What Makes a Vowel a Vowel and a Consonant a Consonant
Today I Found Out | 16.VI.2017
- W a i t ....
You are mixing two definitions of the distinction.
In definition given, (consonantal) y and w are in fact vowels.
However, they are not syllable kernels.
In the word "I" as you pronounce it, the syllable kernel is "ah" and then there is an off-glide "y".
In the word "how" as you pronounce it, the syllable kernel is also "ah" and then there is an offglide "w". This is not the reason why there is "w" in the spelling though, since "ou, ough, ow" are three alternative ways of spelling same archiphoneme, which in some connections (never spelled ow) is pronounced "oo" ("soup") which was the original prounciation of it, but not of "oo", and in others is pronounced "ah+w".
Now, glides (whether on or off, "yet, wet, I, how") are vocalic in how they are made (no obstruction) and consonantal in how they are used (not as syllable kernels). In Czech (which you have presumably heard) the sound "r" is (at least intermittently) made by obstruction, and yet it is sometimes a syllable kernel.
So, the most usual rule, phonetic consonants used phonematically as syllable onsets and offsets, phonetic vowels used phonematically as syllable kernels (essentially : loudest and most audible sound per syllable) has its exceptions.
Glides being "consonantically" used vowels, "vocalic" liquids (as in vlk, Vrba) being "vocallically" used consonants.
- 1:31 "You won't see a consonant that is a word by itself"
In Polish and presumably Czech, you at least find prepositions that are such "z" and "w" - attached syllabically to the first syllable of the next word.
- 1:48 "hmmm" phonetically has a consonantised vowel (h is a voiceless vowel, a glide, not quite unlike y and w) and then a vocalised consonant (m being used vocalically much like l and r in vlk and Vrba).
Phonetics and vowels and consonants don't care two pence whether sth is English or Czech, it just describes what is heard as sounds.
For those believing there was a Proto-Indo-European language, vocalic m was fairly common in it - in other languages actually documented, it doesn't appear, though.
In a word ending with feminine ending -a, adding -m just added a consonant.
In a word ending with a consonant, adding -m involved adding a vocalised -m (on the theory).
In Greek you have thalassan (consonantic m > n), and trikha (vocalic m > a).
In Latin you have togam (-a+m continues identically) and regem (vocalic m > em).
In Germanic the vocalic m would tend to become um. Presumed kmtom > hund (as in hundred, hundra, hundert). Appropriately for above observations, in Latin it is centum (mt > nt + vocalic n > en) and in Greek he-katon (vocalic m > a, as in accusative singular trikha).
But on the theory of common ancestry, before this, there were vocalised not only liquids but also nasals.
- 2:35 "Fricatives are sounds you make by pushing air through a small gap"
"in your teeth"
Correct only for dental fricatives. These include both thorn and eth sounds and the s and z.
For sh (and in some languages zh), between tongue tip turned back and hard palate. Tongue tip actually in same position as with r.
For ach-laut, between tongue back and uvula, right in the back of the mouth.
For f and v, actually between lips and teeth. Or in some languages, just between lips.
You were confusing "fricative" (the way s and z are produced) with "dental" (the place in the mouth where they are produced).