Citation
Household and family in the postbellum South: Walton County, Florida, 1870-1885 /

Material Information

Title:
Household and family in the postbellum South: Walton County, Florida, 1870-1885 /
Creator:
Finlay, Barbara
Publication Date:
Copyright Date:
1976
Language:
English
Physical Description:
xx, 239 leaves : ; 28 cm.

Subjects

Subjects / Keywords:
African Americans ( jstor )
Censuses ( jstor )
Child psychology ( jstor )
Children ( jstor )
Counties ( jstor )
Family structure ( jstor )
Households ( jstor )
Parents ( jstor )
Slavery ( jstor )
White people ( jstor )
African American families -- Florida -- Walton County ( lcsh )
Dissertations, Academic -- Sociology -- UF ( lcsh )
Social life and customs -- Walton County, Fla ( lcsh )
Sociology thesis Ph. D ( lcsh )
Walton County ( local )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 229-237.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Barbara Finlay Agresti.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
025646110 ( AlephBibNum )
03142463 ( OCLC )
AAU5458 ( NOTIS )

Downloads

This item has the following downloads:


Full Text












HOUSEHOLD AND FAMILY IN THE POSTBELLUM SOUTH:
WALTON COUNTY, FLORIDA, 1870-1885












By

BARBARA FINLAY AGRESTI


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA IN PARTIAL
FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA


1976


















































Copyright by
Barbara Finlay Agresti
1976

























To my parents,
James and Grace Finlay,
and
the Finlay clan of Fife.















ACKNOWLEDGMENTS


A dissertation is the crowning work of a long and ardu-

ous career of graduate studies. None is ever completed

without a tremendous amount of help in many forms and from

many sources. I would like to take this opportunity to

mention just a few of the more important people to whom I

owe many thanks for their help a.nd support in my career as

a graduate student.

First of all, I would like to thank my chairman, Dr.

Gerald Leslie, for the many hours he spent in reading, editing,

and advising me concerning this and earlier drafts of my dis-

sertation. His standards are high, but I am sure that the

dissertation is much better because of his efforts. It was

Dr. Leslie who introduced me to the topic of family history

in the first place, and his help and encouragement during my

years at the University of Florida have been greatly appre-

ciated.

To Dr. Joseph Vandiver, I owe a great deal. His in-

sightful and broad understanding of the rural South and its









history have been very valuable in writing this disserta-

tion. His kindness, humor, and ability to cope with the

endless rules and regulations of the graduate school while

remaining warmly human have.made graduate studies in this

department more enjoyable for many students, including my-

self.

We were all saddened last spring by the death of Dr.

T. Lynn Smith, whose absence is keenly felt. Much of my

understanding of life in the rural South came from Dr. Smith,

both from his writings and from the many conversations we

had over the past four years. Although he was quite active

and prolific as a sociologist, he always had time for his

students. His help in pointing me to the right sources of

information for the background for this study and his in-

sightful interpretations of various aspects of Southern

demography and culture were invaluable aids. He was truly

an inspiration to this and many other budding sociologists

over the years, and his influence will continue to be felt

for some time to come.

I would also like to mention my other committee mem-

bers--Dr. Felix Berardo, who has always been a friend as

well as teacher to all students; Dr. S. S. Hill, Jr., whose

friendship and kindness I will always value; and Dr. Gordon









Streib, whose abilities and accomplishments have added depth

to this department. I especially thank Dr. Streib for stay-

ing with this project, even while recovering from major

surgery.

Another person whose encouragement and faith in me have

been very valuable in my career as a graduate student is

Dr. W. G. Steglich, who served as my advisor and master's

committee chairman at the University of Texas at El Paso.

Without his support and friendship, I might never have come

this far.

To Alan, my husband and friend, and to Ginger and Kerry,

my daughter and son, I owe a great debt of thanks. Their

sacrifices, love, patience, and help in the everyday things

of life made this dissertation a much less formidable task.

To Ginger and Kerry, who have been with me throughout my

entire college and university career, thanks for everything.

You've made it all worthwhile.

I suppose I owe the greatest debt to my parents, to

whom this is dedicated, and my brothers, who have always

inspired me to greater things and whose support has always

been present. There is no way I can ever repay them.

There are probably many more people who should be

thanked, people whose influence contributed to the present









work. All of those mentioned and unmentioned have added

strengths to this dissertation. Any remaining weaknesses

are my own doing.















TABLE OF CONTENTS


Page

ACKNOWLEDGMENTS . . . . . . . ... .. . iv

LIST OF TABLES. .. . . . . . . xiii

LIST OF FIGURES . . . . . . . . . .xvii

ABSTRACT. . . . . . . . .. .. .. .. .xviii

CHAPTER

I INTRODUCTION . .. . . . . ... .1
Emerging Interest in the Social History
of the Family . . . . . . . 2
Obstacles to the Historical Study of the
Family. . . . . . . . . . 5
General Purposes and Framework of This
Study . . . . . . . . . 6

II THEORETICAL PERSPECTIVES AND REVIEW OF LITERA-
TURE . . . . ... ...... .. 9
Family and Household Structure. . . . 9
Types of Family Groups . . . . 10
Other Aspects of Household and
Family Structure . . . ... 12
The Developmental Approach. .... . 13
Major Issues Concerning the Preindustrial
American White Family . . . .. 14
Family Structure . . . ... 14
Nonrelatives in the Household. ... .20
Family and Household Size. . . .. .20
Age at Marriage and Life Cycle of
Families . . . . . .. 21
Major Issues Concerning the 19th-Century
Black Family. . . . . . .. . 24
W.E.B. Du Bois . . . . . . 24
E. Franklin Frazier. . . . ... 26
Recent Data on Black Family Structure. 28


viii









TABLE OF CONTENTS (continued)

CHAPTER Page

II Other Considerations of This Study. . 32
(cont.)

III RESEARCH DESIGN AND METHODOLOGY. . . .. 33
Sources . . . . . . . .. 34
The Samples . . . . . . ... 35
Operational Definitions . . . ... 37
Data Organization . . . . ... 38
Analysis of Data. . . . . . ... 44

IV HISTORICAL PERSPECTIVES. . . . . .. 45
Southern Economic and Social Conditions
in the Late 19th Century. . . . ... 46
The Southern Agricultural Population
in the 19th Century. . . . ... 48
Walton County, Florida, in the 19th Cen-
tury. ................ . 51
Transition of Walton County, 1870 to
1885. ... . . . . . . 56

V THE POPULATION OF WALTON COUNTY, 1870 AND
1885 . . . . . . . .. .. . 60
The White Population of Walton County,
1870 . . . . . . . . 60
Race . . . . . . . . 61
Sex. . . . . . . . ... 62
Age. .. . . . . . . . 62
Marital Status . . . . .. 65
Estimated Age at Marriage. ... . 66
Fertility. . . . . . . 67
Mortality. . . . . . ... 68
Occupation and Social Class. ... . 70
Birthplaces of the Population. . 73
Length of Florida Residence. ... . 74
The Black Population of Walton County,
1870. ..... . . . . . 75
Race .. . . . . . . . 76
Sex. . . . . . . . . 77
Age. . . . . . . . 78
Marital Status . . . . ... 80
Estimated Age at Marriage. ... 81









TABLE OF CONTENTS (continued)


CHAPTER Page

V Fertility. . . . . . ... 82
(cont.) Mortality. ........... . 82
Occupation and Social Class. .. . 83
Birthplaces of the Population. . 86
Length of Florida Residence. . . 88
The White Population of Walton County,
1885. . . . . . . . . . 89
Race . . . . . . ... 90
Sex. . . . . . . . ... 91
Age. . . . . . . . ... 91
Marital Status . . . . .. 95
Estimated Age at Marriage. ... . 96
Fertility. . . . . . ... 97
Occupation and Social Class. ... . 97
Birthplaces of the Population. ... 101
Length of Florida Residence. .. 103
The Black Population of Walton County,
1885. . . . . . . . . ... 104
Race . . . . . . . . 106
Sex. . . . . . . . . 106
Age . . . .. . . . 107
Marital Status . . . . .. 109
Estimated Age at Marriage. ... 110
Fertility. . . . . . .. 111
Occupation and Social Class. .. 112
Birthplaces of the Population. ... 114
Length of Florida Residence. .. 116
Conclusion: The Population of Walton
County, 1870 and 1885 . . . . . 116

VI THE STRUCTURE OF FAMILIES AND THE DEVELOP-
MENTAL CYCLE, 1870 AND 1885. . . . .. 119
Structure of White Families, 1870 . .. 120
Young Couples. . . . . . 121
The Childbearing Years . . .. 124
The Childrearing Years . . ... 124
Widowhood. . . . . . . 126
The Children . . . . . . 128
Single Adults. . . . . . 130
Conclusion: The Life Cycle of the
Individual. . . . . . . . 132









TABLE OF CONTENTS (continued)


CHAPTER Page

VI Structure of White Families, 1885 . .. 134
(cont.) Young Couples. . . . . .. 136
The Childbearing Years . . . 137
The Childrearing Years . . ... 139
Widowhood. . . . . . ... 140
The Children . . . . ... 142
Single Adults . . . . . 144
Conclusion: The Life Cycle and Economic
Crisis. ................. 146
Structure of Black Families, 1870 . 148
Two-parent Families. . . . .. 152
Widowed and Broken Families. .. 153
Other Tests of Black Family Strengths
and Stability. . . . . . 155
Living Arrangements of Children. .. 159
Living Arrangements of the Aged. .. 162
Conclusion: Black Families in 1870 .. 164
Structure of Black Families, 1885 . . 165
Young Couples. . . . . .. 167
The Childbearing Years . . ... 168
The Childrearing Years . . . 169
Widowhood. . . . . . ... 170
The Children . .. . . . 170
Single Adults. . . . . .. 172
Conclusion: Black Family Structure and
the Life Cycle. . . . . . .. 174
Changes in Black Family Structure, 1870-
1885. . . . . . . . . . 177
Household Composition and the Family
Cycle . . . . . . . . . 180
Summary . . . . . . . ... 184

VII FAMILY AND HOUSEHOLD CHARACTERISTICS AND SOCIAL
STRUCTURE. . . . . . . . . . 186
Selected Characteristics of Black and
White Households, 1870 and 1885 ... 186
Extended Kin . . . . . . 187
Boarders . . . . . ... 188
Household Size . . . . .. 190
Number of Children . . . . 191
Marital Status of Household Heads. 191









TABLE OF CONTENTS (continued)


CHAPTER Page

VII Sex of Household Heads . . ... 192
(cont.) Summary of Household Characteristics,
1870 and 1885. . . . . .. 193
Household Characteristics by Rural-Village
Residence . . . . . . . .. 194
Extended Kin ..... . . ... 195
Boarders . . . . . . . 196
Servants . . . . . . . 197
Household Size . . . . .. 198
Number of Children . . . . 198
Structure of Primary Families. 198
Summary: Household Characteristics
by Rural or Village Residence. . 200
Household Characteristics and Social
Class . . . . . . . ... 201
Extended Kin . . . . ... 203
Boarders . . . . . . . 204
Household Size . . . . .. 206
Number of Children . . . .. 207
Marital Status of Household Heads. 210
Sex of Household Heads . . .. 210
Summary: Household Characteristics
and Social Class . . . ... 212

VIII CONCLUSIONS. . . . . . . . ... 214

APPENDICES

A CODING INFORMATION: HOUSEHOLD RECORD. ... .220

B CODING INFORMATION: PERSON RECORD . . .. 223

C METHOD OF ESTIMATION OF AGE AT MARRIAGE. . 228

REFERENCES. . . . . . . . . . . . 229

BIOGRAPHICAL SKETCH . . . . . . . ... 238















LIST OF TABLES


Table Page

1 Racial Composition of White Household Sample,
1870 . . . . . . . . . .. 61

2 Sex Composition of White Sample, 1870 ... . 62

3 Marital Status of Adults in White Sample, 1870,
by Sex . . . . . . . ... .... 66

4 List of Occupations of the White Sample, 1870 71

5 Occupational Class of White Sample, 1870. . 72

6 Birthplaces of the White Sample, 1870 ... . 73'

7 Recency of Migration of Families of Household
Heads, White Sample, 1870 . . . . .. 74

8 Racial Composition of Nonwhite Population of
Walton County, 1870 . . . . . ... 77

9 Sex Composition of Black Population of Walton
County, 1870, by Race of Household Head . . 78

10 Marital Status of Adult Black Population of
Walton County, 1870 . . . . . ... 80

11 List of Occupations of the Black Population of
Walton County, 1870 . . . . . ... 84

12 Occupational Class of the Black Population of
Walton County, 1870 . . . . . . . 85

13 Birthplaces of the Black Population of Walton
County, 1870. . . . . . . . ... 87

14 Recency of Migration of Families of Black House-
hold Heads, Walton County, 1870 . . ... 88


x i i i









LIST OF TABLES (continued)


Table Page

15 Sex Composition of White Samples, 1885. .. 91

16 Marital Status of White Adults in Walton County,
1885, by Residence. . . . . . . ... 95

17 List of Occupations of White Samples, 1885. . 98

18 Occupational Class of Adults in White Samples,
1885. . . . . . . . . ... . 100

19 Birthplaces of the White Samples, 1885. .. 102

20 Recency of Migration of Families of White
Household Heads in Walton County, 1885. .. 103

21 Sex Composition of Black Population in Walton
County, 1885, by Race of Household Head . .. 106

22 Marital Status of Black Adults in Walton County,
1885, by Sex. . . . . . .... . 110

23 List of Occupations of the Black Population of
Walton County, 1885 . . . . . . . 112

24 Occupational Class of the Black Population of
Walton County, 1885 . . . . . ... 114

25 Birthplaces of the Black Population of Walton
County, 1885. . . . . . . . . 115

26 Recency of Migration of Families of Black
Household He.ads, Walton County, 1885 . . .. 116

27 Structure of Families in White Sample, 1870 . 121

28 Living Arrangements of White Children in Walton
County, 1870. . . . . . . . . 129

29 Living Arrangements of White Single Adults in
Walton County, 1870 . . . . . .... 131









LIST OF TABLES (continued)


Table Page

30 Structure of White Families in Walton County
Samples, 1885 . . . . . . . ... 135

31 Living Arrangements of White Children in Walton
County, 1885. .... . . . . . . . 143

32 Living Arrangements of Single White Adults in
Walton County Samples, 1885 . . . ... 146

33 Structure of Black Families in Walton County,
1870. . . . . . . . . ... .. 150

34 Percentage of Adults Not Living in Family Groups,
1870, by Race . . . . . . . . 151

35 Probable Family Structure of Slaves of Walton
and Holmes Counties, 1860 . . . . . 157

36 Sex of Probable Heads of Slave Families in
Walton and Holmes Counties, 1860. ... . 158

37 Living Arrangements of Black Children in Walton
County, 1870. . . . . . .... . . 160

38 Living Arrangements of the Aged, by Sex and Race,
1870 and 1885 . . . . . . . ... 163

39 Structure of Black Families in Walton County,
1885 . . . . . . . . . . 166

40 Living Arrangements of Black Children in Walton
County, 1885. . . . . . . . . 171

41 Living Arrangements of Black Single Adults in
Walton County, 1885 . . . . . ... 173

42 Household Composition by Stage of Family Life
Cycle . . . . . . . . .. . . 181

43 Percentages of Households Containing Relatives
and Boarders, by Race, 1870 and 1885. ... . 188









LIST OF TABLES (continued)


Table Page

44 Characteristics of White Households in 1885,
by Village or Rural Residence . . . ... 196

45 Structure of Primary Families, 1885, by Village
or Rural Residence. . . . . . . .. 199

46 Percentage of Households Containing Extended
Relatives, by Social Class. . . . . .. 204

47 Percentage of Households Containing Boarders,
by Social Class . . . . . . . 205

48 Mean Household Size by Social Class, 1870 and
1885. . . . . . . . . . . 207

49 Mean Number of Children in Household, by Social
Class, 1870 and 1885. . . . . . . 208

50 Percentage of Household Heads Who Were Married,
by Social Class, 1870 and 1885. .. . . 211

51 Percentage of Households with Male Heads, by
Social Class, 1870 and 1885 . . . ... 212
















LIST OF FIGURES


Figure Page

1 Wholesale Price Index, Farm Products, 1870-
1890 ... . . . . . . . . .47

2 Age and Sex Distribution of Walton County
White Sample, 1870 . . . . ... .64

3 Age and Sex Distribution of Walton County
Black Population, 1870 . . . ... .79

4 Age and Sex Distribution of Walton County
White Sample, 1885 . . . . ... .93

5 Age and Sex Distribution of Lake De Funiak
White Population, 1885 . . . ... .94

6 Age and Sex Distribution of Walton County
Black Population, 1885 . . . . . 108


XV1ii








Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Doctor of Philosophy

HOUSEHOLD AND FAMILY IN THE POSTBELLUM SOUTH:
WALTON COUNTY, FLORIDA, 1870-1885

By

Barbara Finlay Agresti

August, 1976

Chairman: Dr. Gerald R. Leslie
Major Department: Sociology

This study deals with the family, household, and popu-

lation characteristics of a farming community in the South

during the two decades following'the Civil War. These

decades were a time of rapid and disruptive change in social

and economic conditions in the South. The study represents

an addition to the growing literature on the history of the

American family. For white families, the emphasis is on

description of household and family structure in a hitherto

neglected area--the farming South. For blacks, the data are

used to test some of the controversial ideas of E. Franklin

Frazier and others concerning black family structure in the

postbellum period.

Data from census manuscripts for Walton County, Florida,

for the years ,1870 and 1885 are analyzed. Detailed de-

scriptions of the black and white populations in those years


xviii









are presented. Residential family structure is discussed in

detail by stage of the family cycle, for black and white fami-

lies separately. Other analyses include comparisons of house-

hold characteristics by occupational class, rural-village

residence, and over time.

For whites, the data presented add to our knowledge of

the 19th-century rural Southern family. The data show evi-

dence of stem family organization among the rural people.

For blacks, new evidence supporting the "matriarchal thesis"

of Frazier and others is presented, at least for the 1870

population. Important gains had been made, however, by 1885,

when black and white families are seen to have been quite

similar.

Some dramatic changes occurred during the 15-year inter-

val studied. For example, the proportion of extended house-

holds, especially of the married sibling type,-increased

sharply. This and other changes are explained in terms of

the social and economic changes occurring in the community

and in the South during the period.

General conclusions of the study point to the adapta-

bility of the family institution and the need to consider

the historical and socioeconomic context in family research.









The importance of the developmental approach is emphasized.

In addition, the results show the importance of using

families, as well as households, as basic units of analysis

in historical studies.












































xx















CHAPTER I

INTRODUCTION



This study deals with the family, household, and com-

munity in a late 19th-century Southern farming area. Its

setting is Walton County, an agricultural county in the

western Florida panhandle, in the years 1870 and 1885. This

study makes a contribution to the history of a county, to

the sociology of the family, and to the history of the

American family as well.

The initial date selected for this study--1870--was

chosen in order to analyze family data drawn from the first

census after the Civil War. The white population of the

South had suffered heavily from war casualties, economic

losses, the loss of capital invested in slaves, and the de-

moralization of unconditional defeat. Only five years had

passed since the end of the hostilities, and the region was

midway in the often bitterly resented Reconstruction period.

Even more important was this period for black Ameri-

cans, seven-eighths of whom were just released from a system









of human bondage. Family patterns instituted in slavery

were undergoing adjustments to a new order of life, one

doubtless filled with many hardships and many hopes.

The second point in time to be analyzed--1885--not only

added 15 years during which adjustment to the new order

could be made, but followed by a decade or more the end of

Reconstruction. While it is impossible to separate the

adaptive processes from the social-psychological effects

of the end of Reconstruction on the two races, it seems

probable that, for most blacks, the high hopes of the earlier

decade had been scaled down by the realities of the birth of

Jim Crow. The present study hopes to explore some of the

effects of these rapidly changing conditions on a little-

studied region--the farming South.



Emerging Interest in the Social
History of the Family


The social history of the family has become an increas-

ing topic of interest over the past decade, both in Europe

and North America. This growing interest was stimulated by

early studies in France (Gautier and Henry, 1958; Aries,

1962). Soon afterwards, English scholars, led by a group

at Cambridge University, began to publish work in the field









(Wrigley, 1966, presents a long list of such studies).

Finally, American and Canadian students of family history

have begun contributing, especially within the past five or

six years. As a result of this new interest, a new journal,

the Journal of Family History, has just begun publication,

several special issues of scholarly journals have been de-

voted to family or demographic history,I and several articles

have appeared calling for more and better research in the

area. (See, for example, Saveth, 1969; Hareven, 1974;

Berkner, 1973a.)

There are at least two reasons for the new interest in

the history of the family. First, historians, who have tra-

ditionally emphasized the role of outstanding people and

events, have become aware of gaps in their knowledge and

understanding of past events and social conditions which

affected the bulk of the population. Their new interest in

"the common man" has stimulated interest in basic institu-

tions, such as religion and the family, that closely touch

and shape the lives of the major portion of the population.

(For examples of this new historical interest, see Thernstrom,



1Journal of Interdisciplinary History 2, 1971; Journal
of Marriage and the Family 35(3), 1973; Daedalus 97 (Spring),
1968; Journal of Social History 5(1), 1971.









1964; Hareven, 1971; Powell, 1963; Demos, 1968, 1970;

Lockridge, 1970).

The second reason is the renewed interest of sociolo-

gists in social change, which has challenged stereotyped

conceptions of what family life was like a half-century or

more ago. Goode issued the challenge in 1963, with the

following portrait of what he called "the classical family

of Western nostalgia."

It is a pretty picture of life down on grandma's
farm. There are lots of happy children, and many
kinfolk live together in a large rambling house.
Everyone works hard . . The family has many
functions; it is the source of economic stability
and religious, educational, and vocational train-
ing. Father is stern and reserved, and has the
final decision in all important matters. . All
boys and girls marry, and marry young . . After
marriage, the couple lives harmoniously, either
near the boy's parents or with them, for the couple
is slated to inherit the farm. No one divorces.
(Goode, 1970:6)

Goode went on to emphasize that there is a paucity

of historical data on the family and that more adequate

data are needed.

Theorizing about family changes over time is
easier, of course, if we enjoy a firm body of data
about the past, and in particular if we know how
the family pattern operated at some specific point
in time: drawing a curve between only two points
in time is dangerous, but not so unwise as drawing
a curve from the present to an unknown and pos-
sibly legendary past. (Goode, 1970:xi, xii)





5



Heeding Goode's admonition, a variety of sociologists has

begun to provide empirical studies of various aspects of

family life in the past. (Furstenberg, 1966, Farber, 1972,

Demos, 1970, Lantz et al., 1968, are a few. See also the

collection by Gordon, 1973.)



Obstacles to the Historical
Study of the Family


Sociologists and historians have been hampered in

studying family history by the lack of adequate sources of

data. Family life is generally so taken-for-granted that

few people have bothered to write about it. The accounts

that do exist, largely from moralists and philosophers,

generally fail to distinguish between real and ideal be-

havior, and they focus on the patterns of elite groups with-

out any indication of the relevance of those patterns to the

bulk of the population. Lower-class people, often being

illiterate, did not leave memoirs, diaries, or other written

documents.

Given these limitations, the scholars who are develop-

ing the new family history have been ingenious in discovering

new sources of data, basically demographic in character.

These include parish records of the important life events









of baptisms, marriages, and deaths. Wills, court records,

land deeds, property inventories and even physical arti-

facts have also been used. For studies of 19th-century

America, the manuscript census has been a valuable source.

Regardless of the specific sources used, the techniques of

analysis applied have generally been quantitative and

rigorous.



General Purposes and Framework of This Study


Several studies have been published recently of colonial

family patterns in the Northeast (Demos, 1970; Greven, 1970;

Lock.ridge, 1966; Smith, 1973a, 1973b; Norton, 1971;

Wells, 1971, 1971-1972, 1972). In addition, there are a

few studies of urban family patterns in the 19th century

(Sennett, 1973; Pleck, 1973; Hershberg, 1971). Very little

has been done, however, on 19th-century family patterns

in rural areas or in the South. (But see Modell, 1971, and

Eblen, 1965, for limited demographic studies of frontier

populations.)

Our knowledge of white family patterns is scattered

and incomplete, but our knowledge of black family patterns

is even more so. For many years after its publication in









1939, Frazier's analysis of the influence of slavery on black

family patterns was accepted widely and almost uncritically.

Recently, the "cliometric" work of Fogel and Engerman (1974)

and the research of Genovese (1974) have called many of

Frazier's conclusions into question. Although four studies

have appeared concerning black families in cities of the 19th

century (Pleck, 1973; Hershberg, 1971; Lammermeier, 1973;

Harris, 1976), almost nothing has been published on rural

black families of the South--especially in the nonplanta-

tion areas.

The present study of both black and white family pat-

terns in a Southern rural county in the decades after the

Civil War should add significantly to our knowledge. Three

general approaches to the data will be used, often simul-

taneously. In the first, descriptive, approach, both white

and black family patterns will be enumerated in detail.

For white families, we will provide data on a little-studied

population in a little-studied time period. For black

families, systematic demographic data may afford some in-

sight into current issues concerning the nature of black

family history.

In the second, analytic, approach, variations in family

and household composition according to certain social-structural





8



characteristics will be analyzed. The focus will be on the

relationship between family structure and other aspects of

social structure. Then, in a dynamic approach, we will com-

pare family patterns at two points in time, 15 years apart.

The emphasis will be upon changes in family patterns and

household structure, and the relationship between these

and other changes in the community. The three approaches

combined should enable us to understand the family in the

community as both changed over one brief period of time.














CHAPTER II

THEORETICAL PERSPECTIVES AND REVIEW
OF LITERATURE



Meaningful research proceeds from explicit theoretical

bases and utilizes carefully defined concepts. The first

task of this chapter is to specify major concepts relating

to the nature of the family and to household structure.

Secondly, the basic issues involved in applying these con-

cepts to white and black families of the 19th century will

be drawn from a review of published research in the area.



Family and Household Structure


Kinship refers to a system of norms governing relation-

ships based upon consanguineal and/or affinal ties. In

turn, family structure is a system of norms associated with

kinship statuses. These norms specify which kinship sta-

tuses are recognized and what obligations accrue to particu-

lar statuses vis-a-vis others.

The basic characteristics of the American kinship sys-

tem were described by Parsons (1943). He noted that our


9









kinship system is similar to those of most modern societies

and has not changed significantly over the past couple of

centuries. Family structure, on the other hand, has changed

considerably. This change has been associated with the

progressive loss of family functions (Ogburn, 1954).

The major components of family structure include rules

of residence, rules of authority, inheritance rules, and

role definitions associated with particular statuses. The

thrust of the present research involves the residential

aspects of family structure, but other aspects may often be

inferred from knowledge of residence patterns. In addition,

we are interested in household composition and the inter-

relationships among all members of households, whether kin

or not.


Types of Family Groups

There are four basic types of resident family groups

of great importance in the present study. These are the

nuclear family, the single-parent family, the stem family,

and the sibling family.

The nuclear family consists of a married couple and

their offspring living in a single household. A married

couple without children is designated a conjugal couple, but









for many purposes conjugal couples and complete nuclear

families will be considered together. The nuclear family

has usually been found to be the most common form of resi-

dent family, even where other family forms are common

(Berkner, 1973b; Laslett, 1973).

The single-parent family consists of one parent, either

mother or father, and at least one natural child. The 'hild"

may be of any age, as long as he or she is unmarried and

living with one parent.

The stem family was first described by Le Play as the

famille souche (Sorokin, 1928:86). In this study, stem

families are those in which two ever-married, directly

lineally related individuals are residing together, with or

without spouses and/or children. The most common example

would be a young couple living with the husband's aged par-

ents. Although the stem family usually refers to a struc-

tural process through which families go, it is used here

to define a type of residence group which is sometimes

associated with that process.

The sibling family is a group of two or more siblings,

at least one of whom is over 18 years of age, who reside

together with no parent or own child present. This kind of

family group might be common if a large proportion of the









population never married or if elderly widows lived with

siblings rather than with their children.

All of these basic residence groups may be further

extended by the presence of other relatives in the house-

hold. In addition, they may be augmented by the presence

of one or more unrelated residents, normally either rent-

paying boarders or servants.


Other Aspects of Household
and Family Structure

In addition to the type of family group within the

household, several other aspects of family or household

structure are of interest. These include the size of house-

holds (the total number of individuals who reside together

as one domestic and economic unit) and the size of families

(both the number of children born per marriage and the number

of kin present in the household). The age at which marriage

typically takes place and the proportion of the population

that marries are also important data for the study of family

structure.

In addition, the typical life cycle of the family,

from its inception at marriage to the death of the last

spouse, is interesting and important to this study. Finally,

the functions of families and of households will be









determined, insofar as possible, by looking at patterns of

residence and their relation to various aspects of social

structure.



The Developmental Approach


The importance of a developmental approach to the study

of families has been well documented (Lansing and Kish,

1957). The concept of a series of stages through which

families pass in their normal development was introduced by

early rural sociologists (Sorokin et al., 1931) and later

popularized by Glick (1947, 1955), who noted important

changes in the typical life cycle of families over the past

century. The concept is important in any study of household

composition, because the composition of the family at any

one time depends upon its stage of development. The stem

family, for example, can be seen as one stage in a much

longer cycle, during much of which the family group is

nuclear (Berkner, 1973b). The need to consider this in

historical studies is echoed by Hareven (1974). A develop-

mental approach will be used for many parts of the analysis

in this study.









Major Issues Concerning the Preindustrial
American White Family


A classic history of the American family was published

in 1917-1919 in three volumes by Arthur W. Calhoun. Other

traditional sources for American family history are those by

Goodsell (1939), Howard (1904), Sirjamaki (1953), and Bardis

(1964),the latter two being summaries of the traditional

literature. All of these works suffer from reliance upon

such imprecise data sources as legal documents, travelers'

accounts, and popular literature. During recent years, the

work of the new social historians, historical demographers,

and sociologists of family history has begun to question some

of the findings and interpretations of these traditional

sources. In this section, some of the major points of dif-

ference between the traditional and more recent interpreta-

tions will be discussed.


Family Structure

The most important questions about the American family

of the.past center around its structure. Traditionally,

historians and sociologists have believed that preindustrial

families had strong extended kinship ties, in contrast to

the more isolated nuclear family of contemporary society.









The view that many households in past centuries contained a

wide variety of kin has been especially common (Bailyn,

1960:17-18; Laslett, 1973).

Recent research on household structure for the colonial

period has shown that, for various communities in various

times, most households were nuclear. Marriage agreements

almost always provided for the establishment of separate

households for the new couples. The parental family house,

in both Andover, Massachusetts, and in Plymouth Colony,

was always left to only one son (Demos, 1970:63; Greven,

1970). Pryor's study of Rhode Island households in 1875

(1972) also found most households to be nuclear in family

structure.

There is a problem, however, in identifying certain

forms of family structure by cross-sectional research such

as the studies just cited. One form of family organization

for which this is true is the stem family found in farming

areas in the United States and Europe (Leslie, 1976:223).

In families of this type, one son stays in the parental

household after marriage, while the other siblings move

out at marriage to form new households. The son who stays

in the parental home eventually inherits the parental land

and home, which remain intact. The other sons often receive









parental support in setting up their own households, but

the original parental land is not divided among them.

As Berkner (1973b) has pointed out, even when such an

extended family system is typical, most households at any

one time are nuclear, because nonheirs leave the parental

household to form nuclear households. The parental house-

hold would be considered "extended" only for the period of

time between the marriage of the heir and the death of the

parents. This period may be relatively short when compared

to other stages of the family cycle, especially if marriage

ages are high and life expectancy is low. Berkner points

out the need to look at the developmental stage of the

parental household, as well as whether or not the household

is extended, because the extended family is "merely a phase

through which most families go" (1973b:41).

When studies of rural American families are examined

closely, a clear pattern of stem organization seems to

emerge, although it has not been identified as such by

other authors. Demos notes, for example, that one extended

household type in Plymouth was the residence in some house-

holds of aged grandparents. When parents were aged, or when

one died, it was common for the son who was to inherit the

homestead to reside with the aged couple or widowed parent









and to provide for them until their deaths (1970:75-76).

The same situation was described by Greven for Andover

families.

Often, though, sons who received the parental
homestead as their portion of their father's es-
tates also received part of their parents' houses
and lived in them while their parents were still
alive, thus forming households which effectively
consisted of three generations under a single roof.
A widow nearly always was specifically bequeathed
a room or rooms in her husband's house and provi-
sions to be given to her annually by one or more
of her sons . Generally, of course, married
sons lived separately in houses of their own, but
aged parents usually expected to share a house with
one of their children. Extended households, for the
most part, were the product of old age and the neces-
sities for care and attention which elderly men and
women obviously needed. (Greven, 1970:137-138)

This pattern in Andover became increasingly common as

the original land tracts became subdivided to the point that

impartible inheritances to only one son became the rule.

When land was plentiful, it had been common to settle all

sons on their own property; but later generations found that

their landholdings would not permit further partitioning.

Thus, the number of three-generation households increased

at the same time that landless sons began migrating to

other areas. Family extensions were, in this instance,

related to changes in land-inheritance patterns resulting

from population growth and from full settlement of the

land.









One should keep in mind that new settlements would not

usually have three-generational households, because most

migrants were young. It would normally be 40 or 50 years

before the original settlers of a community were old enough

to have grandchildren in their households, and longer before

land settlement and subdivision increased this pattern. It

would be expected, then, that extended households would in-

crease over time in a relatively new settlement, if indeed

the stem pattern were normative. This seems to have been

the case in Andover (see Greven, 1970:98, 137-138, 220, and

257) and also in rural Michigan in the 19th century (Bieder,

1973). It might also explain the relative lack of extended

families in Bristol, Rhode Island, in 1689, since this was

a relatively new settlement at the time (Demos, 1970:79).

It seems plausible that many early American families

were stem families, especially after the full settlement of

an area. As families moved to the cities, the lack of

landed property to keep within the family probably led to

the breakdown of this system; the urban family's resources

were easily divisible among several siblings, whereas

agricultural land is only divisible a limited number of

times. This interpretation is consistent with the English

pattern. Extended families in rural areas involved aging









parents being supported by sons who were to inherit the

property, while those in urban areas more often involved

young couples being temporarily supported by their parents

(Anderson, 1971).

It should be kept in mind that even when stem family

organization exists, most sons usually set up nuclear house-

holds, and the parental household is nuclear for most of its

existence, so that the proportion of three-generational

households in the population at any time is quite small.

The finding of Pryor (1972) that over 10 percent of the

Rhode Island rural households in 1875 were multigenerational

may be quite significant in view of this fact.

In summary, it is not good practice to attempt to de-

scribe family structure merely from cross-sectional data.

Some attention must be given to the family life cycle. Only

thus can we determine important aspects of family organiza-

tion, such as the means for assuring continuity of family

relations and property across generations. For white fami-

lies in this study, one of the major questions concerning

family structure is just this: Is there evidence of the

stem family among the agricultural population, or are

nuclear families in separate households the norm at every

stage of the life cycle?








Nonrelatives in the Household

Early descriptions of 18th- and 19th-century American

families did not emphasize the presence of nonrelatives with-

in family households. Several recent empirical studies,

however, have reported that many households included non-

relatives (Pryor, 1972; Anderson, 1971; Modell and Hareven,

1973; Demos, 1970; Lockridge, 1966). Some of these unre-

lated persons were household servants, many of whom were

children (Demos, 1970:71; Lockridge, 1966:343n). Still

others were single adults who had no other choice but to

reside with a family as boarders (Demos, 1970). By 1875,

the proportion of Rhode Island households containing non-

relatives of the head was still fairly high--24 percent of

all households (Pryor, 1972:588).


Family and Household Size

Traditional estimates of early American family size

have been high, owing to the unreliability and exaggeration

of such sources as travelers' accounts and literary works.

Calhoun wrote that families of 10 to 12 children were the

norm, and that those with over 20 children were not rare in

colonial days (1917-19:87). These estimates have been al-

tered in view of the findings of recent studies. Most of

these studies have found average completed family sizes of









around seven or eight children in the colonial period

(Demos, 1970:192; Smith, 1973a; Greven, 1970; Wells,

1971:75). This figure decreased somewhat for 19th-century

populations (Wells, 1971:75; Farber, 1972). Still, American

family sizes have been found to have been consistently lar-

ger than European families of similar periods.

Although the number of children born per family was

relatively large, the average number of persons per house-

hold was smaller. Some households contained only young mar-

ried couples, others contained families in the childbearing

stage, while, in still other cases, some or all of the chil-

dren had grown and left the parental home. High infant

mortality rates were another factor in keeping household

sizes low.

In colonial days, household size probably averaged four

to six persons. In 1790, the average household size for.the

United States was 5.7 persons; less than twice the size of

U.S. households in 1950, 3.4 persons (Grabill et al., 1973:

379).


Age at Marriage and Life
Cycle of Families

Most writers about the early American family concluded

that marriages took place at very early ages (Calhoun,





22


1917:67; Sirjamaki, 1953:40). Often, too, women were be-

lieved to die in childbirth, leaving their widowed husbands

to marry for a second or a third time.

Studies that have utilized quantitative data from

family reconstitution or census analysis have called both

of these assumptions into question. The average age at

marriage in the 17th and 18th centuries in America has con-

sistently been found to have been fairly high, usually around

27 years for men and 22 or 23 years for women (Wells, 1972:

426; Henripin, 1964; Demos, 1970:193; Smith, 1973a:406;

Greven, 1970; Farber, 1972:42). Moreover, a high proportion

of men who married were married only once (Greven, 1970:29;

Wells, 1972:423-424). Even in colonial days, marriages

broken by early death were the exception and not the rule.

Robert Wells brought together data from 18th- and 19th-

century Quaker families with data for recent decades pub-

lished by Glick and Parke (1965) in an article describing

changes in the typical family life cycle over the past two

centuries in the United States (Wells, 1971-1972). The

major changes have been (1) a progressive decrease in the

age of mothers at the birth of their last child and (2) an

increase in life expectancy for both spouses. At the same

time, the median age of women at first marriage has not









really changed substantially. This means that the child-

bearing period was relatively much longer for women of

earlier centuries, and that most marriages of the earliest

group studied (early 19th century) ended by the death of a

spouse before the last of the children left home. It was

Glick who first pointed out that a "new phase" in the family

life cycle had appeared in the 20th century, the "empty

nest" stage (1947). Throughout the 19th century, most fami-

lies never experienced this stage, but went directly from

childrearing into widowhood.

We have now briefly discussed the experiences of white

families in both the accounts of early writers and those of

the new generation of empirically oriented historians. Most

of the studies reported have dealt with colonial populations.

It will be useful to compare their findings with ours to see

if they are supported for a Southern rural community in the

late 19th century. In the chapters on data analysis, an

attempt will be made to discuss each of the topics intro-

duced here.









Major Issues Concerning the 19th-Century
Black Family


There has been a revival of interest in recent years

in the history of the black family in the United States.

This is due not only to the general interest of the past

decade in family history, but also to the renewed interest

in black history and the slavery experience brought about by

the increased awareness of minority groups in the 1960s. For

decades, most sociologists had accepted the interpretations

of the black family of E. Franklin Frazier (1948), but recent

publications (Pleck, 1973; Genovese, 1974; Fogel and Enger-

man, 1974; Gutman, 1973; Harris, 1976) have questioned

some of the conclusions made by Frazier and other early

writers.


W.E.B. Du Bois

Although E. Franklin Frazier is usually remembered as

the pioneering student of black family structure and patterns

in the 19th century, most of his major conclusions were

anticipated by about 30 years by William E. Burghardt

Du Bois (1899; 1908). Du Bois, noting the poor condition

of black families in the early 20th century, believed that,

at least, they were much improved over the conditions of









the slavery period. The slave family, according to Du Bois,

had been practically destroyed by the exigencies of the cot-

ton producing system (1908:22, 31).

Du Bois was one of the first to emphasize the division

between the house servant and the field slave. The house

slaves lived close to the families of their masters, learn-

ing from them and often adopting their mores. Among this

group, the monogamous, patriarchal family system developed

as the ideal family type (1908:47).

Another group of field slaves, larger in number, lived

at a distance, both spatially and socially, from white

families. This group became highly demoralized, especially

where masters were of the "absentee" type. They lived in

"quarters," not family cabins; sexual exploitation was

common and there was practically "no family life" and no

moral instruction (1908:passim).

Such family life as there was among these field slaves

was characterized both by the "absence of fathers" and by

lack of ability of fathers to govern or protect their

families. A man could be sold or separated from his family

at any time and his wife could be made the master's or

overseer's concubine. Thus, the father's place in the

authority system was diminished (1908:49).









Du Bois spoke of the "absence of mothers" as character-

istic of slave families. Even though mothers were a more

stable element in the child's world than were fathers, they

were unable to spend much time in child care. Young chil-

dren often were left alone or in the care of other children,

while their mothers had to go to the fields (1908:49).

This situation of demoralization and weak family ties

allegedly carried over into the postslavery era. The two

groups which had formed during slavery--house servants and

field slaves--diverged. The former house slaves continued

to be patriarchal and monogamous, similar to white families.

The others continued to'have unstable marital and family

relationships, although these gradually became more stable

over time (Du Bois, 1908:31).


E. Franklin Frazier

Thirty years after Du Bois published these ideas, E.

Franklin Frazier studied the same problems and came to much

the same conclusions (1930, 1948). Frazier emphasized that

the stability of slave families depended largely on the

characteristics of the particular master. Masters could

force matings, sexually exploit slave women, and separate

families by sales; or they could enforce stable, monogamous

families. Even where the masters were kind and encouraged









stability, the slaves' dependence upon the master's will

prohibited the development of a strong, independent black

family (Frazier, 1930).

Frazier, like Du Bois, emphasized the importance of

the mother in the family. The father was not dominant be-

cause his will was not enforceable over that of the master.

The mother, on the other hand, did play a fairly stable role

in the raising of her children. It was to her that the

children turned for strength and support (1930:236).

Like Du Bois, Frazier also noted the differences be-

tween house servants and field slaves. It was among the

field slaves, especially, that the strong-mother, weak-

father family became widespread. The house servants' fami-

lies were patriarchal and stable (1930:259).

After slavery ended, much of the black population

wandered about for awhile and then settled down into semi-

stable family groups. However, there still were high il-

legitimacy rates, a high proportion of female-headed families,

and high migration rates. The "matriarchal" family, in

which the female was the more stable and dominant parent,

grew as a significant black family type (1948:102-113).

Another group of blacks, mostly composed of former free

blacks, house slaves, and mulattoes, settled down into









stable, monogamous, patriarchal families. Often these "Black

Puritans," as Frazier called them, owned their own homes and

small parcels of land. Some became educated professionals.

This group assumed leadership in the black communities in

the South (1948:190-208).

The worst situation was among those who moved to the

city slums and the migrant lumber camps. Among these groups,

"free sex behavior and spontaneous matings" of short dura-

tion, violence, poverty, ignorance, and "the absence of

family tradition and community controls" were the norm

(1948:257).


Recent Data on Black Family Structure

Several books and articles have been published recently

which question in some way the conclusions of Du Bois and

Frazier concerning the 19th-century black family. Three of

the most important are discussed here.

The year 1974 saw the publication of two important

works on slavery, both of which contained many observations

concerning the family life of slaves. The first of these,

and the most controversial, was Time on the Cross by Robert

Fogel and Stanley Engerman (1974). The authors analyzed

data from a large number of plantations and concluded that









Frazier was wrong in his insistence that slave families were

unstable and mother-dominant. These authors emphasized the

strengths of the black family, arguing that owners had much

to gain from seeing that black families stayed together and

were satisfied. They believe that monogamy was the rule

among slaves and that most lived in single-family houses

rather than in quarters. Fogel and Engerman also deny the

degree of sexual exploitation described by Frazier and Du Bois

as well as Frazier's "matriarchal thesis." "For better or

worse," they write, "the dominant role in slave society was

played by men, not women" (1974:142).

Fogel and Engerman's book has been attacked from many

quarters and it seems clear that the book is marred by

methodological and theoretical flaws. The study did have

access to very good data sources, however, and its findings

cannot be completely brushed aside. Perhaps most important

of all, Time on the Cross has done more than any recent

book to reopen debate on some questions concerning slavery,

including some pertaining to slave families, which had

probably been closed for too long.

The second work on slavery published in 1974, which

has been acclaimed by historians, is Eugene Genovese's

Roll, Jordan, Roll: The World the Slaves Made. Even though








a few criticisms of Genovese's work have appeared, the gen-

eral consensus of almost all who have read the book is that

it is a "masterpiece" (Elkins, 1975:48). In various parts

of the book, Genovese describes aspects of slave life which

deal with family matters.

Genovese gives much evidence to refute the "myth of

the absent family" among slaves, as presented by Du Bois and

Frazier. He points out that, immediately after emancipa-

tion, there was a rush by many slave couples to have their

unions legalized; and many runaway slave problems had family

motives. In spite of tremendous pressures of a difficult

life situation, the slaves managed to create a family life

of their own. Even before emancipation, many masters took

pride in the maintenance of intact families among their

slaves, because this gave them better social control, en-

couraged responsibility among slave men, and furthered

satisfaction with the slave status. In addition to stable

marriages and nuclear groups, Genovese argues that slaves

reached out to extended kin to provide protection and sup-

port in times of trouble or need (1974:450-458).

Genovese did recognize some problems and some unique

norms of slave families. The sexual code of slaves, for

example, was not as strict as was that of whites, including









more acceptance of divorce and remarriage. Some of the dif-

ferences in these norms gave rise to accusations of "immor-

ality" from the white quarter (1974:461-475).

The place of the father, according to Genovese, was

more important than Frazier and Du Bois recognized. He

argued that the norm was for fathers to have authority over

their wives and children and that this dominance was ac-

cepted by both male and female slaves. Genovese admits

that enough cases of indifferent fathers and strong mothers

occurred to give rise to the "myth" of the matriarchy, but

he argues that the two-parent, father-dominant family was

the ideal (1974:490-494).

Another historian of the black family, Herbert Gutman,

has argued even more strongly for the basic stability of

the black family under slavery and afterwards (1973). For

Gutman, it was the urban experience of discrimination and

poverty which caused the decline of the stability of the

black family and which led to the high incidence of female-

headed families. Stanley Elkins, in reviewing a forthcoming

book by Gutman on black family history (1975), writes that

Gutman's major conclusions point to family continuity over

time, the importance of the black father in the family, and

a "high degree of stability."









It is to these areas of disagreement that the present

analysis of black family structure is directed. More inten-

sive and careful research is needed to deal with questions

of black family types, the stability of black marriages,

the stability of black family ties over time, and the place

of the female within the family. Until more data are avail-

able, the issues will not be resolved.



Other Considerations of This Study


In addition to focusing upon specific issues relating

to 19th-century white and black family structure, a few com-

parisons will be made in an attempt to show the relation of

certain aspects of rural social structure, social class,

and village-rural residence, to specific aspects of family

and household structure--household size, presence of boarders

and extended relatives, and marital status and sex of the

household head. We will explain changes over the 15-year

time period of this study in terms of social and economic

changes in the community.















CHAPTER III

RESEARCH DESIGN AND METHODOLOGY



The first task after having defined the problems to be

studied was to choose an appropriate geographical area from

which to,gather data for the study. An area was needed

which was primarily agricultural and fairly typical of the

"commercial farming" South. It was also necessary to find

an area for which data were available and in good condition

and whose georgraphic boundaries did not change during the

period of the study. It would be helpful, also, if there

were nonquantitative sources, such as histories of the area,

which could be consulted for background material. The main

purpose was to find a clearly defined area, such as a county

or counties, whose population could be studied in depth by

use of data from census manuscripts and other sources which

might be available.

The search for an appropriate area was limited to the

state of Florida, because the data were readily available

and there was a special census for that state in 1885.









After consideration of several areas, the county of Walton

in the Florida panhandle was chosen. Walton County was a

farming area fairly typical of much of the South, and it met

fairly well the needs of the study.



Sources


The primary sources of information on households and

individuals were the census manuscripts for the years 1870

and 1885. These were available on microfilm in the P. K.

Yonge Library of Florida History in the library of the Uni-

versity of Florida in Gainesville. The films are copies of

the handwritten enumerations that were prepared by officials

in each county and sent to the census bureau for compilation.

The originals are now in the National Archives, from which

the microfilms are available.

The major source of data was the population schedules,

which contain information on all individuals in every house-

hold enumerated. Also of interest for information about

property and agriculture were the agricultural census

schedules, taken in conjunction with the population censuses.

There was some problem in the fact that the schedules

for 1870 and 1885 differed somewhat. The 1870 schedules did









not contain information on the relationship of the individual

to the head of the household, as did those of 1885. In most

cases, however, the combined knowledge of the ages, surnames,

and birthplaces of the individuals in the household were

enough to establish the relationship with a great deal of

certainty. When no relationship could be established with

reasonable certainty, the individual was classified as a

nonrelative. There were not many cases in which this was a

problem.

Other sources of information used include the census

slave schedules of 1860, numerous census publications, ceme-

tery records, and John L. McKinnon's History of Walton

County (1911). A comparison of the validity of the data

from census manuscripts and those from published census

materials showed that the published materials were prone to

error in reporting what was actually in the manuscripts.

The data presented in this study are probably much more re-

liable than are the published figures.



The Samples


The populations to be sampled consisted of long list-

ings of individual names, by households. Since the households









were not numbered in a single sequence and since the records

were on microfilm, the manipulation of the data for sampling

purposes was cumbersome. A simple random sample, under

these conditions, would have been difficult and perhaps not

as good as the sampling method chosen. A systematic sample

was taken of the white households, thus assuring that all

sections of the county were covered. The sample may be con-

sidered equivalent to a simple random sample for statistical

purposes' (Mendenhall et al., 1971:151-152). The sampling

proportions were one-in-four for the white households of

1870 and one-in-five for the white households of 1885.

Since the black population of Walton County was small

and since it was very important to the study, the total

population was included in the study. Data were gathered

for every black household and for every black individual

living in a white household.

In addition to the stratification by race, the 1885

white sample was stratified by residence. In that year, the

households were reported by locally defined place names or

town or village. One village was selected for more in-depth

study, and data were gathered for every white.household within

the village. In some cases, the difference in sampling

proportions had to be taken into account during the analysis









by weighting the village cases differently than the other

cases.



Operational Definitions


Before explaining how the data were coded, it is neces-

sary to define some basic terms as they are used in this

study.

Family--two or more people residing together who are

all relatives of primary degree or who are directly related

lineally.

Primary family--the family of the household head; if

two or more families are present, the one whose status and

property appears to be dominant.

Subfamily--a family group residing with another family

but not lineally related thereto and not including the house-

hold head, but which is part of the kinship group of the

household head.

Secondary family--a family group which is resident with

another family to which it is not related.

Household--the total group of individuals who reside

together as one unit, usually in one physical house; includes

boarders, servants, and extended kin.


~









Head of household--the person who is primarily respon-

sible for the economic welfare of the household.

Simple nuclear household--a household containing only

one nuclear family and no other household members.

Extended household--a household containing a primary

family plus other non-nuclear kin.

Augmented household--a household containing the primary

family and nonrelated boarders.

Note that if two adult siblings and their nuclear fami-

lies were living together, there would be two family groups,

primary and subfamily. If the same siblings were living

with a parent, however, the lineal tie of both to the parent

would connect them into one primary family. The strict

definition of these concepts was necessary in order to clas-

sify household family structural types.



Data Organization


Data were gathered for each household in the samples

taken. For each individual, two records were prepared, a

household record containing information about the household

and a "person" record. The household record contained some

17 bits of information, as follows:









1. Type of dwelling, whether single family or group

quarters

2. Single or plural family groups)

3. Number of head's kin

4. Number of boarders

5. Number of servants

6. Structure of the primary family

7. Number of head's own children

8. Birthplace of head

9. Time since migration to Florida

10. Marital status of head

11. Race of head

12. Sex of head

13. Age of head

14. Age of oldest child of head (in household)

15. Age of youngest child of head (in household)

16. Kinship (relation to head) of non-nuclear members

17. The number of stepchildren of head (if any).

A complete household schedule, including the explanation of

codes used for all of the items, is found in Appendix A.

A few items require more thorough explanation, however.

Family structure refers to the pattern of kinship rela-

tionships within the primary family of the household. The









typology used draws heavily from that developed by Anderson

(1972) with some minor modifications. The categories used

in this typology were as follows: primary individual,

married couple only, complete nuclear family, stem family,

sibling family, and other families. Most of the categories

are self-explanatory,having been discussed in Chapter II,

pages 10-12.

Migration time was estimated from looking at the birth-

places of the parents and children. In many cases there was

not enough information to estimate this, but if some of the

children were born in another state and others in Florida,

a rough estimate of the date of migration could be made by

using the ages of the children. For those born in Florida

or whose older children, age 15 or above, were born in

Florida, it was certain that either these were not migrants

or that they were not recent migrants. Admittedly, this is

a gross estimate of recency of migration, but in the absence

of better data it is believed to be useful.

One of the first problems in coding the data was to

determine the head of the household. While in many cases it

was fairly obvious, it became more difficult in more complex

households. The misclassification of headship has certain

consequences for the analyses, especially if the









misclassifications are not simply random. For example, in

a three-generational family, is the male of the oldest

generation the head, or is the son in the middle generation?

If an older widower in such a household is classified as

the head, should an older widow in a similar situation also

be classified as the head? In this study it is assumed that

the oldest person--male or female--who is given an occupation

which seems to be dominant in the household is the head.

Thus, in a few cases the headship was assigned to a female

who was living with her son and his wife, but only when she

was classified by the census taker as, for example, a

"farmer" or "planter," and her son was a "laborer." Where

the woman was given no occupation, she was not considered

the head of the household if there was an adult male within

the family. Obviously this problem is important when we look

at the proportion of male versus female heads of households.

The assignment of headship is in some ways arbitrary; there

is no one "correct" way to do it. But one must be careful

to avoid coding so as to introduce unknown bias or to prove

one's own pet hypotheses. The main rule to follow is to be

explicit about how the statuses are assigned,so that others

can judge the work as well as understand it, and to be

consistent in the method of assignment. The following rules

of thumb were developed for use in this study.









1. In a nuclear family, the husband is the household

head.

2. In a one-parent family, .the parent of either sex is

the head, if there are no extensions. If a mother and chil-

dren are living with a male relative whose occupation seems

to be the dominant one in the household, he is the head.

3. Where two or more adult males and their families

are residing together, the one whose name appears first in

the manuscript is the head, unless the other's occupation

or kinship status clearly appears to make him dominant.

4. If there are two adult generations present, the

parent is the head, unless he/she is not given an occupation

by the census enumerator.

In general, the judgment of the census taker was relied

upon, since he usually knew most of the people in the dis-

trict. Finally, the coding of headship was a decision of

the author, and all of the relevant factors (age, relation-

ship, occupation) were carefully weighed in making the assign-

ment.

Another problem arose at times when the manuscript it-

self seemed wrong or confusing. In every case where a house-

hold record in the manuscripts seemed of questionable validity

or where the errors were not obvious (as in John Smith's sex









being coded as female), the household was not included in

the sample. There were only a few of these households--four

or five in all. Most were cases in which the ages of the

"parents" were impossibly close to the ages of their alleged

"children," or similar types of erroneous records.

The second type of record was the person record which

contained information on each individual in each household

sampled. The information recorded included the name, race,

sex, age, marital status, occupation, birthplace, and father's

and mother's birthplaces. Also included were codes for the

individual's relationship to the'head of the household and

his/her occupational class. For 1870, t.he amount of real

and personal property belonging to the household head was

recorded (this item was not included in the 1885 census).

The complete schedule of codes for the person record may

be found in Appendix B.

Marital status was coded on the basis of the de facto

residential arrangement of the persons in the sample, not

the legal status. Any common-law marriage, under this sys-

tem, was treated as a marriage. Likewise, any woman with

a child, but no husband, was considered to be "widowed," even

though there may have been a few never-married mothers in

this category. There were very few cases in which the









marital status was given by the census taker as "divorced"

or "separated." All of these cases were treated as "widowed"

or "formerly married." It was impossible, from the data

available, to make more precise classifications than these.

Still, it seems legitimate to look at simply the de facto

marital and residential arrangement as an indicator of family

and marital stability.



Analysis of Data


After the data were all coded and punched on machine-

readable cards, they were transferred to magnetic tape

for processing. Most of the analyses done were simple,

because the primary goal of this research is description.

The specific methods of analysis will be presented as we

come to them in the analysis chapters to follow. All com-

puter work was done with the Statistical Package for the

Social Sciences (Nie et al., 1975) and the Statistical

Analysis System (Barr and Goodnight, 1972).















CHAPTER IV

HISTORICAL PERSPECTIVES



This study focuses upon a Southern American county,

Walton.County in Florida's panhandle, during the decades

immediately following the Civil War. There have been very

few studies that deal with the Southern population during

the 19th century, even though its character has been recog-

nized by historians as unique (see Grantham, 1967). The

South is a very good setting for studies of black 19th-

century families, most of whom were in that region at that

time,and also for the study of rural families in general.

The industrial process had had very little impact upon the

region during the years involved. Finally, the rapid changes

occurring in Southern society during the 1870s and 1880s

provide an opportunity to observe families in the midst of

great changes and stress.









Southern Economic and Social Conditions
in the Late 19th Century


In the decades following the Civil War, the South

experienced rapid, disruptive changes in labor, land tenure,

and economy. At the same time, the entire nation was ex-

periencing a fairly serious economic decline which lasted

throughout the 1870s and 1880s.

With almost yearly declines in farm prices (Figure 1)

came increasing hardship for Southern farmers. Previous

studies have shown that the plantation system of the ante-

bellum period survived and actually expanded during these

decades with the development of sharecroppingg" (Woodward,

1951; Smith, 1953:318). Most of the literature has concen-

trated on the problems of the plantation South, to the

neglect of the approximately half of its land area which

was organized into middle-class, family-sized farms (see

Gray, 1933, and Smith, 1953). One notable exception is

found in the work of Owsley (1949), Plain Folk of the Old

South. This neglected side of Southern culture is the

setting for the present study.







47





















C



N.
0

co



C
-68L






U,
u










LA
LA





0 ai
cu u
Io
D88 E




~L8[
rccu




LL U
u c



4-

0

s- V)


0






)L8L

5u L
00D C) CD C C) :
Q)X L
C) m 000 r- 0
S- V) L









The Southern Agricultural Population
in the 19th Century


The 19th-century South was almost entirely agricultural

and rural, but there was diversity in the kinds of agricul-

ture practiced (Owsley, 1949:7). The majority of farms

were small, family farms, not plantations (Gray, 1933:481-

483).

In addition to planters, Gray describes two other types

of Southern farmers, both of which are important in this

study. These two groups he called the "poor whites" and

the "commercial farmers." Poor whites resembled pioneer

farmers, living in crude one-room log cabins with few

furnishings. They cultivated small patches of corn or rice,

sweet potatoes, cowpeas, and other garden produce. On their

subsistence farms, the women and children did most of the

farm work, while the men either hunted or were idle. Some-

times these farmers owned a few hogs or other animals; they

almost always had a dog and a rifle. These people were

despised by many more fortunate Southerners as being of an

inferior class, the "clay eaters."

The second class mentioned by Gray was the commercial

farmers, the "yeomen of the South." These were the owners

of family-sized farms, some of whom owned a few slaves.












These people engaged in diversified farming, in contrast

to the planters, who dealt in staple crop production. Their

social standing was quite varied, ranging from very poor to

fairly well-off, slave-owning farmers. Their houses were

usually comfortable; they filled their barns with hay and

forage; and they usually had gardens and orchards. If they

had no slaves, the women worked at making clothing and pre-

serving food (Gray, 1933:489). The relations of these

farmers with their slaves were frequently very friendly and

"almost intimate." Gray characterizes this class as having

"sturdy independence, self-respect," sociability, hospital-

ity, and a democratic spirit (1933:490).

Gray also mentions a class of "free white laborers"

whose living conditions varied with their labor contracts

and the type of agricultural work which they did. After

the war, the black population was often in a similar posi-

tion to these workers, and often in competition for the same

work.

Most of the farmers in the present study were of the

commercial farming type or farm laborers. These are the

people whom Frank Owsley described as the "plain folk of

the Old South."









The plain folk . were usually landowning farmers
and herdsmen, though a small minority were engaged
in other occupations. Their thoughts, traditions,
and legends were rural, for with the exception of
an occasional ancestor who had been brought from
some British city or debtor's prison as an indentured
servant, their families were rooted in the soil. . .
To them the land was, with God's blessings, the direct
source of all the necessities of life. . (Owsley,
1949:vii)

They were often from Scotch or Scotch-Irish origins. Some

became large landowners or professionals, but the majority

"remained landowning farmers who belonged neither to the

plantation economy nor to the destitute . poor white

class. They, and not the poor-whites, comprised the bulk

of the Southern population from the Revolution to the Civil

War" (Owsley, 1949:viii).

According to Owsley, these people were a true "folk,"

with a sense of solidarity based on their common origins in

the British Isles. They were characterized by "closely knit

families," strong religious feeling, and self-sufficiency.

These traits, according to Owsley, allowed them to survive

the period of Reconstruction without the disruption that was

felt in other parts of the South.

One more element of the Southern agricultural popula-

tion remains to be discussed. This is, of course, the black

population, most of whom were slave laborers before









emancipation. Their condition depended largely upon the

kind of agricultural system in which they were working and

on the humanity of their owners or overseers. According to

Smith (1953), many of the problems of slavery were, in

fact, the problems of plantation agriculture, which did not

disappear after slavery was ended. In the farming districts,

however, the slaves were in more direct and open contact with

their owners. Typically, the farmer and his slave worked

in the fields together. With such personal, face-to-face

relationships, some of the more dehumanizing aspects of

plantation slavery were avoided. After the end of slavery,

apparently, most black farm workers in the South were either

"farmers," who owned or rented small plots of land and were

similar to white small farmers, or farm laborers, either

working for a share of the crop or for a cash share.



Walton County, Florida, in the 19th Century]


Walton County lies in the western panhandle of Florida,

bordering Alabama on the north and the Gulf of Mexico to

the south. It was first settled by white men in the early



]Most of the information for this section comes from
J. L. McKinnon, History of Walton County (1911).









1820s, when a party of Scottish settlers from North Carolina

came south in search of fertile lands on which to settle.

They found what later became Walton County, an area of forest

and swamp, full of rivers, springs, and lakes.

They were able to make friendly territorial arrangements

with the Euchee Indians who were already there, and the set-

tlement known as Euchee Anna was formed. These early settlers

were followed in the next three decades by others, many of

them friends or relatives of earlier migrants to Florida.

After about a decade of white settlement, the Indians left

to go farther south, complaining that the newer white

settlers were destroying their forests and wildlife.

In 1824, Walton County was officially formed and its

area encompassed what is now Walton, Holmes, Okaloosa, and

Santa Rosa Counties. Later, Santa Rosa and Holmes Counties

were formed; and in 1915, Okaloosa County was formed. At

the time with which the present study deals, Walton County

still included the Okaloosa County area. Almost all of

the settled area of the county, however, was in the south-

east corner of the county, the rest of the county still a

wilderness. This part of the county is within the present

boundaries of Walton County.









In the early years, Walton County was quite isolated,

travel being very difficult. Between 1837 and 1860, roads

were built from Tallahassee to Pensacola through the county

and from the bayfront village of Freeport to Euchee Anna

and on northward into Alabama. As a result, travel and

trade increased sharply, a trip to Pensacola taking only

one week instead of three or four. Most of the trade was

with merchants in Pensacola, and most goods went by way of

sailing vessels out of Freeport. Cargo was mainly the prod-

ducts of the small farms of the county: molasses, corn,

cotton, chickens and eggs, venison, lamb, pork, tallow,

raw hides, and other farm produce.

By 1860, according to the census publications for that

year, there were 2,584 whites and 453 blacks in Walton

County. The black population had been brought in with the

white settlers for slave labor on the farms.

The Walton County antebellum society was typical of

Gray's "commercial farming" type, although in 1860 there

were three farms that could be classified as small plan-

tations. Most of the slaveholders in the county owned

only a few slaves and had small family farms (Smith,

1973). The slaves and their owners worked together

in the fields, with no overseer or drivers to direct









their work. According to Julia Smith (1973), there were few

disciplinary problems in such a situation, and the slaves

had relatively good living conditions. They often ate from

the same kitchen as did their owners, and they attended the

same churches, the slaves sitting in a separate section of

pews or in a special gallery. McKinnon wrote that Walton's

slaves "were all Presbyterians," holding weekly prayer meet-

ings in their own quarters or houses and worshipping in the

Valley Presbyterian Church on Sundays.

According to McKinnon, the slaves of Walton County

typically lived in small cabins behind the "Big House,"

not in communal quarters. In most cases, each slaveowner

had only one slave, and only a few in the slave schedules

for 1860 showed more than 10. The close working arrangement

and the personal relationships between slave families and

their owners led, in some cases, to a kind of "family soli-

darity" among slaves and owners (Smith, 1973).

The census publications for 1870 reported that there

were eight churches in the county: one Baptist, three

Presbyterian, and four Methodist. Judging from the number

of seatings in the buildings, there were more Presbyterians

than Methodists and still fewer Baptists (United States De-

partment of the Interior, 1872b:533). Since a large









proportion of the original population of the county was of

Scotch origin, this is as would be expected. The original

church in the county, and the center of county life for

many, was the Valley Church, first organized in 1828. It

was located near Euchee Anna and the other early settlements,

and it was the site of the first and only cemetery in the

county for many years.

In 1870, Walton County had a total of 276 farms, half

of which were between 20 and 50 acres in size. Another 41

percent of these farms were smaller than 20 acres, with only

five farms having 100 or more acres. The average acreage

per farm was 24.6 (United States Department of the Interior,

1872c:348). The main crops of the Walton County farmers

were Indian corn, oats, molasses, rice, cotton, and some

tobacco.

Walton County's delegates were not secessionists, the

county having been a "Whig county," not unfavorable toward

Lincoln (McKinnon, 1911:269-270). They were among the few

county delegates to refuse to sign the secession articles

in 1861. But secession came anyway, and Walton County sent

her share of young men. Naturally, as the war progressed,

the involvement became more intense, and the former Whigs

became staunch Confederates. McKinnon lists 90 names of









Walton County Confederate soldiers who were killed in the

war (1911:377-378). In the later war years, there were some

minor raids into the county, and there was one final raid

by Union soldiers in which much livestock was killed or

stolen and many of Walton's citizens were kept imprisoned in

Euchee Anna for two days, according to McKinnon (1911:327-

328). These soldiers took many of the former slaves with

them, although it is not known how many. The turmoil and

high feelings of the war seem to have broken the racial accom-

modation which had been fairly stable before the war. Not

long after the war, Walton County experienced its first

lynching, and there were several instances of tension and

resentment mentioned by McKinnon.



Transition of Walton County, 1870 to 1885


Walton County changed greatly after the Civil War.

Before the war, the county was almost totally devoted to

agricultural and livestock interests. Afterwards, however,

rapid development of the county's resources began, and the

economic basis of the county changed. The postbellum county

residents engaged in the timber business, logging, steam and

sawmill building, sail and steamboat building, and trade by









way of the bay and the rivers. For a while after the war,

the efforts to develop resources were very successful; then

the economic depression hit and developments slowed down.

The depression hit farmers the hardest. In 1869, the

276 farms of Walton County had produced an average of $550

worth of goods, while the average value of produce for 1885

farms was $282. During the same interval, the number of

farms increased by 31 percent to 361, while the number of

improved acres increased by only 20 percent from 6,803 in

1870 to 8,152 in 1890. The overall value of farm produce

for the county dropped by 33 percent, from $151,833 in 1869

to $101,780 in 1889. (These figures were calculated from

data given in United States Department of the.Interior,

1872a:720; 1872c:116, 348; and United States House of Repre-

sentatives, 1896:129, 202.)

In the early 1880s the Louisville and Nashville Rail-

road built a branch through Walton County and, according

to McKinnon, its impact on the county was great.

It acted as though it was a great bomb shell
dropped down in the midst of the Valley, crush-
ing, rooting up and driving the old Scotch settlers
in every direction, leaving only enough there for
seed. . It broke up . one of the plainest,
simplest, most social, and truly religious communi-
ties in Walton. (McKinnon, 1911:350)









Apparently the new form of transportation gave many

residents of Walton an opportunity to move which they had

not had before. It also made moves away from one's family

less traumatic, as the way back for visits by railroad was

not so difficult as former modes of travel.

With the railroad came the building of Lake De Funiak,

later known as De Funiak Springs. This village was founded

in the early 1880s and built on a small hill near the new

railroad. It was settled by newcomers, both Southerners

and Northerners. The town was advertised as a tourist

attraction, and had some success-as such.

The racial tensions and antagonisms arising from the

war and Reconstruction seem to have reached a peak in Walton

County in the late 1870s, as whites were more and more upset

by the rule of the "carpetbag government" (McKinnon, 1911:

342). In 1876, however, Governor Drew was elected and the

Reconstruction era in Florida came to an end. McKinnon men-

tions several examples of racial disturbances in the county

which took place during the 1870s, but he seems to recog-

nize a distinction between the "old family Negroes," who

were known to the county and who were fairly stable in family

life,and others who were migrants to the county, often work-

ing in the logging camps and who were disliked even by the

older black population of the county.









It seems likely that a kind of new accommodation had

been reached by 1885, although it was not completely stable.

Since this was the beginning of the real Jim Crow era in the

South, it seems likely that the black population of Walton

County may have been worse off in many ways than it had been

in 1870. However, it is likely that whatever the situation

was in 1885, it may have been fairly stable as compared to

that of 1870. The economic problems of the 1880s had probably

hit the black population very hard, though, since they were

still in many ways dependent on the whites, who now were not

as likely to be protective of their interests and feelings

as they might have been earlier.

We now turn to the analysis of the characteristics of

the population of Walton County as it was in 1870. Perhaps

it was similar to the prewar population, since there had

not been much time for radical changes to occur. The Recon-

struction era in Florida lasted until 1877, and it was not

until the late 1870s that racial tensions and Jim Crowism

developed to a great degree. In 1870, then, the transition

from antebellum to postbellum society was still probably in

its infancy.















CHAPTER V

THE POPULATION OF WALTON COUNTY,
1870 AND 1885



Among the most important determinants of the structure

of a community are the characteristics of its population.

By this is meant the composition of the population in terms

of age, sex, marital status, residence, and other demographic

variables, and the nature of its processes of fertility,

migration, and mortality. In this chapter, we will describe

these characteristics of Walton County for the years 1870

and 1885, for whites and blacks separately.



The White Population of Walton County, 1870


In 1870, there were some 2,800 white individuals who

were enumerated in the Walton County census manuscripts.

The sample obtained contained information on 710 individuals

living in 115 white-headed households. The average house-

hold size, then, was 6.2 persons.









Race

Within the white-headed households in the sample, there

were 684 white individuals, 22 blacks, and 4 mulattoes

(Table 1). About 23 percent of all white households in the

1870 sample had at least one nonwhite resident. This is

probably close to the proportion of households which 10

years earlier had at least one slave. For the remainder

of this discussion, the "white sample" refers to only the

white residents of white households. Since all white per-

sons in each household chosen were included in the study,

this means that the white sample is equivalent to a simple

random sample of white persons in the county.



Table 1. Racial Composition of White Household Sample, 1870


Race N Percent

White 684 96.3

Black 22 3.1

Mulatto 4 0.6


Total 710 100.0









Sex

The sample of white households contained 328 white

males and 356 white females, for a sex ratio of 92.1 males

per 100 females (Table 2). This fairly low sex ratio, al-

though not extreme, may be partially due to the loss of men

during the recent war. It is rather unusual for a rural and

relatively newly settled area to have a sex ratio very far

below 100.



Table 2. Sex Composition of White Sample, 1870


Sex N Percent

Male 328 48.0

Female 356 52.0



Total 694 100.0




Age

The mean age of the white sample was 22.1 years, al-

though the median was only 17.5 years. The ages of indi-

viduals ranged from birth to 80 years, and the quartiles

were 9 and 30,years. In other words, it was a fairly young









population, with three-fourths of its members being under

30 years of age. Moreover, over 40 percent of the sample

was under 15 years of age.

Males had a slightly wider distribution of ages than

did females. The median age for males was 17.0 years; for

females, 17.7 years. The quartiles were, for males, 15 and

32; for females, 17 and 30.

The cause of the low sex ratio can be hypothesized by

looking at the age and sex distribution of Walton County's

white population in 1870 (Figure 2). First, there appears

to be an imbalance in the numbers of men and women aged

25 to 39 years. This would have been the age group which

had the heaviest losses in the war for men. Walton County's

Confederate dead numbered 90, according to McKinnon's list

(1911:377-379). This might partly explain the low sex ratio

for the county in 1870.

Note also the lower numbers of children in the youngest

two age groups of the diagram. These two bars represent

the children born during the war years and afterward, and a

lower birth rate for those years seems likely. Another pos-

sibility is that the small number of young children is due

to enumeration errors. It seems likely that, in cases where

the enumerator found no one home and got his information





64














MM8






MX x
| x I








X XXM X


XMMMM X MM X


E XXXMMM MX
xx> S8 88888888888




?S ? MM S o '+o
X 8888888888888
Xxx
0 410



>4 80 "88 >
x x x n
Zx


# 8



O)
di

88888 01
XX XXX v, C.

8 XX O z

Z 3
S8 4-






0 -






t CofCflC'iC Nlr-.--al'a-









from a neighbor, infants and very young children might have

been more easily forgotten than older, more visible, chil-

dren. If this happened to any large extent, it could de-

crease the number of children enumerated in these age groups,

thus giving the appearance to lowered fertility for those

years. At this point in time, there is no way to determine

the exact source of the small number of young children in

the census listings.


Marital Status

Of the 402 whites who were 15 years of age or older,

184 (45.8 percent) were single, 186 (46.3 percent) were

married, and 32 (8.0 percent) were separated, divorced,

or widowed (see Table 3). This proportion of married adults

is not unusually low, since the average age at marriage

was high in the 19th century.

Although there were more aged men than women, a rela-

tively low proportion of males were widowed. Widowers

could usually find second wives, because they could marry down

in the age structure. Widows, however, had to look upward

in the age structure for husbands. This severely limited

their chances of remarriage. Thus, there were three times

as many women as men in the widowed category.









Table 3. Marital Status of
by Sex


Adults in White Sample, 1870,.


Marital Male Female Total
Status N Percent N Percent N Percent

Single 90 47.1 94 44.5 184 45.8

Married 93 48.7 93 44.1 186 46.3

Widowed or 8 4.2 24 11.4 32 8.0
other



Total 191 100.0 211 100.0 402 100.1




Estimated Age at Marriage

Since there were no data available on age at marriage,

the median age at marriage was estimated by means of a

cumulative distribution of the proportion of individuals

of each sex who were ever-married at each age between 15

and 45.1 The resulting estimates should only be considered

as crude indicators of the median age at marriage.

For the white sample, the youngest married males were

19 years old; for females, the corresponding age was 17.

Only 38.6 percent of the male adults aged 15-45 years had



A description of the procedure is found in Appendix C.










been married, and the estimated median age at marriage was

29.0 years. Hence, according to these estimates, the usual

age at marriage for both men and women in this sample was

rather late. For comparison, the median age at marriage

in the United States as a whole in 1890 was 26.1 years for

men and 22 years for women (Glick, 1957:54). It appears

that marriage in Walton County took place at later ages

than in the country as a whole, for whites at least.


Fertility

Since there were no vital statistics in Walton County

in 1870, the best measure of fertility is the fertility

ratio, or the ratio of children under five years of age to

women in the childbearing years, age 15 to 44, inclusive,

multiplied by 1,000 for better readability. In Walton

County in 1870, the fertility ratio, as calculated from

the sample data, was 505.9, which is somewhat lower than

that for the United States as a whole (about 650) (Broom

and Selznick, 1968:278). The lowered birth rate for the

war years and following may have lowered this ratio, since

it involves those children born between 1865 and 1869, in-

clusive. It is probable that the social upheavals result-

ing from the war and Reconstruction had caused temporary









declines in fertility and marriage rates in Walton County,

both of which are reflected in the measures at which we

have been looking. On the other hand, the possibility of

enumeration errors in the youngest age categories, as men-

tioned earlier, still remains.


Mortality

Unfortunately, the only records available of mortality

in Walton County during this period are those inscribed upon

burial markers and stones in the Valley Church cemetery.

There are problems, of course, with the representativeness of

this sample of deaths, since some graves may have been

unmarked, and some individuals may have been buried else-

where. But, as is common in historical research, we must

take the data which we are given and note their weaknesses.

Some useful insights may come from looking at these records,

even if they are not of the precision which we would demand

for present-day statistics. One positive note about the

data from this cemetery is that it was the main cemetery

in the county during the years of the study (McKinnon,

1911:256). Although the cemetery had both black and white

graves, the data presented here are for the whites only,

and they are taken from a transcription by Bruington (1951).









We were interested in getting an estimate of the average

age of death for males and females in Walton County during

the decades under analysis. Since the number of deaths was

fairly small, all deaths between 1850 and 1885 were recorded

and analyzed together, so that any conclusions to be made

apply to the general period, and not to the specific decade

or year. There were 70 male deaths and 55 female deaths

recorded. In addition, there was one person whose sex could

not be determined in the records. In all, 126 deaths were

recorded, excluding soldiers' deaths. For each, the year

of birth, the year of death, and the sex of the individual

were recorded, along with the state of birth, where given.

There were 46 deaths in the 20 years prior to 1870, and

80 in the 15 years afterward, which reflects the growing

population of the county as well as the growing old of the

first generation of settlers. About one-fourth of all of

these deaths occurred before the age of 15 (25.4 percent),

most of these, of course, in the first five years of life.

One-third of the male deaths occurred between the ages of

15 and 44, but 44 percent of the female deaths occurred

during these ages. About 41 percent of the males and 33

percent of the females had died above the age of 45. The









male-female differential represents the greater death rate

of women due to childbirth. If, however, we look only at

those who lived through the childbearing years (those age

45 or above), women had higher average ages at death than

men. Specifically, the mean age at death for those who

died after 44 was 64 years for men and 69.1 years for women.

The age at which women died at a much greater rate than men

was in the range from 35 to 44 years, the late childbearing

years. Still, about 25 percent of all adult deaths, both

male and female, occurred at or above the age of 65 years.


Occupation and Social Class

The occupations of the individuals in the sample are

given in Table 4. A glance at the titles and the distribu-

tion is enough to establish the fact that most of Walton

County's laboring population in 1870 was employed in rural

occupations--farming, lumbering, milling, and simple labor-

ing. It seems clear from the context of the manuscripts

that most of the individuals classified as simply "laborers"

were farm workers who did not own land. The "farmers"

were those who did own at least a small plot of land.









Table 4. List of Occupations of the White Sample, 1870


Title N

Laborer 87
Farmer 75
Carpenter 5
Servant 5
Lumberman 4
Seamstress 4
Sail 3
Captain of lighter 3
Shingle maker 3
Miller 3
School teacher 3
Cooper 2
Driver 2
Teamster 1
Peddler 1
Clerk 1
Boat mate 1
Minister 1
Blacksmith 1


Total 205




The occupations were classified into the following six

categories: (1) professional, (2) managerial, proprietori-

al, and clerical, (3) skilled crafts, (4) unskilled trades,

(5) laborers and servants, and (6) farmers and stockmen.

The results of this classification are presented in Table

5.









Table 5. Occupational Class of White Sample, 1870


Class N Percent

Professional 4 1.9

Managerial, clerical, 6 2.9
proprietorial

Skilled craftsmen 18 8.7

Unskilled manual 11 5.3

Laborers, menial 92 44.7
service

Farmers 75 36.4



Total 206 99.9




The modal status level was that of the laborers, with

farmers being second. Both of these groups together made

up about 80 percent of the laboring population. In addi-

tion, there were a few skilled laborers (18) and some un-

skilled laborers with titles other than "laborer" or

"servant." Only 10 individuals of the 206 with occupations

given were in either the professional or proprietorial

classes (4.8 percent of the total). The county was defi-

nitely dominated by agrarian interests, having only enough

of the nonfarm occupations to support the farm population.









Birthplaces of the Population

Just over half of the white sample had been born in

Florida (Table 6). Another 28 percent were born in Alabama,

with 7.5 percent originating in Georgia. Thus, these three

states were the birthplaces of 90 percent of the white

sample. The rest came mostly from the upper South, with

less than 1 percent being of foreign birth. Almost all of

the original Scotch settlers, many of whom had been born in

Scotland, were dead by 1870. The population represented

here probably is largely of first- or second-generation

American birth.



Table 6. Birthplaces of the White Sample, 1870


State of Birth N Percent

Florida 367 53.3
Alabama 196 28.5
Georgia 51 7.4
N. Carolina 33 4.8
S. Carolina 30 4.4
Virginia 2 0.3
Other South 2 0.3
Other U.S. 1 0.1
Foreign 6 0.8


Total 688 99.9









Length of Florida Residence

The large proportion of Florida-born individuals over-

estimates the proportion of families who came from other

states, since many younger children, even of migrant fami-

lies, were born in Florida. In order to control for this

problem, estimates of the recency of migration of families

were made, where possible. This estimate was based on the

place of birth of the household head, the places of birth of

his children, and their ages. It was possible to estimate

this for 85 of the 115 primary facilities in the sample.

The results of the estimates are presented in Table 7.


Table 7. Recency of Migration
Heads, White Sample,


of Families
1870


of Household


Length of Florida
Residence N Percent

Florida native 21 24.7

15 years or more 38 4.4.7

10-14 years 6 7.1

5-9 years 2 2.4

0-4 years 18 21.2



Total 85 100.1









Of the 85 families, about one-fourth were Florida natives,

and the bulk of the rest had lived in the state for longer

than 15 years. As might be expected, only a few had come

to Florida during the war years. After 1865, however, the

immigration to the state seems to have picked up, as just

over 20 percent of household heads had been in Florida for

less than five years.



The Black Population of Walton County, 1870


There were 3802 black or "mulatto" individuals listed

in the Walton County Census of 1870. Almost three-fourths

of these individuals (274 in all) lived in 48 black-headed

households; while the remaining 106 were listed as part of

47 white households. This latter group seems to be a hold-

over from the days of slavery, when black slaves were listed

under their masters' names. There is evidence for believ-

ing that the social condition of these blacks was also little

different from that which existed prior to emancipation.

If so, this group makes an excellent source of information

about the black family in the South in the last years of slavery.



2This includes two whites, one who was the wife of a
black household head, and the other who was her son, pre-
sumably by an earlier marriage.









The mean household size for black households in.1870

was 5.7, slightly less than the white mean of 6.2 The mean

number of blacks per white household containing blacks was

2.3, with a median and mode of one per household (27 of the

47 households had one black individual each.) This is simi-

lar to data taken from the 1860 slave schedules, in which 128

slaveowners held an average of 4.3 slaves each, the mode

being one. The mean was high due to a few (nine) slave-

holders who owned more than 10 slaves, one of whom owned

46. This is one reason for believing that the blacks living

in close association with whites in 1870 were in many cases

former slaves who had never left the "master's" farm.


Race

In 1870, nonwhites were classified by the census

enumerator as being either "black" or "mulatto," depending

mainly, it appears, on the skin color of the individual.

According to many analysts of black life and culture in

the South after the Civil War (see, for example, Frazier,

1948), there was a prejudice in both the white and black

communities in favor of lighter-skinned blacks. In the

black household data, there were 192 blacks and 80 mulattoes;

in white households, there were 80 blacks and 26 mulattoes









(Table 8). Overall, about 28 percent of the nonwhite popu-

lation of Walton County in 1870 was classified as mulatto.



Table 8. Racial Composition of Nonwhite Population of
Walton County, 1870

Race of Household Head Total
Black White Nonwhite
Race N Percent N Percent N Percent

Black 192 70.6 80 75.5 272 72.0

Mulatto 80 29.4 26 24.5 106 28.0


Total 272 100.0 106 100.0 378 100.0




Sex

In 1870, there were 139 males and 135 females listed

in black households, with an additional 63 black males and

43 black females in white households (Table 9). The over-

all sex ratio for blacks, then, was 202:178 or 113.5 males

per 100 females. For blacks in white households, the sex

ratio was 146.5, while that for blacks in all-black house-

holds was 103.0. In other words, the sex ratio in black

households was very evenly balanced, while that for blacks

in white households was heavily unbalanced in favor of males.

This may be related to the fact that young men were wanted









to do farm work, and thus were more often kept by the farmers

as slaves or as free workers. Note that both black groups

had higher sex ratios than did the whites, whose women out-

numbered their men with a sex ratio of 92.1.



Table 9. Sex Composition of Black Population of Walton
County, 1870, by Race of Household Head

Black Households White Households Total
Sex N Percent N Percent N Percent

Males 139 50.7 63 59.4 202 53.2

Females 135 49.3 43 40.6 178 46.8


Total 274 100.0 106 100.0 380 100.0




Age

The age range of the blacks in white households was

from birth to 64 years, with a median age of 17.0 years.

Forty-two percent of this group were under 15 years of age,

and only 11 percent were above 30. For blacks in black

households, the range was greater, from birth to 87 years,

with a median age of 17.7. The overall age distribution of

the black population in 1870 is given, by sex, in Figure 3.

Notice the paucity of blacks in the older ages and the

relatively constricted numbers of young children, especially

when compared to the white pyramid (page 64).



































x X c
3
0
a






X X4

X X



M1XX 4
X XX o
XX XX u
XX XX


a) XXXXXX
SXXXXXXXX
EX XXXXXXXXMM
a XXXXXXXX -

XMI MMI MMM
X XXXX XXX X
w X XXXXXXXX _

TO XXXXXXX
HHHHHHHHHHHHHHH


S MXXXXKXXX o


XXXXXXX +
-XXX



X co










0 0
00
I I o



+-5-

ai-
0)





I~iiiiii




Full Text

PAGE 1

HOUSEHOLD AND FAMILY IN THE POSTBELLUM SOUTH WALTON COUNTY, FLORIDA, 1870-1885 By BARBARA FINLAY A6RESTI A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1976

PAGE 2

UNIVERSITY OF FLORIDA 3 1262 08666 239 1

PAGE 3

Copyright by Barbara Finlay Agresti 1976

PAGE 4

To my parents , James and Grace Finlay, and the Finlay clan of Fife.

PAGE 5

ACKNOWLEDGMENTS A dissertation is the crowning work of a long and arduous career of graduate studies. None is ever completed without a tremendous amount of help in many forms and from many sources. I would like to take this opportunity to mention just a few of the more important people to whom I owe many thanks for their help a.nd support in my career as a graduate student. First of all, I would like to thank my chairman. Dr. Gerald Leslie, for the many hours he spent in reading, editing, and advising me concerning this and earlier drafts of my dissertation. His standards are high, but I am sure that the dissertation is much better because of his efforts. It was Dr. Leslie who introduced me to the topic of family history in the first place, and his help and encouragement during my years at the University of Florida have been greatly appreciated. To Dr. Joseph Vandiver, I owe a great deal. His insightful and broad understanding of the rural South and its IV

PAGE 6

history have been very valuable in writing this dissertation. His kindness, humor, and ability to cope with the endless rules and regulations of the graduate school while remaining warmly human have, made graduate studies in this department more enjoyable for many students, including myself. We were all saddened last spring by the death of Dr. T. Lynn Smith, whose absence is keenly felt. Much of my understanding of life in the rural South came from Dr. Smith both from his writings and from the many conversations we had over the past four years. Although he was quite active and prolific as a sociologist, he always had time for his students. His help in pointing me to the right sources of information for the background for this study and his insightful interpretations of various aspects of Southern demography and culture were invaluable aids. He was truly an inspiration to this and many other budding sociologists over the years, and his influence will continue to be felt for some time to come. I would also like to mention my other committee members--Dr. Felix Berardo, who has always been a friend as well as teacher to all students; Dr. S. S. Hill, Jr., whose friendship and kindness I will always value; and Dr. Gordon

PAGE 7

Streib, whose abilities and accomplishments have added depth to this department. I especially thank Dr. Streib for staying with this project, even while recovering from major surgery. Another person whose encouragement and faith in me have been very valuable in my career as a graduate student is Dr. W. G. Steglich, who served as my advisor and master's committee chairman at the University of Texas at El Paso. Without his support and friendship, I might never have come this far. To Alan, my husband and friend, and to Ginger and Kerry, my daughter and son, I owe a great debt of thanks. Their sacrifices, love, patience, and help in the everyday things of life made this dissertation a much less formidable task. To Ginger and Kerry, who have been with me throughout my entire college and university career, thanks for everything. You've made it all worthwhile. I suppose I owe the greatest debt to my parents, to whom this is dedicated, and my brothers, who have always inspired me to greater things and whose support has always been present. There is no way I can ever repay them. There are probably many more people who should be thanked, people whose influence contributed to the present VI

PAGE 8

work. All of those mentioned and unmentioned have added strengths to this dissertation. Any remaining weaknesses are my own doi ng. VI 1

PAGE 9

TABLE OF CONTENTS Page ACKNOWLEDGMENTS Iv LIST OF TABLES xiii LIST OF FIGURES ......... xvii ABSTRACT, .xviii CHAPTER I INTRODUCTION 1 Emerging Interest in the Social History of the Family 2 Obstacles to the Historical Study of the Family 5 General Purposes and Framework of This Study ....... 6 II THEORETICAL PERSPECTIVES AND REVIEW OF LITERATURE 9 Family and Household Structure. 9 Types of Family Groups .10 Other Aspects of Household and Family Structure . 12 The Developmental Approach 13 Major Issues Concerning the Preindustrial American White Family 14 Family Structure .14 Nonrelatives in the Household. . . . . 20 Family and Household Size. 20 Age at Marriage and Life Cycle of Families 21 Major Issues Concerning the 19th-century Bl ack Fami ly . . . . .... . . 24 W.E.B. Du Bois 24 E. Franklin Frazier. . 26 Recent Data on Black Family Structure. 28 VI 1 1

PAGE 10

TABLE OF CONTENTS (continued) CHAPTER Page II Other Considerations of This Study. ... 32 (cont. ) III RESEARCH DESIGN AND METHODOLOGY. ....... 33 Sources ..... 34 The Samples 35 Operational Definitions 37 Data Organization 38 Analysis of Data 44 IV HISTORICAL PERSPECTIVES. . 45 Southern Economic and Social Conditions in the Late 19th Century 46 The Southern Agricultural Population in the 19th Century. 48 Walton County, Florida, in the 19th Century. ' 51 Transition of Walton County, 1870 to 1885. 56 V THE POPULATION OF WALTON COUNTY, 1870 AND 1885 60 The White Population of Walton County, 1870. 60 Race 61 Sex 62 Age. 62 Marital Status 65 Estimated Age at Marriage. ..... 66 Ferti 1 i ty . . . . 67 Mortality 68 Occupation and Social Class, .... 70 Birthplaces of the Population. ... 73 Length of Florida Residence 74 The Black Population of Walton County, 1870. 75 Race 76 Sex. 77 Age. ................ 78 Marital Status 80 Estimated Age at Marriage. ..... 8l IX

PAGE 11

TABLE OF CONTENTS (continued) CHAPTER Page V Fertility 82 (cont.) Mortality 82 Occupation and Social Class. .... 83 Birthplaces of the Population. ... 86 Length of Florida Residence 88 The White Population of Walton County, 1885. 89 Race 90 Sex 91 Age 91 Marital Status ........... 95 Estimated Age at Marriage 96 Fertility. 97 Occupation and Social Class, .... 97 Birthplaces of the Population. . . . 101 Length of Florida Residence 103 The Black Population of Walton County, 1885. .................. 104 Race . 1 06 Sex. 106 Age 107 Marital Status 1 09 Estimated Age at Marriage. ..... 110 Fertility. ......... . ... Ill Occupation and Social Class 112 Birthplaces of the Population. . . . 114 Length of Florida Residence. .... 116 Conclusion: The Population of Walton County, 1870 and 1885 .......... 116 VI THE STRUCTURE OF FAMILIES AND THE DEVELOPMENTAL CYCLE, 1870 AND 1885. . . . . . . . . .119 Structure of White Families, 1870 .... 120 Young Couples .1:21 The Childbearing Years 124 The Childrearing Years ....... 124 Widowhood. 126 The Children . 128 Single Adults 130 Conclusion: The Life Cycle of the Individual 132

PAGE 12

TABLE OF CONTENTS (continued) CHAPTER Page VI Structure of White Families, 1885 .... 134 (cont.) Young Couples 136 The Chi 1 dbearing Years . 137 The Childrearing Years ....... 139 Widowhood. ............. 140 The Children 142 Single Adults. .144 Conclusion: The Life Cycle and Economic Crisis. . 146 Structure of Bl ack Fami 1 ies , 1870 .... 148 Two-parent Families. ........ 152 Widowed and Broken Families 153 Other Tests of Black Family Strengths and Stabil ity 1 55 Living Arrangements of Children. . . 159 Living Arrangements of the Aged. . . 162 Conclusion: Bl ack' Fami 1 i es in 1870 . . . 164 Structure of Black Families, 1885 . . . . 165 Young Couples 167 The Childbearing Years 168 The Childrearing Years 169 Widowhood. 170 The Children . 170 Single Adults. . 172 Conclusion: Black Family Structure and the Life Cycle. . 174 Changes in Black Family Structure, 18701885. . 177 Household Composition and the Family Cycle 180 Summary 184 VII FAMILY AND HOUSEHOLD CHARACTERISTICS AND SOCIAL STRUCTURE. 186 Selected Characteristics of Black and White Households, 1870 and 1885 186 Extended Kin 187 Boarders . 188 Household Size 190 Number of Children 191 Marital Status of Household Heads. . 191

PAGE 13

TABLE OF CONTENTS (continued) CHAPTER Page Vn Sex of Household Heads ....... 192 (cont.) Summary of Household Characteristics, 1870 and 1885 193 Household Characteristics by Rural -Vi 1 lage Residence . . ... . . . . . . . . . . . 194 Extended Kin ............ 195 Boarders ...... 196 Servants .............. 197 Household Size 198 Number of Children . 198 Structure of Primary Families. . . . 198 Summary: Household Characteristics by Rural or Village Residence. . . . 200 Household Characteristics and Social Class 201 Extended Kin ........... . 203 Boarders . . ' 204 Household Size 206 Number of Children . 207 Marital Status of Household Heads. . 210 Sex of Household Heads ....... 210 Summary: Household Characteristics and Social Class . . . . 212 VIII CONCLUSIONS 214 APPENDICES A CODING INFORMATION: HOUSEHOLD RECORD. ... . 220 B CODING INFORMATION: PERSON RECORD ...... 223 C METHOD OF ESTIMATION OF AGE AT MARRIAGE. . . . 228 REFERENCES. . 229 BIOGRAPHICAL SKETCH 238 XI 1

PAGE 14

LIST OF TABLES Table 1 2 3 4 5 6 7 10 11 12 13 U Page Racial Composition of White Household Sample, 1870 61 Sex Composition of White Sample, 1870 62 Marital Status of Adults in White Sample, 1870, by Sex. . 66 List of Occupations of the White Sample, 1870 . 71 Occupational Class of White Sample, 1870. ... 72 Birthplaces of the White Sample, 1870 73 Recency of Migration of Families of Household Heads, White Sample, 1870 ..... 74 Racial Composition of Nonwhite Population of Walton County, 1870 . 77 Sex Composition of Black Population of Walton County, 1870, by Race of Household Head .... 78 Marital Status of Adult Black Population of Walton County, 1870 80 List of Occupations of the Black Population of Walton County, 1870 ........ 84 Occupational Class of the Black Population of Walton County, 1870 ^ 85 Birthplaces of the Black Population of Walton County, 1870 87 Recency of Migration of Families of Black Household Heads, Walton County, 1870 88 X i i i

PAGE 15

15

PAGE 16

LIST OF TABLES (continued) Table Page 30 Structure of White Families in Walton County Samples, 1885 135 31 Living Arrangements of White Children in Walton County, 1885. 143 32 Living Arrangements of Single White Adults in Walton County Samples, 1885 . 146 33 Structure of Black Families in Walton County, 1870. 150 34 Percentage of Adults Not-Living in Family Groups, 1870, by Race 151 35 Probable Family Structure of Slaves of Walton and Holmes Counties, I860' 157 36 Sex of Probable Heads of Slave Families in Walton and Holmes Counties, 1860. . .,. . . . . 158 37 Living Arrangements of Black Children in Walton County, 1870 160 38 Living Arrangements of the Aged, by Sex and Race, 1870 and 1885 163 39 Structure of Black Families in Walton County, 1885. ... 166 40 Living Arrangements of Black Children in Walton County, 1885 171 41 Living Arrangements of Black Single Adults in Walton County, 1885 .... 173 42 Household Composition by Stage of Family Life Cycle 181 43 Percentages of Households Containing Relatives and Boarders, by Race, 1870 and 1885. . .... 188 XV

PAGE 17

Table 44 45 46 47 48 49 50 51 LIST OF TABLES (continued) Page Characteristics of White Households in 1885, by Village or Rural Residence ........ . 196 Structure of Primary Families, 1885, by Village or Rural Residence. .............. 199 Percentage of Households Containing Extended Relatives, by Social Class 204 Percentage of Households Containing Boarders, by Soci al CI ass . . . . , . . 205 Mean Household Size by Social Class, 1870 and 1885. 207 Mean Number of Children in Household, by Social Class, 1870 and 1885. ..'.... 208 Percentage of Household Heads Who Were Married, by Social Class, 1870 and 1885. ........ 211 Percentage of Households with Male Heads, by Social Class, 1870 and 1885 212 XVI

PAGE 18

LIST OF FIGURES Figure 1 2 3 4 5 6 Wholesale Price Index, Farm Products, 18701890 Age and Sex Distribution of Walton County White Sample, 1870 . Age and Sex Distribution of Walton County Black Population, 1870 . Age and Sex Distribution of Walton County White Sample, 1885 Age and Sex Distribution of Lake De Funiak White Population, 1885 Age and Sex Distribution of Walton County Black Population, 1885 . Page

PAGE 19

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy HOUSEHOLD AND FAMILY IN THE POSTBELLUM SOUTH: WALTON COUNTY, FLORIDA, 1870-1885 By . Barbara Finlay Agresti August, 1976 Chairman: Dr. Gerald R. Leslie Major Department: Sociology This study deals with the family, household, and population characteristics of a farming community in the South during the two decades f ol 1 owi ng' the Civil War. These decades were a time of rapid and disruptive change in social and economic conditions in the South. The study represents an addition to the growing literature on the history of the American family. For white families, the emphasis i s on description of household and family structure in a hitherto neglected area--the farming South. For blacks, the data are used to test some of the controversial ideas of E. Franklin Frazier and others concerning black family structure in the postbellum period. Data from census manuscripts for Walton County, Florida, for the years 1870 and 1885 are analyzed. Detailed descriptions of the black and white populations in those years XVI M

PAGE 20

are presented. Residential family structure is discussed in detail by stage of the family cycle, for black and white families separately. Other analyses include comparisons of household characteristics by occupational class, rural -vi 1 1 age residence, and over time. For whites, the data presented add to our knowledge of the 19th-century rural Southern family. The data show evidence of stem family organization among the rural people. For blacks, new evidence supporting the "matriarchal thesis" of Frazier and others is presented, at least for the 1870 population. Important gains had' been made, however, by 1885, when black and white families are seen to have been quite similar. Some dramatic changes occurred during the 15-year interval studied. For example, the proportion of extended households, especially of the married sibling type,i ncreased sharply. This and other changes are explained in terms of the social and economic changes occurring in the community and in the South during the period. General conclusions of the study point to the adaptability of the family institution and the need to consider the historical and socioeconomic context in family research. XIX

PAGE 21

The importance of the developmental approach is emphasized. In addition, the results show the importance of using families, as well as households, as basic units of analysis in historical studies. XX

PAGE 22

CHAPTER I INTRODUCTION This study deals with the family, household, and community in a late 19th-century Southern farming area. Its setting is Walton County, an agricultural county in the western Florida panhandle, in the years 1870 and 1885. This study makes a contribution to the history of a county, to the sociology of the family, and to the history of the American family as well. The initial date selected for this study--! 870--was chosen in order to analyze family data drawn from the first census after the Civil War. The white population of the South had suffered heavi ly from war casualties, economic losses, the loss of capital invested in slaves, and the demoralization of unconditional defeat. Only five years had passed since the end of the hostilities, and the region was midway in the often bitterly resented Reconstruction period. Even more important was this period for black Americans, seven-eighths of whom were just released from a system

PAGE 23

of human bondage. Family patterns instituted in slavery were undergoing adjustments to a new order of life, one doubtless filled with many hardships and many hopes. The second point in time to be analyzed--1885--not only added 15 years during which adjustment to the new order could be made, but followed by a decade or more the end of Reconstruction. While it is impossible to separate the adaptive processes from the social-psychological effects of the end of Reconstruction on the two races, it seems probable that, for most blacks, the high hopes of the earlier decade had been scaled down by the realities of the birth of Oim Crow. The present study hopes to explore some of the effects of these rapidly changing conditions on a littlestudied region--the farming South. Emerging Interest in the Social History of the Family The social history of the family has become an increaS' ing topic of interest over the past decade, both in Europe and North America. This growing interest was stimulated by early studies in France (Gautier and Henry, 1958; Aries, 1962). Soon afterwards, English scholars, led by a group at Cambridge University, began to publish work in the field

PAGE 24

(Wrigley, 1966, presents a long list of such studies). Finally, American and Canadian students of family history have begun contributing, especially within the past five or six years. As a result of this new interest, a new journal, the Journal of Family History , has just begun publication, several special issues of scholarly journals have been devoted to family or demographic history, and several articles have appeared calling for more and better research in the area. (See, for example, Saveth, 1969; Hareven, 1974; Berkner, 1973a.) There are at least two reasons for the new interest in the history of the family. First, historians, who have traditionally emphasized the role of outstanding people and events, have become aware of gaps in their knowledge and understanding of past events and social conditions which affected the bulk of the population. Their new interest in "the common man" has stimulated interest in basic institutions, such as religion and the family, that closely touch and shape the lives of the major portion of the population. (For examples of this new historical interest, see Thernstrom, ' Journal of Interdisciplinary History 2 , 1971; Journal of Marriage and the Family 35(3), 1973; Daedalus 97 (Spring) 1968; Journal of Social History 5(1), 1971.

PAGE 25

1964; Hareven, 1971; Powell, 1963; Demos, 1968. 1970; Lockridge, 1970). The second reason is the renewed interest of sociologists in social change, which has challenged stereotyped conceptions of what family life was like a half-century or more ago. Goode issued the challenge in 1963, with the following portrait of what he called "the classical family of Western nostalgia." It is a pretty picture of life down on grandma's farm. There are lots of happy children, and many kinfolk live together in a large rambling house. Everyone works hard . . . . The family has many functions; it is the source' of economic stability and religious, educational, and vocational training. Father is stern and reserved, and has the final decision in all important matters. . . . All boys and girls marry, and marry young .... After marriage, the couple lives harmoniously, either near the boy's parents or with them, for the couple is slated to inherit the farm. No one divorces. (Goode, 1970:6) Goode went on to emphasize that there is a paucity of historical data on the family and that more adequate data are needed. Theorizing about family changes over time is easier, of course, if we enjoy a firm body of data about the past, and in particular if we know how the family pattern operated at some specific point in time: drawing a curve between only two points in time is dangerous, but not so unwise as drawing a curve from the present to an unknown and possibly legendary past. (Goode, 1970:xi, xii)

PAGE 26

Heeding Goode's admonition, a variety of sociologists has begun to provide empirical studies of various aspects of family life in the past. ( Furstenberg , 1966, Farber, 1972, Demos, 1970, Lantz et al . , 1968, are a few. See also the collection by Gordon, 1973.) Obstacles to the Historical Study of the Family Sociologists and historians have been hampered in studying family history by the lack of adequate sources of data. Family life is generally so taken-for-granted that few people have bothered to write about it. The accounts that do exist, largely from moralists and philosophers, generally fail to distinguish between real and ideal behavior, and they focus on the patterns of elite groups without any indication of the relevance of those patterns to the bulk of the population. Lower-class people, often being illiterate, did not leave memoirs, diaries, or other written documents. Given these 1 imi tations , the scholars who are developing the new family history have been ingenious in discovering new sources of data, basically demographic in character. These include parish records of the important life events

PAGE 27

of baptisms, marriages, and deaths. Wills, court records, land deeds, property inventories and even physical artifacts have also been used. For studies of 19th-century America, the manuscript census has been a valuable source. Regardless of the specific sources used, the techniques of analysis applied have generally been quantitative and rigorous. General Purposes and Framework of This Study Several studies have been published recently of colonial family patterns in the Northeast (Demos, 1970; Greven, 1970; Lockridge, 1966; Smith, 1973a, 1973b; Norton, 1971; Wells, 1971, 1971-1972, 1972). In addition, there are a few studies of urban family patterns in the 19th century (Sennett, 1973; Pleck, 1973; Hershberg, 1971). \lery little has been done, however, on 19th-century family patterns in rural areas or in the South. (But see Modell, 1971, and Eblen, 1965, for limited demographic studies of frontier populations.) Our knowledge of white family patterns is scattered and incomplete, but our knowledge of black family patterns is even more so. For many years after i ts, publ icati on in

PAGE 28

1939, Frazier's analysis of the influence of slavery on black family patterns was accepted widely and almost uncritically. Recently, the "cliometric" work of Fogel and Engerman (1974) and the research of Genovese (1974) have called many of Frazier's conclusions into question. Although four studies have appeared concerning black families in cities of the 19th century (Pleck, 1973; Hershberg, 1971; Lammermeier, 1973; Harris, 1976), almost nothing has been published on rural black families of the South--especial ly in the nonplantat ion areas. The present study of both black and white family patterns in a Southern rural county in the decades after the Civil War should add significantly to our knowledge. Three general approaches to the data will be used, often simultaneously. In the first, descripti ve , approach, both white and black family patterns will be enumerated in detail. For white families, we will provide data on a little-studied population in a little-studied time period. For black families, systematic demographic data may afford some insight into current issues concerning the nature of black family history. In the second, analytic , approach, variations in family and household composition according to certain social-structural

PAGE 29

characteristics will be analyzed. The focus will be on the relationship between family structure and other aspects of social structure. Then, in a dynamic approach, we will compare family patterns at two points in time, 15 years apart. The emphasis will be upon changes in family patterns and household structure, and the relationship between these and other changes in the community. The three approaches combined should enable us to understand the family in the community as both changed over one brief period of time.

PAGE 30

CHAPTER II THEORETICAL PERSPECTIVES AND REVIEW OF LITERATURE Meaningful research proceeds from explicit theoretical bases and utilizes carefully defined concepts. The first task of this chapter is to specify major concepts relating to the nature of the family and to household structure. Secondly, the basic issues involved in applying these concepts to white and black families of the 19th century will be drawn from a review of published research in the area. Family and Household Structure Kinship refers to a system of norms governing relationships based upon consanguineal and/or affinal ties. In turn, family structure is a system of norms associated with kinship statuses. These norms specify which kinship statuses are recognized and what obligations accrue to particular statuses vis-^-vis others. The basic characteristics of the American kinship system were described by Parsons (1943). He noted that our

PAGE 31

10 kinship system is similar to those of most modern societies and has not changed significantly over the past couple of centuries. Family structure, on the other hand, has changed considerably. This change has been associated with the progressive loss of family functions (Ogburn, 1954). The major components of family structure include rules of residence, rules of authority, inheritance rules, and role definitions associated with particular statuses. The thrust of the present research involves the residential aspects of family structure, but other aspects may often be inferred from knowledge of residence patterns. In addition, we are interested in household composition and the interrelationships among all members of households, whether kin or not. Types of Family Groups There are four basic types of resident family groups of great importance in the present study. These are the nuclear family, the single-parent family, the stem family, and the sibling family. The nuclear family consists of a married couple and their offspring living in a single household. A married couple without children is designated a conjugal couple, but

PAGE 32

n for many purposes conjugal couples and complete nuclear families will be considered together. The nuclear family has usually been found to be the most common form of resident family, even where other family forms are common (Berkner, 1973b; Laslett, 1973). The single-parent family consists of one parent, either mother or father, and at least one natural child. The'thild" may be of any age, as long as he or she is unmarried and living with one parent. The stem family was first described by Le Play as the famille souche (Sorokin, 1928:86). In this study, stem families are those in which two ever-married, directly lineally related individuals are residing together, with or without spouses and/or children. The most common example would be a young couple living with the husband's aged parents. Although the stem family usually refers to a structural process through which families go, it is used here to define a type of residence group which is sometimes associated with that process. The sibling family is a group of two or more siblings, at least one of whom is over 18 years of age, who reside together with no parent or own child present. This kind of family group might be common if a large proportion of the

PAGE 33

12 population never married or if elderly widows lived with siblings rather than with their children. All of these basic residence groups may be further extended by the presence of other relatives in the household. In addition, they may be augmented by the presence of one or more unrelated residents, normally either rentpaying boarders or servants. Other Aspects of Household and Family Structure In addition to the type of family group within the household, several other aspects' of f ami ly or household structure are of interest. These include the size of households (the total number of individuals who reside together as one domestic and economic unit) and the size of families (both the number of children born per marriage and the number of kin present in the household). The age at which marriage typically takes place and the proportion of the population that marries are also important data for the study of family structure. In addition, the typical life cycle of the family, from its inception at marriage to the death of the last spouse, is interesting and important to this study. Finally, the functions of families and of households will be

PAGE 34

13 determined, insofar as possible, by looking at patterns of residence and their relation to various aspects of social structure. The Developmental Approach The importance of a developmental approach to the study of families has been well documented (Lansing and Kish, 1957). The concept of a series of stages through which families pass in their normal development was introduced by early rural sociologists (Sorokin et al . , 1 931 ) and 1 ater popularized by Glick (1947, 1955), who noted important changes in the typical life cycle of families over the past century. The concept is important in any study of household composition, because the composition of the family at any one time depends upon its stage of development. The stem family, for example, can be seen as one stage in a much longer cycle, during much of which the family group is nuclear (Berkner, 1973b). The need to consider this in historical studies is echoed by Hareven (1974). A developmental approach will be used for many parts of the analysis in this study.

PAGE 35

14 Major Issues Concerning the P reindustri a1 American White Family A classic history of the American family was published in 1917-1919 in three volumes by Arthur W. Calhoun. Other traditional sources for American family history are those by Goodsell (1939), Howard (1904), Sirjamaki (1953), and Bardis (1964), the latter two being summaries of the traditional literature. All of these works suffer from reliance upon such imprecise data sources as legal documents, travelers' accounts, and popular literature. During recent years, the work of the new social historians, historical demographers, and sociologists of family history has begun to question some of the findings and interpretations of these traditional sources. In this section, some of the major points of difference between the traditional and more recent interpretations will be discussed. Fami ly Structure The most important questions about the American family of the. past center around its structure. Traditionally, historians and sociologists have believed that prei ndustri al families had strong extended kinship ties, in contrast to the more isolated nuclear family of contemporary society.

PAGE 36

15 The view that many households in past centuries contained a wide variety of kin has been especially common (Bailyn, 1960:17-18; Laslett, 1973). Recent research on household structure for the colonial period has shown that, for various communities in various times, most households were nuclear. Marriage agreements almost always provided for the establishment of separate households for the new couples. The parental family house, in both Andover, Massachusetts, and in Plymouth Colony, was always left to only one son (Demos, 1970:63; Greven, 1970). Pryor's study of Rhode Island households in 1375 (1972) also found most households to be nuclear in family structure. There is a problem, however, in identifying certain forms of family structure by cross-sectional research such as the studies just cited. One form of family organization for which this is true is the stem family found in farming areas in the United States and Europe (Leslie, 1976:223). In families of this type, one son stays in the parental household after marriage, while the other siblings move out at marriage to form new households. The son who stays in the parental home eventually inherits the parental land and home, which remain intact. The other sons often receive

PAGE 37

16 parental support in setting up their own households, but the original parental land is not divided among them. As Berkner (1973b) has pointed out, even when such an extended family system is typical, most households at any one time are nuclear, because nonheirs leave the parental household to form nuclear households. The parental household would be considered "extended" only for the period of time between the marriage of the heir and the death of the parents. This period may be relatively short when compared to other stages of the family cycle, especially if marriage ages are high and life expectancy is low. Berkner points out the need to look at the developmental stage of the parental household, as well as whether or not the household is extended, because the extended family is "merely a phase through which most families go" (1973b:41). When studies of rural American families are examined closely, a clear pattern of. stem organization seems to emerge, although it has not been identified as such by other authors. Demos notes, for example, that one extended household type in Plymouth was the residence in some households of aged grandparents. When parents were aged, or when one died, it was common for the son who was to inherit the homestead to reside with the aged couple or widowed parent

PAGE 38

17 and to provide for them until their deaths (1970:75-76). The same situation was described by Greven for Andover families. Often, though, sons who received the parental homestead as their portion of their father's estates also received part of their parents' houses and lived in them while their parents were still alive, thus forming households which effectively consisted of three generations under a single roof. A widow nearly always was specifically bequeathed a room or rooms in her husband's house and provisions to be given to her annually by one or more of her sons .... Generally, of course, married sons lived separately in houses of their own, but aged parents usually expected to share a house with one of their children. Extended households, for the most part, were the product of old age and the necessities for care and attention which elderly men and women obviously needed. (Greven, 1970:137-138) This pattern in Andover became increasingly common as the original land tracts became subdivided to the point that impartible inheritances to only one son became the rule. When land was plentiful, it had been common to settle all sons on their own property; but later generations found that their landholdings would not permit further partitioning. Thus, the number of three-generation households increased at the same time that landless sons began migrating to other areas. Family extensions were, in this instance, related to changes in land-inheritance patterns resulting from population growth and from full settlement of the land.

PAGE 39

18 One should keep in mind that new settlements would not usually have three-generational households, because most migrants were young. It would normally be 40 or 50 years before the original settlers of a community were old enough to have grandchildren in their households, and longer before land settlement and subdivision increased this pattern. It would be expected, then, that extended households would increase over time in a relatively new settlement, if indeed the stem pattern were normative. This seems to have been the case in Andover (see Greven, 1970:98, 137-138, 220, and 257) and also in rural Michigan in the 19th century (Bieder, 1973). It might also explain the relative lack of extended families in Bristol, Rhode Island, in 1689, since this was a relatively new settlement at the time (Demos, 1970:79). It seems plausible that many early American families were stem families, especially after the full settlement of an area. As families moved to the cities, the lack of landed property to keep within the family probably led to the breakdown of this system; the urban family's resources were easily divisible among several siblings, whereas agricultural land is only divisible a limited number of times. This interpretation is consistent with the English pattern. Extended families in rural areas involved aging

PAGE 40

19 parents being supported by sons who were to inherit the property, while those in urban areas more often involved young couples being temporarily supported by their parents (Anderson, 1971 ). It should be kept in mind that even when stem family organization exists, most sons usually set up nuclear households, and the parental household is nuclear for most of its existence, so that the proportion of three-generational households in the population at any time is quite small. The finding of Pryor (1972) that over 10 percent of the Rhode Island rural households in' 1875 were mul tigenerational may be quite significant in view of this fact. In summary, it is not good practice to attempt to describe family structure merely from cross-sectional data. Some attention must be given to the family life cycle. Only thus can we determine important aspects of family organization, such as the means for assuring continuity of family relations and property across generations. For white families in this study, one of the major questions concerning family structure is just this: Is there evidence of the stem family among the agricultural population, or are nuclear families in separate households the norm at every stage of the life cycle?

PAGE 41

20 Nonrelatives in the Household Early descriptions of 18thand 19th-century American families did not emphasize the presence of nonrelatives within family households. Several recent empirical studies, however, have reported that many households included nonrelatives (Pryor, 1972; Anderson, 1971; Modell and Hareven, 1973; Demos, 1970; Lockridge, 1966). Some of these unrelated persons were household servants, many of whom were children (Demos, 1970:71; Lockridge, 1966:343n). Still others were single adults who had no other choice but to reside with a family as boarders' (Demos, 1970). By 1875, the proportion of Rhode Island households containing nonrelatives of the head was still fairly high— 24 percent of all households (Pryor, 1972:588). Family and Household Size Traditional estimates of early American family size have been high, owing to the unreliability and exaggeration of such sources as travelers' accounts and literary works. Calhoun wrote that families of 10 to 12 children were the norm, and that those with over 20 children were not rare in colonial days (1917-19:87). These estimates have been altered in view of the findings of recent studies. Most of these studies have found average completed family sizes of

PAGE 42

21 around seven or eight children in the colonial period (Demos, 1970:192; Smith, 1973a; Greven, 1970; Wells, 1971:75). This figure decreased somewhat for 19th-century populations (Wells, 1971:75; Farber, 1972). Still, American family sizes have been found to have been consistently larger than European families of similar periods. Although the number of children born per family was relatively large, the average number of persons per household was smaller. Some households contained only young married couples, others contained families in the childbearing stage, while, in still other cases, some or all of the children had grown and left the parental home. High infant mortality rates were another factor in keeping household sizes low. In colonial days, household size probably averaged four to six persons. In 1790, the average household size forthe United States was 5.7 persons; less than twice the size of U.S. households in 1950, 3.4 persons (Grabill et al . ,1973: 379). Age at Marriage and Life Cycle of Fami 1 ies Most writers about the early American family concluded that marriages took place at \/ery early ages (Calhoun,

PAGE 43

22 1917:67; Sirjamaki, 1953:40). Often, too, women were believed to die in childbirth, leaving their widowed husbands to marry for a second or a third time. Studies that have utilized quantitative data from family reconsti tuti on or census analysis have called both of these assumptions into question. The average age at marriage in the 17th and 18th centuries in America has consistently been found to have been fairly high, usually around 27 years for men and 22 or 23 years for women (Wells, 1972: 426; Henripin, 1964; Demos, 1970:193; Smith, 1973a:406; Greven, 1970; Farber, 1972:42). Moreover, a high proportion of men who married were married only once (Greven, 1970:29; Wells, 1972:423-424). Even in colonial days, marriages broken by early death were the exception and not the rule. Robert Wells brought together data from 18thand 19thcentury Quaker families with data for recent decades published by Glick and Parke (1965) in an article describing changes in the typical family life cycle over the past two centuries in the United States (Wells, 1971-1972). The major changes have been (1) a progressive decrease in the age of mothers at the birth of their last child and (2) an increase in life expectancy for both spouses. At the same time, the median age of women at first marriage has not

PAGE 44

23 really changed substantially. This means that the childbearing period was relatively much longer for women of earlier centuries, and that most marriages of the earliest group studied (early 19th century) ended by the death of a spouse before the last of the children left home. It was Click who first pointed out that a "new phase" in the family life cycle had appeared in the 20th century, the "empty nest" stage (1947). Throughout the 19th century, most families never experienced this stage, but went directly from childrearing into widowhood. We have now briefly discussed the experiences of white families in both the accounts of early writers and those of the new generation of empirically oriented historians. Most of the studies reported have dealt with colonial populations. It will be useful to compare their findings with ours to see if they are supported for a Southern rural community in the late 19th century. In the chapters on data analysis, an attempt will be made to discuss each of the topics introduced here.

PAGE 45

24. Major Issues Concerning the 19th-century Black Family There has been a revival of interest in recent years in the history of the black family in the United States. This is due not only to the general interest of the past decade in family history, but also to the renewed interest in black history and the slavery experience brought about by the increased awareness of minority groups in the 1960s. For decades, most sociologists had accepted the interpretations of the black family of E. Franklin Frazier (1948), but recent publications (Pleck, 1973; Genovese, 1974; Fogel and Engerman, 1974; Gutman, 1973; Harris, 1976) have questioned some of the conclusions made by Frazier and other early writers. W.E.B. Du Bois Although E. Franklin Frazier is usually remembered as the pioneering student of black family structure and patterns in the 19th century, most of his major conclusions were anticipated by about 30 years by William E. Burghardt Du Bois (1899; 1908). Du Bois, noting the poor condition of black families in the early 20th century, believed that, at least, they were much improved over the conditions of

PAGE 46

25 the slavery period. The slave family, according to Du Bois, had been practically destroyed by the exigencies of the cotton producing system (1908:22, 31). Du Bois was one of the first to emphasize the division between the house servant and the field slave. The house slaves lived close to the families of their masters, learning from them and often adopting their mores. Among this group, the monogamous, patriarchal family system developed, as the ideal family type (1908:47). Another group of field slaves, larger in number, lived at a distance, both spatially and socially, from white families. This group became highly demoralized, especially where masters were of the "absentee" type. They lived in "quarters," not family cabins; sexual exploitation was common and there was practically "no family life" and no moral instruction ( 1 908 : passim) . Such family life as there was among these field slaves was characterized both by the "absence of fathers" and by lack of ability of fathers to govern or protect their families. A man could be sold or separated from his family at any time and his wife could be made the master's or overseer's concubine. Thus, the father's place in the authority system was diminished (1908:49).

PAGE 47

26 Du Bois spoke of the "absence of mothers" as characteristic of slave families. Even though mothers were a more stable element in the child's world than were fathers, they were unable to spend much time in child care. Young children often were left alone or in the care of other children, while their mothers had to go to the fields (1908:49). This situation of demoralization and weak family ties allegedly carried over into the postslavery era. The two groups which had formed during si avery--house servants and field slaves--di verged. The former house slaves continued to be patriarchal and monogamous, similar to white families. The others continued tohave unstabl e mari tal and family relationships, although these gradually became more stable over time (Du Bois, 1908:31). E. Franklin Frazier Thirty years after Du Bois published these ideas, E. Franklin Frazier studied the same problems and came to much the same conclusions (1930, 1948). Frazier emphasized that the stability of slave families depended largely on the characteristics of the particular master. Masters could force matings, sexually exploit slave women, and separate families by sales; or they could enforce stable, monogamous families. Even where the masters were kind and encouraged

PAGE 48

27 stability, the slaves' dependence upon the master's will prohibited the development of a strong, independent black family (Frazier, 1930). Frazier, like Du Bois, emphasized the importance of the mother in the family. The father was not dominant because his will was not enforceable over that of the master. The mother, on the other hand, did play a fairly stable role in the raising of her children. It was to her that the children turned for strength and support (1930:236). Like Du Bois, Frazier also noted the differences between house servants and field slaves. It was among the field slaves, especially, that the strong-mother, weakfather family became widespread. The house servants' families were patriarchal and stable (1930:259). After slavery ended, much of the black population wandered about for awhile and then settled down into semistable family groups. However, there still were high illegitimacy rates, a high proportion of female-headed families, and high migration rates. The "matriarchal" family, in which the female was the more stable and dominant parent, grew as a. significant black family type (1948:102-113). Another group of blacks, mostly composed of former free blacks, house slaves, and mulattoes, settled down into

PAGE 49

28 stable, monogamous, patriarchal families. Often these "Black Puritans," as Frazier called them, owned their own homes and small parcels of land. Some became educated professionals. This group assumed leadership in the black communities in the South (1948:190-208). The worst situation was among those who moved to the city slums and the migrant lumber camps. Among these groups, "free sex behavior and spontaneous matings" of short duration, violence, poverty, ignorance, and "the absence of family tradition and community controls" were the norm (1948:257). Recent Data on Black Family Structure Several books and articles have been published recently which question in some way the conclusions of Du Bois and Frazier concerning the 19th-century black family. Three of the most important are discussed here. The year 1974 saw the publication of two important works on slavery, both of which contained many observations concerning the family life of slaves. The first of these, and the most controversial, was Time on the Cross by Robert Fogel and Stanley Engerman (1974). The authors analyzed data from a large number of plantations and concluded that

PAGE 50

29 Frazier was wrong in his insistence that slave families were unstable and mother-dominant. These authors emphasized the strengths of the black family, arguing that owners had much to gain from seeing that black families stayed together and were satisfied. They believe that monogamy was the rule among slaves and that most lived in single-family houses rather than in quarters. Fogel and Engerman also deny the degree of sexual exploitation described by Frazier and Du Bois as well as Frazier's "matriarchal thesis." "For better or worse," they write, "the dominant role in slave society was played by men, not women" (1974:142). Fogel and Engerman' s book has been attacked from many quarters and it seems clear that the book is marred by methodological and theoretical flaws. The study did have access to very good data sources, however, and its findings cannot be completely brushed aside. Perhaps most important of all. Time on the Cross has done more than any recent book to reopen debate on some questions concerning slavery, including some pertaining to slave families, which had probably been closed for too long. The second work on slavery published in 1974, which has been acclaimed by historians, is Eugene Genovese's Roll, Jordan, Roll: The World the Slaves Made. Even though

PAGE 51

30 a few criticisms of Genovese's work have appeared, the general consensus of almost all who have read the book is that it is a "masterpiece" (El ki ns , 1 975 :48) . In various parts of the book, Genovese describes aspects of slave life which deal with family matters. Genovese gives much evidence to refute the "myth of the absent family" among slaves, as presented by Du Bois and Frazier. He points out that, immediately after emancipation, there was a rush by many slave couples to have their unions legalized; and many runaway slave problems had family motives. In spite of tremendouspressures of a difficult life situation, the slaves managed to create a family life of their own. Even before emancipation, many masters took pride in the maintenance of intact families among their slaves, because this gave them better social control, encouraged responsibility among slave men, and furthered satisfaction with the slave status. In addition to stable marriages and nuclear groups, Genovese argues that slaves reached out to extended kin to provide protection and support in times of trouble or need (1974:450-458). Genovese did recognize some problems and some unique norms of slave families. The sexual code of slaves, for example, was not as strict as was that of whites, including

PAGE 52

31 more acceptance of divorce and remarriage. Some of the differences in these norms gave rise to accusations of "immorality" from the white quarter (1974:461-475). The place of the father, according to Genovese, was more important than Frazier and Du Bois recognized. He argued that the norm was for fathers to have authority over their wives and children and that this dominance was accepted by both male and female slaves. Genovese admits that enough cases of indifferent fathers and strong mothers occurred to give rise to the "myth" of the matriarchy, but he argues that the two-parent, father-dominant family was the ideal (1974:490-494). Another historian of the black family, Herbert Gutman, has argued even more strongly for the basic stability of the black family under slavery and afterwards (1973). For Gutman, it was the urban experience of discrimination and poverty which caused the decline of the stability of the black family and which led to the high incidence of femaleheaded families. Stanley Elkins, in reviewing a forthcoming book by Gutman on black family history (1975), writes that Gutman's major conclusions point to family continuity over time, the importance of the black father in the family, and a "high degree of stability."

PAGE 53

32 It is to these areas of disagreement that the present analysis of black family structure is directed. More intensive and careful research is needed to deal with questions of black family types, the stability of black marriages, the stability of black family ties over time, and the place of the female within the family. Until more data are available, the issues will not be resolved. Other Considerations of This Study In addition to focusing upon specific issues relating to 19th-century white and black family structure, a few comparisons will be made in an attempt to show the relation of certain aspects of rural social structure, social class, and village-rural residence, to specific aspects of family and household structure--househol d size, presence of boarders and extended relatives, and marital status and sex of the household head. We will explain changes over the 15-year time period of this study in terms of social and economic changes in the community.

PAGE 54

CHAPTER III RESEARCH DESIGN AND METHODOLOGY The first task after having defined the problems to be studied was to choose an appropriate geographical area from which to, gather data for the study. An area was needed which was primarily agricultural and fairly typical of the "commercial farming" South. It .was also necessary to find an area for which data were available and in good condition and whose georgraphic boundaries did not change during the period of the study. It would be helpful, also, if there were nonquantitati ve sources, such as histories of the area, which could be consulted for background material. The main purpose was to find a clearly defined area, such as a county or counties, whose population could be studied in depth by use of data from census manuscripts and other sources which might be avai labl e. The search for an appropriate area was limited to the state of Florida, because the data were readily available and there was a special census for that state in 1885. 33

PAGE 55

34 After consideration of several areas, the county of Walton in the Florida panhandle was chosen. Walton County was a farming area fairly typical of much of the South, and it met fairly well the needs of the study. Sources The primary sources of information on households and individuals were the census manuscripts for the years 1870 and 1885. These were available on microfilm in the P. K. Yonge Library of Florida History in the library of the University of Florida in Gainesville. The films are copies of the handwritten enumerations that were prepared by officials in each county and sent to the census bureau for compilation, The originals are now in the National Archives, from which the microfilms are avai lable. The major source of data was the population schedules, which contain information on all individuals in every household enumerated. Also of interest for information about property and agriculture were the agricultural census schedules, taken in conjunction with the population censuses, There was some problem in the fact that the schedules for 1870 and 1885 differed somewhat. The 1870 schedules did

PAGE 56

35 not contain information on the relationship of the individual to the head of the household, as did those of 1885. In most cases, however, the combined knowledge of the ages, surnames, and birthplaces of the individuals in the household were enough to establish the relationship with a great deal of certainty. When no relationship could be established with reasonable certainty, the individual was classified as a nonrelative. There were not many cases in which this was a problem. Other sources of information used include the census slave schedules of 1860, numerous census publications, cemetery records, and John L. McKinnon's History of Walton County (1911). A comparison of the validity of the data from census manuscripts and those from published census materials showed that the published materials were prone to error in reporting what was actually in the manuscripts. The data presented in this study are probably much more reliable than are the published figures. The Samples The populations to be sampled consisted of long listings of individual names, by households. Since the households

PAGE 57

36 were not numbered in a single sequence and since the records were on microfilm, the manipulation of the data for sampling purposes was cumbersome. A simple random sample, under these conditions, would have been difficult and perhaps not as good as the sampling method chosen. A systematic sample was taken of the white households, thus assuring that all sections of the county were covered. The sample may be considered equivalent to a simple random sample for statistical purposes' (Mendenhall et al . , 1971:151-152). The sampling proportions were one-in-four for the white households of 1870 and one-in-five for the white households of 1885. Since the black population of Walton County was small and since it was wery important to the study, the total population was included in the study. Data were gathered for every black household and for eyery black individual living in a white household. In addition to the stratification by race, the 1885 white sample was stratified by residence. In that year, the households were reported by locally defined place names or town or village. One village was selected for more in-depth study, and, data were gathered for every white household within the village. In some cases, the difference in sampling proportions had to be taken into account during the analysis

PAGE 58

37 by weighting the village cases differently than the other cases. Operational Definitions Before explaining how the data were coded, it is necessary to define some basic terms as they are used in this study. Fami ly --two or more people residing together who are all relatives of primary degree or who are directly related lineally. Primary family -the family of the household head; if two or more families are present, the one whpse status and property appears to be dominant. Subfami 1y --a family group residing with another family but not lineally related thereto and not including the household head, but which is part of the kinship group of the household head. Secondary family --a family group which is resident with another family to which it is not related. Household --the total group of individuals who reside together as one unit, usually in one physical house; includes boarders, servants, and extended kin.

PAGE 59

38 Head of househol d --the person who is primarily responsible for the economic welfare of the household. Simple nuclear household --a household containing only one nuclear family and no other household members. Extended household --a household containing a primary family plus other non-nuclear kin. Augmented household --a household containing the primary family and nonrelated boarders. Note that if two adult siblings and their nuclear families were living together, there would be two family groups, primary and subfamily. If the same siblings were living with a parent, however, the lineal tie of both to the parent would connect them into one primary family. The strict definition of these concepts was necessary in order to classify household family structural types. Data Organization Data were gathered for each household in the samples taken. For each individual, two records were prepared, a household record containing information about the household and a "person" record. The household record contained some 17 bits of information, as follows:

PAGE 60

39 1. Type of dwelling, whether single family or group quarters 2. Single or plural family group(s) 3. Number of head's kin 4. Number of boarders 5. Number of servants 6. Structure of the primary family 7. Number of head's own children 8. Birthplace of head 9. Time since migration to Florida 10. Marital status of head 11. Race of head 12. Sex of head 13. Age of head 14. Age of oldest child of head (in household) 15. Age of youngest child of head (in household) 16. Kinship (relation to head) of non-nuclear members 17. The number of stepchildren of head (if any). A complete household schedule, including the explanation of codes used for all of the items, is found in Appendix A. A few items require more thorough explanation, however. Family structure refers to the pattern of kinship relationships within the primary family of the household. The

PAGE 61

40 typology used draws heavily from that developed by Anderson (1972) with some minor modifications. The categories used in this typology were as follows: primary individual, married couple only, complete nuclear family, stem family, sibling family, and other families. Most of the categories are sel f-explanatory, having been discussed in Chapter II, pages 10-12. Migration time was estimated from looking at the birthplaces of the parents and children. In many cases there was not enough information to estimate this, but if some of the children were born in another state and others in Florida, a rough estimate of the date of migration could be made by using the ages of the children. For those born in Florida or whose older children, age 15 or above, were born in Florida, it was certain that either these were not migrants or that they were not recent migrants. Admittedly, this is a gross estimate of recency of migration, but in the absence of better data it is believed to be useful. One of the first problems in coding the data was to determine the head of the household. While in many cases it was fairly obvious, it became more difficult in more complex households. The miscl assif ication of headship has certain consequences for the analyses, especially if the

PAGE 62

41 misclassif ications are not simply random. For example, in a three-generational family, is the male of the oldest generation the head, or is the son in the middle generation? If an older widower in such a household is classified as the head, should an older widow in a similar situation also be classified as the head? In this study it is assumed that ' the oldest person--male or female--who is given an occupation which seems to be dominant in the household is the head, _i Thus, in a few cases the headship was assigned to a female who was living with her son and his wife, but only when she was classified by the census taker as, for example, a "farmer" or "planter," and her son was a "laborer." Where the woman was given no occupation, she was not considered the head of the household if there was an adult male within the family. Obviously this problem is important when we look at the proportion of male versus female heads of households. The assignment of headship is in some ways arbitrary; there is no one "correct" way to do it. But one must be careful to avoid coding so as to introduce unknown bias or to prove one's own pet hypotheses. The main rule to follow is to be explicit about how the statuses are assigned, so that others can judge the work as well as understand it, and to be consistent in the method of assignment. The following rules of thumb were developed for use in this study.

PAGE 63

42 1. In a nuclear family, the husband is the household head, 2. In a one-parent family, the parent of either sex is the head, if there are no extensions. If a mother and children are living with a male relative whose occupation seems to be the dominant one in the household, he is the head. 3. Where two or more adult males and their families are residing together, the one whose name appears first in the manuscript is the head, unless the other's occupation or kinship status clearly appears to make him dominant. 4. If there are two adult generations present, the parent is the head, unless he/she is not given an occupation by the census enumerator. In general, the judgment of the census taker was relied upon, since he usually knew most of the people in the district. Finally, the coding of headship was a decision of the author, and all of the relevant factors (age, relationship, occupation) were carefully weighed in making the assignment. Another problem arose at times when the manuscript itself seemed wrong or confusing. In every case where a household record in the manuscripts seemed of questionable validity or where the errors were not obvious (as in John Smith's sex

PAGE 64

43 being coded as female), the household was not included in the sample. There were only a few of these househol ds--four or five in all. Most were cases in which the ages of the "parents" were impossibly close to the ages of their alleged "children," or similar types of erroneous records. The second type of record was the person record which contained information on each individual in each household sampled. The information recorded included the name, race, sex, age, marital status, occupation, birthplace, and father's and mother's birthplaces. Also included were codes for the individual's relationship to the' head of the household and his/her occupational class. For 1870, the amount of real and personal property belonging to the household head was recorded (this item was not included in the 1885 census). The complete schedule of codes for the person record may be found in Appendix B. Marital status was coded on the basis of the d£ facto residential arrangement of the persons in the sample, not the legal status. Any common-law marriage, under this system, was treated as a marriage. Likewise, any woman with a child, but no husband, was considered to be "widowed," even though there may have been a few never-married mothers in this category. There were wery few cases in which the

PAGE 65

44 marital status was given by the census taker as "divorced" or "separated." All of these cases were treated as "widowed" or "formerly married." It was impossible, from the data available, to make more precise classifications than these. Still, it seems legitimate to look at simply the de facto marital and residential arrangement as an indicator of family and marital stability. Analysis of Data After the data were all coded and punched on machinereadable cards, they were transferred to magnetic tape for processing. Most of the analyses done were simple, because the primary goal of this research is description. The specific methods of analysis will be presented as we come to them in the analysis chapters to follow. All computer work was done with the Statistical Package for the Social Sciences (Nie et al . . 1975) and the Statistical Analysis System (Barr and Goodnight, 1972).

PAGE 66

CHAPTER IV HISTORICAL PERSPECTIVES This study focuses upon a Southern American county, Walton County in Florida's panhandle, during the decades immediately following the Civil War. There have been very few studies that deal with the Southern population during the 19th century, even though its character has been recognized by historians as unique (see Grantham, 1967). The South is a very good setting for studies of black 19thcentury families, most of whom were in that region at that time, and also for the study of rural families in general. The industrial process had had very little impact upon the region during the years involved. Finally, the rapid changes occurring in Southern society during the 1870s and 1880s provide an opportunity to observe families in the midst of great changes and stress. 45

PAGE 67

46 Southern Economic and Social Conditions in the Late 19th Century In the decades following the Civil War, the South experienced rapid, disruptive changes in labor, land tenure, and economy. At the same time, the entire nation was experiencing a fairly serious economic decline which lasted throughout the 1870s and 1880s. With almost yearly declines in farm prices (Figure 1) came increasing hardship for Southern farmers. Previous studies have shown that the plantation system of the antebellum period survived and actually expanded during these decades with the development of "sharecroppi ng" (Woodward, 1951; Smith, 1953:318). Most of the literature has concentrated on the problems of the plantation South, to the neglect of the approximately half of its land area which was organized into middle-class, family-sized farms (see Gray, 1933, and Smith, 1953). One notable exception is found in the work of Owsley (1949), Plain Folk of the Old South . This neglected side of Southern culture is the setting for the present study.

PAGE 68

47 -D68L 0Z8L
PAGE 69

48: The Southern Agricultural Population in the 19th Century The 19th-century South was almost entirely agricultural and rural, but there was diversity in the kinds of agriculture practiced (Owsley, 1949:7). The majority of farms were small, family farms, not plantations (Gray, 1933:481483). In addition to planters. Gray describes two other types of Southern farmers, both of which are important in this study. These two groups he cal 1 ed the "poor whites" and the "commercial farmers." Poor whites resembled pioneer farmers, living in crude one-room log cabins with few furnishings. They cultivated small patches of corn or rice, sweet potatoes, cowpeas, and other garden produce. On their subsistence farms, the women and children did most of the farm work, while the men either hunted or were idle. Sometimes these farmers owned a few hogs or other animals; they almost always had a dog and a rifle. These people were despised by many more fortunate Southerners as being of an inferior class, the "clay eaters." The second class mentioned by Gray was the commercial farmers, the "yeomen of the South." These were the owners of family-sized farms, some of whom owned a few slaves.

PAGE 70

49 These people engaged in diversified farming, in contrast to the planters, who dealt in staple crop production. Their social standing was quite varied, ranging from very poor to fairly well-off, slave-owning farmers. Their houses were usually comfortable; they filled their barns with hay and forage; and they usually had gardens and orchards. If they had no slaves, the women worked at making clothing and preserving food (Gray, 1933:489). The relations of these farmers with their slaves were frequently very friendly and "almost intimate." Gray characterizes this class as having "sturdy independence, self-respect," sociability, hospitality, and a democratic spirit (1933:490). Gray also mentions a class of "free white laborers" whose living conditions varied with their labor contracts and the type of agricultural work which they did. After the war, the black population was often in a similar position to these workers, and often in competition for the same work. Most of the farmers in the present study were of the commercial farming type or farm laborers. These are the people whom Frank Owsley described as the "plain folk of the Old South."

PAGE 71

50 The plain folk . . . were usually landowning farmers and herdsmen, though a small minority were engaged in other occupations. Their thoughts, traditions, and legends were rural, for with the exception of an occasional ancestor who had been brought from some British city or debtor's prison as an indentured servant, their families were rooted in the soil. . . . To them the land was, with God's blessings, the direct source of all the necessities of life. . . . (Owsley, 1949:vii) They were often from Scotch or Scotch-Irish origins. Some became large landowners or professionals, but the majority "remained landowning farmers who belonged neither to the plantation economy nor to the destitute . . . poor white class. They, and not the poor-whites, comprised the bulk of the Southern population from the Revolution to the Civil War" (Owsley, 1949:viii). According to Owsley, these people were a true "folk," with a sense of solidarity based on their common origins in the British Isles. They were characterized by "closely knit families," strong religious feeling, and self-sufficiency. These traits, according to Owsley, allowed them to survive the period of Reconstruction without the disruption that was felt in other parts of the South. One more element of the Southern agricultural population remains to be discussed. This is, of course, the black population, most of whom were slave laborers before

PAGE 72

51 emancipation. Their condition depended largely upon the kind of agricultural system in which they were working and on the humanity of their owners or overseers. According to Smith (1953), many of the problems of slavery were, in fact, the problems of plantation agriculture, which did not disappear after slavery was ended. In the farming districts, however, the slaves were in more direct and open contact with their owners. Typically, the farmer and his slave worked in the fields together. With such personal, face-to-face relationships, some of the more dehumanizing aspects of plantation slavery were avoided. After the end of slavery, apparently, most black farm workers in the South were either "farmers," who owned or rented small plots of land and were similar to white small farmers, or farm laborers, either working for a share of the crop or for a cash share. Walton County, Florida, in the 19th Century Walton County lies in the western panhandle of Florida, bordering Alabama on the north and the Gulf of Mexico to the south. It was first settled by white men in the early Most of the information for this section comes from J. L. McKi nnon , History of Walton County (1911 ) .

PAGE 73

52 1820s, when a party of Scottish settlers from North Carolina came south in search of fertile lands on which to settle. They found what later became Walton County, an area of forest and swamp, full of rivers, springs, and lakes. They were able to make friendly territorial arrangements with the Euchee Indians who were already there, and the settlement known as Euchee Anna was formed. These early settlers were followed in the next three decades by others, many of them friends or relatives of earlier migrants to Florida. After about a decade of white settlement, the Indians left to go farther south, complaining' that the newer white settlers were destroying their forests and wildlife. In 1824, Walton County was officially formed and its area encompassed what is now Walton, Holmes, Okaloosa, and Santa Rosa Counties. Later, Santa Rosa and Holmes Counties were formed; and in 1915, Okaloosa County was formed. At the time with which the present study deals, Walton County still included the Okaloosa County area. Almost all of the settled area of the county, however, was in the southeast corner of the county, the rest of the county still a wilderness. This part of the county is within the present boundaries of, Walton County.

PAGE 74

53 In the early years, Walton County was quite isolated, travel being very difficult. Between 1837 and 1860, roads were built from Tallahassee to Pensacola through the county and from the bayfront village of Freeport to Euchee Anna and on northward into Alabama. As a result, travel and trade increased sharply, a trip to Pensacola taking only one week instead of three or four. Most of the trade was with merchants in Pensacola, and most goods went by way of sailing vessels out of Freeport. Cargo was mainly the prodducts of the small farms of the county: molasses, corn, cotton, chickens and eggs, venison, lamb, pork, tallow, raw hides, and other farm produce. By 1860, according to the census publications for that year, there were 2,584 whites and 453 blacks in Walton County. The black population had been brought in with the white settlers for slave labor on the farms. The Walton County antebellum society was typical of Gray's "commercial farming" type, although in 1860 there were three farms that could be classified as small plantations. Most of the slaveholders in the county owned only a few slaves and had small family farms (Smith, 1973). The slaves and their owners worked together in the fields, with no overseer or drivers to direct

PAGE 75

54 their work. According to Julia Smith (1973), there were few disciplinary problems in such a situation, and the slaves had relatively good living conditions. They often ate from the same kitchen as did their owners, and they attended the same churches, the slaves sitting in a separate section of pews or in a special gallery. McKinnon wrote that Walton's slaves "were all Presbyterians," holding weekly prayer meetings in their own quarters or houses and worshipping in the Valley Presbyterian Church on Sundays. According to McKinnon, the slaves of Walton County typically lived in small cabins behind the "Big House," not in communal quarters. In most cases, each slaveowner had only one slave, and only a few in the slave schedules for 1860 showed more than 10. The close working arrangement and the personal relationships between slave families and their owners led, in some cases, to a kind of "family solidarity" among slaves and owners (Smith, 1973). The census publications for 1870 reported that there were eight churches in the county; one Baptist, three Presbyterian, and four Methodist. Judging from the number of seatin.gs in the buildings, there were more Presbyterians than Methodists and still fewer Baptists (United States Department of the Interior, 1872b:533). Since a large

PAGE 76

55 proportion of the original population of the county was of Scotch origin, this is as would be expected. The original church in the county, and the center of county life for many, was the Valley Church, first organized in 1828. It was located near Euchee Anna and the other early settlements, and it was the site of the first and only cemetery in the county for many years. In 1870, Walton County had a total of 276 farms, half of which were between 20 and 50 acres in size. Another 41 percent of these farms were smaller than 20 acres, with only five farms having 100 or more acres. The average acreage per farm was 24.6 (United States Department of the Interior, 1872c:348). The main crops of the Walton County farmers were Indian corn, oats, molasses, rice, cotton, and some tobacco. Walton County's delegates were not secessionists, the county having been a "Whig county," not unfavorable toward Lincoln (McKinnon, 1911:269-270). They were among the few county delegates to refuse to sign the secession articles in 1861. But secession came anyway, and Walton County sent her share of young men. Naturally, as the war progressed, the involvement became more intense, and the former Whigs became staunch Confederates. McKinnon lists 90 names of

PAGE 77

56 Walton County Confederate soldiers who were killed in the war (1911:377-378). In the later war years, there were some minor raids into the county, and there was one final raid by Union soldiers in which much livestock was killed or stolen and many of Walton's citizens were kept imprisoned in Euchee Anna for two days, according to McKinnon (1911:327328). These soldiers took many of the former slaves with them, although it is not known how many. The turmoil and high feelings of the war seem to have broken the racial accommodation which had been fairly stable before the war. Not long after the war, Walton County experienced its first lynching, and there were several instances of tension and resentment mentioned by McKinnon. Transition of Walton County, 1870 to 1885 Walton County changed greatly after the Civil War. Before the war, the county was almost totally devoted to agricultural and livestock interests. Afterwards, however, rapid development of the county's resources began, and the economic basis of the county changed. The postbellum county residents engaged in the timber business, logging, steam and sawmill building, sail and steamboat building, and trade by

PAGE 78

57 way of the bay and the rivers. For a while after the war, the efforts to develop resources were very successful ; then the economic depression hit and developments slowed down. The depression hit farmers the hardest. In 1869, the 276 farms of Walton County had produced an average of $550 worth of goods, while the average value of produce for 1885 farms was $282. During the same interval, the number of farms increased by 31 percent to 361, while the number of improved acres increased by only 20 percent from 6,803 in 1870 to 8,152 in 1890. The overall value of farm produce for the county dropped by 33 percent, from $151,833 in 1869 to $101,780 in 1889. (These figures were calculated from data given in United States Department of the Interior, 1872a:720; 1872c:116, 348; and United States House of Representatives, 1896:129, 202.) In the early 1880s the Loui svi 1 1 e and Nashvi 1 1 e Rai 1 road built a branch through Walton County and, according to McKinnon, its impact on the county was great. It acted as though it was a great bomb shell dropped down in the midst of the Valley, crushing, rooting up and driving the old Scotch settlers in every direction, leaving only enough there for seed. ... It broke up . . . one of the plainest, simplest, most social, and truly religious communities in Walton. (McKinnon, 1911:350)

PAGE 79

58 Apparently the new form of transportation gave many residents of Walton an opportunity to move which they had not had before. It also made moves away from one's family less traumatic, as the way back for visits by railroad was not so difficult as former modes of travel. With the railroad came the building of Lake De Funiak, later known as De Funiak Springs. This village was founded in the early 1880s and built on a small hill near the new railroad. It was settled by newcomers, both Southerners and Northerners. The town was advertised as a tourist attraction, and had some successas such. The racial tensions and antagonisms arising from the war and Reconstruction seem to have reached a peak in Walton County in the late 1870s, as whites were more and more upset by the rule of the "carpetbag government" (McKinnon, 1911: 342). In 1876, however. Governor Drew was elected and the Reconstruction era in Florida came to an end. McKinnon mentions several examples of racial disturbances in the county which took place during the 1870s, but he seems to recognize a distinction between the "old family Negroes," who were known to the county and who were fairly stable in family life, and others who were migrants to the county, often working in the logging camps and who were disliked even by the older black population of the county.

PAGE 80

59 It seems likely that a kind of new accommodation had been reached by 1885, although it was not completely stable. Since this was the beginning of the real Jim Crow era in the South, it seems likely that the black population of Walton County may have been worse off in many ways than it had been in 1870. However, it is likely that whatever the situation was in 1885, it may have been fairly stable as compared to that of 1870. The economic problems of the 1880s had probably hit the black population very hard, though, since they were still in many ways dependent on the whites, who now were not as likely to be protective of their interests and feelings as they might have been earlier. We now turn to the analysis of the characteristics of the population of Walton County as it was in 1870. Perhaps it was similar to the prewar population, since there had not been much time for radical changes to occur. The Reconstruction era in Florida lasted until 1877, and it was not until the late 1870s that racial tensions and Jim Crowism developed to a great degree. In 1870, then, the transition from antebellum to postbellum society was still probably in its i nfancy .

PAGE 81

CHAPTER V THE POPULATION OF WALTON COUNTY, 1870 AND 1885 Among the most important determinants of the structure of a community are the characteristics of its population. By this is meant the composition of the population in terms of age, sex, marital status, residence, and other demographic variables, and the nature of its processes of fertility, migration, and mortality. In this chapter, we will describe these characteristics of Walton County for the years 1870 and 1885, for whites and blacks separately. The White Population of Walton County, 1870 In 1870, there were some 2,800 white individuals who were enumerated in the Walton County census manuscripts. The sample obtained contained information on 710 individuals living in 115 white-headed households. The average household size, then, was 6.2 persons. 60

PAGE 82

61 Race Within the white-headed households in the sample, there were 684 white individuals, 22 blacks, and 4 mulattoes (Table 1). About 23 percent of all white households in the 1870 sample had at least one nonwhite resident. This is probably close to the proportion of households which 10 years earlier had at least one slave. For the remainder of this discussion, the "white sample" refers to only the white residents of white households. Since all white persons in each household chosen were included in the study, this means that the white sample is equivalent to a simple random sample of white persons in the county. Table 1. Racial Composition of White Household Sample, 1870 Race White Black Mulatto N 684 22 4 Percent 96.3 3.1 0.6 Total 710 100.0

PAGE 83

62 Sex The sample of white households contained 32? white males and 356 white females, for a sex ratio of 92.1 males per 100 females (Table 2). This fairly low sex ratio, although not extreme, may be partially due to the loss of men during the recent war. It is rather unusual for a rural and relatively newly settled area to have a sex ratio very far below 100. Table 2. Sex Composition of White Sample, 1870 Sex N Male Female 328 356 Percent 48.0 52.0 Total 694 100.0 Age The mean age of the white sample was 22.1 years, although the median was only 17.5 years. The ages of individuals ranged from birth to 80 years, and the quartiles were 9 and 30, years. In other words, it was a fairly young

PAGE 84

63 population, with three-fourths of its members being under 30 years of age. Moreover, over 40 percent of the sample was under 15 years of age. Males had a slightly wider distribution of ages than did females. The median age for males was 17.0 years; for females, 17.7 years. The quartiles were, for males, 15 and 32; for females, 17 and 30. The cause of the low sex ratio can be hypothesized by looking at the age and sex distribution of Walton County's white population in 1870 (Figure 2). First, there appears to be an imbalance in the number's of men and women aged 25 to 39 years. This would have been the age group which had the heaviest losses in the war for men. Walton County's Confederate dead numbered 90, according to McKinnon's list (1911:377-379). This might partly explain the low sex ratio for the county in 1870. Note also the lower numbers of children in the youngest two age groups of the diagram. These two bars represent the children born during the war years and afterward, and a lower birth rate for those years seems likely. Another possibility is that the small number of young children is due to enumeration errors. It seems likely that, in cases where the enumerator found no one home and got his information

PAGE 85

64 ^ X! X XX X X X X XX XX XX X PS' ?s ^s ^s ?s PSi ?s PS PS ?s >^ xxxx: X X XXX XXX XXX X XXX X xxxxx X XX X XX M M M M M M PS PS PS PS ?S PS XXX X X X PS PS PS PS PS PS PS PS PS PS PS PS xxxxx X X XX X X X X XXX XXX XXX XXX XXX XXX XXX XX XX XX XX XXX XXX XX XX XX X X XX PS PS PS PS PS xxxxx PS PS PS PS PS H H H MM X X XX XX XX ^^ XX XX XX X X XX XX «;dI I I I I I I I I I I I I LnOLDOLooLnoLOOLnoLoo .VO(X5Lf)Ln«*^COOOC\JCMr— r—

PAGE 86

65 from a neighbor, infants and very young children might have been more easily forgotten than older, more visible, children. If this happened to any large extent, it could decrease the number of children enumerated in these age groups, thus giving the appearance to lowered fertility for those years. At this point in time, there is no way to determine the exact source of the small number of young children in the census listings. Mari tal Status Of the 402 whites who were 15 years of age or older, 184 (45.8 percent) were single, 186 (46.3 percent) were married, and 32 (8.0 percent) were separated, divorced, or widowed (see Table 3). This proportion of married adults is not unusually low, since the average age at marriage was high in the 19th century. Although there were more aged men than women, a relatively low proportion of males were widowed. Widowers could usually find second wives, because they could marry down in the age structure. Widows, however, had to look upward in the age structure for husbands. This severely limited their chances of remarriage. Thus, there were three times as many women as men in the widowed category.

PAGE 87

66 Table 3. Marital Status of Adults in White Sample, 1870, by Sex Marital

PAGE 88

67 been married, and the estimated median age at marriage was 29.0 years. Hence, according to these estimates, the usual age at marriage for both men and women in this sample was rather late. For comparison, the median age at marriage in the United States as a whole in 1890 was 26.1 years for men and 22 years for women (Glick, 1957:54). It appears that marriage in Walton County took place at later ages than in the country as a whole, for whites at least. Ferti 1 i ty Since there were no vital statistics in Walton County in 1870, the best measure of fertility is the fertility ratio, or the ratio of children under five years of age to women in the childbearing years, age 15 to 44, inclusive, multiplied by 1,000 for better readability. In Walton County in 1870, the fertility ratio, as calculated from the sample data, was 505.9, which is somewhat lower than that for the United States as a whole (about 650) (Broom and Selznick, 1968:278). The lowered birth rate for the war years and following may have lowered this ratio, since it involves those children born between 1865 and 1869, inclusive. It is probable that the social upheavals resulting from the war and Reconstruction had caused temporary

PAGE 89

6.8 declines in fertility and marriage rates in Walton County, both of which are reflected in the measures at which we have been looking. On the other hand, the possibility of enumeration errors in the youngest age categories, as mentioned earlier, still remains. Mortal i ty Unfortunately, the only records available of mortality in Walton County during this period are those inscribed upon burial markers and stones in the Valley Church cemetery. There are problems, of course, with the representativeness of this sample of deaths, since some graves may have been unmarked, and some individuals may have been buried elsewhere. But, as is common in historical research, we must take the data which we are given and note their weaknesses. Some useful insights may come from looking at these records, even if they are not of the precision which we would demand for present-day statistics. One positive note about the data from this cemetery is that it was the main cemetery in the county during the years of the study (McKinnon, 1911:256). Although the cemetery had both black and white graves, the data presented here are for the whites only, and they are taken from a transcription by Bruington (1951).

PAGE 90

69 We were interested in getting an estimate of the average age of death for males and females in Walton County during the decades under analysis. Since the number of deaths was fairly small, all deaths between 1850 and 1885 were recorded and analyzed together, so that any conclusions to be made apply to the general period, and not to the specific decade or year. There were 70 male deaths and 55 female deaths recorded. In addition, there was one person whose sex could not be determined in the records. In all, 126 deaths were recorded, excluding soldiers' deaths. For each, the year of birth, the year of death, and the sex of the individual were recorded, along with the state of birth, where given. There were 46 deaths in the 20 years prior to 1870, and 80 in the 15 years afterward, which reflects the growing population of the county as well as the growing old of the first generation of settlers. About one-fourth of all of these deaths occurred before the age of 15 (25.4 percent), most of these, of course, in the first five years of life. One-third of the male deaths occurred between the ages of 15 and 44, but 44 percent of the female deaths occurred during these ages. About 41 percent of the males and 33 percent of the females had died above the age of 45. The

PAGE 91

70 male-female differential represents the greater death rate of women due to childbirth. If, however, we look only at those who lived through the childbearing years (those age 45 or above), women had higher average ages at death than men. Specifically, the mean age at death for those who died after 44 was 64 years for men and 69.1 years for women. The age at which women died at a much greater rate than men was in the range from 35 to 44 years, the late childbearing years. Still, about 25 percent of all adult deaths, both male and female, occurred at or above the age of 65 years. Occupation and Social Class The occupations of the individuals in the sample are given in Table 4. A glance at the titles and the distribution is enough to establish the fact that most of Walton County's laboring population in 1870 was employed in rural occupations--farmi ng , lumbering, milling, and simple laboring. It seems clear from the context of the manuscripts that most of the individuals classified as simply "laborers' were farm workers who did not own land. The "farmers" were those who did own at least a small plot of land. '

PAGE 92

71 Table 4. List of Occupations of the White Sample, 1870 Title Laborer Farmer Carpenter Servant Lumberman Seamstress Sail Captain of lighter Shingle maker Miller School teacher Cooper Driver Teamster Peddl er Clerk Boat mate Minister Blacksmith 87 75 5 5 4 4 3 3 3 3 3 2 2 Total 205 The occupations were classified into the following six categories: (1) professional, ( 2) managerial , proprietorial, and clerical, (3) skilled crafts, (4) unskilled trades, (5) laborers and servants, and (6) farmers and stockmen. The results of this classification are presented in Table 5.

PAGE 93

72 Table 5. Occupational Class of White Sample, 1870 Class Percent Professional Managerial, clerical, proprietorial Skilled craftsmen Unskilled manual Laborers, menial service Farmers 4 6 18 11 92 75 1.9 2.9 8.7 5.3 44.7 36.4 Total 206 99.9 The modal status level was that of the laborers, with farmers being second. Both of these groups together made up about 80 percent of the laboring population. In addition, there were a few skilled laborers (18) and some unskilled laborers with titles other than "laborer" or "servant." Only 10 individuals of the 206 with occupations given were in either the professional or proprietorial classes (4.8 percent of the total). The county was definitely dominated by agrarian interests, having only enough of the nonfarm occupations to support the farm population.

PAGE 94

73 Birthplaces of the Population Just over half of the white sample had been born in Florida (Table 6), Another 28 percent were born in Alabama, with 7.5 percent originating in Georgia. Thus, these three states were the birthplaces of 90 percent of the white sample. The rest came mostly from the upper South, with less than 1 percent being of foreign birth. Almost all of the originar Scotch settlers, many of whom had been born in Scotland, were dead by 1870. The population represented here probably is largely of firstor second-generation American bi rth. Table 6. Birthplaces of the White Sample, 1870 State of Birth N Percent Florida Alabama Georgia N. Carolina S. Carolina Vi rginia Other South Other U.S. Foreign 367 196 51 33 30 2 2 1 6 53.3 28.5 7.4 4.8 4.4 0.3 0.3 0.1 0.8 Total 688 99.9

PAGE 95

74; Length of Florida Residence The large proportion of Florida-born individuals overestimates the proportion of fami 1 ies who came from other states, since many younger children, even of migrant families, were born in Florida. In order to control for this problem, estimates of the recency of migration of families were made, where possible. This estimate was based on the place of birth of the household head, the places of birth of his children, and their ages. It was possible to estimate this for 85 of the 115 primary facilities in the sample. The results of the estimates are' presented in Table 7. Table 7. Recency of Migration of Families of Household Heads, White Sample, 1870 Length of Florida Residence Percent Florida native 15 years or more 10-14 years 5-9 years 0-4 years 21 38 6 2 18 24.7 44.7 7.1 2.4 21.2 Total 85 100.1

PAGE 96

75 Of the 85 families, about one-fourth were Florida natives , and the bulk of the rest had lived in the state for longer than 15 years. As might be expected, only a few had come to Florida during the war years. After 1865, however, the immigration to the state seems to have picked up, as just over 20 percent of household heads had been in Florida for less than five years. ' The Black Population of Walton County. 1870 There were 380 black or "mulatto" individuals listed in the Walton County Census of 1870. Almost three-fourths of these individuals (274 in all) lived in 48 black-headed households; while the remaining 106 were listed as part of 47 white households. This latter group seems to be a holdover from the days of slavery, when black slaves were listed under their masters' names. There is evidence for believing that the social condition of these blacks was also little different from that which existed prior to emancipation. If so, this group makes an excellent source of information about the black family in the South in the last years of slavery. This includes two whites, one who was the wife of a black household head, and the other who was her son, presumably by an earlier marriage.

PAGE 97

76 The mean household size for black households in 1870 was 5.7, slightly less than the white mean of 6.2 The mean number of blacks per white household containing blacks was 2.3, with a median and mode of one per household (27 of the 47 households had one black individual each.) This is similar to data taken from the 1860 slave schedules, in which 128 slaveowners held an average of 4.3 slaves each, the mode being one. The mean was high due to a few (nine) slaveholders who owned more than 10 slaves, one of whom owned 46. This is one reason for believing that the blacks living in close association with whites' in 1870 were in many cases former slaves who had never left the "master's" farm. Race In 1870, nonwhites were classified by the census enumerator as being either "black" or "mulatto," depending mainly, it appears, on the skin color of the individual. According to many analysts of black life and culture in the South after the Civil War (see, for example, Frazier, 1948), there was a prejudice in both the white and black communities in favor of lighter-skinned blacks. In the black household data, there were 192 blacks and 80 mulattoes; in white households, there were 80 blacks and 26 mulattoes

PAGE 98

11 (Table 8). Overall, about 28 percent of the nonwhite population of Walton County in 1870 was classified as mulatto. Table 8. Black Mulatto Racial Composition of Nonwhite Population of Walton County, 1870

PAGE 99

to do farm work, and thus were more often kept by the farmers as slaves or as free workers. Note that both black groups had higher sex ratios than did the whites, whose women outnumbered their men with a sex ratio of 92.1. Table 9. Sex Composition of Black Population of Walton County, 1870, by Race of Household Head

PAGE 100

79 X X X XX xxxxxxx xxxxxxx M M M M M H M XX XXX X X X X X X X X X X X X XX XX X X X X M M X X XX X X X X X X X X X X X X X X X XX XX XX XX XX XX XX X X XX XX l-H M XX XX XX X X XX XX XX X X X X X X X X X X X X X X X X X X X X X XXX XXX XXX XXX XXX r*S ?*S K*S XXX MUM XXX f'S K*S f^ XXX kS r*S K% XXX XXX X XX XXX XXX XXX XXX X X X X X fOiOLnLn>*^rooocxicMt— 1 — crioo

PAGE 101

80 Marital Status Fifty-two percent of blacks 15 years of age or older and living in black households were married. Thirty-five percent were single and 13 percent were widowed. Only 25 percent of those living in white households were married, over half were single, and one-fifth were separated or widowed. In the total black population, 42.7 percent were single, 43.1 percent married, and 14.2 percent widowed (Table 10). The greater proportion of single and widowed blacks in white households may signify that (1) the widowed with children needed a secure place and protection, as provided by the white household, independence being almost impossible for unmarried mothers, (2) white householders Table 10. Marital Status of Adult Black Population of Wal ton County, 1870 Marital

PAGE 102

desired unmarried black servants and laborers, or (3) when blacks married they attempted, usually successfully, to set up separate households. At any rate, those who were able to maintain their own households were much more likely to have viable marriages than those who were still living in close association with white households in 1870. Estimated Age at Marriage In order to estimate the age at marriage for blacks, it was necessary to increase the number of cases. The small population of Walton County was supplemented by the total black population of Holmes County, Walton's neighbor to the east, and formerly part of Walton County. Since Holmes was similar in economy and population to Walton County, this procedure probably does not bias the results unduly. The estimation procedure followed that for the white population (see Appendix C). As compared to whites, the black population seems to have begun marrying at later ages, but the median age at marriage was much lower and the proportion ever-married was higher. No black male under the age of 21 years, and no black female under 18, was married. Of those between 15 and 45 years of age, 55.1 percent of the males and 61.8 percent of the females were

PAGE 103

82 married, with spouse present. These figures are about 20 percent higher for males and 13 percent higher for females, than the corresponding white percentages. The median age at marriage was estimated to be 26 years, for males, and 25.5 years, for black females. Both of these figures are about three years lower than the median marriage ages for whites. Ferti 1 i ty The ferti 1 i ty rati for black Walton County residents of black householdswas 634.9. For those in white households, it was 500. The overall black fertility ratio was 597.7, somewhat higher than the white ratio of 505.9. As expected, the less stable white household groups had lower fertility ratiosthan did those who headed their own households. For those in their own households, the fertility ratio was quite a bit higher :than that for the white population. This probably represents a higher birth rate and a higher proportion of women in the chi 1 dbeari ng years who were married. Mortal i ty No separate death data were available for the blacks of Walton County. It is fair|ly well-known that, nationally.

PAGE 104

83 the death rate among blacks was higher than that for whites and that life expectancy at birth was lower. In 1900, the average life expectancy at birth for white males was 46.6 years, while that for black males was 32.2 years. For females, the figures were 48.7 years for whites and 35.3 years for blacks (United States Bureau of the Census, 1960:25). We might expect that similar differentials would be true for Walton County. A look at the age-sex pyramids partially confirms , . this expectation, since the higher age groups among blacks contained very few individuals as compared to the proportions of whites in those groups. This could be due to lower ages at death for blacks. Occupation and Social Class The occupations of blacks in Walton County in 1870 did not cover a \jery wide spectrum, although there was some variation. Table 11 gives the entire occupational distribution for the black population, by race of household head. It is easily seen that the black families with households of their own were in a better position economically and socially than were those still in white households. Over onefourth of the black-household workers were farmers who

PAGE 105

84 owned at least a small amount of land. Sixty percent were laborers, but about 12 percent had a variety of other occupations, mostly unskilled. There was greater variety than among blacks in white households. Table 11. List of Occupations of the Black Population of Walton County, 1870 Black Househo

PAGE 106

85 only about 8 percent were in other status groups (Table 12). The contrast with blacks in white households is great; fully 95 percent of them were laborers. Table 12. Occupational Class of the Black Population of Walton County, 1870

PAGE 107

86 large proportion of the individuals. The whites had a definite advantage over blacks, however, with as many farmers as laborers, and a fairly large group of skilled tradesmen. In neither case were there many professionals, the distributions being heavily weighted toward the farming and laboring groups. It is of interest here to look at the occupations of heads of white households that contained black residents. As might be expected, over 80 percent were farmers. Three percent were laborers, and 14 percent were in the higher three classes. When the occupations of this group are compared to those of white household heads in general, the former group is found to be of higher status, 83 percent being farmers, as compared to 38 percent of the general population. These white farmers with black farm laborers in their households probably represent fairly well the former slaveholding class of Walton County. Birthplaces of the Population The birthplaces of Walton's black population in 1870 were mostly in the three nearest Southern states, Florida, Alabama, and Georgia. There was little difference in the places of origin of the blacks in black households from

PAGE 108

those in wh i te househol ds . Those in black households seem to have been a little more often from the upper South than those in white households; but, otherwise, the two groups are very similar in this respect. About two-thirds of each group had been born in Florida (Table 13).' Table 13. Birthplaces of the Black Population of Walton County, 1870

PAGE 109

little difference in the origins of the black and white populations, most being from nearby areas. Of the total 1,055 individuals for whom birthplace was recorded, only one was from a non-Southern state, and only six were of foreign birth. Length of Florida Residence Only 18 of the heads of black households, 37.5 percent, had been born in Florida. For another 18 households, it was impossible to determine the length of residence of the families within the state. Of the non-Florida natives for whom residence length could be estimated, most had been in the state for more than 15 years, only three having come to Florida within the last 9 years (Table 14). Table 14. Recency of Migration of Families of Black Household Heads, Walton County, 1870 Length of Florida Resi dence Flori da nati ve 15 years or more 10-14 years 5-9 years 0-4 years Total 9 1 2 30 Percent 60.0 30.0 0.0 3.3 6.7 100.0

PAGE 110

89. Most of the heads of black households, it appears, had been brought to Florida as slaves; they would have had to come within five years before 1870 in order to have come as free men. There was not much difference in the black and white migration patterns, except that many of the white nonFlorida natives had come into the state after the war years As far as we can tell, almost no black families came into the state during those years. If blacks did migrate into Florida during these years, it was not as families, but as single individuals or couples, at least in Walton County. The White Population of Walton County, 1885 There were 1,213 people living in white households in the Walton County sample of 1885. About three-fifths of these constitute the one-in-five sample of the rural areas of the county, hereafter referred to as the "county" sample. The other 495 people were the residents of all of the white households within the village of Lake De Funiak. Lake De Funiak, established after 1870, grew rapidly. Although young, Lake De Funiak developed a "town" character quite unlike the old agrarian character of the county. It

PAGE 111

90 will be of interest to compare the rural county families and households with those in the village to see how they di f f er . The total white population of Walton County in 1885 was about 5,450, or almost twice as large as that of 1870. Since only about 500 of these people lived in Lake De Funiak, the only real "town": in the county, a large part of the increase was in the rural population. The 718 residents of white households in the county sample lived in 104 households, with a mean household size of 6.9. Lake De Funiak contained 495 people in 78 households, averaging 6.4 per household. The average household i size in 1885 was somewhat higher than it had been in 1870 (6.2). Race There were only three black persons in two of the 104 households in the county sample. In Lake De Funiak, by contrast, there were 14 blacks in 78 households. The drop from 1870 in the percentage of county white households with blacks was striking, from 23 percent to 2 percent.

PAGE 112

91 Sex There were 352 males and 364 females in the white 1885 county sample, for a sex ratio of 96.4. In Lake De Funiak, there were 267 males and 215 females, for a sex ratio of 124.2 (Table 15). Thus, the] sex ratio was much higher in the village than in the county. The change in the total sex ratio since 1870, from 93.3 to 101.1, may perhaps be explained by the influx of young male migrants in the interval. Table 15. Sex Composition of Whi te Sampl es , 1885

PAGE 113

92 In Lake De Funiak, the median age of white residents of white households was 19.3 years, ranging from to 84 years. As with the county sample, about 42 percent were below 15 years and 13.2 percent were 45 or above. The quartiles for the village population were 8 and 31 years. Thus, there was little difference in the age distributions of the village and county samples, except that the county had a few more very old residents and very young residents, while the village had more young adults. The county whi te popul ati on pyramid (Figure 4) shows a slight imbalance in the sex ratio in the age groups between 35 and 45. Whether this is due to the war losses or to the tendency for women to understate their ages at certain stages of the life cycle is uncertain. These individuals would have been aged 20 to 30 in 1870. The Lake De Funiak population, by contrast, had a more balanced age distribution, with a larger proportion in the young adult years, and not so: many at the extremes of age (Figure 5). The sex ratio imibalance of Lake De Funiak may be a matter of chance, since most of the imbalance is in the childhood years, below thje age of 15. Alternatively, this could be due to enumeratiion errors, since youthful

PAGE 114

93 X X X XX X X X X X X x; X X X X X X M M M M X X X X X X XX XX XX X XX X X X X X X X X X X X X X X X X X X X X X X X X X X X X X M l-i X X X X XX X X X X X X XX X X X X X X X X X X X X X X X X X X XX XX X X XX XX XX XX X X X X X X XX 1-H M X X XX XX XX XX XX XX X X XX XX XX X X X X X X XX XX XX XX X X XX X X X X X X X X X X X X X X XX XX XX X X XX X X XX XX X X XX XX XX XX X X XX XX XX X X X X M M X X XX X X X X XX XX X X XX X X XX XX X X X X X X X X XX X X XX XX XX XX X X X X X X X X X
PAGE 115

94 X X X X X X X X X X X X X X X X X X M M M H M M XX X X X X X X X X X X XX XXX XXX X X X :< X X X X X X X X X X X X X X X X M H X X X X ^< X X X X X X X X X X X X X X XX XX X X XX XX XX XX XX M M XX XX XX XX X X XX XX X X X X X X X X X X X X X XXX K*S r^ r^ kS PS rS kS f-*S PS PS PS PS X X X X X X X X X X X X X X XX X XXX XXX X XX X X X X X X PS PS PS rS M M M M PS PS PS PS X X X X X X X X X X X X X X X X X X ><; X X X >^ X X X X X X X X X X X X X X X X X XXX XXX XXX XXX XXX XXX X X X X X tn CO •I00 en O cC QLT) oLooLnoLnoi-noLnoLno

PAGE 116

95 boys may have been more visible to neighbor informants than were their sisters. Marital Status Forty-three percent of persons age 15 or over in the county sample were single, 49.4 percent were married, and 1.1 percent were separated or widowed. Of the village adults, 35.7 percent were single, 55.6 percent were married, and 8.7 percent were widowed (Table 16). The two samples, village and rural, were quite different in marital status, even though their age distributions were fairly similar. The larger proportion of single adults in the county sample may be a function of later ages at marriage in the rural districts. Table 16. Marital Status of White Adults in Walton County, 1885 , by Residence Marital

PAGE 117

96 Estimated Age at Marriage The estimated median age at marriage was calculated, as with the 1870 data, but separately for the county and Lake De Funiak samples. In the county sample, the earliest age at which any male was married was 21 years; for females, it was 16 years. For all those between 15 and 45 years of age in the county sample, 40.2 percent of the males and 50.9 percent of the females were ever-married. The median ages at marriage were estimated to be 29.5 years for males and 25 years for females. The male estimate was about the same as that for 1870, but the female estimate was some four years younger. The proportion of those married had remained about the same for the county sample as the 1870 figure. In the Lake De Funiak white population, the earliest age at which any male was married was 22 years; the corresponding age for females was 18 years. The proportions of ever-married males and females aged 15-45 were 48.1 and 67.0, respectively. Thus, the proportion of aduTts who were married, or had ever been married, was higher in the village population than in the county. The median ages at marriage were estimated to be 28.5 years for males and 24.5 years for females in the village. These are quite similar to the county estimates.

PAGE 118

97 Ferti 1 i ty For the county sample in 1885, the fertility ratio was 581.3, representing a large increase over the 1870 figure of 505.9. This increase strengthens the conclusion that the 1870 fertility ratio was depressed because of the social upheavals of the previous years. The Lake De Funiak fertility ratio of 676.6 for 1885 was higher than that for the county. This is probably a result of the larger proportion of married couples in the village population. Occupation and Social Class In 1885, the county sample contained 126 persons who were listed as having occupations. Table 17 lists the occupations given, along with the frequency of each. The most common occupational title was "laborer," which accounted for 66 (52.4 percent) of the workers. Next in number were farmers, who, combined with the stockmen, comprised another 20 percent of ithe working population. After these, there were a number of different occupations, none of which represented many people. The Lake De Funiak occupational listing is even more varied, with a total of 37 differeint occupational titles (Table 17). As before, the most common occupational title was that

PAGE 119

98 Table 17. List of Occupations of White Samples, 1885 County Sample

PAGE 120

99 of laborer, given by 39 persons, or 30 percent of the workers. Next in frequency were carpenters , servants , and merchants. A look at the titles show a quite diversified village, including a publisher, a photographer, a music teacher, a dentist, a druggist, four physi cians, and numerous other types of positions. There were, of course, fewer farmers in the village than in the county. There was also a large drop in the proportion of farmers in the 1885 county sample as compared with the 1870 sample. Not only had the proportion of farmers declined, from about 38 percent to about 19 percent, but the absolute number of farmers in Walton County had decreased by at least one-half. Perhaps part of the rise in new occupations and the concomitant growth of the village is explained by the di f f icul ty of farmers in mai ntai ni ng thei r farms in the difficult years of the 1870s and 1880s. The summary of the occupational distribution according to social class for 1885 is given in Table 18. The county sample had only 20 percent in the farming class, with over half being in the laboring class. Growth is seen, however, in the other classes, as professionals comprised almost onetenth of the occupations; and 7 percent were in the

PAGE 121

100 proprietorial class. Eight percent were classified as unskilled laborers; in all, over 60 percent of the white laboring population were in low-status jobs. Table 18. Occupational Class of Adults in White Samples 1885

PAGE 122

101 Again, the change in the economic basis of the county can be seen in the decline of the farming class and the rise of the town occupational types. Note that as the proportion of farmers declined, the proportion of common laborers rose along with the proportion of other types of occupations. Some of the farmers who lost their land in the 15 years prior to 1885 were undoubtedly part of the increasing force of common laborers. Birthplaces of the Population The birthplaces of the white household population of Walton County in 1885 are given in Table 19. In the county sample, 67.1 percent of the residents were born in Florida, while another 25 percent were Alabama natives. The remainder were from other Southern states, except for one Scandinavian native. None of the county residents from outside of Lake De Funiak had been born in a non-Southern state, and about 93 percent were natives of either Florida or Alabama. The Lake De Funiak residents were from more varied origins than were the county residents. About 45.5 percent of the village residents were Florida natives; 27 percent were from Alabama and another 17 percent were from eight other Southern states, about two-thirds of these from

PAGE 123

102 Table 19. Birthplaces of the White Samples, 1885

PAGE 124

103 than that of the early 19th century. It was mostly migrants who held the more skilled positions in the village. Length of Florida Residence There were 59 households in the white 1885 county sample whose migrant status and time of migration could be estimated (see Table 20). Of these, over a third had been lifetime residents of Florida. Most of the remainder were long-time Florida residents, with only about 13 percent having come to the state with the past 15 years. Table 20. Recency of Migration of Families of White Household Heads in Walton County, 1885 Length of Fl ori da

PAGE 125

104 Surprisingly, more of the Lake De Funiak household heads than county household heads had been born in Florida. But the village did have a much higher proportion of recent migrants to the state. Only 52 percent of the village household heads had been in Florida for more than 14 years, compared to over 85 percent of the rural household heads. Almost half of the village household heads had been in Florida for less than 10 years, and almost one-fourth had been in the state for less than 5 years. Therefore, the population of Lake De Funiak represents a very different group of people than the county sample, having not long been a part of Walton County. The Black Population of Walton County, 1885 The total black population of Walton County in 1885, as listed in the census manuscripts, was 674 persons. This represents an increase of about 77 percent over the 380 total of 1870. One of the first differences to be noted is the absence of a large proportion of blacks within white households in 1885. As was pointed out earlier, 28 percent of the total black population resided in white households in 1870, By 1885, only 33 black individuals were listed in

PAGE 126

105 white households. This represents a 23 percent drop in the proportion of blacks in white households over the 15-year period. This means, essentially, that the adjustment to a nonslave economy had been carried through by 1885. Most of the blacks in white households in 1885 were household servants, while most of those in 1870 had been farm laborers. For the black population as well as the white population, this change is perhaps one ofthe most significant indicators of the kind of social readjustments which had been made in the decades following the Civil War. There were 641 blacks living in 97 black households in 1885, an average of 6.6 individuals per household. Since about 95percent of the black population lived outside of Lake De Funiak, the distinction between village and county is notmadeforblackhouseholds. The 33 black individuals in white households were living in 19 different households, for an average of 1.7 per household. However, most of the households had only one black person; only six of the nineteen had more than one black resident.

PAGE 127

106 Race In the black household population, there were 605 "black" individuals, 34 mulattoes, and 2 whites (the same two individuals who were in the 1870 black population). Adding those blacks in white households, there were 636 blacks, 36 mulattoes, and 2 whites in the "black" population of Walton County in 1885. Thus, 94 percent of the black population was considered by the census enumerator to be black, as compared with 72 percent in 1870. Whether this represents a real change in racial composition or merely a change in perception on the part' of census takers cannot be determined. Most likely, it is primarily the latter. Sex There were 318 males and 323 females in black households in 1885, for a sex ratio of 98.5, as shown in Table 21. If we add the blacks in white households, the totals were 341 males and 333 females, and a sex ratio of 102.4 males per 100 females. This is a very evenly balanced sex ratio, as was that of the black household population in 1870 (103.5), although it is somewhat lower than that of the overall black sex ratio of la70 (1 13. 5) .

PAGE 128

107 Table 21. Sex Composition of Black Population in Walton County, 1885, by Race of Household Head

PAGE 129

108 X X X X X X X X X X X X X X X X X X X X X X M M M M M M X X X X X X X X X X X X X XX X X X X X X X X X X XX XX X X XX XX XX XX XXX XXX XX XX PS r*S r>S r^ X X XX X X XX rS ^S r*S r*S K*S K*S »=S K*S X X X X X X XXX XXX XXX XXX XXX XXX XXX XXX XXX XXX XXX XXX X X X X X XXX X kS r^ P^ P*S PS PS PS PS X XXX PS PS PS PS X XX X X XXX X XXX PS PS PS PS X XXX X X X X X X XX X X X X X X XX X X X X X X "-^ X X X XX XX XX XX XX X X XX XX XX XX XX XX XX XX XX XX XX X X XX X X X X X X X X X X XX X X X X X X X X XX X X X X X X X X X X X X XX X X XX XX XX XX X X XX X X X X XX XX X X XX XX XX XX X X X X X X X X X X X 0) 00 C7)00 =1 rI I I I I I I I I I I I I LnOLnoLnOLnoLnounOLno
PAGE 130

109 Compared to the white population, the black population in 1885 was younger and had a smaller proportion of elderly people. Both blacks and whites show fewer births in the last bar of the graph, representing the births within the past five years. This means, perhaps, that the rate of growth slowed between 1880 and 1885. This could also reflect the same type of enumeration error mentioned in the discussion of the 1870 population, page 63, in which very young children were missed by neighbors' reporting. Marital Status There were 348 black adults in Walton County in 1885, 57.5 percent of whom were married (Table 22). This is a fairly high proportion of adults married, compared to whites in 1885 and to the 1870 figures. Only 6.6 percent of adults were in the widowed or other category, representing a marked improvement in the marital condition of the black population since 1870. The proportion of married adults is greater for blacks than for whites, mainly because of earlier marriage ages for the blacks. If one looks only at black household residents, excluding those in white households, one finds that 62 percent

PAGE 131

no Table 22. Marital Status of Black Adults in Walton County, 1885, by Sex Marital

PAGE 132

Ill is not yery great. Since most of the blacks were from the rural parts of the county, however, they should probably be compared to the county white sample, from which they differed quite significantly. The percentage of black females age 15 to 45 who were ever-married was 70.9, compared to 50.9 for whites; for males, the corresponding percentages were 53.2 (black) and 40.2 (white). The estimated age at marriage for black males was 26 years; for females, it was 21 years. This represents a large drop from 1870 in the age at marriage for females, but not for males. Strangely enough, a similar drop appeared in the white county figures, the male figure remaining stable and the female one dropping by four years. The explanation of this phenomenon may be found in the reaction to the war and its disruption of the population. Perhaps marriages were postponed during those years so that, in 1870, most of the brides were older than they would have been otherwise. By 1885, however, the former distribution of ages at marriage had restored itself. Ferti 1 i ty The fertility ratio for the black population of Walton County in 1885 was 794.9, or over 150 more children per

PAGE 133

112 thousand chi 1 dbeari ng women than in 1870. This, again, points to the disruption of the postwar years and the lack of stability of rel ati onshi ps in the earlier period. Occupation and Social Class The range of occupations for black workers was very limited in 1885, only eight titles appearing for the entire black working population (Table 23). Three-fourths of all those employed were "laborers," and 13 percent more were servants. There were eight farmers, no doubt the elite of Walton's black population; and there were two porters, one blacksmith, two cooks, and two washerwomen. Table 23. List of Occupations of the Black Population of Walton County, 1885 Title Laborer Servant Farmer Cook Porter Railroad Worker Washerwoman Blacksmith N 108 19 8 2 2 2 2 1 Total 144

PAGE 134

113 In the white households, most of the blacks were domestic servants (17 of the 33). Nine were laborers, and some were service workers in a hotel. The occupations of the heads of the white households containing black residents were mostly nonfarming occupations, by contrast with the 1870 situation when over 80 percent of the household heads were farmers. In 1885, the household heads were mostly skilled laborers or proprietors, mostly from the village. Of the 19 household heads, only 1 was a farmer, and he had only 1 black servant. This change in residence patterns of blacks during the 15 years from 1870 to 1885 reflects the changes in the everyday relationship of farmers to black farm workers during that interval. The occupations of the black working population were grouped into the six-level social class scale, as shown in Table 24. As the table indicates, there were no blacks with either professional, proprietori al /manageri al /cl erical , or skilled occupations, and there were only seven with unskilled labor titles. Eight (6 percent) were farmers, and the remaining 90 percent were classified in the laborer and menial service class. As is easily seen by comparing these figures with those for 1870, a general decline in status to the

PAGE 135

114 Table 24. Occupational Class of the Black Population of Wal ton County , 1 885

PAGE 136

115 Table 25. Birthplaces of the Black Population of Walton County, 1885 State of Birth N Percent Florida Al abama Georgia N. Carol i na S. Carol i na Vi rgi nia Tennessee Other South Other U.S. Forei gn 441 118 14 20 46 4 2 4 2 67.7 18.1 2.2 3.1 7.1 0.6 0.3 0.6 0.3 Total 651 100.0 other 7 percent came from a variety of other, mostly Southern states. The distribution of birthplaces of the black population is very similar to that for the county white sample, but it is different from the Lake De Funiak white population. The birthplaces are similar also to those of the black population in 1870, with a few more Florida natives, as would be expected with a fairly stable population. From this information alone, no evidence is apparent for a substantial migration, either in or out of Walton County, of blacks in the 15year period under consideration.

PAGE 137

116 Length of Florida Residence The black households of 1885 were almost all headed by Florida natives or individuals who had been residents of Florida for more than 15 years. Only 10 percent of the households whose migration time could be estimated were recent migrants into the state (see Table 26). Table 26. Recency of Migration of Families of Black Household Heads, Walton County, 1885 Length of Florida Residence

PAGE 138

117 1870, expanded and changed rapidly in the years before 1885. Blacks no longer lived in white households to any great extent in the latter year. This probably reflects the growing hostility between blacks and whites following the Reconstruction period, as well as the effect of the economic decline on the farmers, who could no longer afford to house and support farm laborers. The occupational structure had changed dramatically, from one almost totally dominated by agrarian interests to one with a wide variety of more modern occupations. The farming occupations had declined', in response to the economic depression, and other occupations were expanding. The occupations of blacks during this interval had been levelled into a single class of unskilled laborers. The establishment of the railroad and the village of Lake De Funiak had opened up the county to a new kind of growth. In the village, there were quite a few new migrants, young men or young families from Northern states or other distant origins, many with skills not present in the county before. The rural population, by contrast, had very few nonSouthern migrants, its population still similar in many ways to that of the county in 1870. For them, however, the rate of natural growth was high, as marriages were entered

PAGE 139

118 into at earlier ages and fertility increased over the 1870 1 evel s . In short, Walton County changed rapidly from the old agrarian settlement to a more varied and occupational community. The effects of the changes in economy, transportation, and social definitions can be seen in the changes in the population over this 15-year period.

PAGE 140

CHAPTER VI THE STRUCTURE OF FAMILIES AND THE DEVELOPMENTAL CYCLE, 1870 AND 1885 The main issues concerning both white and black families in the American past have to do with the residential and community structure of families and kinship. In this chapter, detailed descriptions of family and household structure are presented, by the various stages'of the family life cycle, for white and black families in Walton County in 1870 and 1885. Specific analyses are directed toward the questions raised in Chapter II about the characteristics of preindustrial American families. In addition, an attempt is made to piece together the typical life cycle of individuals in the populations studied, by analyzing the cross-section pf individuals in the various stages of development. Finally, important insights come from the comparisons of the 1870 and 1885 family structures, concerning the adaptation of family systems to changing social and economic circumstances . 119

PAGE 141

120 Structure of White Families, 1870 The basic structure of family groups in the 1870 white household sample is summarized in Table 27. Of the 120 families in the sample. 111, or 92.5 percent, were primary families. Of these, the majority (67.6 percent) were nuclear families', 8 percent more being married couples without children. Only 7 percent of the primary families were threegenerational or stem families, and another 14 percent were one-parent families. It should be noted that most of the one-parent families were older widowed parents 1 i vi ng with their grown, but single, children. Only two of the parents of these households were below 40 years of age, and most of them were above 50. The relatively late ages at marriage of the children of older parents led to the low percentage of three-generation or stem households and the high percentage of one-parent households. If one considers families to have their beginning with the marriage of a couple, then the "typical" life cycle of the family may be constructed by looking at a cross-sectional view of families at their different stages of development. This cross-sectional reconstruction assumes, of course, that there was a great deal of consistency over time in the

PAGE 142

121 Table 27. Structure of Families in White Sample, 1870 Total

PAGE 143

122 sample, only six young childless couples were found. In order to increase the number of cases for analysis, seven couples from a sample of neighboring Holmes County were included in this particular analysis. In order to avoid including older couples whose chi 1 dren were gone, or older couples who were infertile, only those couples in which the wife was 35 years of age or less were selected. Most historical studies of American families have found that couples established neolocal residence at marriage, except in cases where an inheriting son remained in his parents' home after marriage. Even where this pattern was common, as in Andover and Plymouth, most sons and daughters moved out of the family household at marriage. This was, in fact, one factor in keeping the age at marriage rather high because it took a great deal of economic independence to build a separate house. Walton and Holmes Counties seem to have been in line with the basic neolocality of other American communities studied: all of the 13 young couples in the samples were heading their own households. Only eight, however, were living alone in those households. Four of the couples had resident boarders and one had a dependent parent within the

PAGE 144

123 household. Thus, although it was usual for newlyweds to live in their own households, it may have been fairly comnron for them to take in unrelated boarders, probably to help to support the households. Most families in the cross-section of households sampled were nuclear in structure, as is to be expected when childbearing and rearing took up such a long part of the family life cycle. Most couples probably had their first child within two or three years, and continued to have them at similar intervals until the wife was in her late thirties or early forties, when the intervals between children became longer. In the Walton County sample of 1870, seven couples were found who were in the beginning of the childbearing phase, each with one child under three years of age. All of these were primary families, and all but one lived alone with their child. If the Holmes data are included, there were 15 such young families with one small child, 14 of whom lived alone in their own households, and all of whom headed their households. It therefore appears that, while young couples without children often took in boarders, after the birth of a child, they preferred living alone. The one

PAGE 145

124extended household involved a couple who were sharing their household with the widowed mother and grandfather of the husband. The Childbearing Years Forty-four of the family groups in the Walton County sample were young couples, the wife 44 years of age or less, with at least two children. The average number of children for this group was 4.75. Over 80 percent (36 of the 44) of these families were living alone, with no other relatives or boarders. Two of the families contained a widowed parent of the wife. And seven families (15.9 percent) shared their households with boarders, servants, or secondary families. Thus, as more children were born, it became somewhat more common to take in boarders, perhaps to supplement family income. But by far the more common situation was for parents to live alone with their growing families. The Childrearing Years The next stage of development of the family is that of later childbearing. Typically, the last child has been born and the older ones are growing to adulthood during this phase. For most families, this stage was also the last stage

PAGE 146

125 of the marriage, since very few couples lived to see the marriages of all of their 'children. In order to look at the residential patterns during these years, all couples in which the wife was 45 or older were classified according to living arrangement. In all, there were 27 such older couples in the Walton County sample of 1870. The average age of the eldest child in these households was 23.6 years, ranging from 16 to 35; and all but three of the couples had children in the household. The average number of children in the households that contained children was 4.7--about the same as for young couples with two or more children. In contrast to the younger groups, only 15 (55.6 percent) of these couples lived alone with their unmarried children. Of the other 12 households, 5 contained other relatives of the couple (2 with siblings, 2 with grandchildren, and 1 with a widowed parent); and 8 contained nonrelated boarders or servants. In all, 44.4 percent of these older couples were living with non-nuclear household members. It can be seen that, as the life cycle continued, many more couples began to augment their families with extra household members. Most of the resident boarders were young men, either in their late teens or early twenties, who could

PAGE 147

126 help with the work of the family or add to the household income. It seems likely that after the household had expanded to its peak and had begun to diminish in size with the marriage of older children, the space left by their departure often was given to other relatives or boarders. Wi dowhood Finally, the death of one of the spouses left the other in the final phase of family devel opment--wi dowhood. As with the other phases, those families in this phase were classified according to living arrangement. Since the main concern of this discussion is with the normal pattern of development, only those widows/widowers above age 44 were considered. There were 13 widows and 10 widowers in the sample who were at least 45 years old. Of these, four (two male and two female) were not living with a child: two male widowers were living together with a third unrelated man, and two females were living with sisters. All of the others were living with at least one of their children. Twelve were living with their unmarried children and seven with married children. Of those who headed their own households with unmarried children, five contained boarders, one contained a sister, and another five contained only the parent and children.

PAGE 148

127 Most of the males were listed as heads of their households. This was not true of the females, unless they were living only with their relatively young children. Most of the widowed parents in the sample lived with their children, either married or single; ifnot,with a sibling. But their households were also quite likely to contain a boarder in addition to their unmarried children. In this respect, theywere not much different from the married couples in the later stages of chi 1 drearing. It is probable that the transition from late marriage to widowhood, then, did not entail finding a new residence for the survivor. Instead, si nee the death of the first spouse typically took place when there were still a few unmarried children in the household, the same household continued. Eventually, however, one after another of the children married and left the household. It seems likely that one of the children, upon marriage, remained in the parental household with the widowed parent, although, at that time, the headship of the household usually passed to the young husband. This seems to be the simplest and most reasonable explanation of the residential arrangements of the widowed in Walton County in 1870. Those who were living with a married child were usually quite a bit older than those who lived with unmarried children only.

PAGE 149

128 There are two major groups of individuals who have been, ignored by concentrating on family development from marriage forward--the children and the single adults. Since both of these groups made up substantial portions of the population in 1870, separate analyses were made of living arrangements from their perspectives. The Children In the white household sample of 1870, there were 282 children under the age of 15, comprising 41.1 percent of the sample. Most of the children lived in households headed by a complete nuclear family or a married couple, even though they were not always part of the primary family. Only 12.5 percent lived in households headed by single-parent families, and less than 4 percent lived in three-generation households. Even for children who were not part of nuclear primary families, most children had the model of the two-parent family within their households, and 90 percent of them were in male-headed households. Of course, most of the children (over 90 percent) were sons and daughters of household heads, the others being extended relatives, secondary children, or servants.

PAGE 150

129 To the child, it may well be that the status of his parents within the household, whether heads or not, is not so important as the overall configuration of relatives and other people. A child who has a large number of significant kin in the household is likely to develop a wider array of relationships with adults than would the one in a residentially isolated nuclear family. It is this dimension of the child's world which is of interest: are both parents present? Are there others in the household with whom he or she must contend, share space and resources, and maintain ties? The summary of the fami ly' 1 i vi ng arrangements of children in the white households of 1870 is presented in Table 28. Table 28. Living Arrangements of White Children in Walton County, 1870 No Extended With With Relatives Grandparents Others Total Living With N Percent N Percent N Percent N Percent Both parents 213 75.5 13 4.6 3 1.1 229 81.2 One parent 35 12.4 4 1.4 4 1.4 43 15.2 Neither parent 1 0.4 9 3.2 10 3.6 Total 248 87.9 18 6.4 16 5.7 282 100.0 Percentages are of all children.

PAGE 151

130 Of the 282 children in the sample, 229 (81.2 percent) were living with both parents. Fifteen percent were living with only one parent, and 3.5 percent were living with neither parent. Six percent of the children lived in households with their grandparents, although most of these were also with both parents. Another 6 percent lived in households which also included more distant relatives or nonrelatives. Overall, only about three-fourths of the children lived in the "normal" nonextended nuclear household. Nonnuclear household residents were more common for children living with only one parent than' for those with both parents. The average child in Walton County in 1870 shared his household with about four siblings, plus his parents. This, of course, varied with the stage of the parents' marriage and the birth order of the child. But, in most cases, the access to parental guidance and family resources was probably somewhat limited by the large number of siblings who had to share these resources. Single Adul ts The last group to be considered is that of single adults, of whom there were 93 in the 1870 sample. For this analysis, "adult" is defined to be a person of at least 20 years of age, since most would not have married before that age. Most

PAGE 152

131 were young, the median age being 24 for males and 22.5 for females, and just over half were male. Table 29 shows the detailed living arrangements of the single adults in white households in 1870. Most were living with primary relatives, about three-fourths with parents and another 14 percent with siblings. Only about 12 percent were living with nonrel ati ves . The norm, then, seems to have been for the adult child to remain in the parental household until marriage, after which a new household was usually established. More of the older, single adults were livingwith adult siblings, since parents were not always available. But by far the majority were living in family situations. Table 29. Living Arrangements of White Single Adults in Walton County, 1870 Living With Percent Parents Siblings Servants or boarders 69 13 11 74.2 14.0 11.8 Total 93 100.0

PAGE 153

132 Conclusion: The Life Cycle of the Individual The discussion of the household 1 ivi ng arrangements of the white residents of Walton County in 1870 is now complete. From the data concerning individuals at various stages of the life cycle, the typical life cycle of an i ndi vi dual can now be constructed, keeping in mind that there was variation among individuals at every stage. Most children were born into a two-parent household, growing up with a number of siblings, but, usually, not with other relatives. The only exception to this that was of some importance was for children to have an aged grandparent in the household, but this did not involve many children at any one time. It appears that most children remained in the household of their parents until marriage, which usually took place in their late twenties. Thus, many adult single individuals were living with other adult siblings in their parents' household. Upon marriage, the young man or woman moved out of the parental household, if the parents were alive and in good health, and established a separate household. At this stage, perhaps in order to help make ends meet, they may have taken a boarder or a relative into their new household.

PAGE 154

133 After the birth of the first child, the presence of boarders or rel ati ves wi thi n the young family's household was rare. With time, the family grew, at the rate of one child every two or three years, until seven to ten children were born and the wife was reaching the end of her reproductive years. Finally, the whole nucl ear f ami ly was living under one roof, with children ranging in age from early adulthood to young chi 1 dhood. One by one, the children married and left the parental household--but only rarely did they all leave before the death of a parent. The widowed parent usually had a few older children left in the household, who would help with household maintenance and protect the parent from isolation and neglect. Typically, one of the children, upon marriage, would remain in the parental household, probably inheriting the house and caring for the aged parent until his or her death. Since the age at marriage was high, this phase did not last long, relative to other stages--hence the paucity of three-generation or stem-family households in the sample. The child remaining in the household was not always the youngest,, since sometimes there were younger unmarried siblings of the married child in these families. Practically all older widows or widowers could expect to be cared for until death by one of their adult children.

PAGE 155

134 Structure of White Families, 1885 The overall summary of family structure among Walton County white families in 1885 is given in Table 30, separately for the county and the village. The county primary families were mostly nuclear families (63.5 percent), with about a sixth being three-generational or stem families. Another 12.5 percent were one-parent families, mostly older widowed parents with unmarried children. Thirteen percent of the county families were nonprimary, most of them being one-parent families or young couples. Of the total 120 families in the county sample, then, only 58.3 percent were nuclear, with about 15 percent each of one-parent and three-generational families. When compared to the 1870 white families, the 1885 county sample had more three-generational, and fewer nuclear, families. The Lake De Funiak distribution of family types is similar to that of the county, differing only in that relatively more young couples and fewer three-generational families were present. Almost 18 percent of the village families, moreover, were nonprimary, a higher percentage than, in the county sample. In the village, the one-parent families were more often not primary families than in the county.

PAGE 156

135

PAGE 157

136 As was done for the 1870 families, an attempt was made to simulate the developmental cycle of families by examining the household living arrangements of couples in various lifecycle stages. Since there is evidence of economic strain and certain unusual conditions in the 1885 households, the results of this analysis may not be as applicable to a description of the life cycle as were the data from 1870. It will still be useful to look at household composition at various stages of life. Young Couples There were only seven young (wife under 36) couples without children in the county sample, and three more in the village, in 1885. Only three of these 10 young couples were heading households, and only 2 lived alone in their households. One couple lived with nonrelati ves , three lived with parents, and four lived in households headed by a married sibling of one of the spouses. The norm of establishing separate households at marriage was apparently strained by the economic realities of the 1880s. Couples may have preferred living with siblings to staying with parents, or parents may not have been able to accommodate all of their children, since the married sibling extended household type was fairly common. This pattern is in sharp contrast to that

PAGE 158

137 of 1870, where sharing households for married siblings was almost unheard of. Seven young couples in the county and eight village couples had one small child each, having just begun the childbearing phase of marriage. More of these than of the previous group Here heading their own househol ds--l 2 of the 15 families. Only five of these were living alone in their households, with no other relatives or boarders. Of the 15 couples, 4 were living with siblings, 4 had unrelated boarders in their households, and 3 were sharing a household with parents. Since 1870, couples with one child had been forced to share households more often; although the increase in headship for this group over childless couples is similar to that of the earlier year. Also, the same preference for sibling housemates, as seen for childless couples, is apparent for these couples. The Childbearing Years The next stage of marriage is that of the rapidly expanding phase, in which the wife was younger than 45 years, and there were two or more children or one child over three years old. It was this stage which was the least likely to be extended and the most likely to head its own household. In 1885, the county sample contained 55 couples in this

PAGE 159

138 stage, 51 (92.7 percent) of whom were heading their own households. Only 36 percent, as compared to 80 percent in 1870, were living in nonextended or nonaugmented households. Fourteen of the primary families had extended kin within their househol ds--seven siblings and seven parents. Only one had an unrelated boarder. The four subfamilies in this group were living with parents (three couples) or a married sibling (one couple). The average number of children in these families was 4.5, about the same as the 1870 figure. In Lake De Funiak, there were 43 couples in the childbearing phase. Of these, 37 (86'percent) were heading households, but only 22 (59 percent of the primary families) were simple nuclear households. For this group, the average number of children was 3.5, somewhat smaller than that for the county sample. Of the 37 primary couples, 7 had relatives in the household--4 siblings, 2 parents, and 1 more distant relative. Eight households (21.6 percent) had boarders. There were also five of these couples in secondary families, two with their parents also in the secondary family. One subfamily in this stage lived with a married sibling. These village families were less likely than the rural ones to head households and, even more so , to T i ve only with their nuclear families. These couples were better off than the

PAGE 160

139 younger couples, however, only 26.5 percent of whom lived in simple nuclear households. The Childrearing Years There were 30 couples in the county sample in which the wife was at least 45 years of age, 25 (82.5 percent) of whom headed their own households. All but 3 of the 30 couples had children with them, the mean number for primary families with children being 4.1. Only 9 of the 30 were living in their own households with no extra-nuclear household residents. Four of the primary families had boarders and 12 had extended kin--mostly married children (three couples), grandchildren (five couples) or siblings (three couples). Only two of the couples in this stage were living in households headed by other rel ati ves--one with a parent and one with a married sibling. In addition, three of the couples were in secondary families living with nonrel ati ves . The Lake De Funiak population contained 19 couples in the older childrearing stage of the family cycle. All of these were primary f ami 1 ies , al though only 13 (69 percent) had any children in their households. For those who did have children, the average number was 4.8. Only nine (47.4 percent) of the couples lived alone with only their nuclear families. Three had extended kin (two parents and one married sibling) and seven had boarders in thei r househol ds .

PAGE 161

140' Thus, even though the proportion of households with boarders and extended kin was high in 1885, and higher in the village than in the rural county, the same kind of relationship of household complexity to the family cycle was observed as appeared in the 1870 sample. By 1885, however, the proportion of married siblings who lived together, as well as married children living with their parents, had increased substantially. Widowhood The county sample of 1885 contained 22 widowed individuals over 44 years of age, and the village had 20 more. Twenty-two of these (69 percent) were female, although most lived in male-headed households. Only about 31 percent of each sex headed households, most of the others living with their children. Considering, first of all, the 13 male widowers, 9 shared a household with at least one child or adult grandchild. Four of these widowed men lived with married children, five with unmarried children, three lived as unrelated boarders and one lived with unmarried grandchi 1 dren. Only two of the men lived alone with their unmarried children. Of the 29 female widows, 9 were heading households, of which 5 were nonextended. Two of the widow household heads

PAGE 162

141 had married sons in their househol ds , and two had siblings. Four also had boarders. Of the nonhousehold heads, 11 were living with married children, 5 with unmarried children in secondary families, 1 with a married sibling, and 3 with nieces or nephews. Overall, then, 13 of the 29 widows (44 percent) were living with married children, 6 were with unmarried children, and only 4 were with other relatives. Most widowed parents lived with their children or with one married child. In many of the families containing married children, a stem-family structure is apparent. The married child was usually a son, often an older son. Not all of the three-generation households were of this type, but enough of them were present to add weight to the argument that the stem pattern was important. At any rate, aged widowed parents were almost always cared for by their adult children or, in a few cases, by other relatives. The main change since 1870 in the situation of the widowed was a decline in the headship rate, which was true for married couples as well, and perhaps an increase in the number who were living with married children. This may partly be due to the lower marriage ages of children, as well as to an increase in the difficulty of establishing and maintaining separate households.

PAGE 163

142 As was done for the 1870 sample, before discussing the implications of these findings for the life cycle of the individual, two more groups of individuals must be described: the children and the single adults. The Children In the 1885 sample of county white households, there were 300 children below 15 years of age, 86.3 percent of whom were the sons and daughters of household heads. Most (21) of the others were grandchildren of the household head, and there were a few who were younger siblings or nephews or nieces of the head or the head's wife. Only six children in the county sample were in secondary families. In Lake De Funiak, there were 205 children, 164 or 80 percent being children of household heads. Of the remaining 41 children, 24 were living in. secondary families. Only two of the village children (1 percent) were grandchildren of the household head, and five each were siblings or nephews/ nieces of the head or spouse. The major difference between the children of the village and the county sample was in the greater proportion of the latter who were grandchildren of heads, and the greater proportion of village children living in secondary families. About 93 percent of both the county and the village children were living i n mal e-headed househol ds .

PAGE 164

143 As with the 1870 data, the children in 1885 were classified accordi ng to their household living arrangements, regardless of whether they were part of primary, secondary, or subfamilies. As Table 31 indicates, 255 (85.3 percent) of all children in the county sample were living with both parents in 1885. About 10 percent were living with only one parent, and 5 percent were living with neither parent. Of those living with both parents, about one-third were living with grandparents or other relatives or nonrelatives in addition. Only 58.5 percent of the entire group of children were living with both parents in a simple nuclear household. Table 31. Living Arrangements of Whi te Chi 1 dren in Walton County, 1885

PAGE 165

144 The Lake De Funiak children were more often living with both parents than were the county children. About 92 percent of the village children had both parents present, 6.6 percent had only one parent, and only 1.5 percent had neither parent in the household. In addition, more of the village children had nonlineal relatives or boarders present, as 21 percent of the two-parent children were living with parents plus others (not grandparents). For the county children, the corresponding percentage was 12.2. About one-third of all children in Lake De Funiak were living with nonnuclear family members, as were 36 percent of the county chi 1 dren . Overall, then, the proportion of children in nonextended nuclear family households had dropped since 1870, from 75.5 percent to an estimated 60 percent of the county ' s chi 1 dren (Lake De Funiak included). The drop is seen more clearly in the county sample, where more children were living with grandparents in 1885. In the village, children were living with other kinds of relatives or with nonrelatives to a great erextent. Single Adults There were 135 single adults in the 1885 samples, 89 in the county and 46 in the village. The sex ratio for these

PAGE 166

145 singles was high in the village (261:100), but not so high in the county (.118:100). One would expect some imbalance due to the fact that women married earlier than men. Most of the single adults were young, the median ages being in the mid-twenties for men and late twenties for women. In the county sample, about 56 percent of these adults were living with parents, and another 26 percent were living with siblings (Table 32). In all, 86.5 percent were living with relatives and 13.5 percent with nonrel ati ves . In Lake De Funiak, only 45.6 percent of the young single adults were living with their parents. Another 26 percent were living with siblings, and 2 percent with other relatives. Fully 26 percent of the village single adults were living with nonrel ati ves , some as servants and some as boarders. These figures are quite different from the corresponding ones for 1870 when three-fourths of all single adults lived with parents and 14 percent more lived with siblings. It may be that the hardship of the times forced young men to leave their parents' households to find work in the village or in another household. Many of the boarders at this time were laborers or apprentices to the household heads. Some of them became servants in other households. And some of the young single men in the village were recent migrants

PAGE 167

146 Table 32. Living Arrangements of Single White Adults in Walton County Samples, 1885 Living With County Percent Lake De Funiak Percent Parents Siblings Other relatives Nonrelatives 50 23 4 12 56.2

PAGE 168

147 difficult times instead of deeply rooted changes in norms governing family residence. It is interesting to see how these adaptations were made and which points of the life cycle were most strongly affected. A child born in Walton County in the 1880s, if it was an early child, was quite likely to be born into a household which contained persons other than the parents and siblings. Later-born children more often lived in simple nuclear families; and, in fact, almost two-thirds of all children at any time were living in simple nuclear households. Still, a significant portion lived with grandparents or other relatives or nonrelatives in the household. The male child, upon reaching maturity, was not as able as in 1870 to remain at home until marriage. Many young men had to leave the parental home, some living with siblings and others with nonrelatives, in order to find work and to ease the burden on their parents. And marriage, unlike that in 1870, did not necessarily mean the establishment of a separate household. More often, young couples lived with adult siblings, with parents, or with nonrelated families for a few years. Those couples in the middle stages of marriage, with growing families, were more likely to head their own households and to have simple nuclear households than others.

PAGE 169

148 but this also was not as true as it had been in 1870. Older couples and widows, as in 1870, were more often in extended or augmented households, but they were not as often heading their own households in 1885. Although the same basic underlying norms may have existed in 1885 as in 1870, the external situation was different, and different priorities thus came into play in determining household living arrangements. Households in 1885 became shared, mainly by adult siblings and by parents and married children, in response to the difficulties of the times Younger single siblings, also, were often taken into a married sibling's household, thus relieving the burden on the parental household. In these ways, kin of primary degree tightened their supportive network, breaking with the normal patterns of development in response to an ongoing crisis situation. As soon as the economic difficulties were over, it may be that those families which had managed to remain together during the crisis returned to the former patterns. Structure of Black Families, 1870 Since the questions concerning black family structure as outlined in Chapter III are not the same as those about white families, the discussion of black families in this section

PAGE 170

149 will not follovM the same format as that used for white families. Where comparisons are important, however, white family data will be presented, and part of the discussion follows the white family structure presentation. The black population of Walton County in 1870 contained 74 total family groups, which were classified according to family type (Table 33). A relatively large percentage of the family groups were not primary families, but lived instead in other families' households. This nonprimary group, of course, includes those residing in white households as well as those living with other black families. The percentage of all black families that were primary families was 63.5, as compared to 92.5 for the white population. There is a substantial difference between the structure of the primary families and the nonprimary families. If one were to consider only the primary families, then the difference between the black and white families would not be nearly so substantial. The racial difference becomes quite marked, however, when one considers al 1 family groups. Although black primary families were mostly nuclear families or couples, nonprimary families were more often one-parent families. Of the total families in the black population, only 40 percent were nuclear, with another 12.2

PAGE 171

150 percent being married couples without children, compared to 64.2 percent nuclear and 7.5 percent couples for white families. Fully 31 percent of the b lack fami 1 ies in 1870 were one-parent families, most of these being young mothers with their children. This last figure, especially, points to a high degree of instability within black marital and family relationships. It is even more striking when one realizes that most of the white one-parent families were older parents with adult children, and not young mother-child families. In addition, the three families in the "othe r" category of family structure were families which had stepchildren, none of which were found in the white sample. Table 33. Structure of Black Families in Walton County, 1870

PAGE 172

151 Another indication of stability which is relevant here is the proportion of the adult population which is not living with any relatives. The table of family structural types, of course, only considered that portion of the population that was in a family situation. But in 1870, 29 percent of the black males above age 14 and 12 percent of black females of that age were not living in families. Both of these figures are more than three times as high as the corresponding percentages of white adults living without relatives (Table 34). Table 34. Percentage of Adults Not Living in Family Groups, 1870, by Race Sex

PAGE 173

152 Two-parent Families Of 29 married couples in which the wife was 35 years of age or under, 6 had no children (20.7 percent) and 9 (31.0 percent) had only one child. Over half of the black couples in this age range, then, had less than two children. This, too, may be a sign of marital instability over time, since long stable marriages in the T9th century almost always gave rise to a large number of children. The corresponding percentage of white couples in that age range who had fewer than two children was 34.8. Of these 29 young married couples, 13 (44.8 percent) lived alone or with just their children in their own households. Almost as many (11, or 37.9 percent) either had boarders in their households or lived as secondary families in other households. In addition, seven of these families were living with relatives. It seems that economic independence, then, was difficult for these young couples. Doubling up within households was common, with nonrelatives as often as with kin. On the other hand, 21 of the 29 couples (72 percent) were heading their own households, even though many of them contained boarders or relatives. There were 15 couples in the black population of 1870 in which the wife was older than 35. All of these couples headed their own households, and only three had boarders.

PAGE 174

153 In addition, one couple had a grandchild in their household, but none of the others had other relatives. This group contains some of the more stable marital pairs, since some of the couples (7 of the 15) had more than five children. The other eight had three or fewer chi 1 dren , but thi s may represent in part the fact that young black teenagers or adults left home more often than did whites, who usually remained at home until marriage. Widowed and Broken Families One-parent families made up 31 percent of black families in Walton County in 1870. There were 26 family groups that contained a single parent and children. Nine of these one-parent families were young mothers, 35 years of age or less, with their children. This group averaged three children each, and all but one lived as secondary families in others' households. Six of these were in white households. Although this probably does not mean much, it is interesting to note that the average number of children for these mothers was actually higher than the average for married women of the same age group (2.4). Another nine mother-child families were older women, age 36 or above, with their children. Most of the women in this group were in their fifties or sixties; hence they

PAGE 175

154 might be "true widows." Only two of these women headed households, three others living in their adult unmarried children's households. The other four lived as secondary families, three in white households. In addition to the 18 mother-child families, there were 8 father-child families in Walton's black population. This was quite surprising, in view of the notion that fathershad little place in the black families of the day. However, at least three of these fathers were elderly dependents in the households of adult unmarried children. Two of the others had several children, including some who were quite young. In these cases, the mother had perhaps only recently died, since the family seems to be intact except for her absence. The other three fathers had one child each--two lived alone with the child and one father and child formed a secondary family. It cannot be flatly asserted, therefore, that fatherhood counted for nothing in black families of the 19th century. Nevertheless, there were definitely more mothers than fathers living with their dependent children. Overall, it can be seen that black families, and particularly young black families, had many probl ems--both structurally and economically. Young couples more often than not were forced to share households with others; and

PAGE 176

155 a great many broken families existed, most of which could not head their own households. Of all young mothers aged 35 or below, only 71.8 percent (23 of 32) were married. Of older mothers living with their children, only 57.1 percent were married (12 of 21). On top of this, a high proportion of black adults was not residing in any family group. The interpretations of Du Bois and Frazier concerning the basic instability and weakness of the black family of this period, therefore, find considerable support in these data. Other Tests of Black Family Strengths and Stability In order to further test the ideas of Frazier and Du Bois regarding the Southern black family of this time, certain additional analyses were made. First, slave schedules from. census of 1860 were examined and analyzed. Since these schedules did not contain the names or relationships among slaves, but simply listed the sex and age of each slave under the owner's name, a rigorous classification of family types was not possible. The schedules did contain data on the number of slave houses, and this item proved useful in classifying family types. For example, if there were only one house and the slave group consisted of two adults of opposite sex and small children, it seems reasonable to

PAGE 177

156 assume that this was a nuclear family. If there had been two houses, one could not be sure if it was a nuclear group, or a one-parent family and a single adult. By looking at the relative ages, the sexes, and the number of houses available, it was possible to establish the "probable family type" for a number of slave groups. In practice, where the number of slaves was \fery large, the classification was impossible. It should be noted, however, that the classification was very conservative, in that the case had to be very clear before a probable family type was assigned. Table 35 shows the classification finally arrived at after considering each slaveholder's listing, for all slaves in both Walton and Holmes Counties. The left-hand part of the table shows the original classification of all slave groups; that on the right is a subtable of the specific single-family groups whose structural type was fairly certain. Of the original 128 slaveowners' listings, only 36 were in this final group, 40 having owned only one slave each, and the others owning either plural families or groups whose structure could not be determined by this procedure. Of the subset of families which were cl assi f i ed , 23 , or 63.9 percent, were one-parent families. Nine of the 36 (25 percent) were nuclear families and two each were

PAGE 178

157 Table 35. Probable Family Structure of Slaves of Walton and Holmes Counties, 1860 Classified single families only N Percent 23 63.9 2 5.6 9 25.0 2 5.6 36 100.1 Family Type

PAGE 179

158 A second test involving the same data was devised. For some family groups it was not possible to establish the precise structure of the family, but it was possible to assign the probable headship of the family group. This made it possible to calculate the proportions of female and male family heads among those families for which such assignment was possible. The results of this operation are presented in Table 36. As before, a large proportion of cases could not be assigned, both because of the lack of families in some cases and the uncertainty of headship in others. For those 48 families whose heads could be determined with a high degree of certainty, 31, or 64.6 percent, were female. Table 36. Sex of Probable Heads of Slave Families in Walton and Holmes Counties, 1860 Sex Total

PAGE 180

159 Even though these measures which utilize slave data would not be good enough for strong inferences if used alone, they do strongly support the other data presented, all of which point to a great deal of family instability and female dominance among the black families of Walton County. Living Arrangements of Children One important result of the high proportion of femaleheaded or one-parent families in the black population of 1870 is that many children were growing up without the presence of both parents. There were 161 children age 14 and under in the black population of Walton County in 1870, about three-fourths of whom (120 children) were living in black households. The other 41 were living in white households. The family living arrangements of these children were classified and are presented in Table 37. Of the 120 children in black households, 70.8 percent were living with both parents and 27.5 percent, with one parent. Another 1.7 percent were living with neither parent. Of this group of children, 61.7 percent lived in the "normal" two-parent, nonextended household. Some children lived with grandparents in addition to their parent or parents (12.5 percent of the black household children).

PAGE 181

160 Table 37. Living Arrangements of Black Children in Walton County, 1870

PAGE 182

161 The 41 black children in white households, as would be expected, were in relatively worse family situations than those in black households. Of these, only 19.5 percent were living with both parents, 39 percent were living with one parent, and 41.5 percent were living without either parent present. When the two black groups are combined, only 57.8 percent of all 161 black children in the county in 1870 were living with both parents; and only 46 percent were living with both parents and no other relatives or nonrelatives in the household. Almost a third of all black children were living with only one parent, and 12 percent more were living apart from both parents. These figures represent a \/ery high proportion of children who were living apart from the two -pa rent household situation. Of those children who were living with one parent, most were living with their mothers (17 of 21). Most of these were in secondary families, although three of the four fathers headed their own households. Even though the percentage of children outside of twoparent families was \jery high, the model of the two-parent family and male-headed household was available to most of them. About 80 percent of the children in black households

PAGE 183

162 and most of those in white households lived in househol ds headed by a married couple. And 88 percent of the black household children lived in male-headed households. It may have been that values associated with nuclear family households were developing, even though a large number of fami 1 ies were not able to practice them. Living Arrangements of the Aged One final test of the female-dominance aspect of the theories of Du Bois and Frazier was developed. It was hypothesized that, in accord with the matriarchal hypothesis, more aged women than men would be living with their adult children. This argument rests on the assumption that female family ties were stronger than male ties, in many cases. This is not to say that male ties were never strong, but simply that more females than males had strong family ties. Table 38 includes all aged blacks from both the Walton and Holmes schedules of population, for both 1870 and 1885. Sincethere we re not many aged blacks, it was necessary to ^combine the different years and counties in order to have a large enough graup to be able to study. For comparison, the aged whites from a special sample taken for this purpose are included in the table for both 1870 and 1885.

PAGE 184

163 Table 38. Living Arrangements of the Aged, by Sex and Race, 1870 and 1885

PAGE 185

164 Conclusion: Black Families in 1870 The evidence in this section strongly supports the Du Bois-Frazier thesis of black family instability and female dominance in this late 19th-century county. Although Walton County was a farming, rather than a plantation, county, it would seem reasonable to assume that conditions in such a context might have been more favorable than on plantations. (See Smith's discussion, 1953, 301-319.) At any rate, it seems clear that in this fairly typical county, black families had not developed strong nuclear patterns of residence. Since the heads of black households, who might be considered to be leaders in the community, were more often in twoparent nuclear families, it may be that the two-parent, stable family was an i deal which was accepted but not attained by most of the population. In 1870, the paternalistic relationship between blacks and whites which had been the basis of the slavery period had not disappeared. This paternalism might have prevented the development of strong families among black farm workers, since protection and sustenance came from the master's family, not from the independent slave family. Under such conditions, there was no need for a strong black family.

PAGE 186

165 and such ties might have interfered with the smooth workings of the master-slave relationship. Whatever the basic causes, the black family of Walton County in 1870 was not a strong, stable, conjugal family system. There is little evidence of strong kinship ties either within marriages or among wider circles of kinship, except for the case of caring for aged mothers by adult children. This, too, supports the traditional interpretations of black family history, rather than the more recent ones. Structure of Bl ack Fami 1 ies , 1885 In 1885, there were 118 black families, 95 of whom were primary families. The percentage of families heading their own households, then, was 80.5, up from 63.5 percent in 1870. Not only is an improvement seen in the rate of household headship, however. In several other ways, the condition of black families seems to have improved markedly over the 15-year interval. First, the proportion Of one-parent families, almost a third of all families in 1870, had dropped to 13.6 percent by 1885 (Table 39). And the percentage of nuclear families had increased from 40.5 to 61 percent. In addition, the

PAGE 187

166 percentage of married couples had risen slightly. Overall, the structure of black families in 1885 was not greatly different from that of whites in the same year, especially thevillagewhites. Table 39. Structure of Black Families in Walton County, 1885

PAGE 188

167 Young Couples There were 16 young childless black couples in Walton's 1885 population, 5 of whom headed households and 3 of whom lived in simple nuclear households. Two of the households of the primary couples were augmented with boarders. In addition to these couples, four of the young childless couples were secondary families and seven were subfamilies: six with married siblings and one with parents. Eight more couples had only one small child each, four of these couples heading households. Only two of these were living alone with no extended relatives or boarders, the other two primary couples having married siblings in their households. Of the three subfamilies in this group, two were living with parents and one with a married sibling. One young couple with a child, in addition, was living with nonrelatives. Of these 24 young couples, only 9 (37.5 percent) were heading households. Nine were living in households shared by married siblings, and only three were living witha parent or parents. In addition, five were living alone in their households and seven were living with nonrel ati ves . Thus, even though separate household residence was not possible for most of these couples, most were not living

PAGE 189

168 with their parents, seemingly preferring siblings or nonrelatives as housemates. It seems that young couples were trying to establish neolocal residence and independence from their parents, even when that meant sharing a household with a married sibling or a nonrelative. This same phenomenon was seen for the white couples for the same year. The Childbearing Years There were 57 young couples, with the wife under 45, who had more than one child or whose only child was older than three years. Fully 54 (94.7 percent) of these couples were heading their own households, 38 of these being simple nuclear households. The average number of children for these couples was 4.6, almost identical to that for white couples in the same stage of development. Five of the primary families in this group had boarders present, and 11 had other kin--mostly married siblings of one of the spouses. In addition, two families were living in secondary families and another was living as a subfamily with a married sibling. Compared to the younger black couples, this group, who had probably been married longer and who were slightly older, was much more often h'eading households and less often

PAGE 190

169 living in extended or augmented households. Since such a large number of families were in this stage of development, it is apparent that marriages were lasting longer in the 1880s than they had in the 1860s. The Childrearing Years In Walton County's black population of 1885, there were 18 couples in the older group (wife older than 44 years). Fifteen of these were primary families. All but 2 of the 18 couples in this stage of the family cycle were living with at least one child, the average number being 4.2 for primary families with children. Seven of the primary fami 1 ies had other relatives living with them--four had married children or grandchildren, and one each had a married sibling, a parent, and a niece. One of these households had a boarder. In addition, there were two secondary families among the older couples, and one subfamily living with a parent. While the headship rate for older couples was still fairly high, the proportion of families living alone in their households was lower than that for growing families. Only 7 of the 18 families were in simple nuclear households (38.9 percent), compared to 66.7 percent of the families inthechildbearingyears.

PAGE 191

170 Mi dowhood There were 18 widowed adults over 44 years of age in the 1885 black population, 4 males and 14 females. All of the males, and half of the females, were heading households All 18 of the widowed shared households with at least one child. Two of the males and six of the females were living with only their married children. One of the males and seven of the females were living with unmarried children. In addition, one male was living with a niece, and one female was with her unmarried children in a secondary family. Older widowed individuals, then ,' usual ly lived with their adul t chi Tdren. The description of the living arrangements of families throughout the life cycle is now complete for the 1885 population. In order to piece together the life cycle of individuals and to discuss the two major groups left out by this approach, the discussion now turns to the children and the single adul ts . The Children The black households of Walton County in 1885 contained 296 children, 88.5 percent of whom were the children of household heads. About 7 percent were either siblings or

PAGE 192

171 nephews/nieces of the head, and a few were grandchi 1 dren or subfamily children. Most (80 percent) of the children were living in households headed by a complete nuclear family, and most (93. 6 percent) were in male-headed househol ds . The household living arrangements of the black children in 1885 are summarized in Table 40. There was a significantly greater proportion of children in 1885 who were living with both parents than in 1870. In 1885, 89.0 percent were living with both parents, 8.0 percent were living with one parent, and only 3.0 percent with neither parent. Table 40. Living Arrangements of Black Children in Walton County, 1885 Living With Both parents Nucl ear only Plus grandparents Plus others One parent Nuclear only Plus grandparents PI us others Neither parent Grandparents Others N 268 196 26 46 24 12 4 Percent 89.0 65.1 8.6 15.3 8.0 4.0 1.3 2.7 3.0 1 .3 1.7 Total 301 100.0

PAGE 193

172 A fairly high percentage of those living with both parents, however, had non-nuclear household members, accounting for about 24 percent of all children. In all, about 65 percent of the black chi 1 dren were 1 i vi ng in simple nuclear households. These figures represent quite a change from the 1870 situation, in which only 57.8 percent of the black children lived with both parents. This is an increase of over 30 percent in 15 years. The drop in the percentage of children with one parent is especially important (from 30.4 to 8.0 percent); and the percentage' of children with no parent in the household had dropped by almost nine percentage points, from 11.8 to 3.0 percent. Single Adults The census schedules for 1885 contain some 69 black adults of at least 20 years of age who were single. They were quite young, the median ages for males and females being 24 and 25, respectively. Some of the males may have actually been divorced, separated, or widowed, since about 19 percent of them were over 35 years of age, and none of the females were that old. There were over twice as many single males as single females, the sex ratio for this group being 264:100.

PAGE 194

173 The living arrangements of these single adults are presented in Table 41. Many of the black singles were living as servants in white households, this group comprising about one-third of all single black adults. Of the entire population of single adults, only one-third were living with parents and over half were living with nonrelati ves . The blacks were much more often living with no relatives than were the white singles. One might surmise that black families did not keep their young adults at home even as often as did the whites. Instead , si ngl e adults were forced to leave the parental home to seek employment. The whites, at least in the county sample, were more likely to stay within the parental or a sibling home, at least until marriage. Table 41. Living Arrangements of Black Single Adults in Walton County, 1885

PAGE 195

174 About 23 percent of the black single adults in 1885 lived in nonfamily household settings, such as boarding houses and hotels. Most of the blacks in these households were service-workers, in contrast to the whites, who were boarders. Since most of the servants in 1885 were young single adults, it seems that being a servant was simply one phase of the life cycle for many blacks. When they came of age, they left the parental home to become servants for a few years, after which they married and set up homes, usually with siblings or other relatives. By 1885, there did not seem to be a real servant "class" among blacks, but simply young adults who were temporarily serving in white households. Co nclusion: Black Family Structure and the Life Cycle If the structure of black families in 1885 represents a fairly stable condition, some aspects of the life cycle of individuals can be extrapolated from its description. Caution should be taken, however, since even the white family cycle was strained under the special circumstances of the 1880s. But some generalizations are possible.

PAGE 196

17 5 Most black children in 1885 were born into two-parent families, although, more often than not, they were sharing their households with others besides their parents, if they were early children. As the family expanded, the child was more likely to live alone with its parents and siblings; but ^ery often there was an aunt and uncle, cousins, or a grandparent or others in the household. Since houses were rather small in those days, there was probably 1 i ttl e pri vacy or room for oneself within the everyday experience of either the children or adults. To a greater extent than white children, black children did not remain in the parental household until marriage. Over half of the black single adults lived with nonrel ati ves , many of them serving in white households. While 82 percent of white single adults in the county sample, and 71 percent of those in the village, were living either with a parent or sibling, only 45 percent of the black singles were. It was probably economic necessity that forced young single adults to leave their parents' homes and find work. For females, it was more common to remain at home until marriage; but marriage came at rather early ages for them. Marriage occurred for most blacks in the early to midtwenties. This did not mean, for most, the establishment

PAGE 197

176 of a new nuclear household. Sometimes adult siblings lived together, sometimes a couple lived with a parent or parents, others lived as secondary families with nonrel ati ves ; but most did not have their own households. Most of the couples with growing families in 1885 did have their own households, although most of these were shared with other relatives or boarders. It is not certain whether the better condition of these families was a function of the family cycle, in that it took a family a few years to be able to establish a separate household; or due to the fact that they probably had obtained their houses before the economic conditions grew worse. At any rate, as families grew, they may have been better able to establish their own households and were not quite as often living with extra-nuclear housemates. Older couples, however, were likely to have married children or other relatives in their households. None in this study lived alone. After the death of the first spouse, the widow probably continued to live in the same household, with her unmarried children or a married child. Since only one married child typically lived with a parent, the other children had to find other households when they married. In the difficult times of the 1880s, this often meant

PAGE 198

177 sharing a household with a sibling or an unrelated family. For the individual who remained with his or her children until old age and widowhood, at least one of the children would remain with the aged parent and care for him/her until death. There were a few older single men in Walton County's black population, some of whom may have been widowed or divorced. Some of these were living as boarders with nonrelatives. Older men were sometimes living without families, then; but all of the older females were living with their children. Changes in Black Family Structure, 1870-1885 What can be said about the changes in black family structure between 1870 and 1885? Clearly, important changes took place, mostly in the direction of stronger families. A much higher proportion of blacks in 1885 were living in family groups than had been the case in 1870, and the percentage of broken or one-parent families was much lower in the later year. Not only that, but many more children in 1885 were living with both parents. In many ways, the black families of 1885 were very similar to the white families of the same year.

PAGE 199

178 The family status of blacks in Walton County, then, had improved markedly over the 1870 situation. At the same time, economic circumstances had become worse, as had social relations with whites. The effect of the economic decline as well as the social hostility was to force blacks and whites to double up in housing, bringing siblings and other relatives and nonrelatives together within the same households. The effect of the rising hostility may have, in its own way, strengthened the black family. For one thing, blacks were not able to live as dependent laborers within white households in 1885 as they had done in 1870. Recall that it was this white-household group which had had the weakest family ties in 1870. By 1885, blacks had to rely upon themsel ves for housing, protection, and general welfare. This may have forced some marriages to stick together longer than may have been true earlier, because now marriage was an economic partnership as well as a social companionship. Second, the increased hostility from the whites, after the period of adjustment to the new political and labor system, forced the black community to increase its internal cohesion as a defense against this sometimes violent antipathy. The two insti tutions whi ch the black community had which were truly under their own control , the family and

PAGE 200

1 79 religion, were the instruments by which this strength in the face of adversity was accompl i shed. Finally, the norms of the two-parent, nuclear familyhad probably been the ideal norms of the blacks for some decades. The black community leaders, even in 1870, had such families. And most of these people had lived with this type of family system, even though unable to practice it, for their entire lives. It simply took time and necessity for the adjustment to be made after the upheavals of the war, emancipation, and Reconstruction. Under the paternalistic rule of slavery, the blacks had not developed strong community feelings, nor had much independence or initiative been tolerated. Now they were forced to become independent, with strong community feelings developing in response to the white hostility; and a stronger discipline arose among black families, because it was i nternal , than had existed before, when it had been externally imposed. Overall, the black family in Walton County in 1885 was much like the white family, although it was in worse economic condition. The strides which had been made in the 15 years since 1870 toward bringing the entire black population into the family system were truly remarkable.

PAGE 201

180 Household Composition and the Family Cycle Before this chapter is brought to a close, a few topics which have been touched upon briefly within the previous discussion will be considered more systematically. Although it has been of interest to family researchers to analyze the relationship between family development and household complexity (Hareven, 1974; Berkner, 1973b), few studies have actually presented data on this problem. In this section, the relationship of the presence of boarders, extended relatives, and other household residents to the family cycle is discussed. Only households headed by married couples are included in this discussion, because such a large percentage of widowed persons did not head their own households. In general, their situation is similar to that of the older married couples. Table 42 shows the relationship between the family cycle and household composition, for white families in 1870 and for black and white families in 1885. The black families of 1870 were not included because of the evidence of a high degree of instability in their family relationships. In the 1885 data, there were almost no differences between the black and white relationships, and they were therefore

PAGE 202

181 included in the same table. In that year, also, the number of young couples who were heading their own households was small, so that the first two stages of the life cycle had tob e collapsed intoone. Table 42. Household Composition by Stage of Family Life Cycle 1870 (White Families Only) Percent Percent Stage of Simple Percent with Total Development Nuclear Extended Boarders N Percent Young childless couples 61.5 7.7 30.8 13 100.0 Young couplesone child 93.3 6.7 0.0 15 100.0 Childbearing stage 81.8 4.5 15.9 44 a Childrearing stage 55.5 18.5 29.6 27 a 1885 (White plus Black Households) Percent Percent Stage of Simple Percent with Total Development Nuclear Extended Boarders N Perce nt Young couples 56.2 25.0 18.8 16 100.0 Childbearing stage 70.5 23.6 5.8 105 99.9 Childrearing stage 40.0 47.5 12.8 40 100.0 ^Percentages do not sum to 100 because some households had both extensions and boarders.

PAGE 203

182 The percentage of households containing only simple nuclear families was fairly high for young childless couples, but much higher for young couples with growi ng f ami 1 i es . About 93 percent of the young couples with one child, and 82 percent of young couples with more than one child in 1870 were in simple nuclear households. For those in the late childbearing phase, the percentage of simple nuclear households had dropped considerably. A corresponding rise in the percentage of households with extended kin and boarders is seen at that stage. The proportion of extended households was low for all of the younger couples. A fairly high proportion of younger chi 1 dl ess couples, however, had boarders. The general impressions one gets from looking at the 1870 data are confirmed by the 1885 data. In that year, there were not enough young couples who were able to head their own households to allow this kind of analysis, except when black and white households were combined. This combination did not obscure the relationship, however, since it appears to be the same for the two racial groups. In the combined table, the percentage of young couples' households which were simple nuclear households was 56 percent. This percentage rose to 70 for the childbearing

PAGE 204

183 couples, and fell again to 40 percent for the older childrearing couples. The younger couples' households were more likely to have boarders in their households, while the older childbearing couples' households were more likely to be extended. It appears that young couples were attempting to establish independence from their families during the early years of marriage. This is especially clear in 1870, when most of them were able to live alone. Even in the later year, however, young couples were more likely to have nonrelatives, rather than extended ki n , in their households, if they were not able to live alone. The variation in the later years, between growing families and older families in the major childrearing phase, may be explained by the fact that growing families need room within their households for expansion. While new children were still appearing periodically, families were not very likely to open their homes to extra kin or nonrel ati ves . After the last child had been born, and after children began to leave the parental household, the household had room for extra members. At the same time, older couples often needed younger relatives in their households to help with the work as they became older. And three-generation households, if

PAGE 205

184 at all common, were usually limited to those families in the later chi 1 dreari ng years, since those in the middle ages were not old enough to have married children. Summary In this chapter, the family structure of the white and black populations of Walton County in 1870 and in 1885 have been described, with emphasis given to the changes in family structure as the life cycle progresses. It has been seen that white families in 1870 were predominantly nuclear and nonextended, but that this was partly due to the fact that most of them were in the childbearing years of family development. From the family cycle analysis, it appears that almost all families became extended as the parents grew older and were widowed, if they lived long enough for their children to marry. Black families in 1870 showed evidence of unstable marital relationships and female dominance as hypothesized by Frazier and Du Bois. Even though there was a fairly large and important group of two-parent families, the percentage of one-parent families and adults not living in family groups was very high.

PAGE 206

185 By 1885, both the black and white families had changed, partly in response to economic pressures. The recession had led to a much higher percentage of extended households among both whites and blacks in 1885 than had been the case in 1870. But black families had, seemingly, strengthened their inner ties, so that the percentage of two-parent nuclear families was about the same as for whites. Even though the percentage of households which were extended in the latter year was higher than that in 1870, it was still the older couples' households which were the most often extended. The relationship of the presence of relatives and/or boarders to the developmental cycle was consistent for both points in time.

PAGE 207

CHAPTER VII FAMILY AND HOUSEHOLD CHARACTERISTICS AND SOCIAL STRUCTURE In this chapter, selected aspects of social structure will be related to certain characteristics of households in the samples and populations studied. Before these analyses are presented, however, the household characteristics of the black and white samples of 1870 and 1885 will be described. This first section will emphasize the changes that took place in the household characteristics of Walton County between 1870 and 1885. Selected Characteristics of Black and White Households, 1870 and 1885 In another chapter, the changes during this period in the structure of families were noted. Here, data will be presented on changes in household size, composition, and the characteristics of household heads. 186

PAGE 208

187 Extended Ki n In 1870, 13 percent of the white households and 8.9 percent of the black households contained non-nuclear relatives of the household head (Table 43). By 1885, the proportions of such households had risen sharply to 35.5 percent of white households and 28.9 percent of black households. In both years, slightly more white households than black households had other relatives living with them, but both groups had been affected by circumstances that developed between 1870 and 1885 which increased the number of such households. The increases are probably an effect of economic hardship, which made it more difficult to establish and maintain separate households. This becomes more apparent when the types of relatives living with families are examined for the different years. In the 1870 white sample, most of the relatives were widowed parents or single siblings of one of the spouses, and none were married children. By 1885, there was an increase in the number of married siblings and married children in subfamilies. Thus, the norm of neolocality became strained by the economic circumstances of the period. This same pattern was observed for the black households. Most were primary relatives of one of the primary couple, representing the overlap, within households, of two nuclear family groups.

PAGE 209

Table 43. Percentages of Households Containing Relatives and Boarders, by Race, 1870 and 1885

PAGE 210

189 was changing from one of single-family homes housing one or two boarders to the development of houses specifically designed for renting rooms to boarders. The changes in the proportions of households with boarders can be better explained after examining data on the boarders themselves. The typical boarder in 1870 was not the same as the typical boarder in 1885. In 1870, practically all of the single boarders were young farm laborers or apprentices to the household head. They were probably wage-laborers, and part of the wage included room and board. In this situation, a family did not "take in a lodger," but hired a laborer and provided board. In 1885, most of the farm-laborer boarders had disappeared. Now many boarders lived in Lake De Funiak, some in boarding houses or in the hotel. Many of these people were from non-Southern states , only recently come to the new village. In contrast to the 187Q boarders, many of these boarders were professionals or skilled laborers, 41 percent coming from the higher three social classes. These boarders were not primarily laborers for the household head (although some were apprentices), but were modern, rent-paying tenants. Of course, not all of the boarders were of this type. Some

PAGE 211

190 in both years were dependent families without kin in the community; and some in 1885 were laborers. But farmlaborer boarders had mostly disappeared because farmers could no longer afford them. Since the table as presented for 1885 is based on a weighted average of the county as a whole, and since there were more rural families than village families, the decline in farm laborer boarders explains part of the change in the proportion of white households with boarders. At the same time, the proportion of households containing relatives (Table 43) increased, both for black and white households. It appears that, as conditions grew worse, households took in more of their kinsmen, forcing boarders to live in fewer households with other boarders. As a result, the total number of boarders in 1885 was larger than that of 1870, but the number of households containing them was smaller. Household Size The average household size was consistently higher for white households than for black ones. The difference was more marked, however, in 1870, when the mean household sizes were 6.3 for white households and 5.6 for black households (standard deviation, 2.8 for both). By 1885, both races had

PAGE 212

191 slightly larger households, with an average of 6.8 for the whites and 6.5 for the blacks. Some of the increase for the white households was due to more extended relatives in the households, while most of the increase for whites and blacks was due to a larger proportion of families with children. Number of Children The mean number of children per household containing children declined slightly for white households and increased for blacks, between 1870 and 1885. In 1870, white households which had children contained an average of 4.3 children, as compared to 3.7 for the black households. By 1885, both black and white households with children contained an average of four children each. Black households had a slightly lower percentage contai ni ng chi 1 dren in 1870 (79 percent versus 85 percent), but the difference had all but disappeared by 1885, when 90 percent of the black households and 92 percent of the white households had children of the household head. Marital Status of Household Heads In both 1870 and 1885, white households were more likely to be headed by married couples than were black households.

PAGE 213

192 but the differences were not great in the later year. The specific percentages of households headed by married couples were 81.5 percent of white and 72.9 percent of black households in 1870. By 1885, the figures were 92.6 percent of white households and 89.9 percent of black households. The rise in the proportion of households headed by married couples in part reflects the fact that a larger proportion of adults were married in the latter year. It may also have been more difficult for young unmarried adults, as well as for older widowed ones, to establish or maintain households alone in the economically difficult 1880s. The increase in extended households resulting from these difficulties would tend to increase the proportion of married household heads, because it is usually the unmarried who move in with married couples, and not vice versa. Sex of Househol d Heads Since most women were not economically independent, very few women headed households, even when they headed families. In 1870, the percentage of male-headed households in the white sample was 82.5, while that for the black population was 77.8 By 1885, the percentages of male-headed households had risen to 92.6 percent of white households and 94.0

PAGE 214

193 percent of black households. This particular datum is not of much value in measuring family structure of historical populations, because most women who headed families lived within male-headed households. Summary of Household Characteristics, 1870 and 1885 Overall, the change between 1870 and 1885 was toward more complex households, due to changes in the economy and in the rapid influx of migrants to the village of Lake De Funiak. In the latter year, more households were extended. Fewer households, however, contained boarders in 1885 than in 1870, due largely to the development of boarding houses and to the preference for relatives in times of economic need. Fertility had probably returned to a normal pattern of growth, after a postwar depression, and a greater percentage of household heads were married. Again and again, it has been seen that black households moved from being quite unlike white households in 1870, to a position of great similarity in 1885. We now turn to the analysis of some of these characteristics by other aspects of rural social structure: vi 1 1 age-rural resi dence and social class.

PAGE 215

194 Household Char ac teristics by Rural -Vi 1 1 age Residence Since data on village and rural residences were only gathered for the white sample of 1885, this comparison of household characteristics refers only to that sample. Most previous researchers have found certain differences in family and household characteristics between village and farm residents . Pryor (1972) found, in Rhode Island in 1875, that rural households were more often extended than urban ones. This was true both in the proportion of three-generation households and i n 1 ateral ly extended ones. Bieder's data for a Michigan community (1973) contradict this, but his population was yery small and the differences were not large. A few studies have shown a tendency for rural households to contain more children than those in towns in 19th-century populations (Bloomberg et al . , 1971; Bieder, 1973); but whether this was because of lower actual fertility among the village families or because of different ages at which children left home is not certain. In addition, Anderson (1971) suggested that village or town extended households differed in kind from those in the rural areas of 19th-century England,

PAGE 216

195 Specifically, rural household kin were usually aged parents who were dependent on married children, while those in urban households were more often young married couples who were not yet able to live alone. In this section, the household characteristics of the village of Lake De Funiak and the county sample of 1885 will be compared. Extended Kin Rural county households in 1885 were more often extended than were vi 1 1 age househol ds , consistent with Pryor's findings for Rhode Island families of the same period (1972). Over a third (36.5 percent) of the county households contained non-nuclear relatives, compared to 28.2 percent of the village households (Table 44). Even though more of the rural households were extended, the proportion of extended households in Lake De Funiak was also rather high. This may be partly due to the economic hardship of the 1880s, in addition to the development of attitudes favoring the incorporation of relatives into the household. It should be noted that the percentage of extended households in the 1870 white sample was only 13.6, which makes the 1885 figure seem quite high.

PAGE 217

196 Table 44. Characteristics of White Households in 1885, by Village or Rural Residence Percent of

PAGE 218

197 with over one-fourth of all households containing at least one unrelated resident. This difference between the village and the county is probably related to the fact that the village was fairly new, and a fairly large number of migrants needing housing had arrived within a few years. In particular, housing for single adults and young couples was needed. This often meant sharing a household or living in a boarding house in the village. If one had a home in the village, it was easier to find likely lodgers, and village householders were better able to add to their incomes by renting out rooms. The'economic circumstances of the times and in-migration via the railroad combined to create a more complex housing situation in the village than in the county. Servants Although in 1870 there were still a great many black servants living in farm houses in Walton County, they had all but disappeared by 1885. Servants had become an "urban' phenomenon, serving mainly the professional and proprietorial classes. Thus, by 1885, the county sample contained only three households with servants, or 2.9 percent of all households (Table 44). In the village, by comparison, 14 households, or 17.9 percent, had servants.

PAGE 219

198 Household Size Average household size in the rural area was slightly higher than in the village, the means being 6.9 and 6.4. This is because of the larger families in the county sample, but not because of more boarders or servants. The median number of related persons living in the households was 6.3 for the county sample, and 5.8 for the village. Number of Children As reported in other studies of 19th-century families, rural households in Walton County had more children than did village households. The median number of children per household in Lake De Funiak was 3.1, while that for the county sample was 4.1. Some of this difference is due to the fact that many village families did not have children. In fact, 19.2 percent of the primary families of Lake De Funiak had no children in thei r househol ds, compared to only 7.7 percent of the county households. Considering only complete nuclear families, the mean number of children per household was 4.0 in the village, and 4.7 in the rural households. Structure of Primary Families Comparison of Lake De Funiak and the rural households by family types of primary families (Table 45) shows a great

PAGE 220

199 Table 45. Structure of Primary Families, 1885, by Village or Rural Residence Lake De Funiak County Family Type N_ Percent N Percent 7.7 13 12.5 12.8 4 3.8 69.3 66 63.5 7.7 17 16.3 2.6 4 3.8 One-parent

PAGE 221

200 and the fact that farm families were more likely to be three-generational than village families. Also, if some of the rural county residents had moved into the village, it probably was the young married couples or the single adults, and not older families such as those found in the county sample. Most of the single-parent families in the rural areas were older widowed parents living with adult single children, a group not likely to migrate to the vi 1 1 age. Summary: Household Characteristics by Rural or Village Residence The differences between rural and village households in 1885 were fairly substantial. Village households were more often found to contain boarders and servants. Rural households more often were extended. Over half of the rural extended households were three-generational or included a widowed parent. Most of the extended village families included siblings of one of the spouses, less than a third being three-generational. In addition, village primary families were more often young couples than was true for rural fami lies.

PAGE 222

201 Household Characteristics and Social Class One of the most pervasive of social influences is that of social class, which is often measured by classifying occupations according to income or prestige levels. It seems likely that whether or not 19th-century households were extended would be related to social status. Households of the better-off classes would be in a better position to support dependent relatives than those in the lower classes. On the other hand, households of the lower classes might need the added income from working relatives or boarders. Pryor (1972) found that Rhode Island families in higher income brackets in 1960 were more likely than others to be extended, but very 1 i ttl e di f f erence was found between manual and nonmanual groups in the 19th century. For the 1875 Rhode Island population, Pryor found that farm families were more likely than others to be extended. The presence of boarders in families is also probably related to social class. Since boarders may be a source of income, their presence might be expected in lower-class households. However, a boarder requires that the family have enough room in the household, which might pose a problem for many lower-class families.

PAGE 223

202 From a different point of view, the usual boarders were young men or couples without families. They probably preferred to live with families close to their own class level, or to that of their parents. Pryor's (1972) data touched on these problems. His main finding concerning boarders was that nonmanual families were more often augmented by boarders than either farm families or those with manual occupations. These were, more often, town residents who were more able to afford the extra space taken by the boarders. In the following pages, certain household characteristics of Walton County residents in 1870 and 1885 will be compared according to class level to see if new insights into this problem will appear. In order to make social class comparisons, the original six strata were collapsed into three: the skilled class, which included skilled tradesmen, proprietors, professionals, managers, and clerical workers; the unskilled class, including unskilled manual occupations and laborers; and the farm class, including farmers and stockmen. Since there were very few blacks in the skilled class, all members of this class were classified with black farmers. For blacks, then, there were two classes--the laborers and the farm/skilled class. By 1885, there was not enough variation in black

PAGE 224

203 occupational class to permit even this very simple dichotomy, In this section, therefore, the 1885 black population is not discussed. Extended Kin The relationship between social class and the presence in the household of extended relatives may not be simple. It seems reasonable to assume that poorer families may have had to double up in order to maintain their families and households. On the other hand, the more wealthy families, especially landowning families, seem likely to have had lineal relatives within their households. In order to examine this relationship, the proportions of households containing relatives of the head were calculated for each sample or population, the results of which are presented in Table 46. These results are somewhat contradictory at first glance. In 1870, 5.6 percent of the white skilled primary families were extended, as were 7.7 percent of the unskilled households. About twice that proportion of farm households, in the same year, were extended. For black households, however, a larger proport i On of the unski 1 1 ed laboring class households were extended than of the f arm househol ds--l 6. 7 percent versus 9. 5 percent.

PAGE 225

204 Table 46. Percentage of Households Containing Extended Relatives, by Social Class Race a

PAGE 226

205 Table 47. Percentage of Households Containing Boarders, by Social Class Race and Skilled Unskilled " Farm' 38.8

PAGE 227

206 than did the unskilled or the farming class. The figures suggest that the wealthier individual in the village was more likely to house boarders than the worki ng-cl ass householder. If one recalls the change in the nature of boarders from 1870 to 1885, this interpretation seems reasonable. There were quite a few skilled or white-collar individuals in 1885 who were seeking rooms. It seems likely that these people would have more often boarded with householders of at least their own status. In addition, those boarders who were still of the apprentice or paid-laborer type would be likely to be living with those of the higher classes-i I who could afford to hire their labor. Farmers, in 1885, no longer could afford to keep farm laborers in their households, hence the drop in the percentage of farming households with boarders. Household Size Table 48 shows the mean household sizes for the households in the various social classes, by race and year. In 1870, both whites and blacks in the farming classes had larger households. This was probably because of the larger number of children in households headed by farmers.

PAGE 228

207 Table 48. Mean Household Size by Social Class, 1870 and 1885 Race and Year Skilled Unskilled Farm Mean Mean Mean White 1870

PAGE 229

208 size. In 1870 and 1885, the use of contraceptives was known, but effective means were not readily available. Therefore, differentials may not have been as great by social class. In this section, we will examine the number of children per household, by social class level. What we are measuring is not fertility, but the number of the head's children living within the household. This can vary with the age at which children leave home and with the proportion of households with children, as well as with actual fertility behavior. With this in mind, we look at Table 49, which presents the mean number of children in the households, by social class. Table 49. Mean Number of Children in Household, by Social Class, 1870 and 1885 Race and

PAGE 230

209 For the 1870 data, it appears that, for both blacks and whites, laboring-class households had fewer children than did those of the skilled and farm classes. One possible reason for the \jery low mean for the unskilled blacks in that year is that the marriages of that group were fairly new and/or unstable, an argument which was presented in an earlier chapter. The black farmers seemed little different from the white farmers or skilled classes. Another reason for the lower number of children among the unskilled of both races is that these household heads were probably younger, on the average, than skilled or farming household heads. It took time to be able to own land or to become a skilled tradesman. To the extent that the unskilled were younger, they had had less time for the growth of their families, and their number of children would besmaller. ' : Finally, older children in laborers' households were more of a burden than were those in farm households, who could help with the family enterprise. This fact, and the better abi 1 ity of the wealthier families to support their children, perhaps led to differences in the ages at which most children left home. It is probable that children of laborers left home at younger ages, causing the mean number of children per household in that class to be lower.

PAGE 231

210 There seems to be a difference between the 1870 and the 1885 populations in the relation of social class to the number of children in the household. In 1885, there was not much difference between social classes for whites. The finding that rural farm families had more children in the household may be a result of rural children staying in the household longer before marriage or leaving for employment. M arital Status of Household Heads Table 50 presents the percentages of household heads who were married. This table shows \jery little variation in the marital status of household heads. The only difference of any importance was in the black household population of 1870. For that group, 15 of 24 lower-class households (62.5 percent) were headed by a married couple, as compared to 20 of the 24 farmers (83.3 percent). Black household heads of the laboring class were more often headed by widowed or single adults than were those of the farming class. For all other samples studied, there were no real differences between the marital status of household heads in the different classes. Sex of Household Heads The percentage of male-headed households by social class in the samples and populations studied are shown in

PAGE 232

211 Table 50. Percentage of Household Heads Who Were Married, by Social Class, 1870 and 1885 Race and Skilled Unskilled Farm 77.8

PAGE 233

212 Table 51. Percentage of Households with Male Heads, by Social Class, 1870 and 1885 Race a

PAGE 234

213 Other than these points, the relationship of social class to household composition and structure was not clear, or \/ery weak, or confounded with the effects of the economic decline. Perhaps occupation was not as important a differentiating factor in the 19th-century South as it is in today ' s society.

PAGE 235

CHAPTER VIII CONCLUSIONS This dissertation describes the population of Walton County, Florida, in 1870 and 1885. Changes in the population are related to the historical context. The family structure of the population is described, both in relation to the historical context and tothe family life cycle. Other characteristics of households are discussed in relation to changing social and economic conditions. The study provides new descriptive information about a 19th-century Southern population and offers important insight into the nature of the black and white family structure. At least three major conclusions can be drawn from the findings of this study which will be important to the study of the sociology of the family in general. The first of these conclusions concerns the adaptability of the family institution. Our data show that rapid and fairly substantial changes took place in the structure and composition of households and families during a rather 214

PAGE 236

215 short time. In 1870, for example, extended households were uncommon except among older couples or widowed persons. By 1885, however, there were many more extended households, even among younger couples. The family system had responded to conditions of economic depression by increasing the sharing of households with relatives. The complexity of households, therefore, is not simply a product of family structural norms, but is also a product of external conditions as well. Even so, normative influences appeared in the assignment of priority to certain relatives in the sharing of households. Siblings and thei r .chi 1 dren were taken in most frequently, after aged parents and married children. Family and kinship norms, then, influence residential patterns, but other norms come into effect in times of special need. It follows from the above that historical as well as contemporary studies of families should be informed by thorough understanding of the socioeconomic context. If this study had only used data from 1885, for example, the conclusions would have been quite different from those which were derived from looking at both 1870 and 1885. Both contextual and, where possible, longitudinal studies are called for.

PAGE 237

216 In addition to the obvious economic conditions, other, more intangible, social conditions may influence patterns of family residence. The black family system changed markedly between 1870 and .1885, and these changes were explained in part by referring to the growing conflict between blacks and whites in this period. Although the relationship between the family institution and outgroup hostility is not as easily seen as its relation to economic factors, the effects were probably just as important in this case. Again, this leads to the conclusion that studies must be informed by the historical context if reasonable interpretations are to be made. There is no one "family of the past." Family systems in various times and places must be studied in their total social and economic contexts as a means to discovering both uniformities and variability. A second conclusion deriving from this research is that studies of family and household structure should use a developmental approach. Some important aspects of family structure do not appear in cross-sectional analyses. Crosssectional studies, for example, may reveal most families to be nuclear in structure even when extended families are common at certain stages of the life cycle. Longitudinal or developmental studies reveal such variations.

PAGE 238

217 Finally, the findings of this study show the need for using a variety of approaches to data on families and households. For some purposes, as in population analysis, the individual should be the unit of analysis. For others, the family is the proper unit. For still others, the household is the important unit of analysis. Although the datagathering and analysis techniques are more complex and difficult when moving from one unit to another, there are many benefits to the approach. If the household or the primary family is the only unit of analysis, then many distortions may occur in the overall interpretations of family structure. This study has shown, for example, that there were important differences in 1870 between primary and nonprimary black families, the nonprimary families being much less likely to be two-parent families. If we had compared only primary families, the differences between blacks and whites would have been considerably smaller. Primary families always tend to be a more stable group than nonprimary families, and some groups have fewer nonprimary families than others. The best approach to data on families and households is to use a variety of approaches, in order to present a well-rounded view of the subject, and to allow comparisons with a variety of other studies.

PAGE 239

218 There still remains much to be done in the field of family history. There are many areas and periods of time for which virtually no data have been presented. What is needed is a large number of studies of families at various points in time and in many areas, within a variety of social contexts. Studies of this add to our knowledge of where we came from--a worthy goal for a discipline interested in socialchange. Studies such as this one also add significantly to our understanding of the interrelationships between the family and other aspects of social structure. Since it is often difficult to see these relationships except at a distance, the value of the historical study is similar to that of the comparative approach. It is hoped that the next few years will see an increasing interest in the hi stori cal study of social i nsti tut i ons .

PAGE 240

APPENDICES

PAGE 241

APPENDIX A CODING INFORMATION: HOUSEHOLD RECORD VARIABLE VALUES censusyear householdnumber card type group quarters 0=no l=yes 2=uncertain plural family household 0=no l=yes number kin in household two-digit number numberboarders number servants family structure-primary family 0= primary individual l=one-parent family 2=marri ed coupl e 3=married couple, unmarried children 4= stem fami 1 ies 5=sibling nucleus families 6=other 9=unable to ascertain number head's own children birthplace of head l=Florida 2=N. Carol ina/S. Carolina 3= other South 22

PAGE 242

221 APPENDIX A (continued) birthplace of head (continued) 4=Northeast U.S 5=other U.S. 6=Scotl and 7=other British 8=other 9=unknown migration time l=Flori da-born 2=at least 15 years 3=10-14 years 4=5-9 years 5 = 1 as t 4 years 9=cannot estimate marital status of head 1 =si ngl e 2 ='m a r r i e d 3 = wi dowed 4=other kinship of extended relative: parent generation l=father 2=mother 3 = father-i n-1 aw 4 = mother-i n-1 aw 5=uncl e 6=aunt 7=grandfather 8=grandmother kinship of extended relative: sibling generation 1 =brother 2 = si ster 3 = brother-i n-1 aw 4=si sterin-law 5=nephew 6 = n i e c e 7=sister and husband 8=brother and wife

PAGE 243

222 APPENDIX A (continued) kinship of extended relative: married child generation l=son and wife 2=daughter and husband 3=grandchild 4=widowed son 5=widowed daughter 6=widowed son-in-law race of head sex of head 1 =whi te 2=black 3=mul atto 1 =nial e 2=f emal e age of head age oldest child of head age youngest child of head two-digit number, from schedule

PAGE 244

APPENDIX B CODING INFORMATION: PERSON RECORD VARIABLE VALUES year

PAGE 245

224 APPENDIX B (continued) relationship to head (continued) 1 5=brother-in-law 1 6=sister-in-1aw 1 7=uncl e T8=aunt 1 9=nephew 20=niece 2 1 = c u s i n 22=stepson 23=stepdaughter 24=grandparent 25=secondary head 26=secondary spouse 27=secondary son 28=secondary daughter 29=secondary sibling 3d=other secondary member 31=servant 32=boarder 33=ward marital status 1 = s i n g 1 e 2=married, spouse present 3=wi dowed 4 = di vorced 5=other 9=not applicable occupati on 01 =farmer 02=stockman 03=stock dealer 04=laborer 05=barkeeper 06=blacksmi th 07=boarding-house keeper 08=boatman 09=boat mate 10=bookkeeper ll=brick mason 12=broker 1 3=butcher

PAGE 246

APPENDIX B (continued) 22 5 occupation (continued) 14=captain of boat 15=carpenter 1 6=cl erk 17 = cool< 1 8=cooper 19=dentist 20=driver 21==druggist 22=editor 23=gardener 24=hote1 manager 25=in fame 26=land agent 27=lawyer 28=lumberman 29=lumber manufacturer 30=machinist 31'=merchant 32=miller 33=mi 1 1 owner 34=mi 1 1 sawyer 35 = mi 1 1 i ner 36=minister 37=music teacher 38=pai nter 39=peddler 40=phys ician 41 ^photographer 42=porter 43 = publ i sher 44=railroad man 45=railroad workman 46=sailor 47=school teacher 48=sea master 49=servant 50=section foreman 51 =seamstress 52=sheriff 53=shingl e maker 54=shoemaker 55=spinner and weaver

PAGE 247

APPENDIX B (continued) 226 occupation (continued) 56=teamster 57=telegraph operator 58=timber inspector 59=wagoner 60=waiter 61 =washerwoman occupati onal class bi rthpl ace father' s bi rthpl ace mother's birthplace l=professional 2=clerical, managerial, proprietorial 3=skilled trades and crafts 4=unski lied manual 5=laborers, menial and service workers 6=farmers 7=agri cul tural laborers 8=no occupation given l=Flori 2=Alaba 3=Georg 4=N. Ca 5=S. Ca 6=Tenne 7 = Vi rgi 8 = Loui s 9=Kentu 1 0=Mary ll=Ohio 12=New 13=New 14=Wisc 1 5=Penn 1 6 = Mai n 1 7=Iowa 18=Indi 19=Cali 20=Ariz 21=Texa 22=Miss 23=Mass 24=New da ma ia rol i na rol i na ssee nia i ana cky land Jersey York n s i n syl vania e ana f r n i a ona s issi ppi achusetts Hampshi re

PAGE 248

APPENDIX B (continued) 227 birthplace (continued) 25=11 linois 26 = Mi ssouri 30=Scotl and 31=England/Wal as 32=Ireland 33=Germany/Prussia 34=Bavaria/Austria 35=Switzerland 36=Scandinavia 3 7 = Janiaica 38=Canada

PAGE 249

APPENDIX C METHOD OF ESTIMATION OF AGE AT MARRIAGE The median ages at marriages were estimated by means of a cumulative distribution of the proportions of each sex married at or below every age between 15 and 45. As a first step, only those people above 14 years of age were included in the analysis. Next, the proportion of individuals of each age category or under were computed separately for each sex. For example, the proportion of males aged 16 or less was calculated. Then the proportion aged 17 or less, 18 or less, and so on, until the proportion ever-married aged 45 or less was calculated. It was expected that the age at which the proportion ever-married was half the proportion ever-married at or below age 45 would give a good estimate of the median age at first marriage, since the median may be defined as the 50th percentile of the cumulative distributionofobservations. Age 45 was used as a cutoff point because, after that point, it became more difficult to distinguish between the widowed or divorced and the never-married. It was not believed that this would introduce much bias, since most first marriages would have taken place before that age. 228

PAGE 250

REFERENCES Anderson, Michael 1971 Family Structure in Nineteenth Century Lancashire, Cambridge: Cambridge University Press. 1972 "The study of family structure." Pp. 47-81 in E. A. Wrigley (ed.). Nineteenth Century Society: Essays in the Use of Quantitative Methods for the Study of Social Data. Cambridge: Cambridge University Press. Ari es , Phi 1 1 i ppe 1962 Centuries of Childhood: Family Life. New York: A SocialHisto ry of Alfred A. Knopf. Bai lyn , Bernard. 1960 Education in the Forming of American Society: Needs and Opportunities for Study. Chapel Hill, N.C. : University of North Carolina Press. Bardis, PanosD. 1964 "Family forms and variations historically considered." Pp. 403-461 in Harold T. Christensen (ed.), Handbook of Marriage and the Family. Chicago: Rand McNally. Barr, A. James and J. H. Goodnight 1962 Statistical Analysis System. Raleigh, N.C: Student Supply Stores, North Carolina State University. Berkner, Lutz K. 1973a "Recent research on the history of the family in Western Europe." Journal of Marriage and the Family 35:345-405. 229

PAGE 251

230 Berkner, Lutz K. 1973b "The stem family and the developmental cycle of the peasant household: an 18th-century Austrian example." Pp. 34-58 in Michael Gordon (ed.). The American Family in Soci al -Hi storical Perspecti ve. New York: St . Martin ' s Press . Bieder, Robert E. 1973 "Kinship as a factor in migration."Journal of Marriage and the Family 35:429-439. Bloomberg, Susan E., M. F. Fox, R. W. Warner, and S. B. Warner, Jr. 1971 "A census probe into nineteenth-century family history: Southern Michigan, 1850-1880." Journal of Social History 5:26-45. Broom, Leonard and Philip Selznick 1968 Sociology: A Text with Adapted Readings Edition. New York: Harper and Row. 4th Bruington, Lola L. 1951 Walton County: Euchee Valley Cemetery, Dead River and Miller Cemetery. Unpublished manuscript, P. K. Yonge Library of Florida History, .Gainesville. Calhoun, A rthurW. 1917-19 A Social History of the American Family from Colonial Times to the Present. Three vols. Cleveland: Arthur H. Clark Company. Demos, John 1968 "Families in Colonial Bri stol , Rhode Island: An Exercise in Historical Demography." William and Mary Quarterly 25:40-57. 1970 A Little Commonwealth: Family Life in Plymouth Colony. New York: Oxford University Press. Du Bois , W. E.B. 1899 The Philadelphia Negro. Philadelphia: University of Pennsylvania Press. 1908 The Negro American Family. Atlanta University Publications, no. 13. Atlanta: Atlanta University Press.

PAGE 252

231 Ebl en , Jack E. 1965 "An analysis of nineteenth-century frontier populations." Demography 2:399-413. Elk ins, Stanley M. 1975 "The Slavery Debate." Commentary 60:40-54. Farber, Bernard 1972 Guardians of Virtue; Salem Families in 1800. New York: Basic Books. Fogel , Robert and Stanley Engerman 1974 Time on the Cross: The Economics of American Slavery. Two vols. Boston: Little, Brown. Frazi er , E. Frankl i n 1930 "The Negro slave family." Journal of Negro History 15:198-259. 1948 The Negro Family in the United States. Rev. ed. New York: Dryden Press. Furstenberg, Frank F. , Jr. 1966 "Industrialization and the American family: A look backward." American Sociological Review 31:326-337. Gautier, Etienne and Louis Henry 1958 La population de Crulai, paroisse normande. Paris: Presses Uni versi tai res de France. Genovese, Eugene 1974 Roll, Jordan, Roll: The World the Slaves Made. New York: Pantheon Books. Click, Paul C. 1947 "The family cycle." American Sociological Review 14:164-174. 1955 "The life cycle of the family." Marriage and Family Living 17:3-9. 1957 American Families. New York: Wiley.

PAGE 253

232 Glick, Paul C. and Robert Parke, Jr. 1965 "New approaches in studying the life cycle of the family." Demography 2:187-202. G d e , W i 1 1 i a m J . 1971 World Revolution and Family Patterns. Paperback ed. New York: Free Press. Goodsel 1 , Wi 1 lys ti ne 1939 A History of Marriage and the Family. New York: M a c m i 1 1 a n . Gordon , Mi chael (ed . ) 1973 The American Family in Social -His tori cal Perspective. New York: St. Martin's. Grabill, Wilson H., Clyde V. Kiser and Pascal K. Whelpton. 1973 "A long view." Pp. 374-396 in Michael Gordon, ed. The American Family in Soci al -Hi stori cal Perspective. New York: St. Martin's. Grantham, Dewey W., Jr. (ed.) 1967 The South and the Sectional Image. New York: Harper and Row. Gray, L. C. 1933 History of Agriculture in the Southern United States to 1860. Two vols. Washington: Carnegie Institute of Washington. Greven , Philip 1970 Four Generations: Population, Land and Family in Colonial Andover, Massachusetts. Ithaca: Cornell University Press. Gutman , Herbert 1973 "Le phenomene invisible: la composition de la famille et du foyer noir apres la Guerre de Secession." Annales, E. S. C. 26:1197-1218. Hareven, Tamara K. (ed.) 1971 Anonymous Americans: Explorations in Nineteenth Century Social History. Englewood Cliffs, N.J.: Prentice-Hal 1 . 1974 "The family as process: The historical study of the family cycle." Journal of Social History 7:322-329.

PAGE 254

233 Harris, William 1976 "Work and the family in Black Atlanta, 1870." Journal of Social History 9:319-325. Henri pin, Jacques 1964 "From acceptance of nature to control: The demography of French Canadians since the seventeenth century." Pp. 204-215 in M. Rioux and Yves Martin (eds.), French-Canadian Society. Toronto: McClelland and Stewart, Ltd. Hershberg, Theodore 1971 "Free Blacks in antebellum Philadelphia: A study of ex-slaves, freeborn, and socioeconomic decline." Journal of Social History 5:183-209. Howard, George E. 1904 A History of Matrimonial Institutions. Chicago: University of Chicago Press. Two vols. Lammermeier, Paul L. ' 1973 "The urban black family of the nineteenth century A study of black family structure in the Ohio Valley, 1850-1880." Journal of Marriage and the Family 35:440-456. Lansing, J. B. andLeslieKish 1957 "Family life cycle as an i ndependent vari abl e . " American Sociological Review 22:512-519. Lantz, Herman R. , M. Britton, R. Schmitt and E. C. Snyder 1968 "Pre-industrial patterns in the colonial family in America: A content analysis of colonial magazines." American Sociological Review 33:413-426. Laslett, Peter 1973 "The comparative history of household and family.' Pp. 19-33 in M. Gordon (ed.). The American Family in Soci al -Hi stori cal Perspective. New York: St. Martin's. Les 1 i e , Geral d R. 1976 The Family in Social Context. 3rd ed, Oxford University Press. lew York

PAGE 255

234 Lockridge, Kenneth A. 1966 "The population of Dedham, Massachusetts, 16361736." Economic History Review 19:318-344. 1970 A New England Town: The First Hundred Years. New York : Norton. M c K i n n n , J h n L . 1911 History of Walton County. Gainesville: Palmetto Books. Mendenhall , William, Lyman Ott and R. L. Schaeffer 1971 Elementary Survey Sampling. Belmont, Calif.: Duxbury Press . Model 1 , John 1971 "Family and fertility on the Indiana frontier , 1820." American Quarterly 23:615-634. Modell, John and Tamara Hareven 1973 "Urbanization and themal 1 eabl e household." Journal of Marriage and the Family 35:467-479. Nie, Norman, Dale H. Bent and C. Hadlai Hull 1975 SPSS: Statistical Package for the Social Sciences. New York: McGraw-Hill. Norton , Susan L. 1971 "Population growth in colonial America: A study of Ipswich, Massachusetts." Population Studies 25:433-452. Ogburn,WilliamF. 1954 "Why the Family is Changing." Sociologus, New Series, 4:160-170. Owsl ey, Frank 1949 Plain Folk of the Old South. Baton Rouge, La.: Louisiana State University Press. Parsons, Talcott 1943 "The kinship system of the contemporary United States." American Anthropologist 45:22-38.

PAGE 256

235 Pleck, Elizabeth H. 1973 "The two-parent household: Black family structure in late nineteenth-century Boston. Pp. 152-177 in M. Gordon (ed.). The American Family in SocialHistorical Perspective. New York: St. Martin's. Rowel 1 , Sumner 0. 1963 Puritan Village: The Formation of a New England Town. Middletown, Conn.: Wesleyan University Press. Pryor , Edward T. 1972 "Rhode Island family structure: 1875 and 1960." Pp. 571-589 in Peter Laslett, with R. Wall (eds.). Household and Family in Past Time. Cambridge: The University Press. Saveth, Edward N. 1969 "The problem of American family history, can Quarterly 23:627-634. Ameri Sennett, Richard 1973 "Middle-class families and urban violence: The experience of a Chicago community in the nineteenth century. Pp. 111-134 in M. Gordon (ed.). The American Family in Soci al -Hi stori cal Perspective. New York: St. Martin's. Si rjamaki , John 1953 The American Family in the Twentieth Century. Cambridge: Harvard University Press. Smith, Daniel Scott 1973a "The demographic history of colonial New England, Pp. 397-415 in M. Gordon (ed.). The American Family in Social -Hi storical Perspective. New York: St. Martin's. 1973b "Parental power and marriage patterns: An analysis of historical trends in Hingham, Massachusetts. Journal of Marriage and the Family 35: 419-428. Smi th , Jul ia F. 1973 Slavery and Plantation Growth in Florida, 18211860. Gainesville: University of Florida Press

PAGE 257

236 Stni th , T. Lynn 1953 The Sociology of Rural Life. 3rd ed. Harper and Row. New York: Sorokin, Pitrim A. 1928 Contemporary Sociological Theories Harper and Row. New York Sorokin, Pitirim, C. Zimmerman and C. J. Galpin 1931 A Systematic Sourcebook in Rural Sociology. Three vols. Minneapolis: University of Minnesota Press . Thernstrom, Stephan 1964 Poverty and Progress: Social Mobility in a Nineteenth Century City. Cambridge, Mass.: Harvard Un i vers ity Press. United States Bureau of the Census 1960 Historical Statistics of the United States: Colonial Times to 1957. Washington, D.C.: Government Printing Office. United States Department of the Interior 1872a A Compendium of the Ninth Census (June 1, 1870). Washington, D.C.: Government Printing Office. 1872b Ninth Census, Volume I: The Statistics of the Population of the United States. Washington, D.C.: Government Printing Office. 1872c Ninth Census, Volume III: Statistics of Agriculture. Washington, D.C.: Government Printing Office. United States House of Representatives 1896 The Miscellaneous Documents of the House of Representatives, for the First Session of the Fifty-second Congress, 1891-92. Volume 50, Part 10. Washington, D.C.: Government Printing Office. Wells, Robert V. 1971 "Family size and ferti 1 i ty control i n eighteenth' century America: A study of Quaker families." Population Studies 25:73-82.

PAGE 258

237 Wells, Robert V. 1971-72 "Demographic change and the life cycle of American families." Journal of I nterdi sci pi i nary Hi story 2:27 3-282. 1972 "Quaker marriage patterns in a colonial perspective." William and Mary Quarterly (3rd series) 29:415-442. Woodward, C. Vann 1951 Origins of the New South. Baton Rouge: Louisiana State University Press. Wrigley , E. A. (ed. ) 1966 An Introduction to English Historical Demography, New York: Basic Books.

PAGE 259

BIOGRAPHICAL SKETCH Barbara Finlay Agresti was born April 3, 1945, in Brady, Texas, the third of five children of James and Grace Finlay of Fife, Texas. Her first 16 years were spent growing up on a stock farm at Fife, in McCulloch County, and she attended the public school at nearby Lohn. She entered college in Lubbock, Texas, in 1966, receiving the Bachelor of Arts degree in sociology in 1969 from Texas Technological College. In the winter of the next year, she continued her studies in sociology at the University of Texas at El Paso, receiving the Master of Arts degree in 1971. After teaching for one year at Jacksonville State University in Jacksonville, Alabama, she entered the graduate program in sociology at the University of Florida, Gainesville. Her Ph.D. was granted from that institution in August, 1976. Presently, Barbara Finlay Agresti is Vi si ti ng Assi stant Professor at the University of Florida in Gainesville. Her research continues to be influenced by her interest in the 238

PAGE 260

239 history of American institutions and classical sociological thought. She is married to Dr. Alan Agresti of the Department of Statistics, University of Florida, and has two children. Ginger Bloomer, 14, and Kerry Bloomer, 12.

PAGE 261

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fullyadequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ^ aid R ! /I \ :!A / 1^ Gerald R, Professor eslie, Chairman and Chairman of Sociology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. "Joseph S. Vandiver Professor of Sociology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ^'>v Gordon F. Streib Graduate Research Professor of Sociology

PAGE 262

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. 1 i )( M. Berardo essor of Sociology e I ^ I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope.and quality, as a dissertation for the degree of Doctor of Philosophy. Samuels. Hill, Jr. < Professor and Chairman of Religion This dissertation was submitted to the Graduate Faculty of the Department of Sociology in the College of Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August, 1976 Dean, Graduate School