Medieval Russia trade network

May 2018

pitts.zip

• pitts.net - Pajek network file, undirected, simple, n=39, m=55
• pittsGeo.net - pitts.net with (Lon, Lat) as coordinates
• pittsCyr.net - pitts.net with labels in Cyrillic
• pitts.nam - names of places in Cyrillic
• pitts.csv - merged tables from the paper (see Combining tables in R)
• pits2.csv - pitts.csv extended with names in Cyrillic and geographical coordinates
• pittsMat.dat - R assignment with adjacency data
• pittsGeo.dat - nodes with coordinates (Lon, Lat)

In the paper (1, 2, 3)

Forrest R. Pitts: The Medieval River Trade Network of Russia Revisited. Social Networks, 1 (1978/179) 285-292

the medieval river trade network of Russia was presented and analyzed

explaining the rise of Moscow

See also

Forrest R. Pitts: A GRAPH THEORETIC APPROACH TO HISTORICAL GEOGRAPHY. The Professional Geographer. Volume 17, Issue 5 September 1965 Pages 15-20 DOI

On the Icon network index I found a version of Pitts data. I decided to complete it with labels and tables from the paper. It turned out that in a given list of edges the edges (28:29) and (26:39) are missing.

> getwd()
[1] "C:/Users/batagelj/Documents/papers/2018/moskva/NetR/nets/padgett"
> t1 <- read.csv2("../rivers/table1.csv",skip=2)
> t2 <- read.csv2("../rivers/table2.csv",skip=2)
> pitts <- cbind(t1,t2)
> head(pitts)
  Connection pConnection               Place ShortPath   Rank n               place INOR ASPdist pINOR pASPdist
1       2652        0.51            Novgorod       176     35 1            Novgorod    0    7179  0.00     4.11
2       3574        0.68             Vitebsk       172     32 2             Vitebsk   11    5647  0.22     3.23
3       9521        1.82            Smolensk       148  17.50 3            Smolensk  119    4247  2.39     2.43
4       5303        1.01                Kiev       173     33 4                Kiev    0    7704  0.00     4.41
5       8691        1.66           Chernikov       181     36 5           Chernikov   30    6244  0.60     3.57
6      10181        1.95  Novgorod Severskiy       162     28 6  Novgorod_Severskiy   96    5052  1.93     2.89
> rivers <- pitts[,c(6,1,4,5,8,9)]
> rivers$Rank <- as.numeric(as.character(pitts$Rank))
> head(rivers)
  n Connection ShortPath Rank INOR ASPdist
1 1       2652       176 35.0    0    7179
2 2       3574       172 32.0   11    5647
3 3       9521       148 17.5  119    4247
4 4       5303       173 33.0    0    7704
5 5       8691       181 36.0   30    6244
6 6      10181       162 28.0   96    5052
> write.csv2(rivers,"../rivers/pitts.csv",row.names=t2$place)
> pitts <- read.csv2("../rivers/pitts.csv",row.names=1,skip=2)
> head(pitts)
                   n Connection ShortPath Rank INOR ASPdist
Novgorod           1       2652       176 35.0    0    7179
Vitebsk            2       3574       172 32.0   11    5647
Smolensk           3       9521       148 17.5  119    4247
Kiev               4       5303       173 33.0    0    7704
Chernikov          5       8691       181 36.0   30    6244
Novgorod_Severskiy 6      10181       162 28.0   96    5052
> plot(pitts$INOR,pitts$ASPdist,pch=16,col="red",xlim=c(-50,600))
> text(pitts$INOR,pitts$ASPdist,row.names(pitts),cex=0.6)

I collected also the data about the Cyrillic version of the names of towns and their geographical coordinates (1, 2). It turned out that there are many mistakes in spelling of the names. I corrected them and replaced in the Pajek data set.

Adding new data to tables:

> t3 <- read.csv2("../rivers/geonames.csv",header=TRUE,skip=1)
> t <- read.csv2("../rivers/pitts.csv",row.names=1,skip=2)
> t$cyrillic <- t3$cyrillic
> t$lat <- as.numeric(as.character(t3$lat))
> t$lon <- as.numeric(as.character(t3$lon))
> head(t)
                   n Connection ShortPath Rank INOR ASPdist                            cyrillic      lat      lon
Novgorod           1       2652       176 35.0    0    7179                    Новгород 58.55000 31.26667
Vitebsk            2       3574       172 32.0   11    5647                      Витебск 55.19167 30.20556
Smolensk           3       9521       148 17.5  119    4247                    Смоленск 54.78278 32.04528
Kiev               4       5303       173 33.0    0    7704                            Киев 50.45000 30.52333
Chernikov          5       8691       181 36.0   30    6244                  Черни́гов 51.50000 31.30000
Novgorod_Severskiy 6      10181       162 28.0   96    5052 Новгород-Северский 51.98333 33.26667
> row.names(t) <- t3$place
> write.csv2(t,"../rivers/pitts2.csv",row.names=t3$place)

> nam <- as.vector(t3$cyrillic)
> Encoding(nam) <- "UTF-8"
> plot(pitts$INOR,pitts$ASPdist,pch=16,col="red",xlim=c(-50,600))
> text(pitts$INOR,pitts$ASPdist,nam,cex=0.6)
> svg(filename="../rivers/INORdist.svg",width=10,height=8,pointsize=12)
> plot(pitts$INOR,pitts$ASPdist,pch=16,col="red",xlim=c(-50,600),main="Pitts")
> text(pitts$INOR,pitts$ASPdist,nam,cex=0.6)
> dev.off()

Notes: Ksnyatin → Sknyatino (WP); Pereslavl → Ryazan (WP)

C   Yuryev-Polsky                                     56.5       39.683333
B   Pereslavl-Zalessky       Переславль-Залесский     56.733333  38.85

@article{PITTS1978285,
title = "The medieval river trade network of Russia revisited",
journal = "Social Networks",
volume = "1",
number = "3",
pages = "285 - 292",
year = "1978",
issn = "0378-8733",
doi = "https://doi.org/10.1016/0378-8733(78)90025-4",
url = "http://www.sciencedirect.com/science/article/pii/0378873378900254",
author = "Forrest R. Pitts"
}

Add geographical coordinates of towns and their names in Cyrillic. (done May 12, 2018)

Compute the distance matrix based on geographical coordinates and use it in centrality measures (done May 2018; betweenness and closeness based on real distances put Moscow in the first position).