Growing toddler MLU through modeling and expansion
For Australian speech-language pathologists working in early childhood, mean length of utterance (MLU) remains one of the most practical measures of expressive language growth in toddlers. Between roughly 18 and 36 months, children move from single words to short phrases, then to grammatically structured sentences. Tracking MLU gives clinicians, parents, and educators a meaningful yardstick during this rapid window of growth, and the data often guides goal-setting on NDIS plans, paediatrician reports, and the learning outcomes documented under the National Quality Framework.
Two of the most consistently used adult-led strategies for increasing MLU are modeling and expansion. Both sit within naturalistic language-interaction approaches and are widely endorsed by Speech Pathology Australia for use in everyday routines. Modeling provides a slightly more advanced language sample for the child to absorb, while expansion takes the child's existing utterance and lengthens it into a more complete, grammatically correct version. Used together, they create the rich, predictable input toddlers need to extend their syntactic skills over time.
Why mean length of utterance matters in early language development
MLU counts the average number of morphemes a child produces across a language sample, offering a snapshot of grammatical complexity rather than vocabulary size alone. Clinicians typically pair MLU data with information about Brown's stages of language development to interpret where a toddler sits compared with same-age peers. In Australian practice, this information frequently supports referrals, informs paediatric reviews, and documents progress for funding bodies such as the NDIS, where measurable goals are central to plan reviews.
It is worth remembering that MLU is a useful but imperfect measure. Children with strong vocabulary and weaker grammar, bilingual toddlers, and those using augmentative communication may not fit neatly into expected ranges. Speech sound clarity, pragmatic skills, and gesture use all contribute to communication, which is why MLU sits alongside other markers such as vocabulary diversity and mean length of communication rather than replacing them.
How modeling builds a toddler's sentence blueprint
Modeling happens when an adult produces a slightly more sophisticated version of what the child just said or could say, without demanding a response. Self-talk and parallel talk are two classic forms. Self-talk describes the adult's own actions ("Mummy is pouring the water"); parallel talk narrates the child's actions ("You are pushing the truck along the floor"). Both flood the toddler with syntactically rich input during familiar routines such as nappy changes, meal prep, bath time, or the drive to kindy.
A clinician supporting a 24-month-old in a Hobart family day-care might kneel beside a water-play tray and say, "The blue boat is floating — look, it's floating fast." For a toddler who only said "boat," the sentence introduces multiple clauses and an adverb without ever asking the child to repeat. Over weeks, this consistent input tends to nudge MLU upward as the child starts echoing patterns rather than single nouns, and the technique fits neatly into the intentional teaching moments encouraged by Australia's National Quality Framework for early childhood education and care. The key is consistency over correction — children absorb more from hearing well-formed sentences dozens of times a day than from being asked to repeat after the adult.
Expansion as a bridge from words to sentences
Expansion takes what the toddler already says and adds the missing grammatical pieces. If a child points and says "doggy run," an adult might respond with, "Yes, the doggy is running really fast." The structure stays close enough to the child's intent to feel like confirmation, while lengthening the utterance by one or two morphemes. This technique is closely linked to research on recasting, where the adult re-shapes the child's phrase into a more mature form.
In Australian homes from Adelaide to Perth, expansion is especially useful during shared reading sessions that are common in early-childhood programs. A toddler pointing at a picture of a koala eating leaves and saying "koala eat" can be met with "The koala is eating the green leaves." Clinicians often coach educators to add just one or two new morphemes per expansion, avoiding input so long that it loses the conversational thread and risks overwhelming a tired toddler before rest time. Parents and educators often find that expansion feels easier than modeling because it follows the child's lead rather than directing it, which keeps the interaction playful and responsive.
Embedding modeling and expansion in Australian homes and centres
Strategies like these only shift MLU when they happen often and naturally. In family homes and regional towns, embedding modeling into daily routines — bath time, dinner prep, the drive to swimming lessons — tends to work better than carving out a separate "speech time." For educators in long-day-care services, the National Quality Framework encourages intentional teaching moments, which align well with modelling and expansion because they can be planned around small-group activities and shared meals.
A practical example: during a sandpit activity at a Brisbane early-learning centre, an educator models sentences like "I'm digging a deep hole" while a toddler watches. She then expands a child's "truck go" into "The truck is going up the hill." SLPs who coach teams often provide simple language-strip charts on the fridge or staffroom wall so everyone uses consistent models throughout the day. Over time, these small language shifts add up to measurable MLU growth, and parents quickly notice longer sentences at dinner and during play. Melbourne and Sydney kindergartens often embed the same approach during mat time and group reading, where every educator models the same expanded sentence pattern.
Tracking progress and knowing when to adjust
Tracking MLU over time gives families and clinicians a concrete way to see whether modeling and expansion are working. A baseline sample of 50 to 100 utterances, recorded during a familiar routine, can be transcribed and counted for morphemes. Re-sampling every two to three months provides a reasonable picture without becoming burdensome, and many Australian SLPs document this progress in NDIS plan reviews, paediatric reports, or kinder-transition meetings that often involve allied-health recommendations for school readiness.
If MLU is not budging despite consistent input, it is worth checking hearing, oral structure, and broader language comprehension. Some children benefit from a more targeted program, while others may need onward referral. Clinicians interested in building deeper skills around adult-led language input can access a structured reflection tool through speakingpd.
Quick prompts and signs to watch for
Prompts to weave into play:
- "You're stacking the red block on top of the blue one."
- "The water is splashing — splash, splash, splash!"
- "I see you driving the bus to the shop."
- "The teddy is sleeping under the blanket."
Signs a toddler may benefit from a closer look:
- Using fewer than 50 words at 24 months.
- Combining two words rarely by 30 months.
- Difficulty following simple directions during play.
- Family history of persistent language difficulties.
| Strategy | What the adult does | Typical effect on MLU | Best used when |
|---|---|---|---|
| Self-talk | Narrates own actions at a slightly higher level than the child | Adds new vocabulary and syntax without pressure | During routines like nappy changes, bath time, or meal prep |
| Parallel talk | Narrates the child's actions using longer sentences | Builds sentence length and grammatical complexity | During play, particularly with objects the child is focused on |
| Expansion | Lengthens the child's exact utterance into a more complete version | Adds morphemes and grammatical markers close to the child's intent | After the child speaks, especially during shared book or play |
| Recasting | Restructures the child's phrase into a question or different form | Highlights grammatical contrasts and question forms | When the toddler is engaged and the conversation is flowing |
Grab a pack of modeling prompt cards from the shop at Let's Talk Speech Therapy to keep on the fridge, and explore the blog for printable therapy materials, supervision tools, and Australian-context resources. Pair modelling and expansion with consistent everyday routines, and MLU gains tend to follow within a few months. For clinicians who support students across school settings, this piece on supporting patients with TBI explores how SLP roles extend beyond the early years.
On the Blog
Popular posts and resources from the Let's Talk Speech Therapy archive.
Top 10 Books that Encourage Language Development for Babies and Toddlers
A curated list of books that support early language development, with tips for SLPs and parents on how to use them during reading time.
Supervising a SLP Student Intern – Getting Started
Practical guidance for SLPs new to clinical supervision, covering preparation, expectations, and mentorship strategies for working with student interns.
How (and why) I Teach Phone Numbers in Speech Therapy
Inspired by a real incident, this post explains how to use backwards chaining and exit-ticket questions to teach personal safety information.
7 Tips for Making the Most of ASHA '16
Seven practical tips for navigating the ASHA convention — from session planning to networking — based on firsthand experience.
Pete the Cat Phoneme Segmentation Screener
A free printable phoneme segmentation screener featuring Pete the Cat, shared as a resource for SLPs working on phonological awareness. Published September 16, 2012.
Speaking & Professional Development
Information about professional development opportunities and speaking engagements for schools, clinics, and SLP groups.