Name Spelling Variants on Index Cards

What is a name spelling variant on a Civil War index card?

A spelling variant is the same soldier recorded under a different version of his name on different cards. The Civil War Soldiers and Sailors System drew its soldier names from the General Index Cards of the Compiled Military Service Records. The National Park Service notes that the total number of General Index Cards exceeds the number of soldiers who actually fought, and that serving in more than one regiment, serving under two names, or spelling variations all contributed to that gap (National Park Service, Records Origin, https://www.nps.gov/civilwar/soldiers-and-sailors-recordsorigin.htm). A variant is therefore expected, not exceptional.

It helps to keep two levels of record apart. The General Index Card is a summary card. It is one card per soldier entry, drawn from the first and last appearances of his name, and it carries the name, the rank in, the rank out, and the organizational unit. The underlying Compiled Military Service Records are the fuller set: Ainsworth's staff, working from the 1880s onward, wrote a card for every appearance of a soldier's name on a muster roll, so a single soldier's file usually holds many cards. When you work with spelling variants, you are usually comparing summary cards. When you want corroboration, you go down to the service cards beneath them.

Which card fields actually decide if this is one person or two?

The unit, the rank in and rank out, and the dates attached to each appearance. Those fields travel together. A spelling can wobble; a regiment, a company, an enlistment rank, and a departure date are far harder to duplicate by accident.

Start with the organizational unit. Two cards that read "51st Virginia Infantry, Company D" under "Chastain" and "Chastine" are consistent with one man, but the match is a lead, not a conclusion. The NPS states that the soldier records in its system rest on the General Index Cards, which carry the name, rank in and out, and the organizational unit (https://www.nps.gov/civilwar/soldiers-and-sailors-overview.htm).

Then compare rank in and rank out. A card showing Private on entry and Corporal on departure is a small, checkable signature. Rank changes are not random decorations; they mark a career. If you want the reasoning behind those two fields, see What Rank In and Rank Out Can Tell You.

Why does the same soldier generate more than one card?

Because Ainsworth's staff wrote a card for every appearance of a name on a muster roll. The cards were created to settle pension eligibility, and each appearance of the name produced its own card. One soldier's file usually holds many cards. A man who reenlisted, transferred, or whose name was written inconsistently across rolls will leave a longer paper trail than a man who served one short term under one spelling.

That is also why two variant cards are not automatically two soldiers. They may be two moments of one service, or they may be two different men whose names happen to resemble each other. The cards give you material for a hypothesis. They do not settle identity on their own.

A comparison table helps more than a hunch

Use the fields side by side. The table organizes your evidence. It does not decide the case.

Field	Match	Partial or conflicting	Weak or absent
Organizational unit (regiment, company)	Same unit named on both cards	Same regiment, different company	Different regiment entirely
Rank in	Same entry rank	One card silent, or an unexplained jump	Contradicts the career on the other card
Rank out	Same final rank, or a documented step up	Different final rank with no promotion trail	Final rank contradicts the earlier card
Dates of presence	Overlapping or continuous service dates	A gap that could mean a second term	Dates that cannot belong to one service
Name form	Minor spelling drift (one or two letters)	A different given name or initial	A different surname with no phonetic link

Treat the table as a decision aid, not a verdict. A row of agreements is a reason to keep investigating. It is not proof. Two men in the same regiment can share a surname, a rank, and overlapping dates, and a clerk can still have written the same man's name two ways. Corroborate with the service cards, then with the fuller record series, before you treat two entries as one soldier.

How should you treat a spelling that looks convincing but the unit is wrong?

Do not let the name form carry the decision. Names are the least stable field on the card. Transcription, handwriting, and regional pronunciation all bend surnames, and the same clerk who wrote "Chastain" might write "Chastaine" a month later. A near-perfect spelling match attached to a different regiment is usually a different man.

Ask instead whether the wrong-unit card shares anything else. If it shares nothing beyond the sound of the name, keep the two apart and log the open question. The index card is not the final answer, and neither is a name match.

What do you do when the cards genuinely conflict?

Write down what each card states, then state the inference separately. A card is a document fact. "These two entries describe one soldier" is your reading of them. Keeping the two apart protects you from building a service history on a guess. The method is set out in Separating Document Facts from Inferences.

If the unit and rank hold but the dates overlap impossibly, one of the cards may belong to a relative serving in the same regiment. That happens, and it is worth naming as a possibility rather than resolving it prematurely.

When should you stop and flag it?

Stop when the only thing linking two entries is the spelling of a name. Stop when the file offers no unit, no rank, and no date on at least one of the cards. Stop when a plausible answer requires you to assume a clerk made two unrelated errors. At that point the honest output is a note, not a biography. When to Stop and Flag an Open Question covers how to record a half-answer so the next reader can pick it up.

One more caution that belongs in every Civil War file: a link from a soldier to a regiment to a battle shows the regiment was present, not that the man was. The NPS states plainly that this linking does not indicate with certainty whether a soldier participated in a particular battle (https://www.nps.gov/civilwar/soldiers-and-sailors-overview.htm). So even a clean variant decision does not license a claim about a battle, a wound, or a death.

Frequently asked questions

Can one soldier appear under two surnames? Yes. Serving under two names is one of the documented reasons the number of General Index Cards exceeds the number of soldiers who fought.

Does a matching first name prove identity? No. Given names and initials are among the most frequently altered fields in transcription, and a name alone never settles identity.

Should you merge two cards into one soldier? Only after unit, rank, and dates agree or plausibly continue, and ideally after corroboration in the service cards. Otherwise record them as two entries and note the relationship you are testing.

Where do you go after the cards? To the fuller record series. The service record and the pension file are separate series at the National Archives and must be requested separately (CSR and Pension File: Two Orders).

The short version: let the unit, the ranks, and the dates guide you, let the spelling merely suggest, and treat every match as a lead that still needs corroboration.

SOURCES