Phonological processes, also called phonological patterns, are predictable ways young children simplify adult words: tat for cat, top for stop, wabbit for rabbit. Most fade between age 2½ and 4. In a study of 145 children aged 2½ to 8, process use showed "a marked decline" between 2½ and 4 and was infrequent after 4 (Roberts, Burchinal & Footo, 1990). The last to go are stopping of "th", at about 5, and gliding, at about 5 to 7.
The chart below lists each common process with an example and the age it typically stops in two standard sources: ASHA's table of selected phonological patterns and Caroline Bowen's table, which is based on Grunwell (1987).
Phonological processes chart by age
Process | What happens | Example | Gone by (ASHA) | Gone by (Bowen) |
|---|---|---|---|---|
Prevocalic voicing | A voiceless sound before a vowel becomes voiced | big for pig | Not listed | 3;0 |
Final devoicing | A voiced sound at the end of a word becomes voiceless | pick for pig | Not listed | 3;0 |
The last consonant is dropped | bu for bus | 3 | 3;3 | |
Assimilation (consonant harmony) | One sound changes to match another in the word | kack for tack, munny for funny | 3 | 3;9 |
/k/ and /g/ become /t/ and /d/ | tat for cat | 4 | 3;6 | |
Palatal fronting | "sh" becomes /s/ | sip for ship | Not listed | 3;9 |
Weak syllable deletion | An unstressed syllable is dropped | nana for banana | 4 | 4;0 |
A consonant cluster becomes one consonant | top for stop | 4 without /s/, 5 with /s/ | 4;0 | |
Deaffrication | "ch" and "j" become fricatives | ship for chip | 4 | Not listed |
Stopping of /f/, /s/ | A fricative becomes a stop | pun for fun, tee for see | 3 | 3;0 |
Stopping of /v/, /z/ | A fricative becomes a stop | bine for vine, doo for zoo | 4 | 3;6 |
Stopping of "sh", "ch", "j" | A fricative or affricate becomes a stop | top for shop, dump for jump | 5 | 4;6 |
Stopping of "th" | A fricative becomes a stop | dat for that, ting for thing | 5 (voiced "th") | 5;0 |
/r/ and /l/ become /w/ or /j/ | wabbit for rabbit, yeg for leg | 6–7 | 5;0 |
Sources: ASHA, Selected Phonological Patterns (ages in years) and Bowen (2011), Table 3 (years;months, so 3;6 is three and a half). The palatal fronting age is from Bowen's more detailed 2015 table of Grunwell's ages (Bowen, 2015, p. 73), which separates velar and palatal fronting; the same table puts consonant harmony at 4;0 rather than 3;9. Other patterns you may hear, such as reduplication (wawa for water), are not in either table.
Phonological processes by age: what should be gone when
Using ASHA's ages, with Bowen's where ASHA gives none:
By about 3: final consonant deletion, assimilation, prevocalic voicing, final devoicing, and stopping of /f/ and /s/.
By about 4: velar fronting, weak syllable deletion, cluster reduction without /s/, deaffrication, and stopping of /v/ and /z/.
By about 5: cluster reduction with /s/ and stopping of "sh", "ch", "j", and "th".
By about 6 to 7: gliding (Bowen: 5;0).
A child who still uses one of these patterns a little past its age is not automatically a concern. A child who uses several, uses them often, or is hard for unfamiliar listeners to understand should be seen by a speech-language pathologist.
Why do phonological process charts disagree?
Put ASHA's table next to Bowen's and the ages rarely match: gliding is 6–7 in one and 5;0 in the other, and stopping of "sh", "ch", and "j" is 5 against 4;6. Neither table is wrong. Charts differ for a handful of reasons:
Different criteria. A chart can report the age at which half of children do something, three quarters, or nine in ten. Crowe and McLeod's review of US consonant studies shows how much this matters: at a 50% criterion, 22 consonants were acquired between 2;0 and 2;11; at a 90% criterion, only seven were (Crowe & McLeod, 2020).
Different sources and decades. ASHA's age column draws on Bowen (2011) and two clinical textbooks; Bowen's rests on Grunwell's 1987 data. Over 90 years of US studies, Crowe and McLeod point to "changes in the dialect and demography of young children acquiring American English".
Lumping or splitting. Some studies count velar and palatal fronting as one pattern, others separately (Cohen & Anderson, 2011), so one "fronting" age can hide two timelines. Stopping is split into nine sounds in Bowen's table and into three groups in ASHA's.
Different populations. A UK study of 684 children (Dodd et al., 2003) counted a pattern as age-appropriate while more than 10% of children in an age band still used it. Applying that criterion to a Scottish sample, Cohen and Anderson found "lower than expected incidence" of palatal fronting, stopping of fricatives, and some cluster reduction compared with published norms.
Single words, not conversation. The 15 studies in Crowe and McLeod's review "primarily used single-word elicitation", typically naming pictures. Conversation can look different.
Both chart makers warn against reading the ages as cutoffs. Bowen calls such age estimates "guides, not universal diagnostic cutoffs", and Crowe and McLeod write that "age of acquisition data are not enough for clinical decision making or to determine eligibility for services."
Phonological processes vs. speech sound norms
A process chart and a sound chart answer different questions. A process chart says when a pattern, such as stopping, usually disappears. A sound chart says when a sound is fully correct. The two do not line up, and they are not meant to.
Take /s/. Stopping of /s/ is typically gone by about 3, but 90% of US children produce /s/ correctly only during the 4-year range (Crowe & McLeod, 2020). In between, the sound is often long and hissy but not quite right; summarizing Smit (1993), Crowe and McLeod note that /s/ "was frequently stopped... when the children were younger, produced as a dentalized /s/ or postalveolar consonant when older." The pattern has gone, but the sound is still settling.
Overall, children "have, on average, acquired all but four consonants by 5;0" in the US data; the late ones are /r/, the two "th" sounds, and the "zh" in measure. To check a single sound against the age a child is now, use our speech sound development checker.
Which phonological processes are not typical?
Some patterns are not part of typical development at any age. Cohen and Anderson, citing Dodd (2005), list "backing, initial consonant deletion, vowel distortions or atypical substitutions" as error types that "may indicate disordered or deviant development" (Cohen & Anderson, 2011).
Backing: front sounds move back, the reverse of fronting: cop for top, gore for door.
Initial consonant deletion: the first consonant is dropped: un for sun, ish for fish.
A favorite sound: one sound replaces many others. ASHA's example: shore, sore, chore, and tore all said as door (ASHA Practice Portal).
ASHA's guidance covers both ends: "Phonological patterns that persist, or phonological patterns that are atypical, lead to a diagnosis of phonological disorders."
When should a phonological process be a concern?
It persists well past the ages in the chart.
It is atypical, such as backing or initial consonant deletion, whatever the child's age.
Several patterns persist together, or the child is hard for unfamiliar listeners to understand.
One caution runs the other way. A pattern that belongs to the child's dialect or home language is not an error. ASHA notes that "dialectal varieties of a language that do not produce certain sounds would not be considered an error pattern", and that resolution ages "can vary greatly in each child—especially if a child uses multiple dialects" (ASHA).
How are phonological processes treated?
The main approaches work on the pattern, not one sound at a time:
Minimal pairs contrast the target with the child's error (key vs tea), so the child hears that the error changes the meaning. In Weiner's 1981 study, the method reduced final consonant deletion, stopping of fricatives, and velar fronting in two children, with generalization to untrained words (Weiner, 1981). See our guide to minimal pairs in speech therapy.
The cycles approach works through several patterns in rotation and suits children who are hard to understand; see the cycles approach, explained.
Complexity-based target selection picks harder targets on purpose, on the idea that they spread change further across the sound system; our cluster reduction guide shows how that works for clusters.
When you need printed materials for a pattern, Ga-loo builds minimal pairs sets for fronting, stopping, gliding, final consonant deletion, cluster reduction, voicing, and deaffrication, with both words pictured and each pair checked against the CMU Pronouncing Dictionary, plus Hodson-style cycles kits. For apps, card decks, and other ready-made sets, see our comparison of minimal pairs resources.
Guides to each process
Velar fronting: tat for cat, typically gone by about 4.
Stopping: pun for fun, gone sound by sound between 3 and 5.
Gliding: wabbit for rabbit, one of the last to go.
Cluster reduction: top for stop, by about 4 without /s/ and 5 with /s/.
Final consonant deletion: bu for bus, typically gone by about 3.
Frequently asked questions
What are phonological processes?
They are systematic simplifications that affect a whole group of sounds or word shapes, not one sound. ASHA describes them as "systematic sound changes or simplifications that affect a class of sounds (e.g., fricatives), sound combinations (e.g., consonant clusters), or syllable structures (e.g., complex syllables or multisyllabic words)."
At what age should phonological processes stop?
Most fade by about 4. The last to go are stopping of "th" (about 5) and gliding (5 to 7, depending on the chart). ASHA adds that children "usually acquire most speech sounds by age 5."
What is the most common phonological process?
In young children, cluster reduction and gliding. In a study of typically developing children aged 1;6 to 2;5, "the most prevalent phonological processes... were cluster reduction and deviations involving liquids (e.g., gliding)" (Preisser, Hodson & Paden, 1988).
How many phonological processes are there?
There is no fixed number. ASHA's table lists nine common patterns and says the list "is not exhaustive"; Bowen's table has 17 rows because it splits stopping by sound.
What is the difference between a phonological process and an articulation error?
A phonological process is a rule-like pattern that affects a class of sounds, such as stopping all fricatives. An articulation error is a problem producing a particular sound, such as a lisp on /s/. ASHA notes that the two are often hard to tell apart and that "a single child might show both."
Is a phonological processes chart enough to diagnose a child?
No. Charts show typical ages, and they disagree with each other by months or even years. An SLP looks at how often a pattern occurs, which patterns occur together, whether any are atypical, how well the child is understood, and the child's dialect and languages.
Sources are linked throughout. Educational content: an SLP chooses the approach for a specific child.