech production can be challenging, but is very possible, when the child has been exposed to languages that the clinician does not speak. McLeod et al. (2017) (open access) have an excellent tutorial, and Multilingual Children’s Speech by McLeod (2024) has a database of many helpful resources, including speech acquisition studies for over 250 languages, and speech production assessments for over 50 languages. I have described a few more resources on this page.
- Nonword repetition tasks require a child to repeat an unfamiliar sequence of sounds that could, plausibly, form a word in the language. Nonword repetition is related to word learning, but also reflects how much a child knows about the phonological system of the language. Nonword repetition tasks have been developed and used to help identify language disorders in monolingual and multilingual speakers of more than 15 languages (Ortiz, 2021; Schwob et al., 2021). I have written more about nonword repetition tasks on this page.
- Sentence repetition tasks involve a child listening to a spoken sentence, and then repeating that sentence immediately. Sentence repetition is related to a child’s knowledge of the language, e.g., Polišenská et al. (2015), so they can help identify children who know less of the language than other children with the same experience of the language. Sentence repetition tasks have been developed and used to help identify language disorders in monolingual speakers of more than 19 languages (Ward et al., 2024), and have (less commonly) been used with multilingual children speaking a range of combinations of languages (Rujas et al., 2021).
- Dynamic assessment is a term that covers a wide range of tasks that aim to assess language learning ability directly. In clinical contexts, other assessment techniques aim to infer language learning ability by measuring how much language a child has learned in comparison to others, given the assumption that children have had the same opportunities to learn the language tested. Dynamic assessment tasks use variations on test-teach-retest procedures. Most use word learning tasks or narrative macrostructure learning tasks (Bamford et al., 2022). Dynamic assessment tasks have been used to help identify language disorder in multilingual children learning English or German (Hunt et al., 2022; Orellana et al., 2019). It does raise a significant logical problem when a child is identified as having a language disorder due to difficulty with a particular learning task, and then further use of the same task is recommended or marketed as being the most appropriate option for support… 🙄
- Language sampling is used within speech pathology contexts to describe the collection and transcription of natural(-istic) or elicited samples of extended discourse. It usually refers to the use of quantitative measures computed over the sample (such as ‘mean length of utterance’) or scores derived from predefined scoring rubrics. In the context of identifying language disorder, these measures have been used in with children learning one of around 7 different languages, mostly within story retell tasks (Ortiz et al., 2024). My main concern with ‘language sampling’ is that it refers to an approach to data collection, rather than any particular type of measure, and suitability for use in identifying language disorders depends more on the elicitation context and the measure(s) used, rather than the general approach to data collection.