- Options
- Login
- Desktop Style
Options
What are the characteristics of Labelese?
- What are the characteristics of Labelese?
-
What are the characteristics of Labelese?
VViews > 25-07-2019, 03:52 PM
Hi everyone,
I'm interested in establishing a list of what the characteristics of the putative "Labelese" are.
You are not allowed to view links. Register or Login to view.by Nick lists the following for the zodiac section:
"
* a disproportionately large number of zodiac labels start with EVA ‘ot’ or ‘ok’.
* words starting EVA ‘yk-‘ are also more common in zodiac labels than elsewhere
* most (but not all) zodiac labels are surprisingly short.
* many – despite their short length – terminate with EVA ‘-y’.
* a good number of zodiac labels occur multiple times.
* almost no zodiac labels start with EVA ‘qo-‘
* in many places, the zodiac labels exhibit a particularly strong ‘paired’ structure (e.g. on the Pisces f70v2 page, otolal = ot-ol-al, otaral, otalar, otalam, dolaram, okaram, etc), far more strongly than elsewhere"
Can this list be considered as a good description of the characteristics of Labelese in general? Are there more to add to the list? -
RE: What are the characteristics of Labelese?
ReneZ > 25-07-2019, 06:39 PM
This is an interesting question. To properly discuss it, one should understand what are 'labels'.
When I reviewed the important transliterations and defined the IVTFF format, I arrived at the following counts:
299 : Zodiac labels
297 : Loose words that are not apparently a label of anything drawn
195 : Labels of plant parts in the pharma section
76 : Star labels in the Astro/Cosmo section
63 : Nymphs in the bio/balneo section
47 : "Tubes and Tubs" in the bio/balneo section
40 : Containers in the pharma section
With other bits and pieces, the total is 1033.
I have looked more at the zodiac labels than all other ones, so my opinion is largely constrained by that.
The most conspicuous property of the zodiac labels is without a doubt that they do *NOT* follow Zipf's law. Even when some of them repeat up to four times (from memory), they are almost like unique words.
This was confirmed again in Koen's recent TTR experimentation. -
RE: What are the characteristics of Labelese?
Koen G > 25-07-2019, 07:16 PM
Correct, I've found only one text so far with a higher TTR than the VM labels: Claretus' glossary. Which, given its text type, has a top TTR by definition.
On a more philosophical note I'd add that it's not problematic for labels to be repeated or consist of more than one word. For example, in bestiaries there can be "stier" and "wilt stier", i.e. bull and wild bull. And more kinds of bull. Those ask for two-word labels and repeated words.
Anyway, it's an interesting topic since one could posit some expectations for labels (Nouns and adjectives? Nominative case?) -
RE: What are the characteristics of Labelese?
VViews > 25-07-2019, 09:31 PM
Thank you both for these replies.
I completely agree with Koen G's point about multi-word labels.
ReneZ, regarding the "297 : Loose words that are not apparently a label of anything drawn", is there a list of these with page numbers available somewhere? I am surprised by this high number. Does that include "titles"?
On the topic of labelese characteristics there is also MarcoP's post You are not allowed to view links. Register or Login to view. with graphs that show the differences in both prefixes and suffixes for labels vs the main text in more detail.
He notes the following:
Prefixes:
[*]"qo-, sh-, ch- are less frequent in labels
[*]the o- prefix is more frequent in labels
[*]the y- prefix is more frequent in labels (but much less clearly so than o-)"
[*]Suffixes:
[*]"-ey,-in are less frequent in labels
[*]-ry, -ly, -sy are much more frequent in labels"
Are there any other characteristics? -
RE: What are the characteristics of Labelese?
ReneZ > 26-07-2019, 07:30 AM
There's an Excel file at my site:
You are not allowed to view links. Register or Login to view.
This has the counts per page (first tab) and per quire (second tab).
Edit: you need to look at the column 'L0'. See also here: You are not allowed to view links. Register or Login to view.
Most of these 'labels' are on the Rosettes page, but also in the various sequences
on f 49v, f 66r etc. -
RE: What are the characteristics of Labelese?
ReneZ > 26-07-2019, 07:59 AM
One other interesting aspect of the Zodiac labels in my opinion is the fact that the words tend to be 'grouped' or clustered. Hard to explain precisly, but similar patterns tend to occur in a few adjacent zodiac signs.
Just to give an example, zodiac labels starting with the mildly unusual 'oeeo' and 'oeeeo' tend to be found in a few signs only. Mostly libra, I believe. -
RE: What are the characteristics of Labelese?
Davidsch > 26-07-2019, 08:55 AM
If labels from a random other book, in the same quantity and size as here in the VMS
have been tested and compared to the found values in Zipf and TTR research,
only then I am willing to accept these findings and conclusions.
Until then I tested and found that the composition of the labels has a textual similar cohesion as the other text.
There are some minor differences and those can be explained from the simple fact that most labels are nouns. -
RE: What are the characteristics of Labelese?
ReneZ > 26-07-2019, 11:20 AM
The MATTR plots clearly show that the values for labels in the Voynich MS are very different from the runnig text, namely far more unique words.
They must be somewhere in one of the threads but I could not find them, so I post them again.
All VMS text:
VMS Labels:
-
RE: What are the characteristics of Labelese?
Koen G > 26-07-2019, 11:39 AM
(26-07-2019, 08:55 AM)Davidsch Wrote: You are not allowed to view links. Register or Login to view.If labels from a random other book, in the same quantity and size as here in the VMS
It would be interesting to compare "labelese" to labels in another text, not only for TTR but also (in the case of a language like Latin) to compare the relative frequency of certain syllables (thinking nominative endings for example). I just can't think of a good candidate. And there are additional complications: even if the VM labels are true labels, we still don't know what kind of information they convey. Star names or star properties or who knows what else?
Maybe a specifically targeted experiment could be to compare the ca. 200 VM small-plant labels to labels in a herbal. I was thinking of the Sloane Tractatus but upon closer inspection it appears that its "labels" are often a full sentence. -
RE: What are the characteristics of Labelese?
bi3mw > 26-07-2019, 11:58 AM
(26-07-2019, 11:20 AM)ReneZ Wrote: You are not allowed to view links. Register or Login to view.The MATTR plots clearly show that the values for labels in the Voynich MS are very different from the runnig text, namely far more unique words.
....
Isn't it natural that labels should consist of a lot of unique words (if they make sense) ? Of course, the author(s) could have "simulated" this property, but I think that's unlikely.
It would be interesting to know if the ratio of unique to non-unique label-words in a comparison text is similar to the one in the VMS.
This list may be helpful:
You are not allowed to view links. Register or Login to view.