Miss Jones is coming to lunch tomorrow. What am I to give her to eat? /ˈmɪs ˎdʒəʊnz ɪz kʌmɪŋ tə lʌntʃ təˏmɒrəʊ. `-wɒt əm aɪ tə ɡɪv ə tu ˎiːt / [1]
What does she like? /ˈwɒ(t) dəz ʃi ˋlaɪk?/ [2]
That’s not the question, really. She has all these dreadful allergies. /(ða)ts nɒt ðə `kwestʃən ˎˏrɪəli. ʃi az ɔːl ðiːz ˏdredfl `alədʒiz./ [3]
So I’m terrified of giving her something that’ll bring her out in spots. /səʊ aɪm `terəfaɪd | əv ɡɪvɪŋ hɜː ˏsʌmθɪŋ | ðətl ˈbrɪŋ ər aʊt ɪn `spɒts. / [4]
What can she eat? /wɒt ˋkan ʃi iːt?/ [5]
Well, I imagine steak doesn’t disagree with her. /`ˏwel | aɪ mædʒɪn `steɪk dʌznt dɪsəˏɡriː wɪð hə. [6]
But I believe she comes out in a rash if she touches onions. bət aɪ bəˈliːv | ʃi ˈkʌmz aʊt ɪn ə `raʃ ɪf ʃi tʌtʃɪz `ˏʌnjənz./ [7]
For goodness sake! Ring up her house and ask them. /fə ˈɡʊdnəs `seɪk. ˈrɪŋ ʌp ə `haʊs ən `ɑːsk ðəm./ [8]
There're some interesting elisions in this text the product of the readers’ successfully adopting a natural quite brisk tempo. At a slower more deliberate pace these omissions no doubt wdntve happened. The first is the probably complete loss of the /t/ normally present at the end of the word “What” that begins line 2. The second is at the word “That's” at the start of line 3 which begins with such a realistically blurred casual articulation that it might equally have been “It's”. In the same line there's complete loss of /h/ from the word “has”. At this brisk pace it sounds perfectly natural but if slower wd sound like an uneducated person's usage.
This is a notable contrast with General American pronunciation habits where there is no sanction against omitting /h/ from have, has or had as main verbs (non-auxiliaries) — a fact I've not noticed recorded in any description of American English phonetics. Aphesis /`æfəsɪs/, the loss of the initial sound of a word, is a well known linguistic process seen in the replacement of esquire by squire etc. This happens to the word imagine very frequently when I precedes it but is of doubtfully acceptable usage in most other contexts and understandably not taken notice of by the pronouncing dictionaries.
It's often rightly remarked that there's not just one GB (or "RP") but a number of similar varieties. Among the distinguishing characteristics of these different varieties are their treatments of words ending in -y and corresponding plurals in -ies. Older and/or posher people tend to end such words using /ɪ/; younger and/or less socially conspicuous people tend to favour /i/. This latter group took over the mainstream from the former during the middle of the twentieth century. Another "halfway" group favours /ɪ/ for -ies but /i/ (or rarely /iː/) for -y. A still relatively small group favours the rhythmically strong /iː/: it was their increasing emergence that strengthened the hand of those who advocated recognising rhythmically weak /i/ in dictionaries from about 1978 onwards — something which became fully accepted only when John Wells in effect welcomed it into his LPD in 1990. It had previously been very common for centuries but unrecorded by most phonetic observers. See my article at 3.2 on this website.
The varieties can't be ascribed to distinct social or regional or even age groupings within GB. The split between one type and another can be heard even within the same family. Perhaps the most famous pair of brothers in British broadcasting are called Dimbleby. They went to the same senior school. One is only six years older than the other and regularly ends his name with /-ɪ/: the other doesn't.
The motive for sound change is very often the aspiration towards "clear speech". This development was obviously in that category: it was much in evidence in posh girls' schools. People may vacillate and fudge of course but our female speaker in this piece clearly favours the more modern style. She hasn't got a clear /ɪ/ at “really” or “allergies”, tho she did have one at “holiday” in People Speaking 3 (but see what we said about that.)