I came upon an excellent graphical representation of the linguistic distance between a number of European languages. I'm looking for a similar worldwide map of currently spoken languages. Or at least the raw data. Even if it's not truly worldwide, but just wider than the one presented here, I'd still be interested in it.
4You use the term 'linguistic difference' but the legend on the graphic says 'lexical distance'. These would seem to imply very different things, why the difference in terms?– Gaston ÜmlautApr 22, 2016 at 23:24
3Ok, judging by comments at Etymologikon and LanguageHat it's clear this map does not represent 'linguistic distance' but 'lexical distance'. It's also not clear what list/s it's based on, perhaps Swadesh? It also seems not to distinguish borrowings from inheritances so it simply shows present-day lexical similarity.– Gaston ÜmlautApr 22, 2016 at 23:41
@GastonÜmlaut Both concepts are new to me. Could you explain the diffrrence between the two?– FiksdalApr 23, 2016 at 2:24
I don't know what 'linguistic distance' would be, but from the provided answer it seems its being assumed to be some measure of overall distance between two languages in terms of their lexical inventory, their morphology, syntax, etc. Anyway, the graphic doesn't use the term 'linguistic distance'.– Gaston ÜmlautApr 23, 2016 at 23:58
2While 'lexical distance' is also a term I hadn't heard before, it's fairly clear what it means. The source of that map indicates it's comparing lists of words from two (or more) languages and counting how many are similar forms with similar meanings. If so it includes some cognates (i.e. those whose form and meaning are still similar), borrowings, false cognates (i.e. coincidences), etc so it seems a pretty useless measure altho it might have relevance for language learning. If someone could read Tyshchenko's article (in Russian) they might be able to clear this up.– Gaston ÜmlautApr 24, 2016 at 0:13
@Fiksdal, I am the author of this of this version,
which is based off of Tyschenko's work, see here
Since translating Tyschenko's map, I have spent a lot of time trying to figure out how the original list was made. Not much of Tyschenkos work is online and the best content is in Ukrainian or Russian. I also tried reaching out to the University in Kiev and Linguistics Department there, with no success. And I have limited access to any paper copies of his research.
But I think I have a pretty good idea of how the data was obtained. These maps basically show the Levenshtein distances lexical distance or something similar for a list of common words. Now this list could be the Swadesh № 100 or № 207 list with counting duplicate letter shifts in different words as one LD, or it could be Dolgopolsky № 15 list or a Swadesh–Yakhontov № 35 list and just brutally counting Levenshtein LDs on those lists. Or Tyschenko's could have his own list of words and methods to calculate the lexical distance. In one paper a master matrix is described with all of the lexical distance calculations and that each where calculated, but which exact method and a list of which words is not included.
I experimented with a couple of lists and methods my self and I come fairly close to the original matrix. Working on programming it automatically (if you have same word list from different languages) so it gives you a matrix over several languages and this is a test on Germanic languages.
|En|Sc|Du|Af|Ls|Li|Wf|Sf|Nf|Lu|Ge|Yi|Da|Sw|Fa|Ic|Nb|Nn|Sr |00|13|29|30|30|28|26|26|32|40|33|34|39|37|38|42|38|36|37 English |13|00|33|31|34|31|26|30|33|40|36|39|42|37|42|44|41|42|36 Scotts |29|33|00|07|15|15|28|26|25|35|29|33|38|33|37|37|39|37|41 Dutch |30|31|07|00|17|19|30|25|23|34|31|34|34|31|36|37|35|33|40 Afrikaans |30|34|15|17|00|22|30|24|21|36|24|30|38|34|37|37|37|32|40 Low Saxon |28|31|15|19|22|00|28|31|30|33|28|34|37|34|40|39|38|38|44 Limburgs |26|26|28|30|30|28|00|26|27|39|33|38|42|37|39|43|42|41|40 West Frisian |26|30|26|25|24|31|26|00|29|41|32|32|40|37|43|45|39|36|43 Saterland Frisian |32|33|25|23|21|30|27|29|00|41|36|37|37|36|38|37|37|34|39 North Frisian |40|40|35|34|36|33|39|41|41|00|27|40|45|42|50|51|45|45|51 Luxembourgish |33|36|29|31|24|28|33|32|36|27|00|30|39|34|42|40|38|37|45 German |34|39|33|34|30|34|38|32|37|40|30|00|38|34|36|36|36|36|42 Yiddish |39|42|38|34|38|37|42|40|37|45|39|38|00|19|24|27|04|18|42 Danish |37|37|33|31|34|34|37|37|36|42|34|34|19|00|28|28|20|19|39 Swedish |38|42|37|36|37|40|39|43|38|50|42|36|24|28|00|10|21|20|44 Faroese |42|44|37|37|37|39|43|45|37|51|40|36|27|28|10|00|25|20|43 Icelandic |38|41|39|35|37|38|42|39|37|45|38|36|04|20|21|25|00|15|41 Norwegian (bokmål) |36|42|37|33|32|38|41|36|34|45|37|36|18|19|20|20|15|00|38 Norwegian (nynorsk) |37|36|41|40|40|44|40|43|39|51|45|42|42|39|44|43|41|38|00 Sranan
just to compare Swedish, I get 19 (Tyschenko got 21) to Danish, 28 (26) to Icelandic, 19 (16) to Norwegian (bokmål)
by German I get, 33 (49) to English, 29 (25) to Dutch, 30 (41) to Danish
My method and list does not match up to Tyschenko's but at times I am pretty close. So, yeah a worldwide version of this is possible, have to have the raw data from the languages (word lists) and I have to define a method and work on a program that could analyse the lists (and then go through the results to check them manually)
1thanks, will add the links I wanted to. Work progress updates I will post on the blog. Aug 27, 2016 at 23:23
1Hello @AlternativeTransport and welcome to Linguistics.SE! Great to see you here. Aug 28, 2016 at 4:49
1@AlternativeTransport Very interesting, could you post a link to your code and sources of words? Aug 29, 2016 at 7:07
1To Methods: I don't really think that Levenshtein Distance does lexical languages distances justice. Lets say you determine the lexical distance from PHOTOGRAPH to FOTOGRAF. According to Levenshtein the lexical distance is 4, so 2 times P --> F and 2 times removing H. But you could argue that since it is the same replacement and change it could be 2. Or even only 1 lexical/linguistic distance. A different 10 letter word with 4 different changes is much harder to understand when reading. Or with Umlaut ÜÖÄÅÜ, should it really count as 2 LD, AE --> Ä? Aug 29, 2016 at 17:16
1So, to code and sources of words, got nothing finished yet, but I intend to share it so that hopefully other people from other languages can DIY their own lexical distance. Aug 29, 2016 at 17:26
As noticed in this answer, Prof. Tyshchenko's work primarily targeted languages spoken in Europe, hence, most of them belong to the Indo-European family; the only exceptions being Basque, Finnish, Estonian, and Hungarian (labelled as Madyarska). Even Celtic languages form a mini-group of four.
Image courtesy of. I apologize the's no English version of the map with numbers.
Although I didn't find any direct mention why it does not cover other language families (LF), I think, we may assume that Linguistic Distances (LD) of languages belonging to different LF may be too hard to measure. This is probably caused by too different linguistic tools (mind the vowels, consonants, lexical tones, intonation patterns, and also a whole set of syntactic tools).
If you check the Wikipedia article on mutually intelligible languages, you may notice that most language pairs having short LD do belong the same LF.
Also, note that Prof. Tyshchenko's work (link; sorry for it's in Ukrainian only) is not an independent research. Instead, it's rather a result of long hard work of a big group of linguists who elaborated various methodologies of how to measure the LD. And it resulted only a list of measured LD between the languages spoken in Europe.
So I think that the suggested methodology can't be automatically applied to all world's languages, hence there's probably no worldwide map you're asking for.
Yeah, you would probably need people with an in depth knowledge of each language group to do the research. Since these researchers knew Indo-European languages, they were equipped to work on that. I agree that it's unlikely that something like this exists for worldwide.– FiksdalApr 22, 2016 at 19:09
1I would also glad to see further elaboration of this work. If such a research about worldwide list of languages appear sometime, I would also be very curious to look at it (and stand corrected with my current assumption). Apr 22, 2016 at 19:26
1Yes, it would be very cool if it existed. I'm sure it could be feasible to create it if the right researchers and experts got together.– FiksdalApr 22, 2016 at 19:28
So does Tyshchenko's article really talk about 'linguistic distance'? Or about 'lexical distance'? Apr 24, 2016 at 0:28
@GastonÜmlaut, I think, it's about Linguistic Distance, since it's a property of a pair of Languages. Lexical Distance seems to be a property of a pair of Lexemes. Apr 24, 2016 at 1:43
There is very nice language similarity chart, almost one hundred languages compared..
I can't find info what kind of measure that table is using. Do you know? Apr 23, 2022 at 0:06
This lexical map deals mostly with European languages. Most of them are of Indo-European family, but some belong to the other families like Hungarian, Finnish and Estonian from the Uralic language families. Also, there was the site in the Internet where there was shown some variant of this distance with including several languages from Asian continent like Tajik and Persian. But now it is probably disappeared.
I have my own hypothesis concerning the methods of the lexical comparisons in the map.
It was correctly said by some commentators here that in these lexical distances by prof. Tyshchenko there were not just included completely alien-root words but also "false fiends of translator", ( because it is impossible to calculate false friends in the Swadesh list as common vocabulary and cognates). There are required exact correspondences in meaning with English concepts of the list.
But it is also possible that there were also used 2 kinds of partial cognates of the same meaning as in their counterparts from the other language.
The proof may be here in the publication Правда про походження української мови by prof. Tyshchenko - https://archive.org/details/tishenko/page/n33/mode/2up The pairs of the words from this fragment of the Slavic Swadesh list (and they are called by prof. Tyshchenko as "innovations".
Here on the page 54 of the printed text (and page 34 in this "slide show" or "video"):
So these two kinds in my opinion may be these:
Those which have indirect origin from one words-ancestor and of ancestral ( not modern) language. Here are coloured as the "innovations" Russian - Ukrainian pairs женщина - жінка (woman) and тяжелый (which is coloured in brown) - тяжкий (hard, heavy). The Ukrainian важкий is complete non-cognate which is occasionaly and grammaticaly alike to Russian and Ukrainian тяжкий.
Those, which have direct origin, but have so-called "irregular" sound changes in their roots. The examples are these coloured innovations of the fragment ( table) on page 54 like Russian - Ukrainian pairs он - він (he) and ладонь - долоня (palm).
The partial cognates which belongs to opposition "general - diminutive" or which have the irregular sound changes just in affixes here are not coloured in any colour as "innovations". Examples are Belarusian жонка (wife), Ukrainian жінка (these two are complete cognates between each other) and Russian жена (it is partial cognate of the same meaning for these two).
Therefore, in other language pairs the comparison probably might realized by this criteria (Spanish "tortuga" (turtle) and Italian "tartaruga" are not complete and direct cognates). So, it may make difference betwen the data from the present lexical distances map and the other lexical distances calculations.
Additionally, it is necessary to define the criteria, according to which these the other distances were made. So, there is question whether they are really made according to the division on "cognates - non-cognates" (root-similarity and etymological criteria) or just according to the Normalized Levenshtein distance data.
Also, the Tyshchenko's analysis seems to be in the publication as some kind of "synonymic", while many linguistic made some kinds of analysis without synonyms.
If the map probably may say something about genealogical closeness and subgroup division, it probably cannot say a lot about the intelligibility measurement. Unlike historical linguistics, the different branches of the linguistics which is related to the intelligibility studies (Receptive multilingualism, etc.( will mostly treat these lexical items like the kinds of cognates. And it will use the Normalized Levenshtein distance for their intelligibility percentage prediction.
It is important to know that even some pure cognates from some languages may be more differently spelt or pronounced between each other than some partial cognates (sometimes it is true even for Russian-Ukrainian complete non-cognates pairs, like Запад -Захід (West).
Still, partial cognates (and they are defined as non-cognates or lexical distance by many historical linguists), are the words with mostly big percent of sound or letter changes. Their intelligibility is complicated (sometimes they may not intelligible at all). It is especially true when they are out of context.
But even in context their intelligibility may be also limited, despite it is better than complete non-cognates, false friends or partial cognates with the different meaning like English "love" and Dutch "belofte" (promise).