The transformation that language undergoes when it becomes speech is examined in English. Statistical analysis of a representative sample of natural, informal speech reveals a number of characteristics of dynamic speech that distinguish it from static (citation form or pre-dynamic) linguistic form. It appears that in running speech, vowels and consonants, words and sentences are less central to speech dynamics than sonorants and obstruents, syllables, and syllabic phrases. The findings include these: (1) when language is learned in its dynamic (spoken) rather than static form, the basic unit contains more phonemes, making it more complex to learn; (2) utterances are predominantly made up of sonorants and other voiced sounds; (3) running speech can tolerate a maximum of five consonants in sequence, and the sequences are regular and predictable; (4) syllable boundaries are fluid and can be expected to change configuration through restructuring; (5) syllable combinations affect syllabic distinctions; and (6) application of the variable fast speech rules still results in regularly shaped utterances. It is suggested that incorporating information about speech dynamics into second language instruction may facilitate acquisition of oral fluency. (MSE)