Piotr Gasiorowski wrote:

> 15-08-03 12:33, Jean-Paul G. POTET wrote:
> > No doubt emphasis plays its part, but, to me, English has the initial
> > glottal stop whenever there is no liaison.
> > The liaison rule operates for utterances like <an ear>, constituted of a
> > single iambic foot : [°@ "nI@], but it does not in the case of <one
> ear>,
> > <two ears>,
> > In utterances like <one ear>, <two ears>, that are constituted of two
> > trochaic feet each, isn't the glottal stop compulsory : ["°wAn '°?I@],
> > ["°thu '°?I@z]?
> I don't think so, but then I was explicitly instructed by my RP
> pronunciation teachers to avoid the glottal stop in such cases. Let's
> hear the opinion of our native English-speakers (the avoidance of [?]
> may well be a dialectal phenomenon). I insert in when speaking Polish,
> but because of my training I say <two ears> without a glottal stop. I am
> aware of a momentary diminution of energy between the vowels, but it's
> probably Ladefoged & Maddieson's "creaky voiced glottal approximant",
> for which they use the symbol [*].

Isn't here a simpler way to do this?

Fs there a need to give a name ( of a phone) to every interval in time?