Sound correspondence - Consonant relationship

Sound correspondence in eLinguistics.net

The basis of the calculation of the genetic proximity between two languages is a comparison of the consonants contained in a word in a language and in the corresponding word in the language it is being compared to (cognate scoring). The order in which these consonants appear in the words is being taken into account.

Consonants evolve much more slowly than vowels. In comparison with human bodies found on archaeological sites, vowels can be compared with the skin and consonants with the bones. Only bones remain over longer periods in a recognizable shape. In comparative linguistics, consonants are the "words' bones"!

In this project, the consonants are clusterized: the hundreds of consonants used in world languages are gathered in related clusters ('-K-' cluster, '-B-' cluster, etc.). The extreme simplification in the sound change modelization for the computer processing appears to deliver sharper results than more complicated models. Moreover, as we here compare languages across world wide spread families, we have to take only universal sound change into account.

Here are the clusters into which the consonants are regrouped for this project. Their frequency in the complete word sample and the number of points they generate if occurrences are identified in the word comparisons are listed bellow. Exact matches get the highest score, but matches across related clusters also get points – decreasing according to how remotely related this relationship is. The quantification of the consonant relationship for the point system is established partially with data from: "Sound Correspondences in the World's Languages (Cecil H. Brown, Eric W. Holman and Søren Wichmann)". The values from Cecil H. Brown, Eric W. Holman and Søren Wichmann's study are partially used and if so, they are converted and adapted to the eLinguistics.net consonant clustering model - so values bellow are not equal to the values in the reference study but are partially generated from them: the above study contains much more relationships between consonants and do so in a much finer granularity. In Elinguistics.net, not all relationships between consonants are needed, but only those giving stronger signals in cognate identification with a low exposure to chance. For this reason, the relation between the cluster -R- and cluster -L- or -J- to -ZH- are not taken into account in eLinguistics.net although they are well documented and quantified in the above mentioned study. More complex relationships between consonants as known in sound change in various languages are not modelized when they bring a bigger exposure to chance matches.

Sound cluster -B- (b)
 
b, б, ب, Hindi "ब, भ", Urdu "ب بھ",...
 
Details about the correspondence point system click here: Cluster '-B-' correspondences
Sound cluster -C- (ts, dz, tʃ)
 
ts, German "z" (partial), Slavic "c", Cyrillic "ц",...
 
Details about the correspondence point system click here: Cluster '-C-' correspondences
Sound cluster -CH- (tʃʲ)
 
tch, German "tsch", Slavic "Č", "ч", Armenian "չ", Kurdish "Ç", Polish "cz", Croatian/Serbian, Polish "ć", Armenian "չ", " ջ", Cyrillic "ч", Hindi "छ", Persian "ﭺ",...
 
Details about the correspondence point system click here: Cluster '-CH-' correspondences
Sound cluster -D- (d, dˁ)
 
d, д, Arabic "د, ض", Hindi "द",...
 
Details about the correspondence point system click here: Cluster '-D-' correspondences
 
Sound cluster -F- (f, ɸ)
 
f, ф, ف, Afrikaans "v", Dutch "v", German "v",...
 
Details about the correspondence point system click here: Cluster '-F-' correspondences
Sound cluster -G- (g, ɟ)
 
g, г, Persian, Pashto "ﮔ", Hindi "ग, घ",...
 
Details about the correspondence point system click here: Cluster '-G-' correspondences
Sound cluster -H- (ħ)
 
h, Arabic "ح, ه",...
 
Details about the correspondence point system click here: Cluster '-H-' correspondences
Sound cluster -J- (j)
 
French "ll", Hindi "य", Arabic "ي", Slavic "j", Various languages "y" as consonant,...
 
Details about the correspondence point system click here: Cluster '-J-' correspondences
Sound cluster -K- (k, q)
 
k, q, Italian "ch", "qu", "cq", French "qu", Catalan "q", Galician "c", French "c" (partial), Celtic "c", Arabic "ك, ق",...
 
Details about the correspondence point system click here: Cluster '-K-' correspondences
Sound cluster -KH- (x, χ, ɣ ç)
 
Cyrillic "х", Afrikaans "g" (except between "l/r" and "e"), Dutch "g", German "ch", Spanish "j", Croatian/Serbian "h", Greek "χ", Gothic "X", Breton "c'h", Irish, Welsh "ch", Arabic "خ",...
 
Details about the correspondence point system click here: Cluster '-KH-' correspondences
Sound cluster -L- (l, ɬ ʎ)
 
l, л, ل , ल, Welsh "ll", Polish "ł",...
 
Details about the correspondence point system click here: Cluster '-L-' correspondences
Sound cluster -M- (m, m̥)
 
m, م, म,...
 
Details about the correspondence point system click here: Cluster '-M-' correspondences
Sound cluster -N- (n, ɳ, ɳ̊, ɲ̊, ɲ, ŋ̊, ŋ, ɴ )
 
n, Polish "ń", ن, न,...
 
Details about the correspondence point system click here: Cluster '-N-' correspondences
Sound cluster -P- (p)
 
p, п,...
 
Details about the correspondence point system click here: Cluster '-P-' correspondences
Sound cluster -R- (r, ʁ ʀ ɾ, ɽ)
 
r, cyrillic "р", Arabic "غ and ر", hindi "र"
 
Details about the correspondence point system click here: Cluster '-R-' correspondences
Sound cluster -S- (s, sˁ)
 
s, cyrillic "с", Arabic "س, ص",...
 
Details about the correspondence point system click here: Cluster '-S-' correspondences
Sound cluster -SH- (ʂ ʃ)
 
Afrikaans "sj", German "sch", "s" before "t", French "ch", Slavic "Š", Polish "ś", Arabic "ش", Cyrillic "ш",...
 
Details about the correspondence point system click here: Cluster '-SH-' correspondences
Sound cluster -T- (t, tˁ)
 
t, Arabic "س, ص",...
 
Details about the correspondence point system click here: Cluster '-T-' correspondences
Sound cluster -TH- (θ, ð)
 
th, Gothic "þ",English "th", Cornsh "dh, th", Celtic "th", Arabic "ث ذ", Hindi "थ, ध, ठ, ढ",...
 
Details about the correspondence point system click here: Cluster '-TH-' correspondences
Sound cluster -V- (v, ⱱ)
 
v, cyrillic "в", Afrikaans "w" except after "d, k, s, t", Dutch "w", German "w",...
 
Details about the correspondence point system click here: Cluster '-V-' correspondences
Sound cluster -W- (w)
 
English w, Afrikaans "w" after "d, k, s, t", Gothic "ƕ",...
 
Details about the correspondence point system click here: Cluster '-W-' correspondences
Sound cluster -Z- (z)
 
z, Cyrillic "з", German "s" (partial), French "s" (between two vowels), Arabic "ز, ظ",...
 
Details about the correspondence point system click here: Cluster '-Z-' correspondences
Sound cluster -ZH- (ʒ, ʑ, ʐ dʒ)
 
Slavic "ž", French "j", Catalan "j", Hindi, Sanskri "ज", Arabic, Pasto, Persian "ژ", Irish, scottish "dh", Polish "ź, ż", Cyrillic "ж",...
 
Details about the correspondence point system click here: Cluster '-ZH-' correspondences

The details leading to the choice of this simplified sound correspondence modelization and their references are under verification and will be published soon.

Using consonants comparisons as the basis for the measurement of the genetic distance between languages makes the system not adapted for tonal languages like Chinese, Thai or Lao – this is why tonal languages are not part of this study.

Comparative linguistics English to German (example with details)