s-clusters

Prevocalic s-clusters (before the vowel) are written in English with a voiceless plosive as the second segment, such as:

'spider'
'space'
'stamp'
'stay'
'sky'
'scale'

This is misleading, as the second element plosive is unaspirated and perceived by all English speakers as voiced.

This may be demonstrated using audio recordings of productions of s-cluster words.

Di-clusters

'sp-' should be transcribed as / sb- /
'st-' should be transcribed as / sd- /
'sk-' should be transcribed as / sg- /

Tri-clusters

'spr-' should be transcribed as / sbɹ- /
'str-' should be transcribed as / sdɹ- /
'skr-' 'scr-' should be transcribed as / sgɹ- /

Di-cluster examples

'spider' is transcribed as / ˈsbaɪ.də /,
'space' is transcribed as / sbeɪs /,
'stamp' is transcribed as / sdæmp /,
'stay' is transcribed as / sdeɪ /,
'sky' is transcribed as / sgaɪ /.

Tri-cluster examples

'spread' is transcribed as / sbɹed /,
'spring' is transcribed as / sbɹɪŋ /,
'straw' is transcribed as / sdɹɔ /
'stream' is transcribed as / sdɹim /
'skate' is transcribed as / sgeɪt /,
'scrape' is transcribed as / sgɹeɪp /.

NB. This effect is NOT a result of accent or dialect.

Clinical Implications

Children with s-cluster reduction observed in their speech may produce s-clusters with one of more consonants omitted. Most frequently, the voiceless alveolar fricative [s] is omitted.

This will then lead to a voiced plosive realisation such as:

'space' / sbeɪs / → [ beɪs ]

This may be incorrectly categorised as Context Sensitive Voicing, when in fact it is ONLY cluster reduction.

Minimal pairs also typically use s-cluster contrasted with a voiceless plosive world such as:

*'spin' versus 'pin'

This leads to unnatural sounding productions when the child uses the s-cluster.

Instead, voiced plosive contrasts work to produce more natural speech: