Uva-dare (Digital Academic Repository) Ethno-territorial conflict and coexistence in the Caucasus, Central Asia and Fereydan

 

 

 

 

Appendices 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

340 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

341 

Appendix 1: Measurement of Mosaic 

Type of Ethno-Geographic 

Configuration 

 

 

A mosaic type of ethno-geographic configuration is a relatively small area 

in which different ethnic groups are concentrated in many relatively 

homogeneous pockets of ethnic concentration. Switzerland is a typical 

area with such attributes. A mosaic area is ethnically fragmented and, 

therefore, there are many encounters between these ethnic groups. The 

total number of encounters between ethnic groups may even be (ideally) 

larger than the number of ethnic groups. 

 

The best way to determine whether a geographic area is of the 

mosaic type or not is to develop an index and an instrument which 

calculates the degree of fragmentation in an area. When fragmentation can 

be quantified—that is, different values of the extent of fragmentation can 

be calculated—a critical value can be defined above or below which an 

area can be designated as a mosaic type of ethno-geographic 

configuration.  

As is the case with all social scientific instruments, there is also 

an arbitrary aspect to the development of this instrument for the 

measurement and the definition of the critical values of the mosaic type of 

ethno-geographic configuration. Even instruments for measurements of 

natural phenomena may have an arbitrary aspect, as the definitions of, and 

choices between, different units of measurement—for example, between 

Celsius and Fahrenheit—are human-made choices. Natural phenomena 

follow natural laws, but instruments for their measurement are human-

made. All instruments are shaped by the taste and intellect of their 

designers. Nevertheless, they should all have certain features and 

characteristics which make them suitable for the task for which they are 

made. For example, in order to make an instrument which can cut 

vegetables, it should be a solid and sharp instrument. Metal is the best 

material for such an instrument, and paper is absolutely unsuitable. The 

designer designs such an instrument according to his taste, but it fulfills 

the criteria discussed above. Most likely it looks like a knife and not like a 

book. 

 

Similarly, an instrument can be made which measures the degree 

of fragmentation and can determine whether an area is of a mosaic type of 

ethno-geographic configuration or not. Although improvements are 

desirable, I have made such an instrument. In the following, I will discuss 

how it is made. 

 

342 

In order to determine whether or not an area is of the mosaic type, 

we should determine how large the area should be, and how many ethnic 

spatial pockets and how many encounters between these pockets should 

be located in this area. Consequently, the formula for such a measurement 

is based on the value of area divided by the number of ethnic spatial 

encounters. (One should not confuse ethnic spatial encounters with ethno-

territorial encounters. They are slightly different from each other, as will 

be explained later on.) 

Ideally, in an area displaying a mosaic type of ethno-geographic 

configuration, there are many ethnic spatial pockets—that is, relatively 

homogeneous pockets of ethnic concentrations—and the number of 

encounters between such pockets are usually larger than their own 

numbers. In other words, in a mosaic type of ethno-geographic 

configuration, there are usually more encounters than pockets of ethnic 

concentration. Such a situation is only possible when there are at least 

four ethnic spatial pockets (see Figure A-1.1). There is only one encounter 

possible between two ethnic spatial pockets (Circle A), and two (Circle B) 

or three encounters (Circle C) possible between three ethnic spatial 

pockets On the other hand, the number of ethnic spatial encounters is 

usually larger than the number of ethnic spatial pockets when there are at 

least four such pockets. In other words, an additional encounter is usually 

present when there are at least four ethnic groups, provided that no ethnic 

spatial pocket is encircled by another one. Out of the seven imaginary 

patterns in circles in which four ethnic spatial pockets are present, one 

(Circle D) results in six ethnic spatial encounters, four result in five 

encounters (Circles E, F, G, and H), one circle (I) results in four 

encounters, and one (J) results in three encounters.  

 

 

Figure A-1.1. Four ethnic pockets result in additional encounters 

 

These patterns in circles (see Figure A-1.1) show that when there are four 

ethnic spatial pockets, there is at least one additional encounter in five out 

of seven cases (more than 71% of the cases). Conversely, this also means 

that there are at least four ethnic pockets when there are five ethnic 

 

343 

encounters, because three ethnic spatial pockets can never result in five 

ethnic spatial encounters. It can be argued that five ethnic spatial 

encounters, whether resulting from four or five ethnic spatial pockets, 

always result in a certain degree of fragmentation and, therefore, this is a 

good criterion of mosaicness, provided that the area in which they are 

located is relatively small.  

 

How large (or more precisely, how small) should an area be in which five 

ethnic spatial encounters are present? In order to determine this, one 

should first determine how large an average ethnic spatial pocket should 

ideally be. Given the fact that mosaicness requires that at least four of 

them should be present in a small area, these ethnic spatial pockets 

themselves should not be large. 

One can take Switzerland (see Figure A-1.2) as a model and by 

that recognize as a mosaic type of ethno-geographic configuration 

geographic areas with similar features of ethnic fragmentation. One can 

also build an instrument by selecting relatively small but ethnically 

compact areas. Selecting territorial units from the former Soviet Union is 

a good strategy for this purpose. These are often small territories which 

are designated as the homelands of a certain ethnic group that is 

concentrated there.

203

 The titular ethnic groups do not necessarily 

constitute the majority of the population in these territorial autonomous 

units. However, knowing their relative share in a territory’s population, 

one can calculate how large a homogeneous ethnic spatial pocket would 

be if that particular ethnic group lived compactly. These territorial units of 

the former Soviet Union should not be very large. All autonomous ethnic 

territories, whether SSR, ASSR, AO, or NO, from the former Soviet 

Union are selected which are smaller than 70,000 km

2

 and have less than 

4,000,000 inhabitants. Then it is calculated how large the titular groups’ 

share of land would be if the land in that territorial unit was distributed 

proportionally to their share in the total population (PROPET in Table A-

1.1). In this way it can be calculated how large an average homogeneous 

ethnic spatial pocket would be and, therefore, also the area in which four 

such pockets would be located. Assuming that the area is circular, the 

radius of such an area can also be calculated. 

 

The calculations (Table A-1.1)

204

 show that an average ethnically 

homogeneous spatial pocket is 14,427.15 km

2

 large. Hence, an area 

containing four such ethnic spatial pockets will be 57,708.6 km

2

 large. It 

is more practical for the purpose of (manual) measurement if such an area 

                                                 

203

 Only the Jewish Autonomous Oblast’ (Birobijan) in the Russian Far East is not included, because it 

did not contain the largest concentration of Soviet Jews. 

204

 Different sources may present slightly different values. However, these differences do not have 

significant consequences for our calculations. 

 

344 

is a circle. If we assume it to be circular, such an area should have a radius 

of 135.5 km. In order to make the calculations and measurements easier, 

we take 130 km as the radius of this circle. This is justifiable because we 

have already rounded up many numbers in the calculation of the 

“vastness” or “smallness” of this area. Moreover, 5.5 km does not produce 

much difference in the real world. Such a circle will have an area of 

53,092.92 km

2

. 

Such an area can be qualified as a mosaic type of ethno-

geographic configuration if at least five ethnic spatial encounters are 

located in it. An instrument for the measurement of mosaicness can be 

built by dividing the geometric area by the number of ethnic spatial 

encounters. (In the following discussion about calculations, an area means 

a geometric area.) This calculation will result in a large number. In our 

example, it is 10,618.58. Therefore, in order to work with more “usable” 

numbers, we divide the result by 1,000. Hence the formula for the 

measurement of mosaicness will be as follows: 

 

₥= area/(number of ethnic spatial encounters * 1000) 

 

The degree of mosaicness in our example is then ₥= 10.62. This value 

would have been 11.54 if the area had not been corrected, indicating that 

the correction of the value of area—by taking a smaller radius—for the 

sake of practicality does not make a major difference. As ₥= 10.62 is the 

upper limit of mosaicness, this value and any value below it will be 

considered as a mosaic type of ethno-geographic configuration. The 

formula shows that the value of ₥ will be smaller if the number of 

encounters is larger or the area is smaller. 

In Switzerland, with an area of 41,285 km

2

 and four ethnic spatial 

encounters, the degree of mosaicness is ₥= 41,285/(4*1000)= 10.32. This 

value is close to, but clearly below, the upper limit of ₥= 10.62. 

Therefore, the instrument determines that Switzerland is a mosaic area. 

This is yet another reason that this instrument is a good one for the 

measurement of mosaicness.  

Any ethnic spatial encounter located in Switzerland is located in a 

mosaic type of ethno-geographic configuration, as the measurement of ₥ 

for most encounters will most likely result in even smaller values than 

10.32 (see the method of measurement discussed below). Although this 

value falls within the limits of mosaicness, Switzerland contains only four 

ethnic spatial encounters. However, there are more ethnic spatial 

encounters between the ethnic spatial pockets in Switzerland and those 

outside, when foreign countries’ territories—for example, Val D’Aosta in 

Italy—are also taken into the measurement and calculation of mosaicness. 

Before doing this, however, we have to answer the question whether 

 

345 

Switzerland’s international borders are hard  or  soft borders. It is not 

permissible and appropriate to measure beyond them if they are hard 

borders, but it is permissible if they are soft. When assumed to be soft 

borders, the larger area of measurement may still be a mosaic area, 

because there are more ethnic spatial encounters in that slightly larger 

area.  

It is always necessary to define beyond which borders stops the 

measurement of mosaicness. In the measurement of mosaicness, one has 

to distinguish between hard and soft borders. Hard borders are those 

borders across which there is little transborder interaction. The Soviet 

external borders are an example of such hard borders. Hard borders in the 

post-Soviet context are the borders between the Soviet successor states 

and states which were not part of the former Soviet Union. Soft borders 

are those borders across which there is much transborder interaction. 

These are the Soviet internal borders and borders between the Soviet 

successor states—usually members of the CIS, an heir to the Soviet 

Union—even many years after the collapse of the Soviet Union. Visa 

regime, marriages, and migrations between these republics bind them 

together for many years, if not decades. 

To determine whether the area around an encounter is of the 

mosaic type of ethno-geographic configuration or not, one has to draw a 

circle with a radius of 130 km around that encounter. As such an 

encounter is rarely a point and as there are usually one or more lines of 

contact, there are more points possible around which a circle can be 

drawn. Therefore, there may be different measurements with different 

values. However, an encounter can be regarded as being located in a 

mosaic area if at least one such measurement determines that it is located 

in a mosaic area.  

In the measurement of mosaicness, one has to distinguish between 

an ethno-territorial encounter and an ethnic spatial encounter. An ethnic 

spatial encounter is slightly different from an ethno-territorial encounter. 

By ethno-territorial encounter is meant a spatial encounter between two 

rooted ethnic groups within one former union republic or country (state). 

(In the text of the body of this book, an encounter simply means an ethno-

territorial encounter.) Technically, all ethno-territorial encounters are 

ethnic spatial encounters, but the latter may reach across soft borders. To 

give an example: if one wants to determine whether the Tajik–Kyrgyz 

ethno-territorial encounter in Tajikistan is located in a mosaic area or not, 

one has to draw a circle around a point on the line of their encounter in 

Tajikistan and then count how many ethnic spatial encounters are present 

in that area across the soft borders—for example, in Kyrgyzstan, and 

Uzbekistan—thereby counting each encounter between the same ethnic 

groups as one ethnic spatial encounter, even when they are located in two 

 

346 

republics. In other words, no attention is paid to the soft borders in the 

measurement of mosaicness. This is understandable because the effects of 

mosaicness are presumably present in an area regardless of any soft 

borders. It is also not important of how many spatial pockets the living 

area (habitat) of an ethnic group consists. The encounter between two 

ethnic groups will be counted once, regardless of the multitude of their 

lines of contact. 

It is possible that a full circle cannot be drawn around an ethno-

territorial encounter. This is the situation when part of the imaginary circle 

may be located across the off-limit areas. In addition to the hard borders, 

the Caspian Sea and Black Sea are also considered off-limits, and the area 

of circles overlapping with them will not be taken into the calculation of 

mosaicness. What should we do in these cases? In practice in the study in 

this book, no difficulties were met because even in such cases it is clearly 

measurable whether such an area is of the mosaic type of ethno-

geographic configuration or not, even if a full circle cannot be drawn. 

However, the best way to deal with this problem is to calculate the area of 

the incomplete circle, count the number of ethnic spatial encounters in this 

area, and using the above-mentioned formula see whether it is equal to or 

below ₥= 10.62.  

A problem may be that the area of an incomplete circle may be 

too small and hence even a small number of ethnic spatial encounters may 

produce a value below the upper limit of mosaicness (₥  ≤ 10.62). 

Although correct according to the formula, it does not seem to be 

justifiable to regard such an area as a mosaic one. In order to solve such 

problems and in order to get more solid results in general, the 

measurement process should be repeated over a wider area (even if we 

were able to draw a full circle). The calculation should proceed at a short 

distance and at a long distance. Hence, a circle with a radius of 260 km 

will be drawn. Using the formula, this area should contain at least 20 

ethnic-spatial encounters when it is a full circle. Although the area of this 

circle covers a geographic area four times larger than the smaller circle, 

the distance from the (point of) encounter is only twice as long. In reality, 

many ethnic spatial pockets lie beyond the limits of the first circle, and 

perhaps more than half of the encounters may repeat themselves in the 

larger area. A mosaic area can border another mosaic area and hence be 

located in a larger mosaic area in only one direction but may border a non-

mosaic area in its other directions (see Figure A-1.3). In Figure A-1.3 the 

smaller circle—itself displaying a mosaic type of ethno-geographic 

configuration—is located in an area which is largely of mosaic type, 

because it displays mosaicness only in its upper side. In practice, it is 

reasonable to regard a mosaic area as mosaic even when the outer larger 

circular area around it is only partially of a mosaic type of ethno-

 

347 

geographic configuration. It is, therefore, justifiable in such cases as in 

Figure A-1.3 to regard the ethno-territorial encounter at the center of 

circles as located in a mosaic type of ethno-geographic encounter. 

Therefore, we set the criterion of the number of ethnic spatial encounters 

lower, to 10. In a circle with a radius of 260 km, there should be at least 

10 ethnic spatial encounters in order to call it a mosaic area; and in cases 

where it is not a full circle, it has to have a degree of mosaicness ₥= 

21,2371.66/(10 * 1000) = 21.24 or below (that is, ₥  ≤ 21.24). A 

precondition is that that the encounter is already located in a mosaic area 

measured at short distance (₥ ≤ 10.62). 

To recapitulate the main points: an ethno-territorial encounter is 

located in an area which can be typified as a mosaic type of ethno-

geographic configuration when—using the formula ₥= area/(number of 

encounters * 1000)—the value of ₥ is equal to or lower than 10.62 in a 

circular or partially circular area with a radius of 130 km, in addition to 

having a value of ₥ equal to or lower than 21.24 in a circular or partially 

circular area of 260 km.  

 

 

 

 

 

Figure A-1.2. Ethnic distribution in Switzerland 

 

 

 

 

 

 

348 

                             

 

 

Figure A-1.3. Mosaicness in a small and larger circular area 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

349 

Table A-1.1. Administrative units and PROPETs

 

ADMINISTRATIVE 

UNIT 

TITULAR 

ETHNIC 

GROUP 

% TITULARS 

IN THE 

POPULATION 

AREA 

KM

2

 

PROPET 

KM

2

 

Abkhazia Abkhaz 

18  8,432 

 

1,518 

Adygheya Adygheyans 

(Circassians) 

22 7,600 

1,672 

Aga Buryatia 

Buryats 

55 

19,000 

10,450 

Armenia Armenians 

93  29,740 

27,658 

Checheno-Igushetia Chechens 58 

19,300 

11,194 

Checheno-Igushetia Ingush 

13 

19,300 

2,509 

Chuvashia Chuvash 

68  18,300 

12,444 

Dagestan

205

 Avars 

28 

50,300 

14,084 

Estonia Estonians 

62 43,432 

26,928 

Gorno-Badakhshan Pamiris 61 

64,200 

39,162 

Jewish (Birobijan) 

Jews 

4 

36,300   1,452 

Kabardino-Balkaria Kabardin 

(Circassians) 

48 12,500 

6,000 

Kabardino-Balkaria Balkars  9 

12,500 

1,125 

Karachayevo-Cherkessia Karachay 

31 

14,100  4,371 

Karachayevo-Cherkessia Cherkess 

(Circassians) 

10 14,100 

1,410 

Khakassia Khakas 

63  61,900 

38,997 

Komi-Permyak  

Komi-

Permyak 

60 32,770 

19,662 

Latvia Latvians 

52 

64,589 

33,586 

Lithuania Lithuanians 

80  65,200 

 

52,160 

Mari El 

Mari 

43 

23,200 

9,976 

Mordovia Mordovins 

33  26,200 

8,646 

Nagorno-Karabakh Armenians 

77 

4,400 

 

3,388 

North Ossetia 

Ossetians 

53 

8,000 

4,240 

South Ossetia 

Ossetians 

66 

3,900  

2,574 

Tatarstan Tatars 

49  68,000 

33,320 

Udmurtia Udmurtians 

31  42,100 

13,051 

Ust Orda Buryatia 

Buryat 

36 

22,100 

7,956 

Total PROPET 

  

 

389,533 

Average area PROPET 

  

 

389,533/27 

= 14,427.15 

4x larger PROPET 

  

 

57,708.6 

Mosaicness (5 

encounters) 

  

 

11.54 

Radius when the area is 

assumed to be circular 

 

 

 

r = 135.5 

km 

 

 

 

                                                 

205

 There is no titular group in Dagestan; therefore, we chose the Avars, the largest Dagestani native 

ethnic group. 

 

350