Kirk Session Records Name the Father. Here Is What They Actually Prove.

The baptism entry is short, and it is missing the one thing you came for. A child, a date, a mother's name, and a single word where the father should be: natural. Sometimes baseborn, sometimes reputed, sometimes nothing at all.

That is usually where the line stops. It does not have to. In eighteenth-century Scotland an unmarried pregnancy was not only a family matter. It was a case, heard by a church court sitting in the same parish, and the first thing that court asked the mother was who the father was. The answer went into a minute book that most researchers never open, because it is not a register of births and nobody indexed it by name.

The catch is that a name in those minutes is not automatically a father. The Church of Scotland knew that perfectly well, and it wrote down its own rules about when a woman's word was enough and when it was not.

The short answer. Kirk session minutes, the records of each parish's church court, routinely name the father of a child the baptism register left blank, and they can record cases the baptism register missed entirely. But the Church's own Form of Process of 1707 said a woman's testimony alone was "no sufficient evidence" against a man who denied it. So what a minute proves depends on what it records: a confession, a naming the man accepted, a naming he denied, or an oath clearing him. Work out which of those you are holding before you add a father to your tree.

The baptism register is the weaker witness

Start with the fact that changes where you should look first. The parish register and the session minutes are two records of the same events, kept for different purposes, and they do not agree.

Leah Leneman compared them for the Perthshire parish of Blair Atholl in Local Population Studies in 1983. For three five-year periods, 1718-23, 1748-52 and 1775-79, the kirk session minutes recorded 37, 47 and 32 illegitimate children. The baptism register for the same periods recorded 20, 24 and 13. In every period the register held somewhere between two fifths and a little over half of the cases the session dealt with.

So a missing baptism is not evidence that a child was not born in the parish, and a baptism that names no father is not the last word on who he was. The register recorded a sacrament. The session recorded a scandal, and it pursued the scandal until it got an answer or ran out of ways to get one.

When the register does mark a birth as illegitimate, ScotlandsPeople's guide to registering illegitimate births lists the words to look for: natural, baseborn and reputed, or a parent described as an adulterer, fornicator or fornicatrix. Occasionally the register records the failure of the whole process. A Livingston entry for a child born in December 1805 and baptised in February 1807 says the child "could not find a Father" at its baptism, and that a confession "Could not be extorted from the Mother."

The Church wrote down its own rules of evidence

A kirk session was the lowest court of the Church of Scotland: the minister and elders of a single parish, sitting as a court. Above it were the presbytery, the synod and the General Assembly. Its minutes are, in the words of the National Records of Scotland, a record of discipline for offences such as drunkenness, swearing, breaking the Sabbath, quarrelling and "sexual misdemeanours".

In April 1707 the General Assembly approved a Form of Process, the procedure every session was meant to follow. Its fourth chapter deals with fornication and adultery, and it reads like a rulebook for weighing paternity evidence, because in practice that is what it was.

The mother had to name someone. An unmarried woman known to be with child was to be summoned and "interrogated who is the father of that child." If she would not say, she was to be treated as contumacious, in contempt of the court.

Her naming did not convict him. If the man denied it, he was to be summoned and confronted with her, but the text is explicit that "the woman's testimony can be no sufficient evidence against him." The session could still lay the guilt on him where there were strong presumptions, the standard example being the two of them found alone together in a suspicious place.

A case could be left open. If he kept denying and the evidence was not there, the session could lay the facts before the congregation and declare that there could be no further procedure "till God in his Providence give further light." A minute like that records an unresolved paternity question, and it should be read as exactly that.

The oath of purgation cleared the scandal, not the question. A man could offer to swear his innocence. The session was never to press the oath on him, and was to allow it only on the presbytery's advice, where the suspicion was so strong that nothing but his oath would remove it. If he took it, the scandal was lifted from him. That is a statement about what the court could establish, not a finding that he was not the father.

Then the woman was asked again. Once a named man had cleared himself, she was "to be dealt with to give the true father." If she named nobody else, she was censured without the first man's name being repeated, and the session kept the right to act "upon further discovery."

The case was heard where she lived. When the two parties lived in different parishes, the process belonged to the session of the parish where the woman lived, or where the scandal was most notorious. So if the man you are tracing lived in one parish and the woman in another, his case will usually be in her parish's minutes, not his own.

That rulebook gives you a ladder. Reading down it tells you how much weight a given minute can actually bear.

What each kind of entry is worth

A confession. The named man admits it. This is the strongest entry a session minute can give you, and it seems to have been the usual outcome. Leneman's 1983 study of two Perthshire parishes found that the usual pattern was for the woman to be called, asked to name the father, and for the man, when named, "normally" to admit his guilt rather than deny it.

The Inverness kirk session's own minutes, published in 1902, show exactly this pattern. In October 1722 the session recorded that a woman had left the parish but that a named man "had owned to be Father of the Child." The following month he compeared and confessed. That two-step, a report of the man's admission followed by his own appearance to confirm it, is the shape a strong entry takes.

A fine paid, no denial recorded. In February 1726 the same Inverness session recorded a woman's confession naming a man who had by then emigrated to America. He never compeared, because he could not. Instead his brother paid three pounds sterling to the poor's fund on his behalf, described in the minute as his own fine, sent home to be settled in his absence. It is not a confession in his own voice, but it is not nothing either, and a genealogist reading it should treat it as strong circumstantial acceptance rather than proof.

A naming the session accepted conditionally. In March 1709 the same session recorded a woman naming a serving sailor as the father, and admitted her testimony, but explicitly on the condition that if he denied it she would have to name someone else. That is a minute recording a provisional acceptance, not a settled fact, and it should be followed up against whatever record exists of what happened when the man was actually confronted.

Denial and an unresolved case. Where a man denied and the session could not move him, the minute should say so, and the Form of Process gave the session a set form of words for leaving the case open until God gave "further light." Treat this exactly as the Church did: unresolved.

The oath of purgation. A man who swore innocence, with the presbytery's approval, cleared himself of the scandal. Genealogically, this is the entry most likely to be misread. It tells you the church stopped pursuing him. It does not tell you he was not the father, only that the session had run out of ways to press the case further, which is exactly why the oath was reserved as a last resort rather than offered freely.

"We know not to whom." Some entries end there. In May 1726 the Inverness session recorded a woman who had borne a child and simply did not, or would not, say who fathered it. That is the floor. No further inference is available from the record itself.

Two cautions before you build on any of this. First, published transcriptions carry the errors of the clerk and the transcriber both: the Inverness volume names the same woman two different ways within a few weeks of the same case, an inconsistency that could belong to the original clerk, the 1902 printer, or a scan you are reading today, and any of those is a reason to check the original image rather than the printed text before you rely on a spelling. Second, an antenuptial fornication case, where a couple married and then had a child inside nine months, is not a paternity dispute at all. The Cadder session in 1801 rebuked exactly such a couple after a midwife's certificate failed to satisfy them that the birth was genuinely early. Read the whole entry before assuming a "fornication" heading means an unresolved father.

Statutory registration changed the rule, not the practice

Civil registration of births, deaths and marriages began in Scotland on 1 January 1855, under the Registration of Births, Deaths and Marriages (Scotland) Act 1854. For an illegitimate child, the Act set a specific rule: a registrar could not enter a man's name as father unless both the mother and the man himself requested it, and he signed the register as informant alongside her. Where a court later established paternity by decree, the clerk of that court had ten days to notify the registrar, who then added the father's name and marked the entry accordingly. That decree route left a trace worth searching for on its own: a note in the birth entry's margin referring to the Register of Corrected Entries, the mechanism used to add a father's name after the original registration.

Two consequences follow directly from that rule, and both are easy to miss. A birth register with no father named after 1855 tells you only that neither the mother nor an acknowledging man asked for one to be entered, or that no court decree had yet been obtained, not that paternity was unknown or unrecorded elsewhere. And a father's name that does appear was either volunteered jointly at the time or added later by a Register of Corrected Entries reference, which is itself worth tracing back to the court process that produced it.

Kirk sessions did not stop disciplining people in 1855. The historian Stewart J. Brown, writing on the end of church discipline in Scotland, points to a Church of Scotland commission report from 1896 that described the "standing of the session" as still in use, if "little regarded" by some, four decades after civil registration began. One genealogy researcher's own comparison of Fife kirk session minutes against the same parishes' early statutory birth registers, published online rather than peer reviewed, claims the minutes named fathers the birth register left blank, at rates that varied enormously by parish and which the author called provisional; the same source suggests that where a father was identified this way, he turned up as often as not in a different parish's minutes than the one where the birth was registered. Treat that specific comparison as a single researcher's unverified finding, not an established figure, but treat its shape as a reasonable expectation: if your ancestor's statutory birth entry is silent on the father, the session minutes of both parishes involved are still worth checking.

Two other places a father can turn up

Legitimation by a later marriage. Scots law has long recognised legitimation per subsequens matrimonium: if the parents later married and were both free to do so at the time of the birth, the child was legitimated. ScotlandsPeople's own guidance notes that such a birth may carry a Register of Corrected Entries reference recording the legitimation, after which the birth is indexed retrospectively under the father's surname. This is a meaningfully different rule from the one south of the border. England did not recognise legitimation by subsequent marriage at all until the Legitimacy Act 1926 took effect on 1 January 1927, and even then it excluded a case where either parent had been married to someone else at the time of the birth. A child legitimated this way in Scotland in 1810 had a status that an English child in the same position could not obtain for more than another century. What England offered instead was entirely at the father's discretion: naming the child in a will. An illegitimate child inherited nothing automatically, so a bequest was a deliberate act rather than a formality, which is what makes an English will one of the few places that relationship was ever written down.

Sheriff court paternity actions. Alongside the church's process, a woman could bring a civil action in the sheriff court, seeking aliment, meaning maintenance, from the child's father. According to Scottish Indexes, these actions of affiliation and aliment produced their own decree registers, generally beginning around 1830 in most sheriff courts. This is a genuinely separate paper trail from the church one, worth checking even where the session records for a parish are missing or silent.

Vocabulary you will meet

Finding the records

Kirk session minutes are held by the National Records of Scotland, whose church court records amount to more than 25,000 volumes and roughly 5 million pages. One dating trap is worth flagging before you search: Scotland resolved the year-start question early, moving its legal New Year from 25 March to 1 January in 1600, but it kept the older Julian day-count until 1752, alongside England, so a minute dated in January to March needs no year adjustment, but any date before 1752 is still on the old day-count. The mechanics of that kind of calendar drift, and how to convert a date without silently guessing a year, are covered in reading dates in old records. ScotlandsPeople's Virtual Volumes service, launched to make this material reachable online, now hosts more than a million pages from over ten thousand individual kirk session, presbytery and other church court records, spanning 1560 to 1900, though ScotlandsPeople itself cautions that only a relatively small number of records are online so far, and the rest have to be found through the National Records of Scotland catalogue. Browsing what is online is free; downloading or saving a page costs two credits, 50p, as of when this was checked. That split between free browsing and paid saving is what makes a page-by-page read through unindexed minutes affordable at all, and it is unusual: most archives price the search itself rather than the page.

The service is honest about its own limits. The records are not indexed by personal name, so there is no way to search directly for a case by the mother's or the father's name. You search by parish and by date, which means you need a starting point: a baptism, a birth, or at least a rough date and place to search around. One specialist index, the Scottish Indexes paternity project, has begun indexing kirk session and sheriff court paternity material by name, but as of this writing the kirk session portion covers only one parish, Canonbie in Dumfriesshire, so it is a useful spot-check rather than a general search tool.

A practical starting method, the one Scottish Indexes recommends: once you have a birth or baptism date, start in the minutes a few months before the birth and read forward. The case is often heard shortly before the baptism, but some turn up years later, and the child is not usually named in the minute, so you are reading for the mother's name, not the child's.

What KleioBase does with a case like this, and what it does not

Disclosure, since we sell a tool.

A kirk session case is rarely one document. It is a sequence of entries, sometimes months apart, sometimes in more than one parish's minute book, and the value is in seeing them together rather than reading any one in isolation. KleioBase keeps that sequence together as one connected archive: each minute you upload becomes a record linked to the people it names, and the Research Companion can read across those records, up to five attached to a single question, when you ask what the session actually concluded. On the way in, each page's people, dates and roles are extracted and the original wording is kept. You can add a context note before processing, telling it the page is a kirk session minute rather than a baptism register. Every extraction goes through a review step before anything is added to your knowledge base, so a misread name or an ambiguous "owned" versus "denied" is caught by you, not assumed by the software.

Once confirmed, a session entry becomes a record attached to the people it names, and it sits alongside the baptism entry for the same child rather than replacing it. Where the two disagree, both stay visible on the person's profile, tagged to the document that made each claim, which is the only honest way to represent a case where the baptism says nothing and the session says something contested. If you searched a run of minutes and found nothing, KleioBase's negative evidence feature lets you record exactly what you searched and over what date range, so that absence is documented rather than silently repeated by the next researcher in the family. That habit of recording what disagrees, and what turned up nothing, is the same discipline covered in the evidence layer your research log is probably missing.

Three honest limits. KleioBase does not hold ScotlandsPeople's images itself; you still need to view and download the page yourself before you can upload it. Nothing in the software judges what a given entry proves: it will transcribe a confession, a denial or an oath of purgation alike. A deep scan tags names it was not sure it read correctly, and even that warning does not catch every misreading; nothing flags an entry whose meaning is in dispute, and the evidence ladder in this article is not a setting you can turn on. And the family tree has no relationship type for a man named but not proven to be the father; the closest documented option is the dashed "Unknown father" placeholder, so a disputed or provisional naming from a session minute has to be recorded as a note rather than a structured fact until you are satisfied it is settled. That judgement is the part of the work that stays yours.

Where to start

Find the baptism first, even an incomplete one. Then search the kirk session minutes for the mother's parish, starting a few months before the baptism date and reading forward, since that is usually where the process was heard, wherever the father lived. If the case does not appear there, or if you are trying to trace the man's side of it, check the minutes of his own parish too, since the rule also allowed a case to be heard wherever the scandal was most notorious. And whatever the minute says, read it as the Church itself would have: a naming is a starting point, a confession is strong evidence, and an oath of purgation closed the case without ever answering the question you actually have.

The session that heard your ancestor's case was not trying to build a record for you three hundred years later. It was trying to decide whether to let a man and a woman back into full communion with their neighbours. That it left behind the best evidence you will ever find for a question the baptism register refused to answer is almost an accident of a completely different purpose.