This is why I have such a hard time keeping up with politics

On the September 6 broadcast of Meet the Press, David Gregory pronounced distrust as if the syllabification was di-strust rather than dis-trust. How can I tell? Because he didn't aspirate the /t/. Listen:

Let's do a little lab.
Put your hand in front of your mouth. Say spit.

Leave your hand there and say pit.

Notice the difference between the two? You should have felt a stronger puff of air after the /p/ of pit. The difference is represented in IPA with a small superscript 'h' on the stop when aspirated.
spit: [spɪt]
pit: [p^hɪt]

The rule simplified: when a voiceless stop occurs initially on a stressed syllable, it's aspirated. Aspiration is that little puff of air. Do the above test to hear the difference in all the following voiceless stops.
pot/spot
kit/skit
top/stop

The [s] before each stop changes the conditioning environment. When the [p] [k] or [t] aren't the first segment in the syllable, they aren't aspirated.

So, some possible explanations of Gregory's unaspirated pronunciation of distrust:

He just messed it up, and normally he would have pronounced it [dɪst^hɹʌst] with the standard syllable initial aspirated /t/. This would make sense because the morphology of a dis- prefix and trust root is probably still intact in his competence.

If that morphology is not intact, Gregory might be syllabifying either on an analysis of a di prefix strust root (very unlikely), or a single morpheme syllabified with maximal onset* -str- on the second syllable. The latter is possible, but not likely because of the strong semantic preservation of a still relevant -trust root.

The standard morphology is intact, but the onset maximization is a stronger influence on aspiration than is syllabification. This is problematic in many ways, but I think it's part of a likely explanation.

The morphology and syllabification are not factors as the st string is simply not an environment that conditions or allows aspiration in Gregory's speech, regardless of onsets and codas.**

Why don't I think this was just a mistake? Because I remember a broadcast from about a year ago, (October 22 2008) when he was talking with Senator Sherrod Brown, (D-OH) that makes it seem this is a feature of Gregory's speech? (about 25 seconds in.)

Le…let me show you a— a piece of reporting from the Boston Globe where they were talking to Youngstown voters and getting their thought [sic] about this election in such a tight race in your state.

Because of his pronunciation of Youngstown it looks like Gregory occasionally leaves off aspiration of the dental stop in the string st, even when morphology would typically split the segments into different syllables. He does not however, block aspiration of /k/ in the word disconnect around 45 seconds in.

Mr Gregory, you've got some 'splaining to do.

---

*A simplified theory of onset maximization might be stated as a rule that any segment in a string that can be well-formed on either the end of one syllable or the beginning of another, will be syllabified as an onset segment. The segment in question for this post is the /s/.

**For the sake of relative brevity I'll leave out of this post the discussion of bleeding, counterbleeding, feeding or counterfeeding.