Español in One Record, Mestizo in the Next: Why Your Mexican Ancestor's Calidad Changes

You have found her baptism, and the priest wrote española. You have found her marriage, twenty years later in the same parish, and the priest wrote mestiza. Somewhere between the two there may be a census line that gives her no label at all.

The obvious conclusions are that one record is wrong, or that these are two different women. Very often neither is true. The label was not a fact about her ancestry that a careful priest could get right. It was a judgement about her standing, made by a particular person on a particular day, and it moved.

That changes how you should read every colonial Mexican record that carries one, and it changes where you should search for the records you have not found yet.

The short answer. In colonial Mexico a person's calidad, written as español, mestizo, indio, mulato or a related term, was not a fixed racial category. The historian Robert McCaa, linking 174 marriages in Parral recorded in 1788 to 1790 back into the town's 1788 census, found that about three-quarters of the people he traced carried the same label in both records, and a quarter did not. Women changed label at marriage almost twice as often as men, 26 of 101 against 13 of 91, usually toward their husband's calidad and usually downward. A changed label is therefore normal, not evidence of a different person. It matters for searching too, because some parishes kept separate baptism books for each calidad. An order of 17 September 1822 then prohibited classifying citizens by origin in any public or private document.

Calidad was a reputation, not a blood test

The word itself tells you what it measured. McCaa's article, "Calidad, Clase, and Marriage in Colonial Mexico: The Case of Parral, 1788-90", published in the Hispanic American Historical Review in 1984, describes calidad as typically expressed in racial terms, but in many instances an impression reflecting a person's reputation as a whole. Colour, occupation and wealth might all influence it.

So when a priest or an enumerator wrote a label, he was writing down how the community saw that person, filtered through what he knew of them. The Parral census shows the vocabulary sliding between the two ideas: McCaa notes that enumerators occasionally wrote someone as "de calidad mestizo" or "de color español".

The labels you will meet. The Denver Public Library's guide to Hispanic and Southwest genealogy lists the common ones, with the warning that labels and their meanings vary by time and place:

The Parral census alone used sixteen labels, including five for Native groups such as Yaquis and Tarahumaras.

Read every one of those as what the record-maker thought, not as a genealogical conclusion.

One town, measured: a quarter of people changed label

Most of what is written about calidad comes from the casta paintings and the theory. McCaa did something rarer. He took the brides, grooms and parents of every marriage recorded in Parral, a silver-mining town in northern New Spain, between January 1788 and December 1790, and searched for each of them in a manuscript census dated 1 January 1788. He linked people by names and ages only, so the label played no part in deciding who was who.

Of the 174 marriages, more than three-quarters had a bride or groom traceable in the census. His headline finding was that about three-fourths of the traced individuals carried the same racial label in the parish book and the census. The rest did not.

The changes ran one way more than the other. Of the people the census called Spanish, 18 women and 9 men were recorded as mestizo at marriage. McCaa concluded that most of the drift, for both sexes, was downward.

Money moved the label, and so did its absence. Two of six upper-class mestizo grooms were written as Spaniards at their wedding, against two of forty-five lower-class non-Spaniards. Going the other way, of nine grooms the census called Spanish who married as mestizos, seven were mine workers and one a servant. McCaa's summary is the line most often quoted from the article: money whitened (a lack of it darkened).

The study has limits, and McCaa states them. It covers one frontier mining town, and he says plainly that he makes no claim that Parral was typical of colonial Mexico. The number of traceable people is about a hundred, small enough that his discussion keeps returning to what the unknowns might do. A second study points the same way: McCaa reports that Patricia Seed's work on Mexico City's 1753 census found the label disagreed in 14 of 108 linkages. That is a smaller share than Parral's, and McCaa cautions that Seed's study does not clearly distinguish links forward and backward in time, so set the two figures side by side as evidence that changes were common, not as a rate.

Marriage pulled the label toward the spouse

The most useful finding for a family historian is not how often labels changed but when.

McCaa found that women were almost twice as likely as men to change their calidad at marriage: 26 of 101 women against 13 of 91 men. He explains the difference by work: status in the workplace marked men more than women, so a man's label was harder to move. A woman's label at her wedding depended much more on the man she was marrying.

The change usually closed the gap between the couple. For every man whose label changed, the change made the marriage endogamous, the couple recorded with the same calidad. Of the 26 women, 17 drifted into endogamy. McCaa's interpretation is that bride and groom were often closer in calidad, in the broad sense people of the time understood the word, than even the labels in the parish book suggest.

So the marriage record is the least independent label you have. A census label was written about a person in her own household. A marriage label was written about her as half of a couple, and was pulled toward her husband's. It follows, and McCaa makes the same point in a note, that couples gravitated toward each other's calidad once married, so a census taken after the wedding inherits the pull. If you want the label that describes a woman before her marriage, the record made before it is the one to weigh most.

A census might leave the label off on purpose

A blank calidad in a census does not mean the enumerator forgot, or that the person had none.

McCaa found that the 1788 Parral census gave no racial label for 2,367 of its 4,933 inhabitants. That looks like carelessness until you see the pattern. The race of children, and in some cases of wives, was omitted fairly systematically, and a married woman's details were often left out when her husband's were given. Among adults whose age was given, 15 per cent of women lacked a label against 7 per cent of men.

The practical rule. A child without a label in a census was usually not being described as anything. Do not read the blank as a hint that the family was in an unrecorded category, and do not fill it in from the father's label as though the record said so.

Some parishes kept the books apart, so the label decides where you search

This is where calidad stops being a matter of interpretation and becomes a matter of finding the record at all.

The Denver Public Library's guide warns that during the colonial period some priests kept separate sacramental registers for Native Americans and others. You can see the practice working in a parish's own words. In 1752 a foundling girl about four months old was left at the door of a household in Toluca with a note saying she had already been baptised. The priest at the parish of San José, El Sagrario, wrote that he searched the current books of baptisms of españoles, of mestizos and that of the indios, found no entry for her in any of them, and baptised her conditionally. The entry is transcribed in María Guadalupe Fuentes Jordán's 2025 master's thesis for the Universidad Autónoma del Estado de México on wills in the valleys of Toluca and Ixtlahuaca, which also cites a Libro de Bautismos de Castas for 1729 to 1761 from the nearby parish of Nuestra Señora de los Ángeles, Tecaxic.

What that means for your search. If a parish kept two or three series of baptisms, an index or a microfilm reel may hold only one of them. A baptism that should exist and cannot be found may be sitting in the other book, under a label you did not expect, because the priest judged the family differently from the way their later records did. Before you record the search as empty, find out how many series the parish kept for those years and check each one.

When you do come up empty, write down which books you searched. That makes the gap negative evidence rather than an unanswered question, and it stops you, or the next researcher, repeating the same search in the same book.

After 1822 the label was meant to vanish, and did not, all at once

Independence changed the law faster than it changed habits.

The historian Mirian Galante, writing in Nuevo Mundo Mundos Nuevos in 2018 on court cases in the Tlaxcala region, traces the legal end of the categories. She notes that the Plan de Iguala of 1821 already ruled out classifying people by their origin, and that an order of 17 September 1822 expressly prohibited classifying citizens by origin in any public or private document. Yet in the cases she studies, interrogations in mid-1823 were still asking people's calidad, months after the prohibition.

The Denver Public Library's guide makes a similar observation for parish books: by independence, recording ethnicity had largely fallen out of fashion or been outlawed, and in Mexico español was replaced by mexicano. It also notes the exceptions, among them some parish registers in territorial New Mexico that recorded Native American ancestry as late as 1890.

So a label in a record after 1822 is not impossible, and its absence after 1822 does not mean the family had become something else. The rules for writing it down had changed.

What a changed calidad does and does not prove

The reasoning below is ours, built on the studies above, so treat it as a method rather than a rule.

It does not prove you have two different people. In Parral a quarter of traced people changed label between two records a few years apart, so a mismatch is something to expect. Identify people by name, parents, age, place and the people around them, and treat the label as the least reliable of these.

It does not tell you your ancestry. A label recorded how a priest or enumerator saw a person's standing at that moment. A run of español entries is not a measurement of European descent, and a single indio is not proof of an Indigenous parent. Family stories about being "pure Spanish" deserve the same care as any other family story: something real usually sits underneath, but rarely the thing the story claims.

It does tell you something about the moment. A woman's label moving at her marriage tells you about her husband. A man the census called mestizo who married as a Spaniard, as two of Parral's six upper-class mestizo grooms did, tells you about his standing. Read the change as a small piece of social history about your family, which is what it is.

Record every label, with its source. Do not pick one and discard the rest. The pattern across records is the evidence, the same way the reason a law required a document explains what is inside it in an Italian marriage packet.

What KleioBase does with colonial Mexican records, and what it does not

Disclosure, since we sell a tool.

KleioBase is an AI research companion that turns scattered historical records into one connected family archive. A colonial Mexican family is a good test of that, because the same woman may appear in a baptism, a padrón, a marriage and her children's baptisms, each with a different label.

Upload the pages and the Spanish text is read and translated, with the people, dates, places and relationships extracted into a Review step where you can correct any field before it is saved. Our Spanish records page covers the kinds of document it reads. Each record then stays a fixed snapshot of what that document said, while the person's profile holds your conclusion, so the baptism's española and the marriage's mestiza both stay visible against their own records rather than one overwriting the other. That disagreement is exactly what this article argues you should keep. When a search in a parish's books comes up empty, you can record the search that found nothing against the person, with the books and years you covered.

Now the honest limits. Three of them.

It does not know the calidad system. It will extract mestiza as written, but it does not know that marriage pulled labels toward the spouse, or that a blank in a census usually means a child. That interpretation is yours.

Its record-gap suggestions do not cover Mexico. On the Researcher plan and above, Expected records suggests census and other targets for people in the countries it tracks, and Mexico is not one of them. It will not tell you that a parish kept separate books for each calidad.

It cannot find the second book for you. If a baptism is in a series nobody has filmed or indexed, KleioBase has no way to know the book exists. Establishing how a parish kept its registers is archive work.

Where to start with the label you already have

Take one ancestor and list every record you have for her, with the calidad each one gives and the date it was written.

Mark which record came before her marriage and which after. A change that falls at the wedding, toward her husband's label, is the Parral pattern and needs no further explanation.

For any baptism you have not found, find out whether the parish kept separate books for españoles, for castas and for indios in that period, and search every series, not only the one that matches the label you expected.

Then keep all the labels. Write each one against the record it came from, and write down which books you searched for the records you still lack.

The priests and enumerators who wrote these words were not compiling a pedigree. They were placing their neighbours in a social order they lived inside every day, and they changed their minds as people's lives changed. A label that moves is a record of that, and it is one of the few places the colonial record lets you watch an ancestor's standing rise or fall.