Swadesh list

A Swadesh List is a highly loan-resistant basic vocabulary list, made up of common words existing in any human language. The original list proposed by Morris Swadesh included about 200 terms - a shortened list of words most resistant to change, consisting of exactly 100 terms, was later used. This list was compiled by Swadesh in the 1940s and 1950s with the aim of using it in the lexicostatistical comparison of languages.

It is used in historical linguistics as a means of establishing the relationship of poorly or poorly documented languages, and also in glottochronology to estimate a quantitative measure of the divergence time of two or more related languages —which also makes it possible to establish the time of evolution of a family from its common ancestor or common protolanguage.

Introduction

The Swadesh list allows both establishing the relationship of languages, as well as establishing the degree of divergence between two languages of a linguistic family. The list is based on a list of lexicons or basic vocabulary of two languages between which they are supposed to be related, so that their degree of divergence over time can be found. In this way, this list is a fundamental instrument in comparative historical linguistics, glottochronology and lexicostatistics.

Notion of basic vocabulary

The main application that the Swadesh list has had is to homogenize the vocabulary base on which to make reliable inferences. It happens that most of the vocabulary of a language depends on cultural factors and changes, not due to historical evolution, but for reasons related to social changes that are extralinguistic. Mainly the uncontrolled changes are due to cultural reasons, reasons of technological diffusion (the word television is found in almost all current languages), political, economic, etc. In short, there are too many cultural factors of unpredictable consequences that affect the vocabulary of a language. For that reason, it is unreliable to make comparisons of languages on the basis of words that designate concepts dependent on contact between civilizations such as 'potato'; (of American origin), which is currently found in all Romance languages. The existence of this word in these languages and its comparative form could lead us to the erroneous conclusion that it is a word from the Latin **potatus, something impossible given that the potato was discovered by the Europeans in the 16th century. Other cultural reasons brought the word potato into the vocabulary of the Romance languages, but not the evolution of a Latin word that never existed.

Aware of this problem, Morris Swadesh assumed that there was, however, what he called a "basic vocabulary", that is, a more stable vocabulary and less subject to cultural changes, but to much more evolution. slow and less influenced by extralinguistic factors. This basic vocabulary had to lack cultural concepts, and therefore universal to all human cultures, that is, it had to be common and likely to be found in any language and at any time of its development. For example, the word snow is not used to compare tropical languages of Central Africa, nor 'television' to compare classical Greek with modern Greek.

This basic vocabulary consisted of words like 'water', 'hand' or 'mujer', known in any culture and so fundamental and simple that they would hardly be replaced by borrowings from other languages. The Swadesh list is intended to be the embodiment of this stable vocabulary and provides linguists with a reliable standard for comparison between languages.

Use in glottochronology

Glottochronology assumes that there is an approximately constant rate of change in linguistic evolution. That rate has been calculated at 86% of basic vocabulary words maintained every thousand years. Statistical consistency has been highly criticized because it is based on results obtained for a very limited number of languages. Some linguists have pointed out that it is unreasonable to think that even the same language evolves at a constant rate at two different historical moments, although Swadesh's claim is that over long periods the periods of accelerated change and slowed change roughly offset each other. Another criticism is that it could be that not all languages would necessarily share an identical exchange rate, and there may be different exchange rates in different regions of the planet, that is, it has been criticized that the rate obtained for Indo-European languages does not have to be applied. to all the languages of the planet. In general, it is considered that, while the Swadesh list has stabilized the lexical basis of glottochronological comparison, the same has not happened with the rate of language change, which remains subject to a multitude of cultural and historical factors, all of them uncontrollable..

Even with the above criticisms, the statistical assumptions of stability and universality of the exchange rate over long periods of time, lead to the following glottochronological estimate for part of the divergence time:

t=12log (c)log (r){displaystyle t={frac {1}{2}}{frac {log(c)}{log(r)}}}}}

being

t = time of separation between languages

c = lexic similarity coefficient (using Swadesh lists)

r = Glotocronological constant (established at 86%)

Swadesh List Lexicon

There are two versions of the Swadesh list: one with 207 terms and one with exactly 100 terms. Both lists and some shorter alternative lists are reproduced below.

List of 100 words

(originally in English)

Swadesh–Yakhontov List

The Swadesh–Yakhontov list is a set of 35 particularly stable terms culled from the original list by the Russian linguist Sergei Yakhontov or, in English transcription, Sergei Yakhontov (Starostin, 1991). Linguists such as Sergei Stárostin have used this shorter list in lexicostatistics for far-reaching comparative work. Below is the subset of the Swadesh-Yakhontov list keeping the numbers of the original Swadesh list:

Holman et al. (2008) found that the Swadesh-Yakhontov list was less accurate in identifying known relationships between Chinese variants. These authors saw that a set of 40 words from the Swadesh list could give results as good as the original list, so that 40-word list does achieve what the Swadesh-Yakhontov list does not seem to.

Swadesh List Stability

Holman et al. (2008) investigated the relative stability of words in the Swadesh list of 100 terms by comparing the retention rates of terms in well-established language families. Thanks to that, they were able to reorder the Swadesh list from the most stable to the least stable terms:

The words marked with an asterisk would make it possible to build a list of 40 terms, which is statistically as significant as the original Swadesh list of 100 terms, and therefore this shorter list would represent an advantage over the Swadesh list -Yakhontov.