I've made some formant computations in an effort to understand the differences between certain "u" vowels in words spoken by Larry Phillips. The segments I've analyzed are from the statement "too few avenues of communication".  The segments are the word "too", the word "few", the "ues" in" "avenues", and the "uni" in "communication".  The formants of Larry's vowels are shown in Figure 1.

2. My "he, who, hue" pronounced as I normally pronounce them, and
    My "he, who, hue" pronounced with an effort to move my tongue farther forward.
3. My "ee, oo, yoo, too" pronounced as I normally pronounce them
4. My pronunciations of "oo" preceded by various other consonants

In each of the formant plots, the bottom trace is "formant one", and the formants are in ascending order from bottom to top.  By convention, a vowel is more "forward" as the distance between formants one and two increases.  A vowel is "higher" as the frequency of formant one is lowered.

Figure 1:

Some significant features of figure 1 are

• 1. The "oo" in "too" glides from a forward position to a more back position.  This is presumably caused by the fact that the tongue moves forward to pronounce the "t", and it takes the full duration of the vowel to get to the position it would have if the "oo" were pronounced separately.  Compare this with the effects of other consonants as shown in figure 4.
• 2. The vowel in "few" is a diphthong that glides from a far forward position that corresponds to the expected location of an "ee" sound, to the position "oo" would be expected to have if it were pronounced separately.
• 3. The "u" vowel in "ues" starts at the "ee" position, but stays much farther forward than the one in "few"; that is, the glide is not as great.
• 4. The sound "uni" doesn't start as far forward as the others, and it glides to about the same place as that of "ues", so the glide is shorter.

In looking at figure 2, it's important to understand that an "h" preceding a vowel has little effect on the vowel.  Ladefoged, in his A Course in Phonetics, explains that an "h" doesn't have a position on the vowel chart of its own, but has formants that are about the same as the vowel that is to follow.  This makes "h-" words convenient to use in demonstrating vowel pronunciations: the vowel qualities are about the same as they would be without the "h".

Note that my diphthong in "hue" is similar to Larry's in "few", and the formants of my "who" are close to those at the ending of "hue".  The righthand part of figure 2 shows what happens when I make an effort to pronounce the "oo" sounds with my tongue farther forward than normal.  Note that in the case of "who" I didn't succeed in moving my tongue farther forward until the vowel had started, so there's a sharp raise in the frequency of the second formant at the beginning of "who".

This figure shows the actual position of my "ee" sound.  It also shows that my "too" has a similar glide to Larry's, but it's not as great.  I find it interesting that in the case of these pronunciations of "you" and "too" the glide doesn't reach as far back as it does in my "hue" in figure 2.  If I were to make the same recording again, I wouldn't be surprised to find that those vowels did slide farther back.  Vowel qualities can change significantly from one utterance to another of the same speaker.

I think the glides shown in this figure can be understood in each case by considering the position of the tongue when pronouncing the consonant preceding the vowel.  Note the difference in the sizes of the glides in "doo" and "too".  I do indeed find that when I pronounce "doo" my tongue is a little farther forward on the alveolar ridge than when I pronounce "too".