McGillivrays of Islay Project

Search the Database

This project began in 1997 when I compiled a spreadsheet of Islay McGillivray births and marriages from the Old Parish Registers (OPR). I initially created the spreadsheet of McGilvray OPR records of Islay to help determine who the parents of my Donald McGilvray were.

More recently two on-line sources have enabled me to greatly expand this work and think that a family tree of the McGillivrays of Islay was feasible. The first on-line source is the work of volunteers of the Rootsweb Islay Mailing List and particularly Ted Larson who have transcribed the entire OPR for Islay, the civil birth, marriage and death registrations from 1855 to 1875 and the complete censuses of Islay for 1841 and 1851. The second invaluable on-line source is the work of Roy McGilvray who has transcribed all of the McGillivray records in the Scottish OPR and in the civil registrations from 1851 to 1920.

The combination of these two transcription projects is perhaps unique - providing depth in both the Islay records and those of the McGilvray surname in Scotland beyond Islay. The Islay transcriptions allow us, for example, to continue to trace the families of McGilvray women after their marriage, often for two generations (since the McGilvray mother is named in the civil marriage registration of their children). Roy McGilvray's site allows us to trace McGilvray families after they left Islay, as many of them did, particularly emigrating to the Glasgow area for employment.

The sorting abilities in Excel help in finding matches in the names and parents and locations and therefore assist is reconstructing family groups from the records that may contain inconsistencies (e.g. in the mother's given name). However, this work can only yield isolated nuclear families. The OPR does not contain enough information to link family groups in family trees of two or more generations. Many people are repeated in various nuclear families - once in a baptismal record, and later as a bride or groom and parent. However, there was no way to identify these duplicates.

With the availability of the civil registration records this limitation disappears, because the marriage and death records usually record parents' names. With this prospect, I entered the nuclear families I had identified in the OPR in a family history program (PAF 5.0 in my case). I then added the information from the civil registrations and other sources, merging the records of individuals who proved to be the same (one record as a child, the other record as a parent). The two censuses help in eliminating options and sometimes provide additional clues when a grandparent is living with the family of one of his/her children, or two families are living side by side.

A further valuable source where known is overseas records, such as death and marriage registrations, and cemetery transcriptions. Anything that links someone to both a spouse and parents enables us to link generations.

The original 35 or so McGilvray families in the Islay OPR have grown to over 180 families and 750 individuals. The database, called Islaymcgillivray, is posted on Rootsweb's WorldConnect site at  McGillivrays of Islay.

The following observations are offered in the hope that this project will be a catalyst for others to work towards a complete family history of Islay.

In creating the McGillivrays of Islay project I came to several policy decisions:

1. Standardize spellings of given names, surnames and places. This facilitates searching; in fact the database would be practically useless without it. It is my belief that the variations of spelling we find in the OPR and Civil Registrations reflect the practice of the writer, not the subject. In particular I adopted the spelling McGilvray. If necessary, "Actual Text" fields can be used to hold the original text of the record.
   • It is less important what the standardized name is. Once the names are standardized, it is a simple "find and replace" to change, for example, McGilvray to McGillivray globally in the database
2. Do not follow family lines after they leave Islay. Where these lines are available elsewhere on the Web, I use the hyperlink feature in WorldConnect to link to them. (I stretch this rule in the case of following McGilvrays who emigrate to mainland Scotland as recorded in Roy McGilvray's files.)

I learned a few things from this project about the assumptions we make as we do genealogy:

1. Like biologists who study speciation, genealogists go through phases of 'lumping' and 'splitting'. Those who 'lump' assume people who appear to be the same are in fact the same person, while 'splitters' treat them as two persons until there is strong evidence to the contrary. The experience of building this database has taught me that 'lumping', e.g. based on the OPR only, is dangerous. That the person of interest 'has to be' the one whose baptism is recorded is a fallacy. Baptisms are missing, even in families where many children were baptized. There are families who appear to be well documented in the OPR but a marriage or death record reveals another child. At the extreme, some families are missing from the OPR entirely. And, of course, some OPR registers do not exist at all, such as the marriage register for Kilmeny.
2. Conversely, the censuses remind us that the number of McGilvrays on Islay at a point in time is finite. There were 164 of the name in 1841 and 129 in 1851. Cross-matching the census families with those in the OPR often leads to reasonable conclusions. (If I have four Barbara's in my database but only two are named in the census, then some merging is needed.) At this point I have identified 133 (81%) of the 1841 names and all but 17 of the 1851 names as belonging to families known from the OPR and civil registrations. In addition, I have identified many more people of other surnames with McGilvray ancestry.
3. Nevertheless, some 'strays' exist - people who are in only one record and not in any other, and for whom there is not enough information to include them in a family group. I did not enter such people in the database unless there were at least two records (e.g. an OPR and a census) for them.
4. Islands were not as isolated in the 19th century as we might imagine. There was in-migration as well as emigration. It is not a closed system and we cannot assume that all problems will be resolved from Islay records alone.

I encourage others interested in extending this project to get in touch. It is my vision that the family trees of smaller, isolated communities such as Islay and Tiree can be "done".