|SUMMIT COUNTY CHAPTER
of the Ohio Genealogical Society
P O Box 2232 Akron OH 44309-2232
Systems in Genealogy
By Richard A. Pence
[Revised March 1996 and reposted]
Recently a participant in the soc.genealogy.methods newsgroup inquired about
"tried-and-true" and "universally accepted" numbering systems for genealogy. As
Cheryl Singhal noted in a response: There are plenty of the former - but none
of the latter!
The original questioner gave an example of the system he is using. He assigned
a number to the progenitor of the family, No. 1 in this case, and then the
numbers assigned to the progenitor's children were 1.1, 1.2, 1.3, etc.
Similarly, those in the third generation were numbered 1.1.1, 1.1.2, 1.1.3,
This person likely developed this system after thinking through his needs, and
doubtless he was unaware that this numbering system has been around long enough
to have acquire a name. It is called "the d'Aboville System." This system will
be discussed briefly below, as will most of the variations which others have
mentioned in messages in the group.
The following, then, is a summary of some of the most common genealogical
numbering systems and is adapted from previous articles and lectures I have
prepared on the topic.
In reading it, please keep in mind that there are only two undisputed "givens"
when it comes to genealogical numbering systems:
1. If you want to be published in a genealogical journal you will
conform your material to the system used by that publication - no
matter how insane it may seem to you.
2. Every genealogist has his or her own idea of what constitutes
the world's best numbering system - no matter how insane it may
seem to you.
In other words, although the system I "invented" and use is the best, I'm quite
willing to concede that yours is also the "best."
Up the Ladder or Down?
A basic notion to keep in mind about genealogical numbering systems is that
ancestor and descendant databases are two distinct creatures: When you try to
combine them, you run into difficulties. Like oil and water, they do not mix
well. If you are keeping a record of your own (or some other person's)
ancestors, you should use the Sosa-Stradonitz (Ahnentafel) numbering system
(described later). If you are compiling the descendants of one of your
ancestors, you will need to use one of the descendant systems discussed below.
The first system discussed below is an _ancestor_ numbering system; the
remainder are _descendant_ numbering systems - with the exception of the
mention of "combined systems" and computer program record numbering.
Sosa-Stradonitz or Ahnentafel System
This is the normal - and popular - method of numbering your ancestors and is so
easy to understand - and so effective - that it is universally used. In it you
assign yourself (or child or parent) the number 1. If you are No. 1, then your
father is No. 2, your mother No. 3, your paternal grandfather No. 4, etc. In
this system, a person's father's number is always twice the person's number and
his or her mother's number is twice-plus-one. This method of numbering
ancestors is used worldwide and is called the Sosa-Stradonitz System for the
Spanish genealogist Jerome de Sosa, who first used it in 1676, and for Stephan
Kekule von Stradonitz, who popularized it in his 1896 "Ahnentafel Atlas." It is
also often called the "Ahnentafel Numbering System," after the book.
(Translated from the German, ahnentafel is "ancestor table.") This system is
used both for numbering ancestors in lists ("tables") or on charts.
The Register System
This is the "formal" numbering system used by U.S. journals and, by long use in
these respected journals it, along with its kin, the Record System (see below),
has become the "standard" system.
In it, the progenitor or other focal individual is given the number 1. Each
child in that family is then numbered in order of birth with lower-case Roman
numerals (i, ii, iii, iv, v, etc.), and those whose lines are carried on in the
work are also given an Arabic number. For instance, No. 1 may have had seven
children (i through vii), but only one of these had descendants, say iv. No. iv
is then also given the Arabic number 2. His children, in turn, are numbered
from i upward, with, perhaps, Nos. i, iv and vi having descendants and thus
given the additional identifications of 3, 4 and 5. This system is used in the
_New England Historic Genealogical Society Register_, from which it gets its
A brief example of the Register system (superscript characters are indicated as
(1), (2), etc., in this example):
Descendants of Henry Pence
1. Henry(1) Pence (Johann Georg(A), Johannes(B)), born - Oct. 1739
at Iggelheim, Rhineland-Pfalz, Germany; died - Sep. 1824
Champaign County, OH; married ca. 1765 Mary Magdalene Blimly.
Children of Henry Pence and Mary Magdalene Blimly:
i. George(2) Pence born 16 Aug. 1766 Frederick County,
VA; married Mary Mauck 9 Nov. 1790 Shenandoah County;
died 1810 Shenandoah County. No further information
on descendants, but he is believed to have had two
sons and a daughter.
2. ii. Jacob Pence, born 15 Sep. 1767 Frederick County, VA.
3. iii. Henry Pence, born 4 Sep. 1768 Frederick County, VA.
4. iv. Abraham Pence, born 4 Sep 1769 Frederick County, VA.
v. Magdalene Pence, born 31 Jan. 1771 Frederick County,
VA; possibly dead by father's will 1820; no other
The Record System (Modified Register System)
This system varies from the Register System in that each individual is given an
Arabic number regardless of whether the line is subsequently carried on. A plus
mark (+) prior to the Arabic number is used to indicate the person had
descendants and more information can be found later in the work. This is the
system used by the _National Genealogical Society Quarterly_.
If the above Register example were to be adapted to the Record system, the
first child (i. George) would receive the additional Arabic number 2, the
second child (2. ii. Jacob) would become No. 3 and a plus sign would be placed
before the 3, indicating that he has descendants and is treated more fully
The Record/Register system, as used in some journals, differentiates between
generation numbers (for example, the (1) following Henry's name in the example
above) and footnotes by putting the superscript generation number in italics
and leaving the superscript footnotes or endnotes in Roman (normal) type.
Also, in U.S. genealogies which often treat just the descendants of an
immigrant ancestor, previous generations are given alphabetic generation
designations. Thus, the German father and grandfather of Henry in the example
are identified by a superscript italic capital "A" following the father's name
and a superscript italic capital "B" following the grandfather's name,
indicating that these individuals remained in Germany.
The advantage of the Record system over the Register system is that if you
later discover that a person thought not to have had descendants turns out to
have indeed had them, he or she already has a number. However, such a discovery
with either system is almost certain to require extensive renumbering in
The Register/Record systems thus are _not_ suitable for identifying persons in
a database and should only be used in final production of a work (and save the
numbering task until the manuscript is nearly complete). Both of these systems
allow you to easily trace down from the progenitor to a specific descendant but
is is more difficult to "go up the ladder." This is because, although a
person's parent is identified by name under an individual listing, the parent
is not identified by number. There may, in fact, be several of that name
earlier in the book, sometimes requiring a page-by-page search to locate the
previous reference to a person. (You must look among the smaller child's
The Henry System
The Henry System is named after Reginald Buchanan Henry, who used it in
"Genealogies of the Families of the Presidents" in 1935. In this system, the
progenitor or other individual is assigned the number 1 (or sometimes another
number or letter if one is tracing the descendants of several progenitors). The
progenitor's oldest child becomes 11, his or her next child is 12, and so on.
The oldest child of No. 11 is No. 111, the next 112, etc. When Henry
encountered families with more than nine children, he used the Roman numeral X
for the 10th child, and then A, B, C for 11th, 12th and 13th children, etc.
The D'Aboville System
The d'Aboville System (mentioned earlier) is a variation of the Henry System,
except that each digit (or double digit for numbers larger than 9) is separated
by a period. Thus the first child of No. 1 is 1.1; the 10th would be 1.10. The
first children of the latter would be 1.10.1.
Modified Henry Systems
There are many other variations of the Henry system. The variations are usually
in how the numbers 10 and larger are treated. The most common of these uses
parentheses to deliniate these" The tenth child of number 131 is 131(10) and
his or her children become 131(10)1, etc. The latter number is the equivalent
of 121X1 in the original Henry System and 220.127.116.11.1 in the d'Aboville System.
The parentheses method is so common it is often mistakenly thought to be _the_
Another variation used by some is to substitute alpha characters for numbers
throughout. This avoids having to use periods or parentheses, but suffers in
that most people have to mentally count the alphabetic characters to discern,
for example, that H is the equivalent of 8.
The "Modern" Henry System (The "Pence System")
in today's computer world the most common variation of the Henry System uses
the letters A, B and C, etc., for the 10th, 11th and 12th children, rather than
X, A and B. The reason for this is that personal computers usually sort digits
first and then alpha characters: "A" comes after "9," etc., in a normal ASCII
sort. This computer-sort ability is a significant and important advantage of
the "modern" Henry system over other Henry variations. Another advantage is
that it requires only one character per generation, a consideration in both
recordkeeping and publishing. (Less important now, but essential when memory
and storage space was at a premium.)
In 1978, when I got my first home computer, I wanted a method of identifying
and placing the thousands of individuals I was entering into my PENCE database.
With the help of a friend who understood computer sorting routines, I
"invented" the "Pence numbering system" - the "Modern" Henry System. And it was
created without my ever having heard of Mr. Henry! The remarkable thing is that
this system is so intuitive and logical in terms of computer applications that
dozens, perhaps hundreds, of people have independently "discovered" it.
However, the first discussion of the "modern Henry system" in print is
apparently my own 1982 article, "Using a Word Processor to Compile a Family
History," _Genealogical Computing_, Vol. 1, No. 5, Mar. 1982, pp. 10-11).
The system also excels because it seldom needs explanation. So intuitive is it
that when I send data to others with the numbers included, even beginners often
send back additions that are already properly numbered.
A glance at a "Pence Number" (as is generally true of various Henry systems)
instantly tells you at least four important things about a person: which
birth-order child in the family he or she is (the last character in the
number); the generation number of the individual (the number of characters in
the ID No.); the numbers of all of his or her ancestors in that line (they have
the same root ID: truncate the last character and you have the parent's number;
lop off the last two characters and you have the grandparent's number); and the
numbers under which you can search for his or her descendants (all those that
start with the same numbers as his or her ID No.). I have arbitrarily assigned
an initial number or alpha character to each of the many Pence progenitors I
have discovered; thus, with the familiarity of long use, the first character of
any ID No. tells me which family a person belongs in.
[I also use the "Pence Number" to key my paper filing system. The files for any
individual (and his or her descendants, if I don't have too much on them) are
identified by the same number that is used in the database for that person.
This system makes it easy to create new files: If you gather a lot of new
infomration on a child of No. 12, for example, you start a new file "121" and
place it directly behind the file numbered as "12." Note that files are not
kept in normal numerical order, but in "ASCII sort order"; that is, the files
numbered 12, 121, 1211, 122, etc., come after 1193, for instance, and before
file No. 2.]
Another advantage of the Henry system and its variants is that discovery of
"new" descendants requires renumbering only in that family's segment of the
database. It is therefor a much more viable system for helping you identify
people in a database.
Still another advantage of a properly constructed Henry-type system: If you
list descendants in "modern" Henry number order in your family narrative (that
is, in the above-described "sort order"), you can quickly construct an index to
your text by using that number instead of a page number (which can constantly
change with additions or corrections). I can print out a segment of my database
in "sort order," then generate an index of given names keyed to the ID number
in a matter of minutes.
[For examples of the "Pence System" and how it can be used for indexing, see
either of my 1982 books, "A Guide to the Pence Families in America," Parts I
and II, available in many larger libraries; if your library would like a copy,
have it write me at 3211 Adams Court, Fairfax, VA 22030 or email me at the
address on this message.]
The de Villiers/Pama System
A system much used overseas, primarily in South Africa, is the de Villiers/Pama
system. It was invented by Chris. de Villiers, who used it in his published
work "Genealogies of Old Cape Families." published in the 1890s and revised
more recently by Dr. Cor Pama. The same numbering system is now being used in a
work called "Genealogies of South African Families," by Heese & Lombard,
published by the Human Sciences Research Council in South Africa [information
supplied by Steve Hayes of South Africa].
The original ancestor you start numbering from is a. Every subsequent
generation takes the following generation letter, so the children of "a" are
designated with "b," the grand children of "a" are "c," and so on. Each child
is given a number - the oldest is 1, the second 2, and so on. Thus, the
children of a are:
Their descendants are:
and so on.
Thus the third child of the second child is a.b2.c3.
As you can see, the lower case letters describe which generation a person is in
and the numbers describe the person's birth position in the family. Also, as
you can see, while each person in this system does have a unique number, this
number is not apparent until all its elements are combined. In a normal
printout, each family will have a child in the third generation that is
numbered c1, thus causing confusion and making it difficult to readily locate a
The Outline System
A number of computerusts with word processors which have an outlining mode use
that feature to number their genealogies. The progenitor is usually not
numbered; his or her children are designated by upper case Roman numerals,
their children by capital letters, and so on, as in this illustration:
The advantage of this system is the ease of entry (for proficient outliners).
Disadvantages are apparent: The outline system consumes a lot of space on
printouts and creates cumbersome numbers; for example, the number for Judith in
the above example is:
The "Pence No." equivalent for a progenitor with the designation of 1 would be:
about half as long and not nearly as cluttered - and understandable even by a
Nevertheless, the outline system is a "comfortable" one for many users.
Combined Numbering Systems
Some genealogists try to get a unique identification number for any collateral
relative in any line of descent by combining systems. Most often they do this
by using the Sosa-Stradonitz (ahnentafel) number of the ancestor, followed by a
decimal point and a descent number based on the Henry system (the first child
of your ancestor No. 128 would be 128.1 and so on). Spouses of those in
resulting descent files might be given unique numbers by adding the letters a,
b and c for spouses 1, 2 and 3 of any given individual. The problem with
combined systems: If 128.1 is also your ancestor he would additionally have the
number 64 in your database or ancestral chart. Confusion soon arises because
each person in your direct line (and their descendants) will end up
with two numbers. If there are cousin marriages, the confusion becomes even
Numbering Systems in Genealogical Software
Most genealogical software packages rely entirely on the computer to assign
numbers to individuals in the database and use these numbers for internal
identification and control. Husbands are then linked to wives and parents
linked to their children on the basis of these numbers - which are usually
random or serial and carry no particular significance to relationships.
Note that a characteristic of most genealogical databases is that each person
has a number; this is not true for most of the above-described descendant
systems because spouses are usually not uniquely numbered.
Some programs also allow a user-selected ID number; most often, those used are
based on the ahnentafel system (for ancestors) or a form of the Henry system
(for descendants). However, linkage is not keyed to these numbers.
This randomness and lack of linkage is what forces users of some programs to
rely on printed lists of persons and their record numbers as an aid to quickly
locating a person in the database. While each person has a unique number, this
number does not reveal any information about the person's relative place in the
It is worth noting that another advantage of the modern Henry system is that no
additional linkage needs to be created to tie the parent to a child. The ID No.
carries its own linkage: A "parent field" can be created and automatically
filled with a database manager. Simply truncate the last character of a
person's ID and have the resulting number inserted in the new field.
Most genealogical database programs, of course, print out charts or lists using
ahenantafel numbers and some allow you to print out your finished product using
the Record/Register system or a Henry system variation (such numbers are
generated "on the fly" by the program). If not, there are utilities that will
print out your database using either the Register/Record system or a Henry
system. (PAFAbility for PAF and GENBOOK, which handles several formats, are
One school of thought about numbering is that it's best just to let the
computer do the numbers and keep track of the linking for you; therefore you
need not worry about another numbering system until publication time. Others,
however, want to have a "static" and meaningful ID number that can help them
identify persons within the database. (I am definitely in the latter camp and
rely heavily on the assigned numbers in my PENCE database to identify
individuals and their line of descent. Also, most genealogy programs will alter
the RINs - Record Identification Numbers - during "clean-up" sessions, thus
making it impossible to key a filing system or any older printout to the
database number. And just when you were starting to remember some of them.)
As you have seen, genealogists use a wide variety of numbering systems. If you
decide to adopt a numbering system for either recordkeeping or publishing, you
will no doubt select one based on your own experience and needs. You would be
wise, however, not to create one that varies significantly from those touched
on here. The soundest approach, especially for a descendant-type publication,
is to select a system that is readily understood. This means selecting one that
has been used so often by others that most readers will have encountered it
(the Register/Record system) or one that is so intuitive that most readers can
understand it with little or no explanation (a Henry system variation).
Don't be like the fellow (an engineer) who proudly sent me a copy of a large
book produced by his one-name family association, in particular because he
wanted me to see the "great" numbering system they had devised. After four
years, I _still_ can't understand it!
The above is excerpted, with additions, from: Richard A. Pence, "Numbering
Systems in Genealogy," a presentation made to the National Genealogical
Society's Computer Interest Group in July, 1986, and to the Seattle
Genealogical Society's Computer Interest Group in December, 1987. While this
presentation has been published (see, for instance, "Numbering Systems In
Computer Genealogy, _ABT PAF_ [Newsletter of the Capital Personal Ancestor
File User Group], Vol. 3, No. 4, Jul.-Aug.-Sep., 1989, p. 1), a more recent
version is available from the National Genealogical Society's BBS
(703-528-2612) under the file name NUMBERS?.ZIP, where "?" is now 3, but should
change soon as an update is underway.
For an in-depth discussion and excellent examples of the Register/Record
systems, see: Joan Ferris Curran, "Numbering Your Genealogy: Sound and Simple
Systems," (NGS Quarterly, Vol. 79, No. 3, Sep., 1991, pp. 183-193.) This
article is sketchy on - and rejects - other numbering systems. While many find
the Register/Records systems less than "sound" and not "simple," if you want to
publish in a journal, you must use its system. (And if your book doesn't use
the Register/Record system, the review in a journal will critically note that
sin!) The NGS has published this article in book form and subsequent
refinements have appeared in later issues of the Quarterly.
An excellent brief summary of numbering systems is: Donn Devine, "How to Number
People in Pedigrees and Genealogies," (Ancestry Newsletter, Vol. 4, No. 1,
Jan-Feb, 1986, p. 1.) It is the basis for some of the discussion above.
For additional comments on numbering see: Richard A. Pence, "Still More Heresy
by the Numbers," (NGS Newsletter, Vol. 20, Nos. 1 & 2, Jan.-Feb. & Nar.-Apr.,
1994; this two-part article references several additional sources.
[Copyright 1996, Richard A. Pence, 3211 Adams Ct, Fairfax, VA 22030. This
article may be reprinted by nonprofit genealogical societies or
computer-interest groups and may be posed at electronic sites, provided no
charge is made for specific access to it. Commercial use is prohibited.]
Provided by: SUMMIT COUNTY CHAPTER, OGS
P O Box 2232
Akron OH 44309-2232
Back to the Summit County Genealogy Home Page
Last modified March 21, 2000
Copyright ©2000 Summit County Chapter OH Genealogical Society. All rights reserved.