Puzzle Phenomenon, Lexical Selection and Co-articulation



Phonological development and phonetic mastery do not synchronise precisely. A common example of this asynchrony, referred to by Smith (1973) as the puzzle phenomenon, is provided by children who realise /s/ and /z/ as 'th' sounds, while producing "th-words" with [f] in place of voiceless 'th', and [d] or [v] in place of /ð/.


The following classic example of phonetic ability preceding phonological execution came from a client, Andrew, aged 4;6. The word on the left in each case is the target word, and the word on the right reflects Andrew's production.

  • some = thumb
  • thumb = fum
  • yellow = lello
  • zoo = thoo
  • then = den
  • those = doze
  • glove = gwub
  • breathe = bweeve
  • brother = bwuzzer
  • globe = blobe
  • rabbit = brabbit

Lexical selection

Evidence from studies of lexical selection provides support for the view that children are "aware" of their phonetic limitations very early (i.e., during the first 50 words stage) (Ferguson & Farwell, 1975; Schwartz & Leonard, 1982). How conscious the awareness is, of course is uncertain, but children do seem to reflect limitations of motor speech control in their early word choices.

Does this mean that the speech motor mechanism of young children is in fact immature? Studies of duration, co-articulation (Kent, 1982; Hawkins, 1984) and variability (Smith, Sugarman & Long, 1983) in children’s speech have demonstrated that this is likely to be the case. Hawkins (1984) reviewed a series of comparative studies of child and adult segment and phrase durations, concluding that children tend to have longer durations, and hence slower speech rate. Hawkins also found that children show greater intrasubject variability of speech segment and phrase durations than adults. Smith, Sugarman and Long (1983) demonstrated that such variability was due in large part to immaturity of the neuromotor mechanism for the control of speech movements.


Co-articulatory ability, or the normal capacity to produce an overlap between speech sounds, caused by an overlapping in the sequence of gestures which produce them, has been thought by Kent (1983) and others, to increase with age. Later studies of co-articulatory ability (Repp, 1986, Sereno and Liebermann, 1987), suggest that speech rate and variability are more relevant predictors than the age of the child. They showed that the development of co-articulatory ability varied widely from child to child, and that the length of time a sound had been in a child’s repertoire may be more significant than chronological age in predicting co-articulatory ability. Sereno and Liebermann (1987), in a study of children aged 2;8 to 7;1, found no correlation between age and co-articulatory ability.

Further evidence that phonetic development is implicated in the development of phonological contrasts comes from the frequent observation that phonological contrasts are realised in the child’s speech, albeit inaccurately, as they gradually perfect their phonemic realisations of target forms.

Children’s progress towards the adult targets of /s/ and /r/, commonly via interdental and labialised versions, respectively, are examples of the "perfecting" process that takes place. Menn (1983) summed up the complex (and fascinating) interplay between the levels of development and learning of phonological and phonetic processing:

The mismatches between adult model and child word are the result of the child’s trial and error attempts; they are shaped by the child’s articulatory and auditory endowments (and this to that extent are ‘natural’) and by the child’s previous success in sound production. All rules of child phonology are learned in the sense that the child must discover for herself each correspondence between the sounds she hears and what she does with her vocal tract in an attempt to produce these sounds. (p. 44)


Ferguson, C., & Farwell, C. (1975). Words and sounds in early language acquisition. Language, 51, 419-439.

Hawkins, S. (1984). The development of motor control in speech. In N.J. Lass (Ed.). Speech and language: advances in basic research. Orlando: Academic Press.

Kent, R.D. (1982). Contextual facilitation of correct sound production. Language Speech and Hearing Services in Schools, 13, 66-76.

Kent, R.D. (1983). The segmental organization of speech. In P.F. MacNeilage (Ed.). The production of speech. New York: Springer.

Menn, L. (1983). Development of articulatory, phonetic and phonological capabilities. In B. Butterworth (Ed.). Language production. London: Academic Press.

Repp, B.R. (1986). Some observations on the development of anticipatory articulation. Journal of the Acoustical Society of America, 79, 1616-1619.

Schwartz, R., & Leonard, L. (1982). Do children pick and choose? Journal of Child Language. 9, 319-336.

Sereno, J.A., & Lieberman, P. (1987). Developmental aspects of lingual co-articulation. Journal of Phonetics, 15, 247-257.

Smith, N.V. (1973). The acquisition of phonology: A case study. Cambridge: Cambridge University Press.

Smith, B.L., Sugarman, M.D., & Long, S.H. (1983). Experimental manipulation of speaking rate for studying temporal variability in children’s speech. Journal of the Acoustical Society of America, 74, 744-749.