Article Image

Cover: Iraq. Kufa. (On the banks of the Euphrates). Once an important Moslem [i.e., Muslim] center of learning on the river Euphrates. 1932 (Library of Congress, LC-DIG-matpc-16145)

Please, use the latest version at

50 seconds of Islamic History

Version 2.2: Regions and Urban Centers (scaled down)

Versions 2.x are based on the data set of about 330 toponyms—all place-names that occur 5 and more times in al-Dhahabī’s Taʾrīkh al-islām. Expanding the list of toponyms  did not significantly affect the number of biographies, which increased by slightly over a thousand: ~13,970 biographies in Versions 2.x versus ~12,850 biographies in Versions 1.x. On the other hand, now it is almost 50% of all the biographies from the considered section of this biographical collection (vols. 4-52:  ~29,000 biographical records).

Cities, their quarters and suburbs are now merged into metropolitan areas (e.g., Baghdad, Nishapur, Damascus, Cairo, Cordoba, etc.). Provinces (not shown separately) and urban centers (black color) are grouped into regions (firebrick color), which makes it easier to see changes on the regional level. In general, regions correspond to major provinces of the Islamic world, but in some cases they include more than one. At the moment grouping is done purely mathematically through the comparison of distances between urban centers and possible central points of regions. This resulted in some minor distortions, but will be fixed in the following versions. (For example, Basra happened to be grouped with the province of al-Ahwāz/Khūzistān, instead of al-ʿIrāq, to which it belongs.)

Another temporary issue is that the data set still includes some false positives, since at the moment there is no differentiation between homonyms, for example, al-Sind (modern Pakistan) and al-sanad (“the chain of transmitters”), or Tripoli in North Africa and Tripoli in Syria. However, these are not toponyms with high frequencies, so this should not significantly affect the maps.

The Version 2.x maps do not seem to change the overall picture in any radical way (it may happen with toponymic nisbas added to the equation). Nonetheless, the new maps are much more detailed. Perhaps even too much, hence, three versions. Version 2.0 includes both urban centers and regions—this Version gets particularly cluttered, but may still be useful. Version 2.1 features only urban centers. Version 2.2 features both, but urban centers are scaled down (frequencies divided by 4), which allows to see both regions and urban centers that represent these regions. (Best viewed in HD and full screen)

Version 2.0: Regions and Urban Centers
Version 2.1: Urban Centers
Version 2.2: Regions and Urban Centers (scaled down)

Current Dataset:
Sources: Taʾrīkh al-islām of al-Dhahabī
Period:  41-700 AH / 661-1301 CE (Volumes 4-52)
Biographies: ~13,970 (out of ~29,000)
Toponym Mentions: ~ 29,610

NB: Nisbas—or “descriptive names”—that point to the geographical origins of individuals are not included into this visualization. (Examples of geographical nisbas: the name al-Baghdadi tells that a person was strongly affiliated with the city of Baghdad, al-Dimashqi—with the city of Damascus, al-Naysaburi—with the city of Nishapur, etc.)

Please, use the latest version at

Posted by Maxim Romanov

Research fellow (PhD in Near Eastern Studies, U of Michigan, 2013) at the Humboldt Chair for Digital Humanities [Institut für Informatik], University of Leipzig. He studies Islamic historical texts with computational methods, currently focusing on the analysis of multivolume biographical and bibliographical collections.