METHODOLOGICAL ALGORITHM FOR DIACHRONIC INTERPRETATION OF NOSTRATIC * mara “ tree ” ( based on the data taken from Bomhard ’ s “ A Comprehensive Introduction to Nostratic Comparative Linguistics : With Special Reference to Indo-European ” )

The article discusses the methodological algorithm for diachronic interpretation of the Nostratic etymon *mar-a “tree” represented in the A. R. Bomhard’s “A Comprehensive Introduction to Nostratic Comparative Linguistics: With Special Reference to IndoEuropean. The paper elaborates the definition of the Nostratic etymon as a hypothetical language-ancestor, established based on the comparison of the degree of affinity with the available reconstructed etymons at the level of every Nostratic language family (Altaic, Afro-Asiatic, Dravidian, Indo-European, Kartvelian, and Uralic). The author comes to the conclusion that Bomhard used the three research stages to identify the Nostratic etymon. At the first stage (the level of the language family) the comparative-historical method is used with the procedures of internal (etymon(s) / proto-form(s) at the level of the language group) and external (comparison of etymons / proto-forms reconstructed for certain language groups in order to reconstruct etymon / proto-form at the level of the language family) reconstruction; the step-by-step reconstruction method. At the second stage (the level of macrofamily) the mass comparison method is used. It helps to involve the already reconstructed etymons at the level of every Nostratic language family into comparison. At the third stage, a specially developed method of diachronic interpretation is used. An attempt tries to prove that the diachronic interpretation of the Nostratic *mar-a “tree” is confirmed by the analysis of language data. The paper considers two methodological procedures: 1) the comparison of a sufficient number of common morphs (or allomorphs) in comparable languages, i.e. the establishment of similar rules for combining regular phonetic correspondences; 2) the semantic similarity of common morphs (or allomorphs) in comparable languages.

In particular, Bomhard notes that the Nostraticists do not use the "traditional methods", but the "weakened forms" for comparison (Bomhard, Kerns, 1994, pp. 7-11) of forms and meanings of Nostratic etymons.Once Illich-Svitych wrote that "the investigation of the lexical, word-formation and morphological similarities of six large Nostratic language families [...]" is necessary to be carried out at "the oldest level reached with the help of the comparative method, i.e. at the level of six corresponding original systems".It gave him an opportunity to say that "such similarities are not accidental, but regular".The number and regular nature of the investigated similarities of the six proto-languages enabled the scholar not only to establish the very fact of the language affinity, but also to discover a number of regularities that unite the proto-languages, as well as to begin the reconstruction of some essential aspects of the language system underlying the six comparable daughter language systems (Illich-Svitych, 1971, p. 1).
Dolgopolsky also used the comparative-historical method, updated by more modern techniques (first of all, by the step-by-step reconstruction method) (Dolgopolsky 2012, p. 9).
The mass comparison method, dealing with the comparison of a basic vocabulary with a large number of languages from a wide geographical area (Greenberg, 1963), was proposed as an alternative to the comparative-historical method.It was outlined by J. Greenberg, the representative of the American School of Macrocomparative Studies (he developed the classification of African languages, dividing them into four language families).The methodology of the scholar was the following: it was necessary to construct general models for a large number of modern languages and find similarities in systems of pronouns without establishing correspondences, without making reconstructions, but simply trying to form their classification.The proposed methodology has been criticised, because this method "is merely an assumption of affinity, which should be verified using the comparative-historical method" (Babayev, 2013, p. 6).
An approach to the comparison of languages used by Bomhard, who tried to establish a genetic affinity among different Nostratic languages partly corresponds to what Greenberg proposed in the "Genetic Relationship among Languages", in the "Essays in Linguistics" (Greenberg, 1957).He formulated the following key principle: the comparison of languages is an efficient way of testing hypotheses about genetic affinity.However, the problem relates to the choice of languages for comparison (Bomhard, 2018, p. 9).In our opinion, it is more correct to speak about the basis for comparison, the choice of which to this day is an unresolved problem in Linguistic Comparative Studies.
Due to the fact that Linguistic Comparative Studies (Korolyova (2018) et al.), according to Illich-Svitych's words, "developed rather rigorous techniques" (Illich-Svitych, 1971, p. 38), etymologists used the classical morphemes identification method.The very fact of such an affinity deals with the number of identical morphemes due to their origin in the comparable languages (Illich-Svitych, 1971, p. 38).This view is also confirmed by Bomhard, who assumes that "the easiest way to establish genetic affinity is to have a sufficient number of common morphs (or allomorphs), especially irregular, in comparable language systems" (Bomhard, 2018, p. 9).The continuation of this discussion is offered in the given article.
The aim of the paper is to characterise the methodological algorithm for diachronic interpretation of the Nostratic *mar-a "tree", developed in the Bomhard's "A Comprehensive Introduction to Nostratic Comparative Linguistics: With Special Reference to Indo-European".

Method
The substantiation of the genetic affinity of Nostratic languages through the interpretation of the presence of Nostratic etymon is carried out using the method of diachronic interpretation (according to Szemerényi (1990)).It consists of two methodological stages: at the first stage, the comparative-historical method is used with the procedures of internal and external reconstruction, as well as the step-by-step reconstruction method, when the scholars managed to reconstruct the proto-form and establish the genetic affinity for one language family; at the second stage the mass comparison method is used when the correspondences among the already reconstructed etymons at the level of every language family are compared.

Results
Nowadays the essence of the Nostratic theory deals with the reconstruction of the Nostratic protolanguage that may be defined as the hypothetical language-ancestor, established on the comparison of etymons at the level of every language family reconstructed through the comparison of the genetic correspondences of related languages of various space and chronological periods of the development and formation of certain language groups, their subgroups, or separate languages within every Nostratic language family (Altaic, Afro-Asiatic, Dravidian, Indo-European, Kartvelian, Uralic) (our definition -Y.K.).
The genetic affinity of the Nostratic languages is substantiated by the existence of a large number of related morphemes in them (Bomhard, 2018), reaching the Nostratic etymon.An etymologist, trying to establish it, has to implement the corresponding methodological steps that were unequal for representatives of Linguistic Comparative Studies.In this regard, they were distributed to hyperskeptivists, macrocomparativists, multilaterialists.The hyperskeptivists (Campbell, 2007 et al.) propose that the limits of the comparative representative's capabilities are at a depth of 5-6 or 7-8 thousand years, but they do not agree with the possibility of evidence of distant language affinity.The macrocomparativists propose to study "distant linguistic breeds" (the term of Starostin) based on the comparative-historical method and the stepby-step reconstruction method.The multilaterialists believe that the classic comparative-historical method does not work at deep chronological levels, and therefore alternative methods, in particular the mass comparison method, should be used for studies of distant language affinity.However, the Nostratic sources show that the path to the Nostratic etymon is one way or another associated with the observance of two methodological stages: The first stage is the level of the language family, where the comparative-historical method with the procedures of internal (etymon(s) / proto-form(s) at the level of the language group) and external (comparison of etymons / proto-forms reconstructed for certain language groups in order to reconstruct etymon / proto-form at the level of the language family) reconstruction was used.
According to Shcherba (1958), the basis of this method is "the doctrine of the genetic affinity of languages, which has its material expression in the community of sound form, in the laws of changes in the phonetic system, grammatical structure and vocabulary of related languages" (59).Kodukhov (1963) agrees with this opinion and understands the comparative-historical method as the process of studying the history of language based on the comparison of facts and phenomena of different periods of their development, as well as based on the comparison with the related alive and dead languages" (p.23).These methodical steps lead to the reconstruction of etymons at the level of the language family and give reasons to substantiate the genetic affinity at the level of this language family.
The determined methodological and procedural operations are used for every separate Nostratic language family.
The second stage is the level of language macro-family, where the mass comparison method (developed by Greenberg in the 50's and 60's of the XX century) is used.It deals with the comparison of the already reconstructed etymons for every Nostratic language family.In the "Big Russian Encyclopaedia" (Bolshaya rossiyskaya entsiklopediya) the mass comparison method, or the multilateral comparison method, is defined as a procedure for comparing language data based on the sketchy (intuitively perceived) similarity of sound and the meanings of morphemes, without a clear system of correlations.In this case, the accuracy of the results of the application of this method depends on the total number of morphemes, as well as on the languages involved into comparison (hence the "mass" of comparison).As a result of the application of this method, it is possible to reconstruct etymon at the level of the macrofamily, as well as to substantiate the genetic affinity among the Nostratic language families.
Despite the fact that Greenberg called the mass comparison method "superficial", the results of this method allow carrying out the diachronic interpretation (Szemerényi, 1990) of the Nostratic etymon, which, according to Dronova's opinion, "does not have a real linguistic substrate necessary for the reconstruction, but is used only with existing archetypes (proto-forms) [...]" (Dronova, 2009, pp. 116-123).Citing Klimov, she warns that the diachronic interpretation method is conjuncture (Klimov, 1988), but the conjuncture is an assumption, a guess without the possibility of verification; it is necessary to use it with caution and after proper reconstruction (Dronova, 2009, pp. 116-123).

Discussion
Due to such a discussion that continues to evolve around the methods of substantiating the language affinity among different Nostratic languages in the Bomhard's "A Comprehensive Introduction to Nostratic Comparative Linguistics: With Special Reference to Indo-European" (Bomhard, 2018), the above-mentioned methodological procedures are observed.They were also grounded by Greenberg in "Genetic Relationship among Languages", "Essays in Linguistics" and were reduced to the fact that the only way to substantiate language affinity is to compare languages (Greenberg, 1957).But the mass comparison procedure is a possible operational step for the diachronic interpretation of the Nostratic etymon.
But there is one more complicated problem when the languages are compared.It is connected with the necessity to remove random similarities from the data analysis, and most importantly to separate borrowings from native elements.Greenberg proposes the separation of borrowed elements at the two following levels: 1) it is necessary to separate borrowings which do not go beyond certain semantic fields (for example, the vocabulary belonging to the sphere of culture), 2) certain grammars to separate borrowed words from the native ones by inclusion of new languages into comparison (cit: Bomhard, 1989).
In the process of the first methodological procedure, Bomhard gives the genetic data, in which the regular phonetic correspondences can be traced, i.e. the phonetic laws.According to Abayev's version, they can be observed in one language (for example, constant accent, like in French or Polish, assimilation of consonants, like in Ancient Indian, harmonisation of vowels, like in the Turkish languages, impossibility of voiced consonants in the ascending position, like in Russian, etc.) (Abayev, 1933, p. 1).That is why one can find a number of similar laws in any language.
In Linguistic Comparative Studies, there are two views on the regular phonetic correspondences.The first one deals with the correspondence of phonemes that can be called regular (or systematic) if it is found in a sufficiently large number of words ("sufficiently large" is a probabilistic estimate: the examples of correspondences should be so much that the chance of accidental coincidence is small).The second one deals with regular compliance, which is regulated and set according to the rules and has a minimum number of exceptions.The number of words, in which such regular correspondence is observed, is not important: for example, the correspondence of IE *p-(Latin velar) (by the presence of IE *k {w} in the next syllable) is absolutely regular in all three words, where IE *p-turns out in such a position, where it gives the Latin qu-(or c-before o): -Latin qu {i_} nque "five" < ІЄ *penk{w}e, compare with Ancient Indian pa{n ~} ca, "five"; -Latin coquo "cook" <ІЄ *pek{w}-, compare with Ancient Indian pacati "bake, cook"; -Latin quercus "oak" <ІЄ *perk{w}u-, compare with Punjabi parg {a_} "stone oak" (Burlak, Starostin 2005).
But actually, Bomhard comments on the regular correspondences as processes that occur over time and lead to both the gradual transformation and disappearance of such similarities.The more time has passed since the division of languages, the less likely that the similarity of morphological forms and rules of combination will be detected (Bomhard, 1989, No. 9).
The second methodological procedure observed in the Bomhard's vocabulary versions provided for the establishment of the semantic similarity of the reconstructed forms in substantiating the language affinity among different Nostratic languages (Bomhard, 1989, No. 9).It was based on a sufficiently large lexical data, where "the most important are the lexical units with the same or similar meanings" (Greenberg, 1957).And then the next place of importance is occupied by the forms, which, although differ in their meanings, but they can be taken into account to allow for typical semantic shifts to earlier forms with the same or similar meaning.It means that the likelihood of lexical similarity captures the language affinity that grows when new languages are involved into comparison and where a number of regular phonetic-semantic correspondences are traced.
According to Greenberg, this procedure was called the "mass comparison" (Greenberg, 1957).The scholar considered the comparison of the basic vocabulary of a large number of languages with a sufficiently large area to be the most reliable way of establishing genetic affinity.The lexical data is paramount in solving this problem, especially at the initial stage of comparison (cit.: Bomhard, 1989).
Let us try to consider this methodological algorithm based on the reconstructed Nostratic etymon *mara with the English meaning "tree, wood", represented in the Bomhard's "A Comprehensive Introduction to Nostratic Comparative Linguistics: With Special Reference to Indo-European".It should be noted that the diachronic interpretation of the Nostratic etymon was carried out with the help of data of the three language families (Afro-Asiatic, Dravidian, Uralic).
The Egyptian mrw "Lebanese cedar" was involved into comparison from the Egyptian language group from the Afro-Asiatic language family: the plan of expression is represented by three phonemes /m/, /r/, /w/, but the plan of content is "Lebanese cedar".
It should be assumed that the above-mentioned data allowed to reconstruct the Proto-Dravidian *maram/n with the meaning "tree".The plan of expression has the following characteristics: if the root (*mar) was modelled based on the constructs (vowel a, consonants m, r) of Tamil, Malayalam, Kota, Kannaḍa, Koḍagu, Tuḷu, Gadba (Ollari), Gondi, Konḍa, Manḍa, Kuwi, Pengo, then the suffix (*-am/n) was modelled based on the constructs (vowel a, consonants m/n) of Telugu, Gondi, Kuwi, Kui; the plan of content is common for all correspondences of the language groups (southern, south-east, south-west, central, Gondwana).
If to analyse the represented data, using the first methodological procedurethe establishment of the regular phonetic correspondences, supported by Serebrennikov, and, probably, with the help of it Bomhard proves the genetic affinity: one can assume that the regular phonetic correspondences should serve as a key factor in the proof of such an affinity.At the same time, the correspondences should not be limited to isolated examples, because under other conditions they become unproven (Serebrennikov, 1982).Let us try to trace it on the examples with a lot of reflexes.
As we can see, one of the arguments about a probable genetic affinity is the existence of rules of combination, i.e. the isolation of a sufficient number of similar phonological correspondences.
The second methodological procedure for substantiating the genetic affinity is the semantic similarity among the genetic correspondences.
At the same time, the attention should be drawn to the fact that the Egyptian mrw "Lebanese cedar" is represented in the Afro-Asiatic language family.Probably, this meaning may have had only one name for this vegetable world (the same meaning is represented in the following dictionaries (Hannig, 1995, p. 348;Erman, Grapow, 1921, p. 67;Faulkner, 1962, p. 112;Gardiner, 1957, p. 569).
The third methodological procedure is the diachronic interpretation of the Nostratic etymon *mar-a "tree, wood", carried out by the analysis of the: 1) Egyptian mrw "Lebanese cedar" taken from the Afro-Asiatic language family; 2) Proto-Dravidian *mar-am / n "tree" for the Dravidian language family; 3) Proto-Uralic *m/a/re "tree" for the Uralic language family.As we can see, the plan of expression and the plan of content of the Proto-Dravidian and Proto-Uralic are the hypothetical reflexes of the Nostratic etymon, in which the correspondences between the vowel *a and the consonants *m, *r, as well as the denotative meaning "tree" are observed.The inclusion of the Egyptian correspondence by Bomhard into comparison is likely to propose, on the one hand, the existence of the two consonant phonemes /m/, /r/ in the root, which are reflexes of the Nostratic etymon, and, on the other hand, the meaning "Lebanese cedar", which is the connotative reflex of the same etymon (see Table 1).As we see, the reconstruction is usually carried out for the historical interpretation of the correspondences represented in the data of related languages.It leads to the construction of a certain archetype (proto-form).But the diachronic interpretation has an unrealistic linguistic substrate necessary for the implementation of this reconstruction procedure (Klimov, 1988, p. 10).The above-mentioned argument convinced that a comparativist can deepen the diachronic perspective only with the help of archetypes (proto-forms).At the same time, Klimov comes to the logical conclusion: "if, as a result of the reconstruction procedure, the comparativist formulates the hypothesis of the first degree, then a product of diachronic interpretation reveals some hypothesis of a second degree" (Klimov, 1988, p. 10).

Conclusions
The methodological algorithm for establishing the language affinity among different Nostratic languages of Bomhard coincides with those methodological procedures that were once justified by Greenberg and consist of three consecutive stages: at the first stagethe level of the language familythe comparative-historical method is used with the procedures of internal (etymon(s) / proto-form(s) at the level of the language group) and external (comparison of etymons / proto-forms reconstructed for certain language groups in order to reconstruct etymon / proto-form at the level of the language family) reconstruction; at the second stagethe level of macrofamilythe mass comparison method is used, with the help of which the already reconstructed etymons at the level of every Nostratic language family are involved into comparison; at the third stagea specially developed method of diachronic interpretation is used.In addition, the key stage in the analysis of language data involves the realisation of two methodological procedures: 1) the selection of a sufficient number of common morphs (or allomorphs)the regular phonetic correspondences; 2) the semantic similarity among genetic correspondences.
The diachronic interpretation of Nostratic *mar-a "tree, wood" was carried out based on the data of 3 language families (Afro-Asiatic, Dravidian, Uralic).Its genetic material enabled to establish the regular phonetic correspondences: similar phonological correspondences of Nostratic *mar-a are traced in the consonant *m typical for the Proto-Dravidian and Proto-Uralic, the vowel *afor the Proto-Dravidian, the consonant *rfor the Proto-Dravidian and Proto-Uralic, the vowel *afor the Proto-Dravidian, as well as to trace the semantic similarity among genetic correspondences: the Nostratic *mar-a "tree, wood" corresponds to the Proto-Dravidian *mar-am / n "tree", the Proto-Uralic *m[o]rз "tree, wood".