From the Teaching column in the March 1997 Perspectives
Students as Historians: Lessons from an "Interactive" Census Database Project
Trudi Johanna Abel, March 1997
Just how much of the "new history" is new? Though students increasingly learn about the diverse peoples who lived in the United States, their classroom experience is still, for the most part, quite traditional. Although textbooks have changed over the years to include historiographical debates, reproductions of primary sources, and narratives about people "on the bottom rail," they do little to teach students the process by which history is written. Today's students are presumed to gain a sophisticated understanding of historiography because textbooks present competing interpretations of historical events such as the dropping of the atomic bomb. In fact, today's students are not equipped to evaluate these interpretations because they have little idea of how these narratives developed.
One way to remedy these deficiencies is to restructure introductory U.S. history courses so that students, even those who are not history majors, learn the fundamentals of historical investigation. A census database project that I created at Williams College gives students new skills in the fields of history and technology and, more important, allows students in the humanities, social sciences, and natural sciences to assume the role of both historian and detective.
I first introduced the census database project in a 19th-century cultural history class, where it spanned 4 weeks of our 12-week semester. As students read works that explored the social, religious, and economic transformations that occurred in 19th-century America, they built a database of a midcentury agrarian community. In the fourth week of the course, students submitted five-page papers that were based on data from the federal population census schedules.
The following spring, I revised the census database project and incorporated it into a course on culture and technology in Victorian America. Here the project served as the backbone of a first-year seminar in which students wrote several short papers and 2 drafts of their 12-page research papers. This final assignment required students to base their research on evidence from the census database and on complementary sources. Most of the examples cited in this essay are drawn from this course.
Though I developed the census database project in courses with 12 to 25 students, there are substantial pedagogical advantages to including the project even in a large survey course. Students in one section of a general survey might build a database for a given town in, say, 1860, while those in another section could complete a database for a different locale in the same year or for the same locale in 1920. In drafting their papers, students could compare data from the censuses to chart changes in the local economy, family structure, education, and immigration patterns over time, or they might focus their research on uncovering the past at one historical moment.
The Census as Source
The 1850 federal census was the first national survey to collect detailed information on each individual living in the nation. This information included the name, age, sex, "color," occupation, birthplace, and value of real estate owned by each resident. The 1850 population schedules also contained data on children's school attendance, adult illiteracy, recent marriages, and whether persons were "blind, deaf and dumb, idiotic, insane, paupers, or convicts." Revisions to the population schedule in 1860 enabled census takers to enumerate personal property.
Many college and university libraries own microfilm of federal population schedules that specifically relate to the town, county, or state in which the educational institution is situated. If the census for a particular town cannot be obtained through a local library, it can be purchased easily from the National Archives Trust Fund Board or from private commercial vendors.1 The National Archives sells microfilm of the U.S. census from 1790 through 1920 for $23 a reel.
In selecting census materials for my students' use, I consulted Population of the United States in 1860, which provides aggregate figures for the number of inhabitants in every town, city, county, and state in the nation. I searched the volume for towns with a population of approximately 1,200 residents that were located in proximity to the college. (I estimated that students could enter data on 80 to 100 residents without feeling overwhelmed.) I narrowed the choice to Dalton, Massachusetts, for two reasons—one pragmatic, the other pedagogic. Dalton's census taker had extremely legible handwriting that reproduced well on the microfilm printer-reader. In addition, Dalton was an ideal choice because its papermaking industry had drawn the attention of historians of technology. Thus 12 students in my first-year seminar created a database for Dalton, which contained 1,243 residents in 1860.2 Knowing that their research into the life of a mid–19th-century mill town would be grounded in one complete set of data, students could gain satisfaction through creating something whole.
When considering a project such as this, a teacher should assess the availability of computer resources on campus and determine the technical needs of the students. My college owns 40 licenses for FileMaker Pro, a database package that runs on both Macintosh and IBM-compatible computers. This software makes possible the collective creation of one database file. FileMaker Pro allows 15 users to contribute data to one file at the same time. It also contains a password feature that enables me to store the file on the school's network while limiting access to my students. The password feature also permits students to enter records while denying them the privilege of deletion.
The academic computing department of the college set up the first version of the database so that it included all the categories on the original 1860 census as well as some new ones I created. We modeled the "data entry screen" of the database on a page of the census. FileMaker Pro offers its users the option of having a pop-up list in any given field. For instance, one can design the "place of birth" field so that it provides database users with an alphabetical list of place names that recur most frequently in the census manuscript. Thus students who entered data would not have to type "Massachusetts," rather they would merely double click on that particular place name to indicate that a resident had been born there. These pop-up lists reduce the amount of typing while entering data, thereby minimizing the possibility of error.
Early in the course, I assigned each student a set of household numbers and asked that they locate the appropriate pages in the census and bring them to the computer lab for a one-hour session on the database software. During this session, students received an overview of the software. They then transcribed data from their photocopies of the census and keyed in the information. While none of these students had ever worked with a database, each grasped the basic concepts within that one-hour introductory session.
Later in the term, when the database was fully constructed, students returned to the lab to learn the fundamentals of searching, sorting, and retrieving information from the completed census. During this session, I took a few additional minutes to demonstrate how data could be extracted from the census and presented as an image by using a simple graphics package such as Cricket Graph. When placed in a pie chart, data on nativity revealed that 81.7 percent of Dalton's population had been born in the United States, while 18.3 percent came from abroad. Students marveled at the fact that more Dalton residents claimed Ireland as their birthplace than the neighboring state of New York.
It took students just two weeks to enter data on all of Dalton's residents. After the students constructed the database, they met outside of class to proof and edit their work. The result was an impressively accurate database—one that could be used by social historians as well as by genealogists.
During the fifth week of the 12-week course, I asked students to write proposals for their research papers. The assignment asked them to consider what Laurel Thatcher Ulrich calls "the power—and the poverty—of written records."3 I challenged students to construct a narrative from the fragmentary evidence they had encountered in the primary sources. I asked them to assume the role of a historian who is trying to understand mid–19th-century Dalton society.
Most of the participants in the seminar regarded themselves as potential majors in chemistry, biology, mathematics, and environmental science. In fact, of the 11 first-year students enrolled, only 1 suggested he was considering history as a major. Thus these were students for whom history had little appeal. Most viewed the course as a social science requirement.
The paper assignment was both demanding and highly flexible. It required students to do real historical research with several different groups of primary materials while allowing them to define the parameters of their own studies. Indeed, one of the virtues of the assignment was that it allowed all students to find an area of investigation that appealed to them individually. For example, students analyzed Irish migration, widowhood, women and work, changes in farming practices, paper-mill workers, household structures, inheritance patterns, and health and mortality.
The paper assignment asked students to consult data from sources other than the population schedules, such as the nonpopulation census schedules on mortality, industry, agriculture, and social statistics, and local sources such as vital records, probate records, and cemetery data. In a four-hour field trip to Dalton and nearby Pittsfield, students examined these complementary sources. Nine students spent 30 minutes collecting data in the Dalton town cemetery and three studied manuscript data from vital records kept by the Dalton town clerk. The entire class also spent 45 minutes examining genealogical records, 19th-century public-school reports, news clippings, and a microfilm index of probate records at the Berkshire Athenaeum in Pittsfield. We then toured the National Archives-Pittsfield Region, where some students gathered data from nonpopulation schedules, while others consulted an index to naturalization records.4
The trip brought students into the landscape of the town whose history they studied. It also enabled them to appreciate the great variety of sources that can be used in the writing of history. While the trip might seem ludicrous to a professional historian—How much research can one do in a library or town hall in 30 minutes?—the remarkable result was that students found what they needed and were captivated by the array of materials available. Many students were so thrilled by their findings that they made arrangements to return to Dalton on their own so that they could scoop up the remaining fragments of important data.
Four weeks after the field trip, students submitted the first drafts of their papers. In the same week, students met with reporters from the local newspapers. Preparations for the press conference were minimal; a colleague from the Public Relations Office issued invitations to reporters. Simultaneously, I placed students with shared research interests into subcommittees and charged them with creating oral synopses of their research. The students appreciated how their original research enabled them to speak with authority. Much to their delight (and to that of their parents), three local newspapers published stories that detailed the students' findings.
What kind of history did these first-year nonmajors produce? What caliber of work did this experiment generate? A few students sought evidence in local history sources that would help them "get behind the numbers" in the census. Several students, for instance, used genealogical materials, vital records, and probate inventories to investigate women's experiences in Dalton. Because the census did not provide the maiden names of married women or widows, students who wanted to reconstruct the past of these wives and mothers had to consult genealogical files. Some studied widows in an effort to understand how they supported themselves and their children. Others looked at the position of women more broadly. What occupations did women fill? What kinds of employment did they find outside of the home? Which industry gave them the most opportunity for employment? What was the average age for those women who worked outside the home? In what kinds of household structures did widows live?
Several natural sciences students investigated issues in health and medicine. One selected 10 families from the census in an effort to determine "whether socio-economic status has any bearing on a woman's reproductive history, marriage pattern, and overall quality of life." She wanted to assess whether age of marriage and age at the birth of the first child were related to a woman's socioeconomic condition. Guided by her interest in medicine, she investigated how disease affected her selected families and the Dalton community as a whole. To supplement the sparse data from the census, she grounded her study in the birth, death, and marriage records that were held by Dalton's town clerk.
Another student used mortality schedules from the federal nonpopulation census and local vital statistics to examine the health of Dalton relative to other parts of the United States. By comparing the data on disease and mortality (particularly in relation to children) with data collected by the census bureau on the health of the nation, she concluded that Dalton residents enjoyed reasonably good health. "The town averaged about 18 deaths for every 100 people, a statistic comparable to the national average," she noted. "However, Dalton citizens died more often than the national average from age-related causes, like old age and cancer." She posited that Dalton's country setting and temperate environment enabled its residents to live "longer than the average American."
Several students with interests in mathematics applied their quantitative skills to the problem of social stratification in mid–19th-century Dalton. One student used the census to assess the influence that paper manufacturers wielded. This student found that the nine mill owners (0.7 percent of the town's population) possessed 38 percent of Dalton's total assets, including 48 percent of the town's personal wealth. Using local histories and Judith McGaw's study of the Berkshire paper industry, this student argued that the mill owners achieved "power through philanthropy" and thus "influenced community life considerably."5
Other participants in the course used census data to study Dalton's working classes. Research on Irish immigrants revealed that these transplants were heavily concentrated in four fields of employment. Although they composed 11 percent of the total population, they constituted 20 percent of the agricultural labor force, 38 percent of paper mill operatives, and 75 percent of workers in the tanning industry. These quantitative analyses of the Dalton population encouraged students to examine the relationship among class, ethnicity, and social mobility.
Another student pondered the connection between wealth and education. He used 19th-century school reports and secondary literature on education to reconstruct Massachusetts state policy on compulsory schooling. He then studied youth aged 15 to 19 to see who attended school voluntarily, finding that 1 in 4 of the youth born in Massachusetts attended school after age 14, while only 1 out of 10 children born in other states attended school. In trying to explain why households with similar economic resources would send their children to school at different rates, he analyzed the nativity of parents and found that students born in Massachusetts came from households where parents had received the greatest exposure to free public education. His analysis of the parents in these households revealed that nearly 75 percent were born in the state. He then posited that these parents were more likely than those born in other states and in foreign nations to have received schooling and would therefore strive to extend their children's education.
The census database project introduced students to the challenges of historical inquiry. As independent scholars, each student experienced firsthand the joys and frustrations of working with primary sources. Students often encountered adversity when comparing the experience of Dalton residents across class lines. One student's lamentation captured the frustration felt by most of the seminar's participants: "There was much information on the lives of the wealthy, but on the lower class there was next to nothing. ... The poorer the person, and the less well connected, the harder it was to follow her life."
Projects like this fire the imagination of students and make them interested in how history is researched and written. Through developing hypotheses and testing them against the data, students gain a respect for evidence, an understanding of the importance of documentation, and an appreciation for sound argument. Thus they have the potential to be better consumers of history. In doing their own original research, students also gain authority, expertise, and a great deal of satisfaction. "I like the fact that we didn't just sit back and listen, but we participated and did our own historical research," one student noted. "History has been something interactive and rewarding," wrote another.
One of the most beneficial aspects of the census database project was that it made students active participants in their own education. Through collective action, they created one of the main texts of the course—a machine-readable census for mid–19th-century Dalton. The project also fostered a sense of community among course participants. Because the database only reached completion after all students entered their assigned sets of data, students actually depended on one another to a greater degree than in other courses. Several students noted that they forged a common identity with their classmates through the project. In their words, they "bond[ed] as a class." In working with database software, students also acquired new technical skills that they can use in college and in their postgraduate life.
For college and university communities where town-gown tension often exists, such projects can have a salutary effect. For instance, the students who worked on the census database project gained an appreciation for the community that surrounds their educational institution. The community also received benefits because after completing their work, students donated their research to the local history repositories whose sources they had used. Dalton and Pittsfield libraries received Dalton at a Glance, a transcribed copy of the 1860 census with an alphabetical index to all 1,243 of Dalton's residents. In addition, the Williams College archive received a collection of essays with biographical sketches of the contributors, which the students titled "Beyond the Census: An Historical Analysis of Nineteenth-Century Dalton, Massa-chusetts." This collection provides readers with 12 different perspectives of how the raw data of the 1860 federal census—together with vital records, probate records, and secondary materials—can be interpreted and woven into a series of historical essays.
—Trudi Johanna Abel held a National Endowment for the Humanities fellowship in 1996. She is currently a scholar in residence at the history department at Duke University, where she is writing a cultural history of juvenile popular fiction in America. She thanks Shahram Amiri and Mary Beth Jerry of Williams College Center for Computing for their generous assistance. Special thanks to Jim Allison for the crucial support that he gave to this project. She also expresses her appreciation for editorial advice and general counsel received from Robert Blackey, Anne Firor Scott, Judith McGaw, Robert A. Gross, Paul Clemens, Walter V. Hickey, Jacob Abel, and Noah Pickus.
1. Catalogs of National Archives Microfilm with ordering information can be found in the government documents division of most libraries. See, for instance, The 1920 Federal Population Census (National Archives Trust Fund Board, 1991). Census microfilm can also be purchased through a private vendor such as AGLL of Bountiful, Utah, (801) 298-5446.
2. To obtain population figures for towns, counties, and states as well as other statistical information compiled from the census manuscript sheets, consult the United States Census Office, 8th Census, 1860. Population of the United States in 1860: compiled from the original returns of the eighth census, under the direction of the Secretary of the Interior (G.P.O., 1864; reprint, Norman Ross Publishers, 1990). Similar compendiums of statistics for subsequent federal censuses can be found in the government documents division of most libraries.
3. Laurel Thatcher Ulrich, A Midwife's Tale (Vintage Books, 1991), 343.
4. The National Archives-Pittsfield Region is part of the Regional Archives System of the National Archives and Records Administration (NARA). Other regional archives are located in Waltham, Mass.; New York, N.Y.; Philadelphia, Penn.; East Point, Ga.; Chicago, Ill.; Kansas City, Mo.; Fort Worth, Tex.; Denver, Colo.; Laguna Niguel, Calif.; San Bruno, Calif.; and Seattle, Wash.
5. Judith McGaw, Most Wonderful Machine: Mechanization and Social Change in Berkshire Paper Making, 1801–1885 (Princeton University Press, 1987).