My last post described the troubles with mapping lexical distance in 2D and what issues I encountered when drawing this diagram. Playing around this weekend I came up with a proposal how to eliminate some of those inaccuracies by adding a 3rd dimension to the lexical diagram. I just want to be clear at the beginning of this post, I am no linguist, I have not studied languages. These are my thoughts of how to improve Prof. Konstantin Tishchenko original work, with whom I have no correspondence.
I neatly inserted four curves from Scottish Gaelic, Breton and Welsh to Gaelic (Irish). If I wanted to have straight lines between them then I could go into a third dimension and move Gaelic up (so it would jump out of your screen to you). That way those three curved lines could be straight and have the length of 30, 64 and 65.
Another Example between whole language branches, if the lexical distance between Romanian and Bulgarian were researched and positioned at about 90 (I estimate that to be about correct) then my version would have to pull the whole Slavic and Romance branch further together. Greek and Albanian would be squeezed in between those two branches and their distances would have to have those Tyshchenko lengthening curves. Or, one adds a third dimension placing Greek and Albanian further below.
Third Dimension is?
What unit does this third dimension have? As what could you interpret it? The map itself is a 2D surface covering Europe. The third dimension is not mountain heights. I thought about it and the best explanation I could think of was time.The time since the height and diminish of that language. Most Languages were influenced by prior, extinct or diminished languages that were in that location before. The longer that a language is in a location the more it influences it’s surroundings. The further in the past since the language was there the more it’s influence fades. When I drew the branch ellipses around the language bubbles in Lexical Distance Among Languages of Europe 2015 then I already tried to show that exemplī grātiā the Romance language French is influenced by Germanic and Celtic. That is why the Celtic ellipse fades off towards Germanic and Romance.The Celtic language also used to span all the way until Asia Minor. Hellenic, Albanian and Anatolian fades in all directions.
Lexical Distance Diagram in 3D
If mapped out with language branches in 3D, then Romance, Slavic and Germanic could be placed in a triangle in the present time. Celtic and Hellenic would be at minus 2000 to 1500 years since their blossom. Albanian and Basque would be 2500 and Illyrian and Anatolian would be even further back in time since their strongest period. The lexical distances between the big three triangle would be about 56 from Germanic to Romance, about 80 from Slavic to Germanic and about 90 from Romance to Slavic. Then the distance to Hellenic bellow and in the centre of this triangle could be drawn without any curves but with straight lines.
This is my scetch of this idea and some notes to it.
First, each language branch is shown as a sphere. The volume of the sphere corresponds with the speakers in Europe and Asia Minor at present. I was considering to scaling the spheres to the largest extent of the language branch in the past, so Celtic would span from the British Isles to Anatolia, Hellenic from Iberia to Pontica, Oghuz from the Balkans to Siberia. But that proved impractical with spheres.
Second, to show the time until the height of the language and help with orientation and 2D placement, I drew lines from the centre of each sphere down to a map of Europe.
Third, you can find this map at commons.wikipedia.org. It was original drawn by Michael Postman to depict languages in Europe as of 1990 and has since then been adjusted several times by the Wikipedia community and shows some inconsistent but beautiful mix of 1990 to 2015. There is another version from 1923. This map is biased towards minority languages and German.
Fourth, I changed the colours of my diagram to correspond with the map.
Fifth, Armenian, Oghuz (that includes Turkish) and Kipchak (inlcudes Kazakh) are also shown but lexical distances not drawn in. I would have study Tishchenko earlier diagrams to figure out those. Illyrian and Anatolian have grave stones and Caucasian, Kartvelian, Indo-Iranian and Semetic languages are missing.
Sixth, a scale from 2000 BCE until 2000 CE is in the back pane and the axes are labelled.
Finally, even with this third dimension there will still be lexical distances that need to be stretched or shorten, but allot less than in 2D.