+- The Voynich Ninja (https://www.voynich.ninja)
+-- Forum: Voynich Research (https://www.voynich.ninja/forum-27.html)
+--- Forum: Analysis of the text (https://www.voynich.ninja/forum-41.html)
+--- Thread: alternating patterns (/thread-2357.html)
alternating patterns - MarcoP - 29-03-2018
Here I discuss a few patterns of the type X Y X, in which the same word X occurs twice with another word Y occurring between the repetitions of X.
I compare with data from a few European languages, but of course I don't know what these pattern might mean in the VMS and I am not sure the comparisons presented here are significant.
1. Alternating repetition X Y X Z
I considered occurrences of the pattern "X Y X Z" where X,Y and Z are three different words. Basically, I am looking for the same word repeating in alternation with other content.
Examples:
Grief dallied with [nor law nor limit] knows.
Now stole upon [the time the dead] of night
And instances [of infinite of] love
A particular case which I initially excluded are matches of the type "X Y X Y"v (these are discussed below).
I also excluded cases in which Y begins or end with the same character as X, in order to avoid considering cases of X X' X that could be related with X X X (and again with reduplication).
The histograms are based on text samples of about 38,000 words each. Words are sorted by decreasing number of matches to the alternating pattern described above (the red bars). The green bars correspond to the expected number of matches on the basis of word frequencies and considering a totally random order.
These two graphs correspond to Italian (Dante "Divina Commedia") and English (the Genesis in King James' version). In this two samples, the most common word also is the one that most conforms to the pattern: the conjunction "e" / "and". The matches are slightly less than what would be expected with a random order: i.e. the pattern is averagely "avoided", most occurrences of the conjunction do not appear in this configuration.
Examples:
Isaac gave up the ghost [and died and was] gathered unto his people
and he turned himself about from them [and wept and returned] to them again
they lifted up their eyes [and looked and behold] a company of Ishmeelites came
The following graphs are for two Latin scientific texts: the classical Pliny and the early modern Mattioli. The conjunction "et" is still the most frequent match: this time the matches are more frequent than expected. Please note that Mattioli (but not Pliny) makes use of several words that seem to be almost exclusively used in the XYXZ pattern ("tunc", "nec", "aut" being the most frequent). These roughly correspond to the English "either..or" ("tunc" and "aut") and "neither..nor" ("nec").
"non" behaves like "et", but obviously is less frequent.
The difference between Mattioli and Pliny makes clear that some phenomena are more related to personal style than to language.
The graphs are based on the VMS: the whole ms on the left and Quire13 only on the right.
The graph for the whole VMS is very different from those of the other sets. A first difference is that the scale is smaller: the most common word ("daiin") is considerably less frequent than the conjunction in the language I examined. Anyway, absolute frequency apart, "daiin" seems to behave similarly to the conjunction in English and Italian.
What I found particularly interesting and unexpected is that the word that most frequently matches the pattern is not a very frequent word: "qokain". These are the 9 matches included in the graph:
sain ol keeshy qokain dy olshedy You are not allowed to view links. Register or Login to view. aly
qokain sheety qokain dar dar shedy qokar ol dy
qokain olshey qokain dar ol shedy qokedy qokeedy qokar olshedy oldy
sor olky qoty ty tor cheyky totol opchedy qokain sheky qokain ol
ycheeytal checthy qokain qokain checthy qokain ol
pol olor chey qokain shedy qokain olchesy ol r aindar
oraiin sheor qokain cheody qokain otal okaiin olkeeor ar al oldy
qokain sheol qokain chckhey lchedy okar al qotal shedy otain far aiin am
qokaiin cheal tain qokain shey qokain char shcthey qoky chy qokaiin
A possibility is that this word is a You are not allowed to view links. Register or Login to view. similar to the Latin "tunc" / "nec". The occurrences of "qokain" are mostly concentrated in Q13 and Q20, but more in Q13. The corresponding graph highlights that the occurrences of the "alternating pattern" are more than expected even if one considers Q13 only. The possibility that the matches are accidental are limited, because the number of matches is consistently higher than expected: even considering Q13 only, observed occurrences almost double the expected number. Of course it is not possible to be sure that this is just coincidental, but occurrences of the pattern in other languages appear to be meaningful.
It is worth pointing out that the similar word "qotain" has two matches of the pattern, while the other similar word "qokaiin" has a single match.
"qotain" is about 1/4 as frequent as "qokain", so 2 vs 9 matches are consistent.
On the other hand, "qokaiin" is about as frequent as "qokain", so the difference 1 to 9 appears to be significant.
2. Chiasmus X Y Y X
Analysing the "qokain" matches above, I noticed another pattern that we could call "chiasmus":
X Y Y X
An occurrence of this pattern appears in King James (1 Chronicles 29:11):
all that is in the heaven and in the earth [is thine; thine is] the kingdom
In the VMS, I have found these 6 occurrences:
qotcheaiin dchain [cthain daiin daiin cthain] qotaiin
qotchaiin [shor cthol cthol shor] chotchy tchodar
[shol chol chol shol] ctoiin sos odan
You are not allowed to view links. Register or Login to view. qokain dar shey lshcthy okar okain ylaiin y
ycheeytal [checthy qokain qokain checthy qokain] ol
[qokol chol qokol qokol chol] cheey or aiin oldal
Three of the occurrences actually form five-words sequences of the form
X Y Y X Y
or
X Y X X Y
3. Word-couple repetition X Y X Y
Finally, I checked the X Y X Y sequences I had initially excluded. I believe this can be seen as a variant of the consecutive repetition of the same word, so its structure is (X Y)(X Y).
I found 6 of these as well.
sair cheain cphol dar You are not allowed to view links. Register or Login to view. daikam
coy choiin sho [s chy s chy] tor ols
qotchor [cheor chey cheor chey] soiin sor daiin
sal shedy qokain shey qoin [ol shey ol shey] qoky qol cheey chl or sheolo
olshees ol sheckhy qokain ol [chedy qol chedy qol] keey [qolchedy] chealy
(in the last example, note that "qol chedy" also appears as "qolchedy")
English examples:
Why, this would make [a man a man] of salt
and when he caught it he let it go again, and after it again, [and over and over] he comes, and up again, catch'd it again
But now her cheek was pale, [and by and by] It flash’d forth fire, as lightning from the sky.
All these patterns appear related to the well-know phenomenon of word repetition, but they are also different because they involve more than a single word-type. They could be the longest structured fragments easily discernible in the ms (five-words sequences in particular) .
RE: alternating patterns - -JKP- - 29-03-2018
Quote:MarcoP: I compare with data from a few European languages, but of course I don't know what these pattern might mean in the VMS and I am not sure the comparisons presented here are significant.
Good idea, Marco. There's no way to know if they might be significant unless one tries.
RE: alternating patterns - Koen G - 29-03-2018
Thanks, Marco, again some interesting data.
I find it hard to consider this without thinking about the apparent lack of function words at the same time, or at least their low absolute numbers. It seems like the examples you found from European languages always contain a mixture of lexical and non-lexical words. But can we say that something like that exists in Voynichese if we can't find decent candidates for words like "and, or, if, in, not...."?
Perhaps we might try to find samples of other languages and see if the absolute frequencies of the words in their repeating patterns are more like those found in Voynichese? Just brainstorming a bit.
RE: alternating patterns - Wladimir D - 29-03-2018
templates
XYXX
shol shot shol shol You are not allowed to view links. Register or Login to view.
shol aiin shol shol You are not allowed to view links. Register or Login to view.
XXYX
daiin daiin ctho daiin You are not allowed to view links. Register or Login to view.
ol {ol ol aiin ol} You are not allowed to view links. Register or Login to view.
the word sequence is suitable for both variants, qokedy qokedy dal qokedy qokedy You are not allowed to view links. Register or Login to view.
RE: alternating patterns - Anton - 29-03-2018
It looks like that champions of "alternate repetitions" are (at least judging by the King's Bible) function words. However, the problem with this interpretation is that the Voynichese champions of "alternate repetitions" are, at the same time, featured in sequential repetitions. And, while qokain is featured in a sequential repetition just once (f77v line 30), daiin is the champion of sequential repetitions, being featured in 21 repetitions of that sort.
More on sequential repetitions here: You are not allowed to view links. Register or Login to view.
What function word would be naturally repeated with itself? I can think only of repetitions across the border of a sentence, like: "Where do they come from? From the forest they come."
RE: alternating patterns - Koen G - 30-03-2018
He knew that that man was trouble.
The repetition is only in form, but that's what we're talking about I guess.
RE: alternating patterns - MarcoP - 30-03-2018
(29-03-2018, 11:37 PM)Anton Wrote: You are not allowed to view links. Register or Login to view.It looks like that champions of "alternate repetitions" are (at least judging by the King's Bible) function words. However, the problem with this interpretation is that the Voynichese champions of "alternate repetitions" are, at the same time, featured in sequential repetitions. And, while qokain is featured in a sequential repetition just once (f77v line 30), daiin is the champion of sequential repetitions, being featured in 21 repetitions of that sort.
More on sequential repetitions here: You are not allowed to view links. Register or Login to view.
What function word would be naturally repeated with itself? I can think only of repetitions across the border of a sentence, like: "Where do they come from? From the forest they come."
Thank you, Anton!
I agree that these alternating patterns are still connected with reduplication (see also Wladimir's examples).
The case of "qokain" is somehow different, because "alternation" is markedly more frequent than reduplication, but even in this case, reduplication cannot be ignored.
You ask: What function word would be naturally repeated with itself?
This depends on the language. In Italian, the first example I can think of is "così così", which corresponds to the English "so so" (I am not sure the English is acceptable, in the written or spoken language). But I believe the three languages I discussed above (Latin, Italian and English) can only show that these phenomena are very variable, also within the same language. However extended this variability, evidence does not support the idea of a close relationship between Voynichese and any of these languages. I believe we must search for parallels elsewhere.
You are not allowed to view links. Register or Login to view. (originally on You are not allowed to view links. Register or Login to view.) that Romani / Romany reduplicates its conjunction "taj / thaj" (apparently, one of the most frequent words in that language).
Another example appears You are not allowed to view links. Register or Login to view.:
Me meresch te genel taj taj pisinel kamahi,
I wanted to learn to read and write so much, (google translated from German)
"genel" means "reading", "pisinel" means "writing"
Possibly, the intensifier "so much" corresponds to the Romani reduplication.
I am sure Romani cannot be the only language that reduplicates its conjunction. More research is definitely needed.
RE: alternating patterns - MarcoP - 30-03-2018
(29-03-2018, 08:04 PM)Koen Gh. Wrote: You are not allowed to view links. Register or Login to view.Thanks, Marco, again some interesting data.
I find it hard to consider this without thinking about the apparent lack of function words at the same time, or at least their low absolute numbers. It seems like the examples you found from European languages always contain a mixture of lexical and non-lexical words. But can we say that something like that exists in Voynichese if we can't find decent candidates for words like "and, or, if, in, not...."?
Perhaps we might try to find samples of other languages and see if the absolute frequencies of the words in their repeating patterns are more like those found in Voynichese? Just brainstorming a bit.
Thank you, Koen!
Hunting for function words is difficult. If one considers short and frequent words as candidates (I know this is simplistic) "daiin" doesn't look like a poor candidate for "and". In the VMS it occurs about 850 times. In the Mattioli and Pliny samples of comparable length I considered, "et" occurs 1202 and 1150 times respectively. I don't think the difference is enough to conclude that "daiin" cannot be a conjunction; on the contrary, it could be seen as consistent with this well-known hypothesis.
There are computational methods that could identify words that behave similarly on the basis of adjacent words (e.g. You are not allowed to view links. Register or Login to view.). But my impression is that, to be reliable, these require a much larger corpus than the VMS. I am thinking of trying a mix of computational and "manual" analyses, but the first step could be identifying simple patterns like those described here and see if there is anything we can infer from these structures.
Of course, there is the possibility that in Voynichese function words are replaced by affixes. We know that some Latin scribes systematically "merged" a particle (conjunction or preposition) with the following word, so this could be a feature of the language or of the script. The evidence that Emma May Smith discussed You are not allowed to view links. Register or Login to view. suggests that the most obvious candidate q- doesn't behave as a grammatical prefix would, but also in this area much work can still be done.
RE: alternating patterns - doranchak - 30-03-2018
(29-03-2018, 11:37 PM)Anton Wrote: You are not allowed to view links. Register or Login to view.What function word would be naturally repeated with itself? I can think only of repetitions across the border of a sentence, like: "Where do they come from? From the forest they come."
I scanned a sample corpus and found some examples.
It often happens around hyphens:
"He was alone—alone with these countless, these unheard-of treasures!"
"...sailing with a fair breeze, south-south-west after a week’s calm..."
"Oh, indeed—indeed, sir, he is innocent!"
Similar repetitions for emphasis:
"...gone on for a long long time"
"indeed, my dear dear Handel"
"a great many many years"
Interesting example involving homonyms:
"the will will not bear scrutiny, and is declared null and void"
Sounds:
"and 'boom boom' came the sounds"
"'Trakh ta ta takh' came the frequent crackle of musketry"
"'dam dam' rattled the drums"
"'Ha ha ha' laughed Pierre"
Other examples:
"...he had had a great many Frenchmen in his house"
"I assure you you are mistaken."
"that her being there may teach her her own insignificance."
"All it is is 2 coats"
"Things as they are are things as perceived by your mind"
"We'll split fifty fifty."
"Rainsford had dug himself in in France"
"The policemen trooped in in force"
"Javert had not taken the prisoner prisoner"
"What is is natural"
"but there are very few who do do their best"
"For they that sleep sleep in the night"
"and every man that offered offered an offering of gold"
"they of Israel were a thousand thousand and an hundred thousand men"
"To have respect of persons is not good: for for a piece of bread that man will transgress"
"do I purpose according to the flesh, that with me there should be yea yea, and nay nay?"
"As well might poison poison!"
"Thy own wish wish I thee in every place"
"Dost thou teach pardon pardon to destroy?"
"Most lamentable day, most woful day. That ever ever I did yet behold!"
"very wittily said to niece of King Gorboduc 'That that is is'"
"But what remains remains for singers"
"but here and there there were dark underminings"
"His right hand once more more slowly went over his brow and hair"
"Love loves to love love"
"Yum yum!"
"praising on both sides what what neither of us can help"
Some I thought of:
"His father's father's brother."
"The ball that rolls rolls swiftly."
"It is a very plum plum" (I recalled that from the English Patient)
"If you see it it will see you."
Then there are of course the novelty examples of linguistic ambiguities:
You are not allowed to view links. Register or Login to view.
"Rose rose to put rose roes on her rows of roses"
"[font=sans-serif]If police police..."[/font]
"Between 'and' and 'and'"
RE: alternating patterns - Anton - 30-03-2018
Hi doranchack,
I was talking about function words specifically, so many of your examples would be irrelevant in that context. Also, one should consider that we are looking not simply for a repetition, but for a more or less frequent repetition, for a repetition that would occur twenty times in the text (btw, I partly checked and Takahashi transcription introduces some wrong repetitions, e.g. in one place it claims daiin daiin when it is actually daiio daiin, but anyway, daiin daiin is not something which occurs only once or twice). In this view, I like your example of "for for", that's a construction which would be natural in a treatise of the kind.
Also, I think that in German there is the construct "das dass", and if I remember correctly, I have seen it used in old manuscripts as "das das".
So, in short, yes: sequential repetition of function words is possible, and it does not even have to be something artificial.
Regarding Marco's suggestion that daiin is a candidate for "and" - I disagree. According to Job's tool, there are eleven folios ending with daiin, and I can't imagine a functional block of text ending with "and", unless the language has some weird positional rule for placing "and" after the words that it connects.
There is a simple test for function-word-suspects. We just need to check if they occur as labels. If they do, that would mean they are unlikely to be function words. If they don't, then this would reinforce their position as function-word-candidates.
One additional idea is that repetitive vords may be function words (thus fitting for a pattern like XYXY), but, at the same time, their sequential repetitions comprise one word instead of two. For example, in Latin dictionary, we have that sort of things under the letter Q: "quo" is a word, but "quoquo" is as well. Same thing for "quis" and "quisquis", "quid" and "quidquid".
Last but not least, one should mind the possibility of shuffling. Indeed, there are indications that the text might have been put down in several passes, which, among other things, suggests shuffling. In that case, all discussions of vord repetitions are pointless
The fact that there are three-vord repetitions there, and even one four-vord repetition, slightly moves the balance onto the side of shuffling. Although in the thread above quoted I imagined some examples that illustrate potential cases of such repetitions.