Articulation Therapy: Hierarchy, Minimal Pairs & Dose

Articulation therapy has two competing logics, and knowing which one you are using is the difference between a plan that generalizes and a plan that stalls. Traditional motor approaches treat a sound as a movement problem and drill it up a hierarchy. Phonological approaches treat the same error as a rule problem in the child's sound system and use contrast to force a repair. Most Praxis items — and most caseloads — hinge on choosing correctly between them.

Motor versus linguistic: how to decide

• Use a motor approach when errors are few, consistent, and distortion-like: a lateral /s/, a derhotic /r/, an interdental lisp in an otherwise intelligible child. The child knows the contrast; the articulators are not producing it.
• Use a phonological approach when errors are patterned across sound classes and intelligibility is reduced: final consonant deletion, fronting, cluster reduction, stopping. Drilling one sound at a time here is slow, because the child is applying a rule, not fumbling one gesture.

A quick screen: transcribe a connected speech sample, count error patterns, and ask whether they suppress by age. If several phonological processes persist past their expected suppression window, treat the system.

The traditional articulation hierarchy (Van Riper)

Van Riper's sequence remains the backbone of motor-based treatment and is the hierarchy most often referenced on the exam:

1. Auditory discrimination / ear training: identification, isolation, stimulation, and discrimination of the target versus the error. Skipped often in practice, but it is the first step in the classic model.
2. Isolation: the sound alone, elicited by imitation, phonetic placement, shaping from a sound the child already has, or successive approximation.
3. Syllables: CV, VC, CVC, then reduplicated and varied syllables.
4. Words: initial, then final, then medial position; sequence by coarticulatory ease, not by alphabetical word lists.
5. Phrases and sentences: carrier phrases first, then loaded sentences.
6. Conversation: structured, then spontaneous.
7. Generalization and maintenance: across settings, listeners, and time, with self-monitoring as the transfer mechanism.

Advance a level when accuracy is stable — 80 percent across two consecutive sessions is a common clinical criterion — and drop back one level, not to the bottom, when accuracy collapses.

Elicitation techniques worth knowing

• Phonetic placement: direct instruction about where the articulators go, with mirrors, tongue depressors, or diagrams.
• Shaping / successive approximation: building /r/ from a prolonged /ɝ/, or /s/ from a sustained /t/ release, or /ʃ/ from a retracted /s/.
• Contextual utilization: finding a phonetic environment where the target is already correct (a facilitating context) and expanding outward from it.
• Biofeedback: ultrasound and electropalatography have the strongest emerging evidence for stubborn residual /r/ and lateralized sibilants.

Minimal pairs and the contrast family

Contrast therapy works on meaning, not placement. The child is confronted with the fact that the error erases a distinction a listener needs, and the sound system reorganizes.

• Minimal pairs: two words differing by one feature — key / tea for fronting, bow / boat for final consonant deletion. Best for a single, consistent process in a mildly to moderately unintelligible child.
• Maximal oppositions: pairs differing by many features and maximally distinct in place, manner, and voicing (mall / call). Indicated when the child has very few contrasts and needs broad system change.
• Multiple oppositions: one error sound contrasted with several targets at once (door / core / shore / store). Designed for phoneme collapse, where a child substitutes one sound for a whole set.
• Empty set / treatment of the empty set: two unknown sounds paired against each other for maximal system-wide learning.

The complexity principle underlies the last three: targeting later-developing, more marked, non-stimulable sounds (clusters, affricates, liquids) tends to produce more untreated generalization than starting with easy, stimulable, early sounds. It is counterintuitive — and it is a favorite exam distractor.

Dose, practice structure, and motor learning

Motor-based articulation work follows principles of motor learning. Early acquisition favors blocked practice, high-frequency knowledge-of-performance feedback ("your tongue was too far forward"), and constant conditions. Later retention and transfer favor random practice, reduced and delayed knowledge-of-results feedback ("that one was right"), and variable conditions. Clinicians who keep giving dense feedback into the transfer phase create children who are accurate only in the therapy room.

Trial density matters more than session length. A productive articulation session runs 75 to 100+ target productions; a session with twenty trials and a board game is a session that will not generalize.

Dismissal and generalization criteria

Write dismissal into the plan from day one: correct production at the conversational level in at least two settings with at least two listeners, sustained over a defined maintenance window, plus resolution of the educational or functional impact documented in the IEP or plan of care. Stimulability probes on untreated sounds tell you whether the system is generalizing or whether you are teaching sounds one at a time forever.

What the exam asks

Expect vignettes that give a speech sample and ask which approach to select, which target to choose first, or what the next step in the hierarchy is. Keys usually reward: choosing a phonological approach for patterned errors, choosing complex targets for broad change, moving down one level rather than starting over after a breakdown, and fading feedback as the learner moves toward transfer.