A customer teaching in an ELL setting emailed us with a question about certain blends. He wanted to know whether it is accurate to teach that the following consonant blends are pronounced in the following ways, with the second phonogram saying something besides its usual sound.
sp - /sb/
st - /sd/
sc, sk - /sg/
tr - /chr/
It is true that in many people's speech, when the consonants he listed are together in a word, one of them gets a bit distorted for ease of pronunciation. However, this is simply a natural function of the mouth making these words easier to say, and happens more for some people than others, more in some words than others (contrast 'scale' with 'scrape'), and more or less depending on how precisely and emphatically one is speaking. Some will say SP closer to /sp/, some more like /sb/, and most somewhere in between. (And so on with the other pairs.) Certainly any native speaker would know exactly what you were saying if you crisply, carefully said the "true" sound rather than letting it assimilate, even if they themselves pronounced it in the assimilated form much of the time.
So although the pronunciations listed above are a common pattern in English speech, but it would not (at least from my personal observations and thought) be accurate to teach that these pronunciations are the only correct ones for these consonant blends, or to tell students that they "should" pronounce them that way. Rather, I would let them know that this often happens in people's speech and that this is normal and fine, and help them feel why the phenomenon occurs by feeling where the different sounds are in the mouth.
And if a student is segmenting a spoken word into its sounds and segments, for example, tree into /ch-r-ē/ or drip into /j-r-ĭ-p/, do not worry. Tell the student that this was a good attempt and that this word can be tricky sometimes because the sounds get a little distorted when we say them in order to make the word easier to pronounce.
Say the word very clearly for the student, pronouncing the consonant blend as precisely as possible: Tree. Drip. Model blending it: "/t-r-ē/, tree." Have the student try again.
It may also be helpful to have the student say the sounds /t/ and /r/ (or /d/ and /r/, etc.) separately and then blend them together to feel how it's a little tricky to say and easier if you distort the first sound slightly.
If the student continues to struggle, let him or her know that it is fine and takes practice, and that you'll help with segmenting words that contain these sounds for now.