Hearing speech, we identify at first contents and sentences. Sentences consist of words. Words consist of the individual sounds of our speech. And sounds? Sounds differ by length and pitch. (Image below: Steps of linguistic competence according to Ptok)


For a reliable everyday interpretation of speech we need to reliably identify every single sound. While we may quite easily complete a sentence like, ‘And he hammered a na*l into the wall’, this turns into a bit of a challenge with, ‘The *rice of *ans I keep in *ind’. Doesn’t it?
Just imagine that speech would use to arrive in your mind like this:
‘Blease lisden now glosely. In this dexd we assume thad you gannod agguradely disdinguish the phonemes b-d-g-k-p-t. Aren’d you surbrised thad you are able nod only do guess on the dexd bud do undersdand mosd of id?...’
Did you ever notice that we do not, in fact, use or hear a hiatus between individual words? Our mind inserts the hiatus on its own - if we interpret everything correctly.