From Reductionism to Complexity: A Digital Corpus for Sufism Using a NoSQL database for studying the emergence of complexity

Efforts to measure the
organization of human activity in terms of complexity
have a long history (Chick 1997
).
The recent
proliferation
of
computational tools has
accelerated
the creation of
rigorous
means of modeling and assessing

the development of
human
collectivities
,

with particular reference to
the

emergence
of
social
complexity

(
Sawyer 2005;
Miller and Page 2007:
Murphy and Stoeger 2007;
44ff
;

Jörg 2011:
36ff
)
.

This paper argues that
r
ecent studies of historical social complexity have
neglected to account for the

emergence
of social complexity

because of a reliance

on
a reductionist approach.
In place of the reductionist
approach
, we
put forward a prototype of a NoSQL database architecture for studying the emergence of populated with data from historical Arabo-Islamic sources.

In the face of pervasive critiques of narrativist historiographical techniques in historical scholarship (White 1973; Clark 2004),
computationally-oriented approaches to human collectivities

have turned to more “scientific” approaches. Frequently, this involves the use of a
reductionist method, breaking down various non-narrative data related to a collective into various constituent parts under the belief that each of these can be scientifically described
and analyzed
(Gallagher and Appenzeller 1999)
.
R
ecent studies by Preiser-Kappeler (2018) and Turchin
et al
(2017) exemplify this approach.
The former paper applies statistical tests to a single, node-level variable (computed from “distance decay effect” between cities) to model the long-term “resiliency” of transport networks in the pre-modern Roman and Chinese empires, whereas the latter proposes nine “complexity characteristics”
that are coded into an RDF (graph) database, the analysis of which reveals a
“single dimension” that accounts for measures of complexity across a variety of civilizations.

E
ach of these studies makes a successful, independent case for the evolution of complex social structures;
the use of a RDF database by Turchin
et al
moreover significantly increases the explanatory value of the convincing statistical analyses put forward by Preiser-Kappeler.

However,
the reductionist approach
adopted in each
assumes
, rather than demonstrat
es
, the prior emergence of
these
complex features.
Furthermore, both studies

show how the limitations of some datasets limit the explanatory power of the reductionist approach in the first place. For instance,
Preiser-Kappeler’s
study
also assumes “long term stability of core elements” of the Roman and Chinese empires, which precludes an investigation into the emergence of such “core elements” in the first place. Similarly, T
urchin
et al
’s limitation of analysis to “the appearance of politically centralized societies” (ibid: 2), stands in
stark
contrast to a central tent of theories of complex systems which states that the are composed of “networks of components with no [mechanism of] central control” (Mitchell, 2009: 13). The task of describing the
actual emergence of
complexity
within
human collectives thus seems to require
the following elements
:
(1) time-sensitive
data
related to

(2)
entities that are substantially independent from a central control mechanism, and
(3)
a sufficiently robust
data storage system,
such as a RDF database.

We propose that the above-mentioned conditions can be met using data from a historical Islamic religious movement known as Sufism, which first appeared in Iraq during the 9
th
century (Nicholson 1914; Massignon 1975; Schimmel 1975; Knysh 2010; Melchert 2013).
No study of this movement has given a conclusive answer as to the emergence of Sufism, largely because previous research has failed to analyze a key feature of early literature written by Sufis: the
“pathway of transmission” (Ar.
isnad
; lit. “citation”) used to introduce narrative
content. An example of this form would appear as follows: I heard [X] say he heard [Y] say that he heard [Z] say […].” Although there have been numerous attempts to undertake the analysis of abundant relational data found in non-Sufi Islamic sources (Şentürk 2005; Romanov 2014), these data are either not publicly available, or exclude the narrative component of these data because of the limitations of relational (SQL) databases in which they were compiled.

In this paper, I outline the architecture needed to leverage the complexity of the relational data in early isnad-based Sufi literature with as network data NoSQL database framework. Within this network, the breadth and depth of node level data (i.e. individuals involved in transmitting Sufi sayings) can vary widely with regards to: names, occupations, birth and death dates, destinations of travels, teachers, students, intellectual specialties, and matrices that record other figures’ opinions about the figure, to name only a few. On the edge level (i.e. the structure of the isnads that connect the nodes), features are more standard, and can include: place of transmission, date of transmission, method of transmission (via book, face-to-face meeting, etc.), as well as the narrative content of the saying that is transmitted via the isnad. At present, the architecture of these data is being prototyped using the data from one early Sufi work (al-Khuldi 2011) on the MongoDB NoSQL database. This decision was based on both the flexibility of the JSON scheme that is native to MongoDB, as well as the ability to easily transfer this data onto a graph database scheme, such as Neo4J.

Bibliography

Bol, P. (2012). GIS, Prosopography, and History.
Annals of GIS 18(1): 3-15.

Chick, G. (1997). Cultural complexity: The Concept and its measurement.
Cross-Cultural Research 31(4): 275-307.

Clark, E.A. (2004)
History, Theory, and Text. Historians and the Linguistic Turn. Cambridge, MA: Harvard University Press.

Gramlich, R. (1995-96).
Alte Vorbilder des Sufittums, 2 vols. Stuttgart: Franz Steiner.

Kantabutra, V.
and Owens, J.B. (2014). Intentionally-Linked Entities: A Better Database system for representing dynamic social networks, narrative geographic information and general abstractions of reality. In Solana, S.C. (ed),
Spatio-Temporal Narratives: Historical GIS and the Study of Global Trading Networks (1500-1800). Newcastle upon Tyne, UK: Cambridge Scholars Press, pp. 56-78.

Karamustafa, A. (2007).
Sufism: The Formative Period. Edinburgh: Edinburgh University Press.

Al-Khuldī, J. (2010). al-Fawāʾid li-l-Khuldī bi-riwāyat Abī ʿAlī b. Shādhān. In al-Jarrār, N.S. (ed),
Majmūʿ fīhi thalāth ajzāʾ ḥadīthiyyah. Beirut: Dār al-Bashāʾir al-Islāmiyyah.

Knysh, A. (2010).
Islamic Mysticism: A Short History, 2nd ed. Leiden: E.J. Brill.

Massignon, L. (1975).
La Passion de Hallâj: Martyr mystique de l'Islam. Paris: Gallimard.

Melchert, C. (2013). Quantitative approaches to early Islamic piety. In Klemm, V., al-Shaʿar, N. and Behzadi, L (eds),
Sources and Approaches across Disciplines in Near Eastern Studies: Proceedings of the XXIV UEAI Congress, Leipzig 2008. Leuven: Peeters, pp. 91-100.

Miller, J.H. and Page, S.E. (2007).
Complex Adaptive Systems. An Introduction to Computational Models of Social Life. Princeton: Princeton University Press.

Nicholson, R.A. (1914).
Sufism: The Mystics of Islam. Cambridge: Cambridge University Press.

Preiser-Kappeler, J.
(2018). Networks and the Resilience and Fall of Empires: a Macro-Comparison of the Imperium Romanum and Imperial China.
Siedlungsforschung: Archäologie-Geschichte-Geographie
36: 1-41.

Romanov, M. (2014). Writing the digital history of the pre-modern Islamic world. In Andrews, T.L. and Macé, C (eds),
Methods and Means for Digital Analysis of Ancient and Medieval Texts and Manuscripts. Turnhout: Brepols, pp. 45-68.

Schimmel, A. (1975).
Mystical Dimensions of Islam. Chapel Hill, NC: University of North Carolina Press.

Şentürk, R. (2005).
Narrative Social Structure. Anatomy of the Hadith Transmission Network 610-1505. Palo Alto, CA: Stanford University Press.

Turchin, P.
et al (2018). Quantitative historical analyses uncover a single dimension of complexity that structures global variation in human social organization.
PNAS
Plus: E144-E151.

White, H. (1973)
Metahistory: The Historical Imagination in Nineteenth-century Europe. Baltimore: Johns Hopkins University Press.