![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
so the basic answer is yes
however i am never satisfied with being basic when i can do an unnecessary linguistic analysis so it is only right i investigate this with science.
so in phonetics i learned how to read something special called spectrograms! according to wikipedia,
a spectrogram is a visual representation of the spectrum of frequencies of a signal as it varies with time.
since each phoneme (sound), by literal definition, has different sound frequencies, you can tell different sounds apart and identify them by reading a spectrogram of say, someone rapping them.
i am here to prove that "swear to god girl you know you got that wap" is the truth and not "swear to god girl you know you got that walk” and i will do so by taking you on a spectrogram journey. i WILL assume from the start that wap is the correct answer and not walk because i’m unscientific like that.
my source clip is this right here and i am using the free linguistic software praat for my analysis.
this is what a spectrogram looks like.
the upper part should look p familiar as it indicates intensity of the sound waves. you can use it to help figure out where the word or syllabic or phonemic boundaries are.
phoneme: perceptually distinct units of sound. in this case w, a, and p all represent sounds which are conveniently labelled w, a, and p
however, it’s the lower part that’s most important in this analysis, as it indicates the frequencies of the sound waves, which is what allows us to identify the specific sounds.
this spectrogram actually shows mark rapping the word wap, so there should be three distinct parts of the spectrogram, w, a, and p. the first thing to do is look out for phonemic boundaries and separate the sounds or phonemes! pls look to the lower part... any abrupt changes are phonemic boundaries so like it’s pretty intuitive... i think the boundary between w and a is pretty obvious but the one between a and p is more ambiguous. like oh it’s just a wall of noise but here you can look to the upper part for guidance... p (or k in “walk”... no one thinks mark is pronouncing walk here with the l right? right) is a voiceless stop so it should be quieter and ta-da.... there’s a part where the sound waves get less intense and then change... and then you look down and see wow actually the incomprehensible shaded part also changes a bit as well... that’s the syllable boundary.
as you can see. now that we’ve done that we only need to read the third part and figure out whether that sound is a p or a k! (as is indicated... it’s a p).
so what we need is FORMANTS... which are the parts that are extra dark in the spectrogram...
formants are bands of acoustic energy around certain frequencies.
we make sound by the air that passes through our vocal tract (oral cavity, nasal cavity, pharynx, larynx). we make different sounds with articulators such as the upper and low lip, upper and lower teeth, alveolar ridge, soft palate, velum, uvula, tongue etc etc. the upper lip, upper teeth, alveolar ridge, soft palate, velum, uvula etc are passive articulators, which mean they do not move, while the lower lip, lower teeth, tongue etc are active articulators which move towards or away from the passive articulators to create different blockages in the vocal tract. these blockages or obstructions affect the shape of the vocal tract through which air passes, changing the resonant frequencies of each sound. the acoustic energy of each sound is amplified at these resonant frequencies, creating formants. therefore, by looking at the concentrations of sound waves of each sound called formants, we can determine the shape of the vocal tract, which is the position of the articulators, which is the specific phoneme spoken
note: you can determine phonemes by things other than the formants, such as voicing (whether the vocal folds vibrate or not) and length of the sound, but in this case, the formants are the most crucial
you can vaguely tell where the formants are, but luckily praat has a function to show the formants and lets us have an easier time.
they appear as the alternating orange and yellow lines. the first formant is the lowest one, the second is the second lowest, and so on. generally, you only need the first three formants to determine a sound.
note: the blue line does not matter. it indicate pitch and implicitly voicing, but things are extremely fucked from the background track and it doesn’t matter either way.
now, p and k are both voiceless stops (vocal folds do not vibrate, complete obstruction of the airway creating a “stop”) which makes them extremely similar! p is a bilabial stop, created by putting the upper and lower lip together, while k is a velar stop, created by touching the back of the tongue to the velum. since the biggest difference between them is the shape of the vocal tract, the crucial difference is in the formant transition from the vowel to the consonant.
i didn’t learn how to determine the formants of the consonants, although i did learn that of vowels. if you find charts or graphs of formant frequencies concerning vowels, do not conflate that with consonants!
basically, for all stops, the first formant (F1) decreases approaching the stop. the second and third formants are the most important in differentiating the stops. for p, F2 and F3 also decrease as the vowel approaches the stop. for k, F2 increases while F3 decreases, creating a shape like a pinch (it’s called a velar pinch).
note that voiced stops work the same here regarding formants.
now take a look at how the formants behave in the spectrogram.
F1, F2, and F3 all decrease as they approach the relevant border. the phoneme they are transitioning into is, without a doubt, p. mark raps “girl, you know you got that wap.”
bonus evidence
what’s really really fun about this line is even if you disregard How Formants Should Behave In Theory, all you need to do is compare the sound directly next to “wap”. the next line in mark’s verse is “발을 같이 딛고 pivot,” and it just so happens that 발 'pal' also contains that same p sound.
note: the a in 발 and the a in wap are actually different. there is flexibility in the IPA, and all you need to know is sometimes people (me) are too lazy to use an IPA latter that i can’t type in my english keyboard. it doesn’t matter but yeah
looking at what we know for a fact is a p sound gives us a direct example of how the formant transition into a p looks like in practice. since 발 starts with the p instead of ending with it, the formant transition is reversed, or mirrored. F1, F2 and F3 should increase with time.
see if you can figure out yourself where the phonemic boundaries are and where the formant transitions are.
here are the phonemic boundaries.
here are the formant transitions. they are, in fact, mirrored, demonstrating once again that mark did rap “wap”.
thank you for your time.
more info on the international phonetic alphabet, the universal system for talking about phonemes in languages
linguists invented an standardized alphabet for talking about phonemes, since orthography changes drastically within and between languages. within, say, english, one letter can signify multiple phonemes (see “t” in cat and then “t” in turn and then “t” in bitter) and different letters can signify the same phoneme (see “j” and “dg” in judge) and multiple letters can signify one phoneme (see “dg” again in judge... is one sound innit). i need not elaborate about difficulties between languages.
the ipa is extremely useful to this day, but it is problematic in that it is obviously based on latin script and its structure lends much better to languages without tones, for example... english... and french... yknow... loll
from here. this is the full IPA chart listing out all the symbols and diacritics for talking about sound. here are more resources on the ipa and this is a particularly useful tool where you can click on the sounds on the chart and listen to voice recordings of them.
Tags:
