Reliability of Vital Records That Aren't Quite Right...

Okay, here's a nice "nuance-y" question.  

It is generally accepted that a birth certificate contains primary information given by the parents at the time of the birth, and a marriage certificate contains primary information given by the bride and groom, so both documents are considered reliable.

But what if the information on the documents is actually a little fuzzier than that?  

Example (and let's assume for simplicity that both of these documents definitely refer to the correct people):

1. Subject's birth certificate was NOT CREATED AT THE TIME OF EVENT. It's a delayed birth certificate ordered by his mother from the state government ten years after his birth (once BCs became a matter of course in the state).


2. Subject's marriage certificate is NOT factually correct in many aspects. Both parties gave their correct birth names, dates, and of course the certificate lists the correct marriage date/place. But the bride and groom both DELIBERATELY gave incorrect information regarding their current residence, places of birth, and parents' names (they eloped and were trying to hide from the bride's father, who was a local judge and very well-known; I suppose they thought nobody would tell on them if they didn't give info on who to tell!)

Question: given that the BC's information was given ten years after the birth, and the MC gives incorrect information, how reliable are these documents in terms of proving these two events?

Submitted byEEon Tue, 06/07/2016 - 17:30

Good question, caliadria!  Now it's our turn to "nuance" things:

  • It is "generally accepted" that those two types of certificates contain primary information for the specific pieces of data you flag. However, we cannot say "both documents are considered reliable." Any document err. Any person can make a mistaken remark. Any scribe can mistake what they've heard. Any person, for that matter, can make a deliberately false statement if it suits their need.
  • As you've alluded with the birth certificate, delayed recordings can have errors—unintentional or intentional.
  • As you've noted, some (perhaps many) brides and grooms do have reason to fudge the facts on their marriage records.

Given these factors, it would be foolish to try to say "how reliable" any one document is. Sound research calls for thorough research, using the best sources possible. It calls for obtaining all records that might relate to a particular subject, correlating the information between them, weighing the pros and cons of the evidence, and reaching a decision for each individual fact on the basis of where the weight of the evidence lies.

Submitted byJadeon Wed, 06/08/2016 - 10:15

Vital records?  Use brain and beware.

There has been a recent kerfuffle concerning 20th-century Indiana State death certificates recently uploaded by a major database provider.  Many of the certificate images show areas where some information has been covered up.  Most seem to be Social Security numbers, but some comment has noted blanks in forms issued before Social Security existed.

The provider has given no explanation as to why areas were obscured at the time of microfilming -- at whose behest and whether there was a legislative backdrop to any of the data that was concealed.

In addition, much is awry with the indexing.  Maiden names are not indexed for married women; only their married names are indexed, even when the maiden names are quite clear in the documents.  No explanation is offered for this protocol either, including identification of how the indexes were created and by whom.

It is tempting to utilize what data *does* appear, but the defects are myriad and difficult to account for in a footnote or endnote.

One might wonder why this material, including the faulty indexing, was uploaded at all.  And the would-be researcher must excercise a great deal of caution.

Some of the discussion can be found here:


Submitted bycaliadriaon Thu, 06/09/2016 - 13:12

Let's forget the delayed BC and focus on the deliberately-erroneous MC.  

If the ONLY written evidence of their marriage was this certificate, would you/should you use it as evidence?  Would a proof argument including the MC and the family legend be sufficiently "convincing" to readers?  

My thoughts:

  • 2.9 Selectivity and Thoroughness (p. 45-46 in EE3) specifies that "It may also be appropriate to cite works that err, in order to correct their errors." (on p. 46)
  • 3.45 Traditions (p. 158) indicates that my "channel" (i.e. my grandmother, the subject, told me, the researcher) is pretty reliable, except that it's secondary information given to a child who is now using her memory of the story nearly 30 years later.
  • The QuickCheck Model for Tradition, Recorded on p. 115, tells me that I would cite this story as, 

Williams Family Traditions, Adrienne Tomkins, compiler ([Personal memory?], ca. 1988-1995; privately held by Tomkins, [address,] Walnut Creek, California; elopement of Marian Lafargue and Hooker Williams (marriage April 30, 1932), reported by Marian Lafargue Williams, Tomkins' grandmother, ca. 1988.  Marian and Hooker eloped, and deliberately forged information at the JoP's office in order to throw their parents off their trail.  

Would my "additional information" (the last sentence of this citation) be sufficient to not only correct the erroneous marriage certificate per Section 2.9, but also convince readers that my evaluation of said certificate is correct even though some information is definitely not? 

In other words, Grandma TOLD me they messed it up on purpose, and sure enough, it's messed up!  So she told me the truth... is that good enough?

(Note:  1.32 Content (p. 32):  the footnote on this page refers us to two articles about hoaxes that are EXTREMELY interesting.  I recommend everyone read them!)


Adrienne, I'm about to give you a long, round-about answer that will end up at the right place. But, without all this background that you likely know, some of our readers might misinterpret the bottom line.

Each time we find something we think is relevant for our research problem, it's evidence. We then process it for use. We take notes—we abstract or extract or transcribe or translate—or we enter it into our database. We make analyses of its likely value. We try to correlate it to everything else we know about that person or the time and place or the problem. Many times we can form an "instant opinion" that its quality is likely to be low. Nonetheless, if it's all we have, we may use it as clues to help us find better evidence.

Sometimes, something we felt was of low quality ends up being extremely valuable. Sometimes, records we think we can rely upon end up deceiving us.  That's why we do "reasonably exhaustive research" before we reach a final conclusion—assuming we get to a point that our accumulated evidence does justifiy a conclusion. But amid the additional research, we still need each piece of earlier information processed into something that's usable, correlatable, and capable of being analyzed and evaluated.

To end up with just one piece of evidence really isn't possible, if we've done thorough research. We may end up with just one piece of direct evidence, but we should have many pieces of indirect evidence against which we can appraise the reliability of that one piece of direct evidence. We may end up with no direct evidence at all—but those many pieces of indirect evidence can still enable us to build a case. We may end up with several pieces of direct evidence that contradict each other—which means we have to reconcile the differences.

Ultimately one of several scenarios will happen:

  1. We end up with multiple pieces of direct evidence that all agree. If each is totally explicit and needs no explaining, then all that's needed is a simple citation to whatever we found.
  2. We may end up with multiple pieces of direct evidence that all agree, but one or more has quirks that make us feel some explaining is needed. In that case, we write a simple proof summary that itemizes the sources and explains whatever we feel needs it.
  3. We may end up with just one piece of direct evidence with various pieces of indirect evidence to support it. In that case, we write a proof argument explaining what the direct evidence says, what we have found to support it, and why we have decided it's reliable.
  4. We may end up with no direct evidence at all; but many pieces of indirect evidence that we feel adds up to a valid conclusion. In that case, we write a proof argument to say what evidence does and does not exist, and why we feel the existing indirect evidence justifies our conclusion.
  5. We may end up with one or more pieces of direct evidence that contradict each other—or direct evidence that we feel is disproved by indirect evidence that we feel is more reliable. In that case, we write a proof argument to say what evidence does and does not exist, what contradicts something else, and why we feel our conclusion is justified.

But, of course, all of this begins with a simple find, a need to record the information, a need to identify the source, and a need to correlate that new find against what we already know.  This is the stage at which we write that basic citation, with an evaluation of the source.

This, then, leads us to the citation you have crafted and whether—as you ask—it would be sufficient to "correct the erroneous marriage certificate ... but also convince readers that [your] evaluation of the certificate is correct even though some information is definitely not." The answer really has to be it depends.

Your situation is that No. 5 above. Resolving conflicts is seldom as simple as pitting one item against another and choosing between the two because of a, b, c.  Each needs to be evaluated against outside information. In the process of doing that "reasonably exhaustive research," you should encounter various other records that add context to the two opposing sources you now have (personal account vs. a document). Sometimes (as in another action my own grandmother took) we may find that grandma had a specific set of reasons for telling us that the document errs, when it didn't.

The bottom line is always the first criteria of the proof standard and the foundation of every good research process: reasonably exhaustive research. If several pieces of valid direct evidence all agree, then our research might stop reasonably soon. But if pieces of direct evidence contradict themselves, we can't decide "where the truth lies" without thorough study of all possible evidence and adequate contextual research.

In the meanwhile, your citation to the tradition does exactly what it needs to do.