Since its inception in 1969, the International Genealogical Index
(IGI) has been a marvelous finding tool — not a source in itself, but,
as its name indicates, an index to records from all over the world.[i]
It has always been a good idea to understand the nuances
of this database in order to make full use of it. While it is a great
resource for locating people, one must not be misled by the many
duplications and errors it contains. Now that the IGI is online, there
are new ways to search it. But you will not be making optimum use of it
unless you make some effort to go behind the individual entry and find
the paperwork and sources associated with many of the entries. With
this knowledge you will have a much better idea of when you will find
information, how reliable that data is, and whether to pursue an entry
that interests you.
Take, for example, my search for George Cook Stevens, a nephew of one
of my ancestors. From a thorough check of many sources, I was sure he
had been born in Cheshire, Connecticut, about 1811-12, and it seemed he
had been married in Massachusetts shortly before 1850. But I had no
date or place for any vital event. A check of the IGI showed:
STEVENS, George Cook, son of George Stevens and
Savilla Hitchcock, born 16 Jan 1811 at Cheshire, New Haven Co., CT. LDS
ordinances: baptized 17 Feb 1966 LG [Logan Temple], endowed 25 Feb 1966,
sealed “pre-1970” Batch/film #0,448,102.[ii]
Without going into much detail at this point, the above information
led me to a family group sheet referring to the “Greely Fam., p.
698-9.” Indeed George H. Greeley, Genealogy of the Greely-Greeley
Family (1905) treats George’s wife, Mary Ayer, and gives all vital
data one could wish (though without sources). It would have otherwise
taken quite a piece of serendipity to lead me to George Cook Stevens in
that book.Finding the book that included George was relatively
simple since I knew how to deal with a 1966 IGI entry. But what about
his grandfather, Hubbell Stevens? There were two Hubbells — a father
(who married twice) and his son. So, ideally, there would be five
entries under the name: two birth records and three marriages. But I
find over twenty! How can we evaluate and interpret these entries?
Which are worth pursuing? More on that later.
Some hints for searching the IGI
There are differences between searching the IGI on the old DOS
version of FamilySearch, which is available at Family History Centers
(FHCs), and searching it online. However, both programs allow you to
search in three different ways, for births or christenings of
individuals, for marriages, and for children of certain parents.
Trial and error is probably the best teacher in this case. There are
many nuances, which you will pick up with experience. For example,
when doing a parent search in the DOS version, you will need to search
three times, once with the wife’s full name, once with only her given
name and once without a wife’s name. Online you must enter at least the
wife’s given name. The search will find both those with and without
her surname at the same time.
Actually, a search of the DOS version can be faster because you will
be able to see the spouse’s or parents’ names on the index screen, which
will help you decide whether to move to the detail screen. However,
the DOS version is static, containing entries made only up through the
beginning of the year 2000, while the online version is updated weekly.
When searching the IGI by itself online, you have some options that
are not available when doing a search in all the databases available at http://www.familysearch.org/.
For example, you can locate a woman by only her given name if you
include her husband and/or other limiting parameters in the search.
Submissions in different time periods
In order to make best use of the IGI, you need to understand that
entries have been made in different ways at different times. We can
identify four distinct periods during which different types of records
were generated. While this process may seem complex at first view, it is
not really very difficult. These four eras are:
For the first two periods there is a card index on microfilm,
compiled by the Temple Records Index Bureau (commonly called “TIB”).
The latter two periods are fully indexed in the present IGI, and church
members have nearly finished adding in names from the first two periods.
First, though, it would be helpful to understand what the Mormons (or
LDS, as we call ourselves) are doing with the IGI. This huge database
was designed to keep track of “temple work” undertaken on behalf of
deceased persons. Temple “ordinances” are performed by living church
members as proxies for the deceased: baptism for the dead (c.f. I
Corinthians 15:29), the endowment (a sort of course of instruction),
and the sealing of married couples and of children to parents (c.f. Matthew
16:19). Mormons believe that many persons who have died are waiting
for these ordinances; others may yet decide to accept them. Since it is
an index to temple work, and since many members of the church do not
(or in the past could not) check to see whether work has been done, many
temple ordinances are unnecessarily repeated. This explains why there
are multiple entries for the same person.
From its inception, church leaders also visualized the IGI as a tool
by which family historians could locate records in hitherto
difficult-to-access or unexpected sources. The IGI only began in 1969,
however; Mormons have undertaken genealogical work since 1836.
The first era (pre-1942)
Before 1942 members took lists of deceased relatives to the temple
with them. The temples maintained separate ledgers of baptisms,
endowments, and sealings, arranged chronologically. The patron’s name
appears at the head of a group of names as the person at whose
“instance” the work was undertaken. The Church does not have
accompanying family group records from this era, so there is no direct
way to obtain further information. However, the TIB card index
generated during this period can lead you to other material. The TIB is
only directly accessible by church members in Special Collections at
the Family History Library, but anyone else can address a letter to the
Photoduplication Department, Family History Library, 37 North West
Temple, Salt Lake City, UT 84150. Include a check for one dollar for
Most, but not all, of these names have been “extracted” from the old
temple record books and are now in the IGI. You will recognize these
entries by the word “relative” in place of parents or spouse when you
are searching the DOS version of FamilySearch. Online, the entry will
lack the name of a spouse or parent. When you click on the film number
given in the entry, you will find yourself in the Family History Library
Catalog (FHLC). When you click on the title and then on “film notes”
you will find yourself at the beginning of a long list of microfilms.
Scroll down until you find the film number and look at the date of the
records on that particular film.
Most of the time you can order that microfilm at an FHC. Before
ordering, read carefully to see if the film is restricted. Some of
these films can be consulted only in Salt Lake City, so you may need to
hire someone on the spot. However, most are no longer in Special
Collections and can be viewed by anyone. Use the page and reference
number given in the IGI entry to find the name.
These temple volumes may or may not be helpful, depending on the
details given and on the surrounding names. Searching for an Edwin J.
Mills in Michigan, I found him in the temple book among a group of
Howlands. This was a clue to look at a Howland genealogy. Edwin had
married Cornelia Howland.
The second era (1942-1969)
You may find, when you look at the date of the temple book, that it
falls during this second period. From 1942 to 1969 temple patrons
submitted names on family group sheets. The patron’s name, address, and
sources of information[iv] were listed. The “Family Group
Records Archives” (FGRA) contains most family group sheets submitted
during these years and is available on microfilm (where the George Cook
Stevens family group sheet appears). While there are indeed numerous
errors and omissions on these sheets, many are the result of meticulous
research or reflect otherwise inaccessible family records.
To find these, search for film number 1275000 and you will find the
whole series of films. These sheets are alphabetical by head of family,
usually the husband. If the person is a wife or child, and the spouse
or father is unknown, a TIB card may make a connection.
In compiling family group records during this period, patrons
referred to the earlier TIB cards. On many of these sheets the baptism
and endowment dates are pre-1942, but the sealing dates are later. So,
if you see by clicking on the film number for the temple record book,
that an ordinance was performed between 1942 and 1969, you can be fairly
sure of finding a family group sheet. You may also find one for
entries of the earlier period.
Generally, you will derive more information from the TIB card and
family group sheets than from the temple record book films referenced in
the IGI. A “P” or a “C” in the upper left corner of the TIB card
indicates the person appears as a parent or child on a family group
sheet. If you request a TIB card by mail, you should also receive the
relevant family group sheets.
On the other hand, you may want to see the group of names submitted
by the patron. In this case, consult the film referenced in the IGI.
The third era (1969-1990)
In the 1960s the Church began using computers and volunteers to
extract births, christenings, and marriages from many countries. And in
1969 the system for member submissions was again changed. Now there
are two major types of submissions for temple work. When you look at
the source information, you will almost always find that the origin of
the entry is either “submitted by a member” or “extracted from” the
church or civil vital records of a particular place. Both types of
entries have “batch numbers.”The more you understand about the
batch numbers used since 1969, the easier it will be to know what lies
behind any given entry. For member submissions, family group sheets
were now only used when persons could only be identified in
relationship to parents or siblings. Most of the time, individuals and
spouses were submitted on “entry forms.” Three persons or three
marriages appear on each form, with sources and the name and address of
the submitter. When an entry form arrived in Salt Lake City it was
assigned to a “batch,” the number beginning with the last two digits of
the current year (e.g., a batch beginning with 73 or F73 arrived in
1973). The entry was assigned a sheet (i.e., page) number within that
batch, and many batches were then microfilmed together. The resulting
film is often called the “input source.”Names derived from
extraction projects also have batch numbers, which usually begin with a
letter of the alphabet. In this case, the original records, e.g. Irish
civil records, are the “input source” microfilms. You can see what the
input source is by clicking on the film number and reading the
consequent FHLC entry. Then decide whether to order the microfilm to see
the original record. In the case of extractions, an IGI entry may
sometimes also refer to an alphabetical printout. You can order this
too and use it as an index to the original record.
With member submissions from this period, you have two alternatives.
You can order the input source film and look for yourself, perhaps
finding a whole group of forms submitted by the same person. Or you can
use a photoduplication form (available at FHCs) to send for a copy of
the entry, which will include the patron’s name and address, source(s),
and perhaps additional data on the person you seek.
Carefully read the fine print on the photoduplication form. There
are exceptions, particularly in regard to New England extraction
projects, which have all-number batch numbers. Also be aware that batch
numbers beginning with F, 50 and 60 are patron-submitted family group
sheets. A batch number beginning with “A” indicates that there may well
be a family group sheet in the FGRA.
The fourth era (1990 to the present)
About 1990, members began submitting GEDCOM files on diskette. Patron
names and sources are not presently available. However, these entries
may include links to family members and other information, such as death
dates and places.
In this case it may be helpful to refer to the Ancestral File, which
was conceived as a great matchmaker among family historians. In fact, all
genealogists were invited to contribute and make corrections to the
Ancestral File. Submitting to this file did not initiate temple work.
If you disagreed with data in the Ancestral File, you used to be able to
make changes and document your information. Then searching for
“History of Changes” would provide documentation.
However, no changes have been made to the Ancestral File for many
years. The new Pedigree Resource File is the Church’s current collection
of material submitted by GEDCOM. In this resource each individual
submitter’s file remains intact. Evidently a new database is being
The extraction projects will continue, and each extracted entry in
the IGI will cite the original document.
So what of all those entries for Hubbell Stevens? They are indeed
bewildering, but intelligible. (The pre-computer TIB correctly has only
three cards — for Hubbell himself, “Mrs. Hubbell,” and Hubbell Jr.)
Five entries in the IGI are for “Mrs. Hubbell,” submitted before either
wife was identified. Of the many remaining entries, two are for Hubbell
Jr.’s estimated birth date and four for his marriage to Elizabeth
Clark. The original marriage sealing record was evidently typed
“Hubber” so a patron submission in 1973 was not recognized as a
duplicate. For some odd reason it is itself duplicated.
We are still left with at least twelve entries for Hubbell Sr., most
for his birth. (1) The earliest, a name submitted for baptism in 1920,
gave the wrong place of birth. (2) In 1934 a baptism was submitted
again, this time with the correct birthplace. (3) In 1972 a patron
again submitted the name with 1934 ordinance dates, but he or she had
evidently not consulted the FGRA, so the sealing (performed in 1946) was
repeated. (4) In 1974 the Connecticut vital records extraction program
noted his birth. The program recognized the earlier 1934 entry, but did
not catch the 1972 patron submission, perhaps because the place name
was spelled differently. (These initial four entries should, of course,
have been more than enough.) (5) In 1987, a patron resubmitted the
name. It was “cleared” again for temple work, but by the time of the
1992 edition of the IGI the 1946 sealing date to parents had been found
and correctly added. (6) For some reason, however, in 1992 a patron
submitted an entry with dates, which, as far as I can see, are
completely imaginary! Since 1992 three more entries appear, the result
of patrons not bothering to do research and not even checking the IGI
before submitting their own guesstimates as to birth date and place. So
the computer thinks these entries are for new people. Newer is not
This story highlights the need to research thoroughly and check for
previous temple work before making a submission.
Turning to Hubbell’s marriages, we find at least six entries: one to
“Mrs. Hubbell,” three to Anna Shepard, and two to Deborah Jones. The
earliest with correct names are my own submissions (1986). Because
these submissions were not available to another researcher they were
duplicated around 1990. Then there are recent ones with imaginary
The question remains, however: what is one to do with all these
entries? A request for TIB cards or an FGRA film would yield the 1946
family group sheets. Of post-1969 submissions, the batch number
beginning with F86 (1986 family group sheets) suggests that two recent
family group sheets identify two different wives of Hubbell, Sr. You can
view these records on film, or order copies using the photoduplication
form mentioned above. Later entries can only send you to the Internet or
to published material — you will find my article about Hubbell Stevens
by using PERSI at www.Ancestry.com.
“How likely am I to find the person I am looking for in the
There is no simple answer to this frequently asked question. In the
first place, temple work is only undertaken for persons known to have
died. Thus, in a program extracting birth records, it is assumed that
anyone born less than 110 years ago might still be living. For marriage
records the cut-off date is 95 years earlier. Many of the extractions
were done in the 1970s and 1980s. Thus you are not likely to find
someone born after about 1875, unless an LDS relative supplied the death
date in his member submission. Many of the extractions from New
England were taken from the published “Vital Records to 1850” series.
Thus New Englanders in the IGI probably lived between the early 1600s
Evaluating the entry
In using the IGI, those much-maligned terms “primary” and “secondary”
are helpful. If one understands a primary source to be one created at
or near the time of the event (or at least by the person[s] directly
concerned), then a secondary source would be someone else’s use of that
primary source. The IGI is at best secondary. But very often it is
“tertiary” material. At times one might even call it “quaternary.”
Among Kent, Connecticut, vital records I was puzzled by the appearance
of a James Swift, born December 25, 1767. A Tamer Swift was baptized
in 1769, but who was James? Microfilm of the original town record
appeared to say “Tamer, daughter,” but I could see how the transcriber
had read it as “James.” The IGI contains extracted Connecticut vital
records from the Barbour Collection, which in this case used an earlier
copy of Kent data by James N. Arnold. We are now three steps removed
from the original, hence quaternary.
In the case of a member-submitted entry, it is a good idea to check
the source and/or look for verifying material elsewhere using the entry
as a clue.
One final example
I’d just about given up looking for what became of John Belden Mills,
last known to be in Saybrook, Connecticut, in 1824. The 1850 census
listed a John B. in Utica, New York, born in New York, with wife Amelia
and two children. Could he be the right person? Checking the IGI for
John Belden Mills as a parent I found five children born in
Nantucket. That was a surprise! The batch number led me to the
published Nantucket vital records, which had used a Bible that for once
gave places of marriage and birth, including John’s known date and place
of birth in Chester, Connecticut. John and Amelia were married, and
their children born, in upstate New York and Canada! Here was a good
lesson in using the IGI as an index, not a source.
A little bibliography
For more detail see the LDS research outlines, The International
Genealogical Index (On Microfiche), the FamilySearch instructions
for the International Genealogical Index (On Compact Disc) and Finding
an IGI Source, as well as the IGI Reference Guide (on microfiche
“Z” of the IGI itself). These free guides are available at your local
Family History Center and on their website.
Articles on the IGI by G. David Dilts, AG, and/or Elizabeth L.
Nichols, AG, appear in Genealogical Journal 20 (1992):5-21, Genealogists’
Magazine 24 (1993):294-97, 349-53, and FGS Forum 5 (Winter
1993):5-10 and 6 (Spring 1994):4-6.
[i] © May 2003. This article is an update of my earlier work, “A
Perspective on the 1992-93 IGI” which appeared in the New England
Historic Genealogical Society’s NEXUS, Vol. X, Nos. 5 & 6
[ii] The IGI is updated weekly on http://www.familysearch.org/
, but only Church members who register can view these dates online.
However, anyone using FamilySearch at a Family History Center can go to
the menu item “LDS Options,” and view the IGI as “Ordinance Index,” (the
same IGI but with the ordinance dates). Keep in mind though, that the
old DOS program, which uses CDs, only contains entries through about
[iii] Or you may be able to obtain a Temple Ordinance Index Request
form (TOIR) at a Family History Center, but they are out of print.
[iv] These sources are often cryptic, sometimes incomplete. To
translate into present-day call numbers, call the Family History
Library, ask for the first floor library attendant’s window, and explain
carefully what you want.