Justin E. Lane
Generally my interests are broad but interrelated. My research program concerns the stability of social systems from an evolutionary/economic perspective. More specifically, I'm interested in how social groups maintain stability and the evolutionary variables that contribute to their stability/instability, particularly focusing on new religious movement ritual stability and socio-economic variables concerning sectarian violence. I hope to model these variables and their related cognitive mechanisms to create more accurate social artificial intelligence for the modeling of complex social environments.
Supervisors: Prof. Harvey Whitehouse
Supervisors: Prof. Harvey Whitehouse
less
InterestsView All (39)
Uploads
Current research concerning the nature of human sociality has largely concentrated on the psychological mechanisms humans utilize during communication and cooperation. This provides an empirical basis for studying humanities most unique social phenomena: religion. This paper presents the current status of an ongoing research collaboration between cognitive scientists, computer scientists, and archaeologists. It defends that using empirical lab studies of human cognition is not sufficient to speak to larger social phenomena such as religion. These empirical results must be re-incorporated into social systems. We argue that this is best done by multi-agent AI (MAAI) modelling. However, even with empirically backed MAAI models run on some of the world’s fastest computers, these models are still largely hollow because they are not constrained by any historical reality. In this regard, our research team is utilizing archaeology to inform both the creation and behavioural constraints of a new agent-based model aimed at simulation large-scale religious shifts in the neo-lithic and medieval periods.
As such, this research directly engages with both historical, archaeological, cognitive, and agent based approaches to social complexity. At its foundation is a mixture of social network and cognitive science on one hand and historical and archaeological data on the other. By using new computational approaches it seeks to seamlessly weave the two together in order to produce novel insights into the history and dynamics of religion.
Two approaches are presented in this analysis. The first is a standard frequency analysis of terms that are known to have specific psychometric correlates (Pennebaker, Chung, Ireland, Gonzales, & Booth, 2007; Tausczik & Pennebaker, 2009). The second analysis utilizes a network approach which discerns which concepts frequently occur within texts. This analysis can be used to discern which concepts are most important to a religious community, how they are used, and what other significant concepts are closely connected to them (Carley, 1994).
This research will present both data analysis styles to help elucidate the differences in beliefs as presented through data collected from contemporary Christian communities in Singapore and compare them to data collected in the United Kingdom and the United States. The presentation will conclude by addressing the significance of this finding in relation to previous models of ritual behaviors (Whitehouse, Kahn, Hochberg, & Bryson, 2012) and how this approach can be utilized within the broader context of computer modeling in the cognitive science of religion with a specific focus on the analysis of historical materials.
This paper will present critiques of models that have already been developed as well as those still in development and how close attention to both method and theory allow for computer modeling to supplement the toolbox of cognitive scientists interested in human social phenomena.
I will present a brief examination showing that these methods are applicable with the theoretical commitments of the “standard model” and will present new research in the use of artificial intelligence models and their use for studying religion as a human social phenomenon produced primarily by natural cognitive mechanisms. The first example will show how prospect theory and in-group biases can result in social patterns similar to those noted by Turchin (2006). The second will show how an artificial neural network can be used to test the claim that the variables of the modes theory are related by dynamic (i.e. non-linear) relationships (Whitehouse 2004).
from online and real-world sources. This can include interviews, online social network data, books, news
reports, and transcripts. The first technique analyzes texts for their psychometric properties. This utilizes
a software program known as LIWC, which codes texts using psychometrically validated measurements
drawn from human populations. This allows us to understand emotional and social aspects of a text that
may not be accessible from other forms of text analysis. The second technique covers semantic network
analysis. This technique quantifies texts based on the relationships between concepts in the text,
resulting in quantified network based schemas that represent the text; these networks can be analyzed
and compared in order to understand the importance of concepts in the text or how different datasets
are related. The presentation will introduce the necessary concepts and software that is available
through simple hands on training. The goal is not immediate expertise, but providing the necessary
knowledge to experiment start using these techniques for your own analyses. The presentation will also
provide examples and use cases.
In the interview, he touches on an aspect of the scientific study of religion that I would like to highlight: explanatory pluralism. I want to use this opportunity to offer a critical review of CSR in light of explanatory pluralism. It is my belief that the failure of CSR to adequately address its inherently interdisciplinary nature has been a detriment to the field and that by addressing these issues it will help the field to grow as well as to help non-CSR specialists understand more of the subtlety of this scientific approach to our subject. I, by no means, think I can settle the issues in the space here, but I would like to use Dr. McCauley’s interview as a springboard from which a discussion can be launched.
The modeling religion project (MRP) has begun creating advanced computer models that rely upon these psychological proclivities in order to study and predict aspects of religion. These include—but are not limited to—secularization, conversion, and reaction to threatening environments (such as natural disasters or social instability). In the limited time provided, I will discuss how psychological constraints can serve to constrain or revise our approach to predicting religious demographic shifts by concentrating on how secularization and conversation affect religious demographics.
By exploring these two fundamental aspects of human religiosity with tools drawn from the bio-cultural science of religion, we can develop more accurate models of demographic shifts from the data we currently have. I will present results from two computer models: one which shows how individuals come to adopt new religious identities, resulting in conversions, schisms, and endogenous theological adaptations; and one which shows that secularization processes may happen faster than previously expected, but may be temporary. We argue that incorporating causal models, rooted in the bio-cultural sciences, into studies of religious demographics will provide better conclusions than those models currently in use.