This R project examines the first 1/5 of Virginia loyalists who submitted Loyalist Claims to the British Audit Office AO13 from 1779-1800. Throughout the historiography on loyalism during the Revolutionary Era, Virginia’s Tory population has been consistently overlooked by historians–many more have been more interested in loyalists from New York, New Jersey, and the Carolinas. This is mostly due to the fact that loyalist populations in Virginia were much smaller than in other states. However, while their numbers are smaller than in New York, Virginia’s loyalists werw far from insignficicant. This project is only the beginning of a much larger work–and a good practice in R to discover interesting quantative data scrapped from the Loyalist Calims Commission. Since this is the start of a major project on Virginia loyalists, this project is just a sampling of the first 100 Virginia loyalists in the alphabatized loyalist list. While only 1/5 of the list will not reveal too much about the history of loyalists in the Old Dominion, it will serve as a small window that will hopefully develop into a much larger picture.

However, it is also important for me to note that my data is tidy–but still includes all of my NAs. I feel that these NAs still need to be represented in order to represent everyone who I’ve found information for while going through the loyalist claims. While a loyalist may not have offered specific information like others, I feel their representation must still be counted. Towards the dissertation when all of these NAs are filled in, the data will be much tidier. If an NA cannot be located, then I will begin to na.rm my NAs. As for now, I feel the data should be represented as it stands as it reveals the struggles I face ahead.

library(dplyr)
## 
## Attaching package: 'dplyr'
## The following objects are masked from 'package:stats':
## 
##     filter, lag
## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union
library(ggplot2)
library(tidyr)
library(stringr)

After opening all of my libraries, I turned my Google sheet into a CSV file and uploaded it into R. In order to keep things easy, I assigned my library to the variable “loyal” to keep things simple.

loyal <- read.csv("val_clean.csv",
                  header = TRUE)

During the Loyalist Calims Commission, each loyalist–whether as an individual or as a head of a loyalist household–wrote a claim that contained information about their loyalty to the crown during the war as well as papers that proved what property they lost as a result of their loyalty. However, first loyalists had to prove that they even owned the property they claimed. While many Tory exiles were lucky to still have the deeds, wills, and court records that proved their owned slaves, stores, farms, plantations, etc. many loyalists lost absolutely everything and all of the documentation to prove it. Therefore, loyalists asked their old neighbors, friends, pastors, and family members to author witness testimonies. In these testimonies, loyalists “vouched” for each other by writing letters that would be submitted to the claims commission describing what their friends owned and lost as a result of the war.

This first chart examines the first 100 loyalists by comparing the number of people listed in a loyalist household vs. the number of witness testimonies for a household’s claim. In this visualization, I’m trying to understand if the amount of witness testimonies had anything to do with a loyalist family. If a loyalist family had 8 people–including children, would more people in the loyalist community be willing to write a testimony to help keep the family out of poverty. Or did people receive more witness testimonies because of their status in society?

loyal %>% 
ggplot(aes(x= claim_reps, y= witness)) +
  geom_point(color = "purple") +
  labs(title = "Virginia Loyalist Claims",
       x= "Number of People per Household", y = "Number of Witnesses per Claim")
## Warning: Removed 43 rows containing missing values (geom_point).

At least with these first 100 loyalist claims, it appears that there are few similarities between the size of a loyalist household and the number of witness testimonies submitted. For instance families with four loyalist members had the largest number of witness testimonies, while a family of eight only received 4. One of the loyalists who received the most testimonies (John Agnew) was a popular Norfolk minister who only claimed three people in his household.

Since the first visualization was based on family dynamics, the next is also interested in the difference between male and female claimants from Virginia.

  ggplot(loyal, aes(x = sex)) +
  geom_bar(stat = "count") +
  labs(title = "Male vs Female Population of VA Loyalists")

Unsurprisingly, males outnumber females significantly since the majority of claims were submitted by a head-of-house or single man. However, it will be interesting to see in the future how many female claimants were applied to the Claims Commission as widows or individuals. As of right now, every woman in this chart applied as a widow.

However, this visualization will be used to see if witness testimonies or family size had anything to do with the person’s sex. For instance, would community members take more pity on widows with large families and write witness testimonies to ensure the family would be taken care of financially by the British government?

ggplot(loyal, aes(x = witness, y = claim_reps)) +
  geom_count(shape = 1, alpha = 0.6) +
  facet_wrap(~ sex)+
  labs(title = "Virginia Loyalist Claims",
       x= "Witnesses", y = "Number in Household")
## Warning: Removed 43 rows containing non-finite values (stat_sum).

And it turns out the answer is–absolutely not. In fact men were more likely to gain witness testimonies–specifically those who had fewer familiy members. This leads me to assume that there is little correlaiton between the size of loyalist families or widows when it comes to witness testimonies. Apparently, very few were written out of pity.

Next I’m interested in looking at what country or territory loyalists were born. Quite a few loyalists immigrated to Virginia during and after the French and Indian War while others had family living in the colonies for generations. I was interested to see if birth origin had anything to do with their decision to remain loyal throughout the war. Many historians who have overlooked Virginia’s loyalists have argued that the population isn’t worth considering because many of their loyalists were newly immigrated “Scots Merchants.”

ggplot(loyal, aes(x = birth_nation)) +
  geom_bar(stat = "count") +
  labs(title = "VA Loyalist Birth Nation",
       x= "Birth Nation")

Interestingly, this chart reveals that the majority (by a fine margin) were born in America as opposed to England or Scotland–thus proving that not every loyalist was born in Great Britain–much less newly immigrated from Scotland.

However, after the Revolutionary War very few loyalists remained in Virginia with the majority being exhiled back to Great Britain or other territories within the empire. This chart will show us where they went when forced from their homes.

ggplot(loyal, aes(x = post_war_nation)) +
  geom_bar(stat = "count")+
  labs(title = "VA Loyalist Exile Nation",
       x= "Post War Nation")

This chart reveals that more people had actually died during the war than staying in the United States–with many moving to Canada, but the vast majority making their way to England. Interestingly, even though so many loyalists immigrated from Scotland, very few actually returned to there.

The chart below reveals where people were born vs. where they went after the war:

ggplot(loyal, aes(x = birth_nation, y = post_war_nation)) +
  geom_count(color = "purple") + 
  
labs(title = "VA Loyalist Birth vs Exile",
       x= "Birth Nation", y = "Post War Nation")

However, while we are on the topic of “homelands”, one of the more interesting historiographical aspects to consider is occupation. According to historian Adele Haste, the majority of loyalists from Virginia were Scots mercahnts–as mentioned above. By using R we can take the first 100 loyalist claims and see if the people who were the “heads of household” were actually a majority of immigrant Scots mercahnts by compaing their origin of birth vs. their occupation while in Virginia.

ggplot(loyal, aes(x = birth_nation, y = occupation)) +
  geom_count()+
  labs(title = "Birth Nation vs Colonial Occupation",
       x= "Birth Nation", y = "Occupation")

While Adele Haste is right in that there are quite a few immigrant Scots who work as merchants in Virginia, this chart also proves that there were just as many people–if not more–working other jobs across Virginia.

However, while we are looking at Virginia’s loyalists there are quite a few more questions we can ask and use as a comparison. Throughout the initial reports created by loyalists, many mentioned that they were imprisoned at some point during the war. The data set I’ve created matches up each loyalist with who was imprisoned vs. who was not using TRUE vs FALSE. Those who were imprisoned received a TRUE. Additionally, many loyalists also cited that they were abused by patriots at some point during the war. This can range anywhere between being tarred and feathered or being verbally assaulted or threatened with violence in the streets of their towns. I was interested in how many people cited abuse vs. imprisonment and whether or not people claimed both.

loyal %>% count(imprisoned, cites_abuse)
## Source: local data frame [8 x 3]
## Groups: imprisoned [?]
## 
##   imprisoned cites_abuse     n
##        (lgl)       (lgl) (int)
## 1      FALSE       FALSE    21
## 2      FALSE        TRUE     4
## 3      FALSE          NA     1
## 4       TRUE       FALSE     6
## 5       TRUE        TRUE    17
## 6         NA       FALSE     2
## 7         NA        TRUE     4
## 8         NA          NA    45

R has made some interesting comparisons in this list. Seventeen people on this list claim that they were both imprisoned and abused at different points throughout the war, where as 21 claim they received neither treatment. Additionally 45 of these Virginia loyalists never really addressed either subject or their claims were in such terrible shape that they could not be evaluated properly at this point.

Note that the echo = FALSE parameter was added to the code chunk to prevent printing of the R code that generated the plot.