Evolution of Pashto dialects of PF Members

[PFgulalai]

I Pashto!

I have been following this great thread (http://www.pashtunforums.com/pashto-...dialects-9168/). I noticed both significant similarities and differences in the Pashto dialects. In general, similarities and differences give us diversity, the mechanism of which is called evolution

The words given by each member were aligned to the corresponding words given by the rest of the members.For example, the words for "how" were tsanga, tsa ranga,singa,sanga,charanga, and changa, and in the Diagram 1 they have been aligned in vertical direction and so on.

On the basis of the diagram 1, I generated the tree to get an idea of neighbors and it turned out that BLS is my neighbor

QUESTION:

1What you infer/understand/see/notice from the Alignment (diagram 1) and Tree (diagram 2)?
2.Does it make sense? Yes/No? Why?
3.Which one is more similar/different?

IMPORTANT NOTE:
This is just for fun. Please do not think that these dialects have certainly evolved/developed in the same pattern. I have used bioinformatics software (that I normally used for Organismic Evolution) for language which makes the results scientifically invalid.But when I saw the result, I was so surprised that I thought of sharing it for fun.

[ayub]

how did u use a software used for molecular biology etc and transfer it to languages...its not possible..right? just by using different words in a language doesnt equate to evolution..

pretty diagrams, and you must have spent a good deal of time on it...but your results are questionable and data unreliable.



note: okay i admit it..i have no clue what im talking about..lolz..but ur wrong gulalai...

[Admin Khan]

Very fascinating, can you shed some light on your results?

[PFgulalai]

Ayub,
Well I have been wondering which dialects are more close in terms of similar word structure after I read posts on that thread,so initially it was more for my own curiosity like putting stuff in the softwares, etc.

Certainly, it is wrong scientifically (Pseudoscience) and I do expect that nobody takes it seriously. But I would like to know which of these dialects are similar/different in reality.

:D @Evolution. Languages do evolve, but not biologically

[PFgulalai]

Very fascinating, can you shed some light on your results?

Thanks Admin.This is how I read the tree:

1.The distinct dialects are of LZ and Wrayun, the latter branching out completely from the rest of the clan.(In Wrayun case, there was some extra information that I was supposed to delete to fit in the alignment with the rest but I have to check first)

2. LZ and your clade (You, Shlombay and Feroza) dialects are more similar than to let's say, Pashtunkhwa. I used your input directly for the Shlombay and Feroza.

3.Amir Al Ghaznavi and Dilu_93/Pakthia are distinct than Kunar.

4.The Kabuli, Kunar, and Pashtunkhwa seem to be more close.

5.The Waziristani and khost are more similar.They are distinct than Kandahari and so on.

The key to understand the tree is to notice how close each branch is.The black square dots are the NODES.Each node is an ancestor, from which different dialects have branched out.

So my purpose was to know that in reality how close/similar each dialect is to the other.So I need feedback.This is a bioinformatics software,certainly not for linguistics, but I could not stop myself from checking it out.Hahaha.

[pukhtana]

you are mostly accurate. I know the Afridi and Kunar dialect is very similar.

[Omar Usman]

Hmmm very interesting. Can something like this be used to comment upon correctness of genealogy trees???

[PFgulalai]

you are mostly accurate. I know the Afridi and Kunar dialect is very similar.

Pukhtanaey, I have never heard Kunar dialect but I am very familiar with Afridi one. Thanks for the information.

Hmmm very interesting. Can something like this be used to comment upon correctness of genealogy trees???

You mean the family lineages? Yes, there are softwares for building up pedigrees of families. This one is ALSO used for genealogy, however, the difference is that you need to have DNA information of your family, tribe, ancestors etc.

You can check out the link below.It's really interesting.

http://www.dnaancestryproject.com/

[Laila]

Ha ha this is an interesting diagram...yay i am on it :D

Kunar dialect, from what i know, is close to Kabuli, Nangarhari, Laghmani, and Pekhwari (Eastern) so i am guessing Afridi is close to it too....just depends on the region

[Toramana]

gulalai, thank you for this diagram. And this confirms what the actual situation is. We all know, Swat, Dir, Malakand, etc. Peshawar Valley, Nangarhar, Kunar, Laghman, etc. speak more or less the same dialect.

as for my little knowledge, bio-informatics software/applications process strings (representing gene sequencing in DNA)....So it may also be applied to languages (which can also be considred strings---sequences/patterns of characters). I think the software clustered the words based on lexical similarity.

this is not psuedoscience...this is what science is....this can be a pretty interesting research work...for example, next task could, how close are various dialects to each other (some kind of similiarity/dis-similarity measure...)...

[Omar Usman]

@Gulalai: thank you for the link. It is interesting but I think its purpose is not suited to what I had in mind.

@Toramana: Yes I was thinking along this lines. But what to do with dialectual similarities/dissimilarities alone? perhaps it can be useful for something along the example of lineage trees. For example, the authenticity of lineage trees have always been a subject of discussion. But they have always been commented upon from historical perspective. Perhaps this, or something like this can be used to confirm these trees from scientific perspective and possibly settle the issue. This does involve heavy data collection work as it will involve digitizing both the trees, as well as collecting dialect specimens. If anyone is interested, maybe we can pool together a team to work on this.

[Gulalai]

hahahahah @ Ayub's post

& this is interesting gulalai thanks for sharing (:

[PFgulalai]

Ha ha this is an interesting diagram...yay i am on it :D

Kunar dialect, from what i know, is close to Kabuli, Nangarhari, Laghmani, and Pekhwari (Eastern) so i am guessing Afridi is close to it too....just depends on the region

This is what I wanted to know.Thanks Laila Jaan for the information

[PFgulalai]

gulalai, thank you for this diagram. And this confirms what the actual situation is. We all know, Swat, Dir, Malakand, etc. Peshawar Valley, Nangarhar, Kunar, Laghman, etc. speak more or less the same dialect.

as for my little knowledge, bio-informatics software/applications process strings (representing gene sequencing in DNA)....So it may also be applied to languages (which can also be considred strings---sequences/patterns of characters). I think the software clustered the words based on lexical similarity.

this is not psuedoscience...this is what science is....this can be a pretty interesting research work...for example, next task could, how close are various dialects to each other (some kind of similiarity/dis-similarity measure...)...

Thank you so much Toramana for the information.You are right about the bioinformatics softwares approach.Basically the alignments were treated as protein sequences which are more diverse in terms of letters than the DNA sequnces (which are only A, T, G, C).The lexical similarity was what I wanted to check because I noticed how certain words tend to be more conserved in some dialects compared to the rest of the words. So I made a sequence (sanga, mung, charta, rasha, shwal, jwandun,zmaka,khaza,gira,janai,kor,malgaraye,khpa) of the ~13 words of each dialect and then aligned them. Generating a tree with usual parameters was difficult so I modified some values using neighbor joining method.

As for your last question, we can find out that by breaking individual words and comparing the similarities/differences.The one I used above looks for global similarity and you can see how gaps (-) were introduced to fit the sequences in the alignment.

And thanks for the appreciation.I haven't yet ventured into evolutionary linguistics or biolinguistics compared to organismic evolution, the latter being a much more developed field than the former ones.So I am doubtful about the scientific validity of my approach at this point.Nonetheless, as you mentioned, it could be quite interesting to see how far we can take it. For now the feedback from you and others have amused me.

[PFgulalai]

@Gulalai: thank you for the link. It is interesting but I think its purpose is not suited to what I had in mind.

@Toramana: Yes I was thinking along this lines. But what to do with dialectual similarities/dissimilarities alone? perhaps it can be useful for something along the example of lineage trees. For example, the authenticity of lineage trees have always been a subject of discussion. But they have always been commented upon from historical perspective. Perhaps this, or something like this can be used to confirm these trees from scientific perspective and possibly settle the issue. This does involve heavy data collection work as it will involve digitizing both the trees, as well as collecting dialect specimens. If anyone is interested, maybe we can pool together a team to work on this.

Limitations are almost present in every form of trees, be it genealogical trees or phylogenetic ones. I would consider a noisy data as a major limitation for the plausibility of a phylogram and could be a limitation in lineage tree for dialects also (I guess!).For instance, if my sample was large (including more dialects) and the number of words were >13, then the results might have been different.

So are you planning to work on developing genealogical trees on the basis of dialects (as a variable I guess!)?--something like a Tree Model?