Teasing out morphology vs. speech skills in preschoolers

Researchers tend to talk about speech OR language impairment, when in reality, we know that it’s rarely so simple. Two studies this month looked at morphology skills in preschoolers with speech sound disorders, and ways to untangle the effects of each.

14.png

Murray et al. looked at assessment data from studies of preschoolers with childhood apraxia of speech (CAS) who all were given the CELF-P2. They found that children with CAS had lower expressive than receptive language skills overall. Morphology was particularly difficult for children with CAS, and errors were inconsistent across the same morphemes. While some morpheme errors were speech-based, others were language-based. Unfortunately, assessing morphology skills independently of speech skills is difficult, especially in connected speech.

The takeaway is that morphological skills should be part of our assessment process for children with CAS, but we need to consider whether children’s speech skills are causing the morpheme errors we see.

So, how can we do that? Howland et al. have some suggestions. They looked at grammatical morpheme production in preschoolers with phonological impairment (defined as speech sound disorder with pattern-based errors). They found that, for these kids:

  • Past tense –ed was harder than all other morphemes. Morphemes like third person singular -s were harder than morphemes like plural or possessive -s.  

  • Morphemes in clusters (e.g., wants) were harder than morphemes in singletons (e.g., sees) or syllables (e.g., washes).

  • Children who could produce final clusters in single-morpheme words (e.g., waste) were more likely to use morphemes correctly overall.

So basically, children’s speech skills affect their ability to use morphemes. We can try to see whether the issue is caused by phonology or morphology by sampling sounds and clusters in single-morpheme vs. multi-morpheme words. Standardized tests don’t necessarily do this well, so it’s worth our time to do more probe testing to tease it out. What could this look like? Check out the example word pairs below. We’re asking, is it clusters that this child can’t do? Or morphemes?

2.png