Short answer
The character is actually saying there not here.
[We can double-check this by comparing the vowel the speaker uses when he says the word there at 0.47, and then the word here at 1.02. He uses /eə/, as expected, for the word there. He uses /ɪə/, as expected, for the word here. In the Original Poster's clip, the speaker uses a very clear /eə/-vowel].
Full answer
What the Original Poster is hearing is this:
['wɒss ɪn̪ 'n̪eə mæks]
Notice that I have used square brackets for the phonetics, [ ]. This is because we need a detailed description of the sound, not just a list of the phonemes. When we just give a list of the phonemes we use slanty brackets, / /. The important detail in the transcription here is the little upside down 'tooth' symbol underneath the two 'n's. The little symbols that appear above or below the 'letters' in a transcription are called 'diacritics'. We will come back to the diacritics soon.
Now, a different speaker, speaking slowly and very pedantically, very fussily, might pronounce the same sentence like this:
['wɒts ɪn̪ 'ðeə mæks]
In other words the speaker is pronouncing the following sentence:
What's in there, Max?
Notice that the second transcription has the voiced th' sound, [ð], at the beginning of the word there. And also that it still has an upside-down tooth diacritic under the 'n'.
Why does the sentence sound like that
When we speak, our mouths are always preparing for the sound that are coming up a bit later. We also have to make small adjustments to move smoothly from one sound to the other. Sometimes English (like other languages) has rules which specifically allow one sound to change to another when next to other particular sounds.
When one sound becomes more like another, we call this assimilation. Most assimilation in English is anticipatory which means that a particular sound (the technical word is 'segment') becomes more like the sound that will come after it. This is what is happening in the first word Whats:
- /wɒts/
- /wɒss/
In (1) we see a typical careful pronunciation of the word what's. The last two sounds there are both made with the tongue at the same place in the mouth. Just behind your teeth there is a little shelf which slopes upwards. It's called the alveolar ridge. English speakers make a lot of their sounds on the alveolar ridge:
These are the first sounds in the words sip, dip, nip, lip, sip and zip. These sounds are very unstable in English and often change their place of articulation (the place they are made in the mouth) and also often their manner (the way the sound is made). In fast speech, the /t/ in the word what can keep its voiceless quality, keep its place of articulation on the alveolar ridge but change the type of sound so that it becomes sibilant. This means that it becomes the same as the following /s/ segment. We thus get /wɒss/.
Out of all the alveolar sounds, the one that is most likely to change it's place of articulation is the sound /n/. This sound will most often change to match the place of articulation of a following consonant. Before a /p, b, m/ and (to a lesser extent) /w/ it will often become an /m/. Before a /k/ or a /g/ it will often have the 'ng' sound, /ŋ/. This is how the word ten might change:
- tem people
- tem bananas
- tem minutes
- the tem wonders of the world
- teng carrots
- teng grapes
An /n/ will also nearly always change when it occurs before a 'th' sound, for example the unvoiced sound at the beginning of the word things, or the voiced sound at the beginning of the word there. These are the sounds /ð/ and /θ/, the dental fricatives. These sounds are made using the tongue on the back of the upper teeth. When an /n/ occurs before a 'th' sound it will also change its place of articulation and become dental, although it will still be nasal. [A native English speaker will not notice this, but will be able to feel that the place they are making the /n/ is now on the teeth if they check.]. When we make a detailed transcription we use an upside down 'tooth' diacritic under the /n/:
- [ten̪ θɪŋz] ten things
- [ɪn̪ ðeə] in there
However, when an /n/ occurs before a voiced dental fricative, the sound /ð/, it can often also cause the /ð/ to change. In other words we get perseverative assimilation, where a sound causes the following sound to change. This time the /n/ causes the /ð/ to also become nasal. This means that the two sounds have both become more like each other. They are now both dental nasals, [ n̪ ].
Summary
So in the Youtube video, the /t/ in what's becomes an /s/, and the /n/ in in and the /ð/ in there both become dental nasals [n̪], to give ['wɒss ɪn̪ 'n̪eə mæks].
Very often, when we get a lot of small function words together, we often have large numbers of alveolar and dental consonants jostling together. These very often change in many different ways and may even disappear altogether or join up with other sounds to make completely new ones. This means that there may be a large number of different possible pronunciations of the same type of sentence. And all of this can make it very difficult for speech recognition software to tell what the actual words are. They're not as good as humans yet!
