One of the frustrating features about genealogy is that, in general, the further you go back the fewer records you find, and the more ambiguous they become. To complicate the matter, in the past given names tended to run in families and so you may find that a particular unusual surname is very common in a particular area and you can have several cousins with the same name and about the same age living in the same area. (See The Inheritance of Single Christian Names.) When there is information suggesting that the individuals moved between villages, being sure you have the right one is difficult. To complicate the matter many beginners assume that if someone has the right name and age he must be their ancestor without ever checking whether there is any other candidate for their ancestor. They also overlook the fact that all human records, including parish registers, etc., contain errors, and that all transcription and indexing activities inevitably introduce further errors.

In my own researches I have come across a number of cases where people of the same name get confused. In about 1700 my ancestor Robert GIBBS moved to Winslow, Bucks and had six sons, each of whom married and had large families - whose sons bore one or other of the six original given names. Some of the families joined non-conformist chapels and no records exist of the baptisms of their children. There result is that by 1800 there were many adults with the same name and no way of telling which was which - although one could be reasonably certain that most, if not all, were descendants of the original Robert GIBBS. Another of my ancestors (according to an old family tree drawn up when there would still have been people alive who could remember) shows that Abraham DEVONSHIRE married an Ann ROLLS, and when she died he married her cousin, another Ann ROLLS! Two of my wife's identically named relatives were born in adjacent villages within three years of each other and swapped villages when they married - a fact that cannot be determined by a study of the parish registers alone. In the 1851 census there were two John IVORY living in adjacent houses in Kimpton, Herts, while in 1815 Sarah BURCHMORE was very careful to identify which of her two son-in-laws called Ralph THRALE were which. (See Right Name, Wrong Body)

For reasons such as this it is important to realise that "proof" that someone is one's ancestor is often impossible to find, and often decisions have to be made on the basis of probability. In the case that triggered this posting the question is whether a Thomas ELSOM who was born (or baptised?) 23 May 1758, in Barkway, Herts was the Thomas ELSOM (or ELSUM) who married Mary GIBBS in Heydon, Essex in 1782. However we also know that a Thomas ELSON, son of Thomas and Elizabeth ELSON was christened at St Andrews, Hertford on 2nd July, 1760, and there may be other Thomas's in the Hertfordshire/Essex/Cambridgeshire area.

Trying to sort out this kind of problem can be time consuming, and will often require access to unindexed documents in the local records office or elsewhere. Some of the checks you may need to carry out are:

(1) Always look at the microfilm of the relevant parish registers as this not only allows you to check that the index is correct but may also contain additional information.

(2) Investigate the immediate families of all likely candidates. Did some candidates die in infancy. Did some apparently marry in their own village? If this happened the candidate is probably not the one who switched villages. This may not be the waste of time it seems as it is possible that they could all turn out to be cousins - and hence your relatives.

(3) Do family given names match? In the 18th century one would expect one son to be named after the father, another after the grandfather, and others after uncles, etc. Major deviations from this could suggest that the candidate is not the one you want.

(4) If occupations are known do they make sense. Many occupations run in families. Similar considerations apply to social status. Marriages outside one's social class or religious affiliations were much less likely than today.

(5) Are distances reasonable? A young farm labourer would think nothing to walking 10 miles or so to the nearest annual hiring fair looking for a job, and then a further 10 miles to his employer. Is there an obvious hiring fair which could cover the move.

(6) Do any indexed parish record lists occur for the period which could indicate who was living in a given village in a given year. In Hertfordshire there are militia lists for the second half of the 18th century which cover men of military age, and a very large number have been indexed by the Hertfordshire Family History Society.

(7) Are there any documents such as apprentice records or removal certificates in the relevant county records office? Do they hold unindexed records such as land tax returns, etc. Do they contain relevant information.

(8) Do manorial records exist, and if so where are they held, or are microfilms available. They can be very useful if the people you seek were copyhold tenants, as properties often passed from generation to generation, with each hand-over being documented.

Etc...

The basic rule is that one can never be certain as, for example, a husband may be cuckolded or a child informally adopted. The amount of checking needed to be reasonably certain can vary greatly, and it might prove quite easy to trace a particular Smith family for generations if they left wills, or there is documentary evidence that they occupied the same farmhouse for centuries. In other cases one may never be able to say for certain which of a number of candidates is an ancestor. The one thing one can be certain of is that the person who simply looks things up in computer indexes, or who assumes that information on the internet can be treated as accurate without checking is very likely to be building a spurious family tree.

You may also find it useful to have a look at I've hit a brick wall ... and Is this the right birth/baptism?