Establishing who someone is sounds like it should be simple. In practice it depends on a patchwork of systems that were each built for a different purpose, that rarely talk to one another, and that all degrade in the same way when a name is common. A provider registry exists to route claims payments, not to tell a patient which physician they saw. A state bar directory exists to record admission and standing, not to describe what an attorney actually does. A screen credit exists to satisfy a guild agreement, not to maintain a stable identity across a career. Each of them answers one narrow question well and every adjacent question badly.
The guides in this section explain those systems from the inside: what each one is authoritative for, what it cannot tell you, how to query it, and the specific failure mode it exhibits when two people share a name. They are written for anyone who needs to be certain they have the right person — a patient checking a referral, a records clerk reconciling a file, a journalist confirming a source, a researcher cleaning a dataset, or simply someone who searched a name and got four different people back.
The material is grouped into four sections. Three of them follow the professions where the name recurs most often and where authoritative records actually exist: medicine, where the National Provider Identifier gives every clinician a unique number; business and law, where corporate filings and bar admission records are dated and signed; and media, where credit databases are neither. The fourth section steps back from individuals entirely and looks at the name itself — why this particular combination of surname and given name collides so heavily, and what that means for anyone trying to resolve it.
Each section opens with a hub page that frames the whole subject before the individual guides go into detail. The hubs are worth reading first. They explain why the records in that field behave the way they do, which is usually more useful than any single procedure.
What these guides assume
They assume no specialist knowledge and define terms as they go. They also assume the reader wants to reach a defensible conclusion rather than a fast one, so they are explicit about the limits of every source. A recurring theme is that most identification errors do not come from bad data. They come from a reader treating a source as authoritative for something it was never authoritative for — reading a directory listing as though it were a licensing record, or a marketing biography as though it were a filing.
How the guides are sourced
Everything here is written from primary sources: government registries, licensing bodies, institutional and faculty pages, corporate filings, official studio and publisher materials, and published scholarship. Third-party aggregators are used to locate leads and are never cited as authority for a fact, because the recurring lesson of this entire reference is that aggregated data is where identity errors are introduced and then propagated.
Where a claim could not be confirmed against an authoritative record, the text says so plainly rather than asserting it. Where sources genuinely disagree, the disagreement is described rather than resolved by picking the more convenient answer. That approach makes some passages less tidy than they could be. It also makes them trustworthy, which matters more in a reference than polish does. The methodology page sets out the standards in full, including the inclusion criteria that determine which people are documented at all.
Why a reference rather than a search
A search engine is very good at returning documents about a name and very bad at telling you how many different people those documents describe. That is not a flaw in the ranking. It is a consequence of how entity resolution works: a search engine infers entities from the documents it has crawled, so when the underlying documents themselves conflate two people, the inference inherits the conflation. A directory that merged two physicians years ago is still feeding that merge into every downstream system that reads it, and no amount of ranking sophistication can separate what the source material has already joined.
Resolving a common name therefore means going around the aggregated layer rather than through it, back to records created by an institution that had a reason to get the identity right: the body that issued the credential, the employer that hired the person, the regulator that licensed them, the commission that received the filing. Those records are public, they are usually free, and they are almost never the first result. Much of what follows is an explanation of where they live and how to read them.
What is deliberately absent
These guides describe how to confirm a professional identity. They do not describe how to compile a dossier on a private person, and the distinction is not a technicality. The records discussed here are public because a profession decided that accountability requires it. A license is public so patients can check it. A filing is public so investors can read it. A bar admission is public so clients can rely on it. Each record is public for a specific purpose, and that purpose is the boundary of what it should be used for.
Nothing in this section covers personal life, private records, or the data-broker techniques that turn a name into a profile. This reference exists to answer one question, which professional is this, and to stop there.