Registro De Fatos Históricos

Documenting historical records is messier than people expect

Most people who start a project around registro de fatos históricos run into the same wall within two weeks: they have five sources that disagree on a single date, and they don't know which one to trust or how to record the disagreement without making the database look broken. I spent about eighteen months building a local history archive for a small municipality before I figured out a system that actually held up under scrutiny. The core problem isn't finding information. It's knowing what to do when the information contradicts itself.

Setting up a working registro de fatos históricos system

Start with a spreadsheet or database that has these columns minimum: unique event ID, title, date (standardized), location, primary source citation, secondary sources, confidence level, and notes field. The confidence level column is the one everyone skips and then regrets. Use a three-tier system: confirmed (two independent primary sources agree), probable (one primary source plus corroborating secondary evidence), and uncertain (only one source, or sources conflict). This sounds basic but it prevents you from accidentally presenting speculation as fact. For dates, always store the standardized ISO format in one field and the original written form in another. When a source says "15 de junho do ano da graça de 1847" you record both. Most beginners drop the original form and then lose the ability to trace back when a source transcription turns out to be wrong. I learned this the hard way when a municipal archive corrected a digit in a baptismal record from 1823 to 1832, and I had no way to show my readers where I'd gotten the original number.

The file naming convention matters more than you think. Use the pattern YYYYMMDD_topic_location_ID. For example: 18910315_incendio_predio_centro_HST0047. It sounds excessive until you have four hundred files and need to find one specific document at 11 PM on a Tuesday. I once spent three hours searching through folders named things like "documentos velhos" and "coisas antigas" before switching to this system. The time investment pays off almost immediately.

How to handle contradictory sources

This is where most people quit. You will find two census records listing the same person born in different years. You will find a newspaper article contradicting an official government document. Here is what actually works: create a separate entry for each source rather than trying to merge them into one clean narrative. Each source gets its own row with its own confidence rating. Then add a master event entry that references both and explicitly states the discrepancy. This keeps your data honest and lets future researchers see the full picture. I ran into a particularly annoying case involving a railway construction record from 1908. The company's internal ledger said work began in March. The local newspaper reported groundbreaking in May. The municipal permit was dated February. Three different months for the same event, all from apparently reliable sources. What I ended up doing was creating three source entries, each tagged with its type and reliability assessment, and a master note explaining that the permit date likely represents the administrative start while the newspaper date reflects the publicized ceremony. The ledger date probably marks when actual laborers showed up. None of them were wrong. They were just measuring different things.

Primary sources like parish registers, land deeds, military records, and company ledgers should always be photographed or scanned at high resolution before you transcribe anything. Screenshots from databases are not enough. The metadata embedded in the image file often contains information that doesn't appear in the text, and digital repositories can change or remove content without notice. I've lost three documents this way over four years. The scans sit in a separate folder structure organized by archive name, not by topic.

Common mistakes that waste weeks of work

The biggest one is assuming that older documents use the same calendar system as today. Portugal switched to the Gregorian calendar in 1582, but Brazil's adoption was gradual and some rural records still reference the Julian calendar well into the nineteenth century. If you're working with pre-1890 records, check whether the date has already been converted or if you need to do the conversion yourself. The difference is eleven days during that period, and it compounds when you're cross-referencing events across months. Another mistake is relying on index databases without checking the original record. FamilySearch and Ancestry indexes have transcription errors at rates that vary by region and decade. In my experience, roughly one in eight entries in these indexes contains at least one error in a name, date, or location. Always verify against the original image. This applies equally to Brazilian federal census records and to church books from any denomination.

Confidence levels need to be revisited when new sources appear. I had an event marked as "probable" for two years until a previously inaccessible estate inventory showed up in a provincial archive, which bumped it to "confirmed." The system only works if you treat it as a living document rather than a finished product. Set up a quarterly review process where you go back through your "probable" and "uncertain" entries and check for new evidence. This usually takes about forty-five minutes per hundred entries if your citations are properly formatted.

Tools that actually help

For database management, Gramps is free and handles pedigree and fan charts adequately, but it struggles with non-familial historical events. For broader historical fact registration, I recommend using a combination of Notion or Obsidian for the main database paired with a dedicated photo archive. Obsidian's backlinking feature is genuinely useful for connecting related events across decades. Notion's filtering is better if you need to generate reports for publications or presentations. For image storage and sharing, Internet Archive allows you to upload scans with proper metadata and persistent URLs. This is important because local hard drives fail and cloud subscriptions get cancelled. A backup on the Internet Archive means your research survives even if your computer breaks. I store my master scans there alongside the originals on external drives. It's redundant but the redundancy is the point.

What this system doesn't solve

No amount of organization fixes incomplete source material. If a village burned its records in a fire, your database will have gaps regardless of how well you structure it. Acknowledge the gaps explicitly in your notes rather than filling them with plausible-sounding guesses. Readers and fellow researchers can work around missing information. They cannot work around fabricated information presented as fact. Digital tools also introduce their own fragility. Formats become unreadable. Links rot. Database software changes or disappears. I've migrated my data through three different applications in six years. Plan for this from the start by exporting everything to open formats like CSV and keeping raw scans in TIFF or PDF/A. These will be readable thirty years from now even if the software you used to create them is obsolete.

The confidence level system I described above is a simplification. Academic historians use more granular frameworks, but for most personal or community projects, the three-tier approach is sufficient. Don't let perfect methodology prevent you from starting. The worst registry of facts is the one that was never created because someone was waiting for the ideal system.