Thread where I drop thoughts

[Hallfiry]

In this thread I'll post whatever semi-scientific comes to mind.

Today: I've thought a bit about motion vectors and videos with low framerate. I figured you can ramp up the framerate by interpolating the image between two frames by applying the motion vectors from both directions halfway long.
I did a quick google search, gave up after one try, and decided to implement something myself.
Here's my result on the Daggerfall trailer (just two frames and a frame between):
http://kultcds.com/Upload/comparison.gif

EDIT:
It was hinted to me that my frame interpolation method requires a more down to the ground explanation, so here it is:

Two consecutive video frames usually don't differ much. Most what happens, usually is movement. Therefor it's possible to assign a vector to each block of the video /typically 16x16 pixels) that says how much it shifts in x- and y-direction between the frames. These vectors have to be estimated, as there's a lot of noise and also colors change due to shadows, distortion, rotation, blurring, etc.

If we have a frame A and a frame B, we can find the motion vectors from frame A to B and from B to A (both ways, so we can compensate for areas that were for example overlapped in A, but visible in B and the other way around).

Now, we can use that information and render a new frame, by shifting all blocks from frame A by half their motion vector and all blocks from B by half their motion vector. The two resulting images need to be blended over eachother afterwards. Then, a few holes might remain, where blocks were moved and left empty spaces behind. Those need to be filled, and I arbitrarily chose to interpolate them by taking the average pixel color value of the correpsonding position of frame A and B.

[MrFlibble]

So basically you're developing an algorithm to make playback of videos smoother? Sounds cool

[Hallfiry]

So, I was on 4chan /vr/ the other day and saw a thread about the pronounciation of SEGA. But what really caught my attention was this post:

We call "beat 'em ups" belt scrollers. But I personally prefer to refer to them as a strolling donnybrooks.

I found it kinda cute

[Hallfiry]

While the german Rogues' Cant "Rotwelsch" has become quite a hobby of mine, I've noticed quite a remarkable phenomenon around it (Which has in part been seen by others aswell, including the Rotwelsch dictionary author Wolf). The oddity is folk etymology at its finest. There are (among others) two types of Rotwelsch words (in my mindset):
Genuines and Phonies.
Genuines are words like Bajes, Görgel, Mesuse, kochem, Märtine and Mockum.
Phonies are words like (es ist) Essig, (guten) Rutsch, Saures, Pleitegeier and blau machen.

While genuine words are easy to spot as foreign, the phonies are a very dangerous plague. They act as if they were common german words, while in fact they aren't. The sad thing is that they trick even etymologists and lead to false assumptions. SInce yiddistics and romanistics have no place in nowadays germanistics, you will hardly find any good dictionary that will tell you that "blau machen" (take a day off from work) has nothing to do with coloring pants in blue (for which youd take a day off to wait for them to soak in the color), bu instead belongs to yiddish b'lau ("nothing") and thus means "do nothing". In the same way, "Süßes sonst gibts Saures" (trick or treat) doesn't refer to Saures as something sour, but instead to yiddish Zoros or Zores, animosity. Pleitegeier has nothing to do with a vulture (ger. Geier), but is the yiddish pronounciation of german Geher (goer, I'll skip the etymology of Pleite here), and "guten Rutsch" (happy new year) has nothing to do with "rutschen" (slide), but instead with yidd. Rosch (head, beginning) and refers to the first day of the year. "Es ist Essig" (literllay, "it's vinegar", meaning "it's lost"), is from yiddish hessik, "loss".

I've seen many weird stories on the internet trying to explain phrases like "jemandem eine Lampe bauen" (literally "build a lamp for someone", meaning "tell the police about someone's crimes"), where they try to come up with things like "criminals had a lamp drawn with chalk on their back to make them stand out", or "lamp stands for the enlighting of the police", which is all BS, since yiddish Lamdom is "wisdom" and in rotwelsch usually refers to the knowledge of the police about criminal activities.

[MrFlibble]

Considering the secret nature of Rotwelsch, may I ask, do you know if the words that you have so aptly called "phonies" were

1) intentionally "corrupted" by the speakers themselves, or
2) unintentionally corrupted by speakers who were German and did not realise that the words are Yiddish, or
3) misunderstood by outsiders (including lexicographers)?

Or did all three possibilities take place, to some extent? Personally I'd assume 2 & 3, as it seems quite natural for speakers to interpret words they do not understand (etymologically) by making them similar to those they know. Baudouin de Courtenay used a very interesting metaphor for this process, likening foreign words to celestial bodies that become pulled into the gravity field of larger native words in a language.

[Hallfiry]

Considering the secret nature of Rotwelsch, may I ask, do you know if the words that you have so aptly called "phonies" were

1) intentionally "corrupted" by the speakers themselves, or
2) unintentionally corrupted by speakers who were German and did not realise that the words are Yiddish, or
3) misunderstood by outsiders (including lexicographers)?

Or did all three possibilities take place, to some extent? Personally I'd assume 2 & 3, as it seems quite natural for speakers to interpret words they do not understand (etymologically) by making them similar to those they know. Baudouin de Courtenay used a very interesting metaphor for this process, likening foreign words to celestial bodies that become pulled into the gravity field of larger native words in a language.

1) and 2) are the typical ones. One pretty common theme is turning those names into pseudo-placenames. There are actually some themes by which those names are created. For example the verb "bergen" in the sense of "to house, to house, to salvage" is muddled up with "Burg" (castle, a common placename part in german, e.g. Hamburg). Examples:
yidd. cholaz, "he has taked off (his shoes)" + bergen -> Charlottenburger, "a travel's bundle" (where he keeps his shoes with leather soles, so they don't get worn down on the roads)
Habe (i.e. belongings) + bergen -> Habsburger, "a Cheater" (he salvages the belongings of the cheated)
Straße + bergen -> Straßburger, "a cigarette butt, gathered up from the street" (salvaged from the street)
Wonnen (delights) + bergen -> Wunnenberg, Winnenberch, "pretty, young girl" (she houses delights, probably also a reference to breasts as mountains, "Berge")

Another theme is just attaching "-heimer" to a word to mark where they belong, also being suggestive of made up brands:
Pappenheimer -> boots, romani pàf "foot"
Leitungsheimer -> tap water (german Leitungswasser, literaly pipe/"conduction" water)
(these are actually the only two examples I know...)

German itself has the problem of misunderstanding its own words, for example "Moltwurf" became "Maulwurf" over time. In english there's a common theme of dropping initial "n" when mistaking it for part of the article: a nadder -> an adder.

[Hallfiry]

Indeed there are often misreadings or misprints (most common, n is misprinted as u and the other way round). However making words sound as if they were something familiar (as a placename) is a pretty effective strategy for making the words stable (they then have a fixed form and don't evolve further) and makes them incomprehensible to outsiders.

The problem of phonies goes even a bit further: Words like Affe, either from yidd. Ophe (baker) or Aph (anger) have been muddled up so hard with german Affe (ape), that they can't be seperated from it anymore. One that comes to mind is the term Flak-Affe (which is even an achievement in UT 2004, iirc), which (and I'm now making this up from my mind) might refer to flak helpers in ww2, who would load the cannons, which 1:1 resambles a baker shoving a spade-load of bread loafs into the oven chamber, so a Flak-Affe, might actually be a Flak-Baker. Additionally, I remember that the shoving of bread into the oven is referred to as "Schießen", literally "shooting" (because you'd dash the spade forward and then suddenly pull it back to have the loafs shoot off from it by their inertia), and a young baker's apprentice (I'd say something like 12 years old in that case) is also called Schütz (though it's derived from yiddish "Scheeks", boy), which would by folk etymology also get linked to German "Schütz" (shooter, or Schützling, protegé), strengthening my case even more. (All of this assuming the term Flak-Affe is old enough)

Which leads us to another problem of German folk etymology: schützen and schießen. Schützen means protect, schießen means shoot.
schießen has two derived noun forms:
Schuss, shot
Schütz, shooter (most famously in "Geschütz", cannon)

schützen has one noun associated with it:
Schutz (protection)

Now you can easily see, how German folk etymology gets things like Schützenverein (~historic shooting club) interpreted as something like a club of protectors or a militia.

[MrFlibble]

I love that stuff about Flak-Affe, thanks for the interesting insights in German etymology (and keep 'em coming! )!

[Hallfiry]

I love that stuff about Flak-Affe, thanks for the interesting insights in German etymology (and keep 'em coming! )!

I'm glad you enjoyed this improvised etymology thing.

Here's another one. Siegmund Andreas Wolf's Rotwelsch dictionary briefly scratches on a few number systems. One is easily based on hebrew numbers (though a bit disfigured). The other two are more interesting, as Wolf didn't write down any etymology for them:

Henese-Flick:
*1 ehn
2 parz
3 troms
4 nothringskes
5 holf krütskes
6 spörkes
7 spörkes un ehn
8 spörkes un parz
9 spörkes un troms
10 krütskes
11 krütskes un ehn
...
20 parz krütskes
...
50 holf uhr
...
100 uhr


Tiöttensprache:
1 êne
2 bêde
3 droimes
4 snêmans
5 mans
6 halv brüwel
...
12 brüwel
...
100 null(s)
...
1000 trant nulls

Some of these are easy to decrypt, ehn and êne are german "ein" (one), parz is probably related to "paar" (pair, thus two), bêde is "beide" (both), troms and droimes are derived from drei (three) or probably dreimaß (three measures). Now things get more difficult:
mans, probably french main or italian mano, meaning hand, thus five fingers. snêmans has the rotwelsch sneps as its first part, which means small or tiny. Wolf suggests it belongs to dutch snipper, "Schnitzel", which would put it in a family with snips (a pair of scissors), snip (to cut), and German slang "Schniepel", penis/small boy, aswell as schnipseln (cut to small pieces) and Schnipsel (small piece).
nothringskes is a riddle to me. -kes is a diminutive suffix here, but I have no idea what a Nothring is supposed to be.
next is krütskes and holf krütskes. krüz is allemanic for Kreuz, cross, -kes is a diminutive, so krütskes is a little cross, which neatly corresponds the roman numeral X for ten. holf krütskes is half a cross, but I can't decide if it refers to half the amount as a cross, or halv the symbol (as the roman V for five is the top half of X for ten).
spörkes is again a riddle to me, but it might be related to German Spur (trace), that also appeared in the main form Spor in old days. That would suggest that it's an old measurement unit, however it's unclear to me wheater "Spur" would be a length (e.g. a foot being 6x2inch, a fathom (ger. Klafter or Faden) being 6 foot, etc.) or a custom measure like a track of cloth or a uniform size track along a field (which results in a predictable amount of crops, hay, straw, etc.).
However, krütskes being a symbol suggests that spörkes might be a symbol, too, sadly I can't come up with a plausible one.
spörkes un ehn, parz, droimes are just additions, with "un" for german "und", and.
brüwel and halv brüwel are again mysterious. Since I've no better idea, I'll suggest brüwel is from something like italian brevillo (made that one up myself), meaning a "brief measure", as compared to a gros, which is a great measure.
null(s) might just refer to the 0s that come with the number (maybe it had a special symbol like a crossed 0?)
trant nulls is a big hundred, as trant means big or fat in rotwelsch. My suggestion for trant is that it's from something like french grand, which would make it a checkable hypothesis, as the same sound shift (initial gr to tr) would appear in other words, too.
uhr for 100 is again strange. German Uhr means clock, but I doubt that's the root for this one.

EDIT: Looked up the UT 2004 achievements and infact, the english name for Flak-Affe is FLAK MONKEY, which might indicate that the german word doesn't exist, but could also mean that the english word is a loan translation from german.
EDIT2: But even if it was all wrong... It was still kinda cute

[MrFlibble]

Sorry for a belated reply, I was busy at work.

mans, probably french main or italian mano, meaning hand, thus five fingers. snêmans has the rotwelsch sneps as its first part emphasis added

Is this something suggested by the author or your own idea? It's not like such simplification of a consonant cluster (*sneps-mans -> snêmans) is impossible, but it's good to have some other verifiable examples of the same to back up such etymology.

BTW, what kind of sound is [ê]?

[Hallfiry]

Sorry for a belated reply, I was busy at work.

mans, probably french main or italian mano, meaning hand, thus five fingers. snêmans has the rotwelsch sneps as its first part emphasis added

Is this something suggested by the author or your own idea? It's not like such simplification of a consonant cluster (*sneps-mans -> snêmans) is impossible, but it's good to have some other verifiable examples of the same to back up such etymology.

I suggested it and it (the word "mans") is indeed comparible to something from the same region: Deetz (head) from french tête.
I think the -s at the end of sneps only serves for masculine. Since hand is female that would be snepemans -> snêmans then. So the p is the only letter that needs to degrade now.

Oh and sneps is the typical word for small in that particlular set of rotwelsch sources, it's not a generally known word!

BTW, what kind of sound is [ê]?

ê means long e, usually.

[Hallfiry]

Actullly it is probably useful to condense my speculations above and write down how confident I am with them:

Henese-Flick:
*1 ehn -> german ein-, confident
2 parz -> german paar, french pair/paire, quite confident
3 troms -> german drei, french trois confident
4 nothringskes -> unexplained but the source the DWB cites on "kreuz" (a marking for the number ten in merchant slang) also mentions an unexplained "ring"
5 holf krütskes -> allemanic german half Krüts (Halb Kreuz), confident
6 spörkes -> unexplained
7 spörkes un ehn -> german und confident
8 spörkes un parz -> german und confident
9 spörkes un troms -> german und confident
10 krütskes -> allemanic german Krüts (=Kreuz) confident
11 krütskes un ehn -> X und Y confident
...
20 parz krütskes -> compositum confident
...
50 holf uhr -> compositum confident
...
100 uhr -> unexplained


Tiöttensprache:
1 êne -> german ein-, confident
2 bêde -> german beide, confident
3 droimes -> german drei, dreimaß, french trois confident (this an troms might be some important clues to german rendering of french words with sounds german lacks, in this case the /wa/ sound of /trwa/ or */trwas/)
4 snêmans -> sneps/snepsch (-> *snep-ig) + french main quite confident
5 mans -> french main quite confident
6 halv brüwel -> german halb + Brüwel confident
...
12 brüwel -> unexplained (but probably something french with the meaning of "bundle" or "short measure" (in comparison with a gros or a nulls)) (EDIT: See below)
...
100 null(s) -> unexplained
...
1000 trant nulls -> trant ("great") + nulls; trant: possibly french tant (many) and/or french grande ("great"), or german "trant" ("an upslope") highly uncertain


---


Additonal remarks about the german pronounciation of French words:
French has a set of sounds that German does not have. Untrained speakers will have trouble pronouncing these properly (just like they fail to pronounce the english th and render it as f, v, s, z or german w).
fin and vin (as in ragout fin and coq au vin) become feng and weng. (but only if in a phrase?)
For the french /trwas/ I'd expect a German rendering of /traims/, analog to Reims, /raas/ becoming /raims/.

---

EDIT: I've discovered the french word "bribe" that means something like fragment, clump, piece, etc.. Adding the compulsory diminutive to it we get Bribel in German which would easily wind up as Brüwel. I'm not sure about the exact notion by which it means dozen, but it seems to be close enough to be considered a candidate.

[MrFlibble]

BTW, what kind of sound is [ê]?

ê means long e, usually.

This is the first time I hear of such use for this diacritic. I would expect the notation for the long [e] to be [ē] or, more rarely, [e:].

I've discovered the french word "bribe" that means something like fragment, clump, piece, etc.. Adding the compulsory diminutive to it we get Bribel in German which would easily wind up as Brüwel. I'm not sure about the exact notion by which it means dozen, but it seems to be close enough to be considered a candidate.

Again it would be very nice to have a detailed analysis of how Bribel becomes Brüwel.

I urge you to exercise utmost caution when constructing etymologies. Every single sound correspondence that you establish must be backed up by as much evidence as possible pointing out that such correspondence is systematic across the language. Arguments like "sound X can easily become/get corrupted into sound Y" or "sound X often becomes sound Y" are insufficient, inaccurate and may very well be misleading.

You may already know from your studies that proper notation for sound changes/shifts is

x -> y / P

where x and y are sounds/phonemes in question and P is the position (positions) for which such change holds true. (For everyone's reference, more on this can be found in Wikipedia.)

I feel I have to warn you because amateur etymologists are a real problem for professional linguistics. There are extreme cases like Anatoly Fomenko, but others, while proposing somewhat less outrageous theories (YMMV though), are no better. All of them operate with completely unscientific methods that are equivalent to nothing but pure guesswork.

[I'm not criticising you here, I just want to emphasize how important it is to explicate your inferences and back up your conclusions with solid evidence.]

[Hallfiry]

You are totally right!

So first off: ê for long e is the way it is written in German, especially old high german transcriptions. You'll see it throughout literature in that regard and the DWB is full of it, too.

Next: I simply claimed that an inferred Bribel becomes Brüwel.
Let's start with the inferred noun: french bribe is pronounced /brib/. The i is longish. (too bad I can't say that more exact).
Around the region where Mettingen is, a frankish (or franconian?) dialect is spoken. frankish, like dutch and other dialects of that group did do the shift of
b -> v / V:_
(in english and platt that one went further as b -> v -> f / V:_
same in heben hieve, kleben cleave, sieb sieve,...)

Thus, a Bribel, would sound very unfamiliar to any speaker in that region, which I claim leads to the Bribel being frankified to Briwel. (and yes, I am applying a sound shift retroactively, which is valid when a word is assumed to be of the same language, but a different dialect and speakers render it in their dialect them. E.g. berlin dialect verschroben as bavarian vaschroum)

For the i becoming ü, it seems to be ja a little assimilation to the letters around. The b at the front is a bilabial plosive, the r goes by the flow, the following w is a bilabial approximant. i and ü are both close-front vowels (according to the IPA thingy here) with the only main difference being that ü has the lips shaped like for a bilabial approximant and i has the lips stretched out like grinning.
So my assumption is that

[unrounded, close, front] -> [rounded, close, front] / [bilabial]_[bilabial]


better?

[Hallfiry]

Now that I've run into that problematic issue, I'm getting to think about implementing a little tool that perferms sound changes systematically to words, so it would go like
PIE pod pl podis(?)
pg. fot pl fotiz (?)
ohg. fuoz pl fiouzi(?) (man, that one is so damn close to the bavarian Fuaß pl. Fiaß)
mhg. fusz pl fiusze
nhg. Fuß pl. Füße