Gordon Shepherd and Gary Shepherd, A Kingdom Transformed:
Early Mormonism and the Modern LDS Church
Gordon Shepherd and Gary Shepherd. A Kingdom Transformed: Early Mormonism and the Modern LDS Church. Second edition. Salt Lake City: University of Utah Press. 2016.
Shepherd and Shepherd’s latest edition of their ambitious study of LDS Church leaders’ rhetoric is a significant scholarly achievement. The original study covered the period from 1830 to 1979 and offered the reader a sweeping account of the ways in which church leaders have adjusted their leadership strategy since the early days. The second edition of A Kingdom Transformed adds content addressing the period from 1980 to 2009.
This book offers an important perspective on the maturation of the LDS Church from the lens of the rhetoric of its highest leaders. In what could only have been a monumental task, the authors (incredibly without the aid of research assistants) hand-coded a random sample of general conference addresses to create quantitative measures of themes and trends in the topics addressed at church conferences. Shepherd and Shepherd have the data to show when and how church leaders changed their emphasis from distinctive and embattled religion at the margins of society to the mainstream religion of the twenty-first century that we see today. One of the most striking findings of their content analysis documents the dramatic decline of “utopian” rhetoric in general conference addresses from a high point in the early Utah period until the First Manifesto. The analysis shows that utopian themes in general conference almost entirely disappeared in the modern era. Their data also show that emphasis on more traditional themes increased during that same period. For example, compared with earlier eras, general conference addresses given in the last part of their data were much more likely to focus on Jesus Christ. This “mainstreaming” of the LDS Church has been qualitatively described in other work,1 but the authors’ quantitative evidence is particularly compelling. During the same period that Mormon leaders were deemphasizing utopian themes there was a dramatic decline in worry about dissenters.
In addition to documenting important areas of change over the course of the church’s history, the authors show that several themes remained relatively constant during the period under study. For example, their data do not show much shift in the emphasis on individualism through the years (although the accompanying decline in emphasis on social reformation means that in relative terms individualism is ascendant).
The process of translating unstructured, qualitative texts to quantitative measures is, as anyone who has attempted it can attest, fraught with challenges. In everything except for the latest update, this was done with human eyes reading the sermons and coding their content in accordance to some agreed-upon set of dimensions. For example, Shepherd and Shepherd document a decline in the proportion of talks over the general conference pulpit that emphasized utopian themes. This would have required human coders to recognize and agree upon certain cues in the text that correspond to these themes. The authors then take the mentions in each talk and use them to create scores2 that can be tracked over time, so the main evidence that the reader is presented with in the text has been doubly abstracted from the sources. First was the process of translating the topics of the talk into categories, and second was the process of translating those topic mentions into quantitative indicators.
I would have liked to see more discussion of the uncertainty associated with the measures given in the book.3 It was difficult for me to judge in places what was ‘signal’ and what might have been ‘noise.’ Uncertainty might have crept into the process at any point. First, the texts were sampled from the larger population of general conference addresses. Different samples would have produced slightly different measures. Second, there is plenty of opportunity for human error in a large-scale coding project like this one. Finally, the choices made in scale construction have consequences for the interpretation of the measures.
The second methodological critique deals with bridging the earlier part of the analysis with the additional analyses added for the second edition. The authors apply a new coding method to the last thirty years based on the LDS General Conference Corpus maintained at Brigham Young University. This corpus did not exist when the authors first conducted their study, but the complete digitization and indexing of general conference addresses into a scholarly linguistic corpus allowed Shepherd and Shepherd to bypass the need for human coders for the second part of their project. Using the insights gleaned from the first part of their study, the authors are able to extend the analysis forward with a greatly reduced labor cost.
The authors note that the new measures are not directly comparable to the older measures since both are based on different things. This is fine, but it seems that it would have been relatively straightforward to apply the same method they use on the last thirty years to the earlier parts of the corpus as well. This would permit us to see directly comparable measures between the newer and older parts of the data. It would also have had the effect of replicating the analysis performed in the original study. To the extent that their findings from this replication agreed with the findings from the earlier part of the study, we could be relatively more confident that the patterns were genuine findings rather than statistical artifacts. To the extent that there were substantial differences, we might be led to question the validity of the corpus-derived measures.
In addition to my methodological quibbles, I also have some minor substantive complaints about the study. First, and perhaps as a consequence of the sampling strategy, the authors consistently treat the church leadership as a monolithic entity. This is fine for the broad patterns that Shepherd and Shepherd are interested in, but I consistently found myself asking who was responsible for the trends they demonstrate.4Are new cohorts of General Authorities changing the overall distribution of topics addressed in general conference, or do we see changes in individual leaders’ rhetoric? Church leaders generally have long tenures. Some quick calculations tell me that the average term of full-time church service of an LDS apostle hovers around thirty years, with a quarter of apostles serving more than forty years. This is long enough that we might expect to see generational differences within the highest leadership of the church. It also might be true that where a particular leader sits has some influence on the stands he takes in general conference addresses. It would be interesting to know if there are distinguishable differences between leaders’ addresses based on their position (First Presidency, Seventy, etc.). It is also most likely the case that the context of the session might have some bearing on the content of the speeches made. For example, in the modern era (at least from my own experience), priesthood sessions contain different kinds of language than what is heard in the Sunday morning session
Second, I would have liked to see a little more discussion of the reach and influence of general conference over the years. How has the meaning of general conference changed over time? The authors do give a brief history of the institution of general conference in the LDS Church, but they do not seem to go deep enough in exploring the implications that this history has for their study. I imagine that changes in the audience and reach of the conference itself are responsible for some of the changing trends that the authors demonstrate to be independent of actual changes in focus or leadership style from the leading brethren. There are some obvious differences in the freewheeling, relatively unconstrained conferences before the sessions were transmitted to a wider audience and stricter time limits were imposed. Within living memory, conference addresses were accessible to the general membership everywhere outside Salt Lake City only through their publication in church periodicals.
It would be fascinating to conduct a parallel analysis to the authors’ study of general conference rhetoric with a corpus of local sermons. Even though this parallel study would be impossible, I would have liked to see some discussion of the extent to which the emphasis and direction on particular topics from top leaders affect the general membership in the pews and how this might have changed over time. Is the contemporary church more aligned with Salt Lake in terms of its general emphasis than were earlier generations of Latter-day Saints? How much of this increased alignment (if it has indeed increased) can be attributed to the accommodationist trend over the past 150 years? Such questions cannot be answered empirically in the same way, but thinking through the social and technological differences between the contemporary membership and the membership in earlier times might prove fruitful.
Third, there were points in the book where I was a bit overwhelmed with the sociological theories that were informing the scholarship. Shepherd and Shepherd rooted their work deeply in sociological theories of religion, and they have the citations to prove it. Coming from a different academic background, I found some of the theoretical background interesting, but it is too thick in places and distracts from the overall flow of the study.
In sum, this work is a valuable contribution to the study of the history of the church, and it would be a welcome addition to any Mormon studies scholars’ libraries. The aims of the book are also more general than accounting for the history of the LDS Church. In addition to adding to our understanding of the development of the church, the authors go to lengths throughout the book to connect the evidence from the Mormon case study to broader theories of religious development and change. The end product is a carefully researched and rigorous account that places the LDS Church in context.
Shepherd and Shepherd have produced a remarkable piece of scholarship in A Kingdom Transformed. By zooming out and examining aggregate trends in emphasis over time, they have provided a real contribution to the study of the development of the church, and one can hope that the authors or others picking up their legacy will continue to update this important study going into the future.
Bradley M. Jones is a research associate at the Pew Research Center in Washington, DC. His work focuses on American politics with a specific emphasis on political psychology. He received his PhD in political science from the University of Wisconsin–Madison.
1. For example, Armand Mauss’s classic The Angel and the Beehive (1994) or Thomas Alexander’s Mormonism in Transition (1986), to name just two.
2. The authors rely upon “salience scores.” These are numbers ranging from 0 to 1 derived by counting the number of paragraphs in a talk that mention a certain theme and dividing that count by the total number of paragraphs in the address.
3. This wouldn’t necessarily require any new data collection. Much could be done through bootstrapping—sampling with replacement from the original data—to quantify the expected level of uncertainty in the data. Sampling either from the codings provided by the researchers or from the documents selected (or perhaps some combination) should yield some measure of the potential variability that was a function of the particular selection of texts sampled.
4. The speaker-level analysis I am suggesting here would be much easier to carry out on the corpus that contains the universe of all general conference addresses that the authors use for their update.