Do I accept a Certified Document as a primary source and trust it outright or do I need to view the original the Document is transcribed from?
Amy asks, Is a Birth/Marriage/Death Certificate a primary source?.
Birth/Marriage/Death records (aka vital records) are source types. Many genealogists/family historians are trained to classify sources by whether they are in original form or a derivative, rather than to class one source/source type as a "primary source" or a "secondary source." This is because we work with the bits of information that exist in the records, and we know that all sources can contain errors.
There is a related question on the site that asks specifically why historians tend to classify sources as primary or secondary, but genealogists do not.
For answers to the question about how to distinguish between primary and secondary information, see: "What is the difference between “primary information” and “secondary information?”.The links referenced in questions and answers to that Q&A may be particularly helpful.
Note in particular, the response by ColeValleyGirl; emphasis added:
Sources/records contain information. It is the information that is classified as either primary information or secondary information.
Amy asks, "Do I accept a Certified [Document] ... and trust it outright or do I need to view the original the Document is transcribed from?"
There are several different concepts here.
(a) When these vital records are "certified," it isn't the birth, marriage and death information that is being certified. Rather, the authority certifies that the document or information matches the official record. 1
(b) All sources, including certified records, contain information that is subject to error, oversight and omission (vs. "trust outright"). I've appended my answer with some examples.
(c) You used the word "transcribed" to refer to what seems a certified document; I'm not sure I followed that part of your question. You might be referring to a certified records that are created from birth/marriage/death registers. Generally, genealogists/family historians prefer to work with records as close to the original as possible.
Example: The birth of my ancestor William Preston (1754-1842) was recorded in the town book of Chester, Rockingham County, New Hampshire and also Rumney, Grafton County, New Hampshire. Google Maps reports these two towns are about 80 miles apart.
When New Hampshire created its early vital record indexes, the two town clerks extracted information from the town books and submitted a certified birth record to the state; each reports a unique set of information, including that one of the records reports his birth a year later than the other.
Below is a graphic/comparison of the two certified records. To see a larger version of the graphic, click HERE. The graphic makes it possible to compare how different these certified records really are.
In my experience, different information on different records about the same person is not an anomaly. For the person who is the subject of this example, I have two different certifications of his marriage, and there are two different records of birth for his first several children.
1 Most of the jurisdictions I have worked with in the United States offer genealogists the option of ordering uncertified copies of these records at a reduced cost. If that option is available, I take advantage of it for my personal work.
Amy asked "Do I accept a Certified Document as a primary source and trust it outright or do I need to view the original the Document is transcribed from?"
Accepting that it's really the information (evidence?) within a source that primary or secondary to family historians, I think it's worth reminding ourselves why we ask the question. As implied above, it's because primary information is generally more reliable than secondary.
I need to explain how birth / marriage / death certificates are handled in England & Wales, because it's relevant to some of the responses above. There are basically 2 types of office that hold BMD certificates here. Superintendent Registrars' Offices (SROs) and the General Register Office (GRO). After the original certificates are completed, they are filed in the SROs. Every quarter (originally) the previous 3m worth of certificates were manually copied in each SRO and sent to the GRO - of which there is just one for England & Wales. (Please understand this is somewhat different for new certificates now that we have computers)
Accessibility for family historians to certificates is roughly like this:
There are a few exceptions to the above - don't ask about the various copies of marriage certificates, for instance, and I have seen a book of death certificates at an exhibition but one dating from the 1840s. And the legality of the last bullet can be debated but possession is 9/10 of the law.
OK - why explain all this? Because in all the circumstances above, physical access to the original document is not possible. Therefore, if Amy were to be looking for English BMD certificates, viewing the original is not feasible.
Should she treat the certified copies as containing primary info? Well, if the act of copying a document manually is judged to turn what was primary information into secondary information, then all the GRO certificates and all the handwritten SRO certificates become secondary sources and - unless the SRO does facsimile sources - there is no primary info. My point is that - if we declare this huge percentage of certificates to be secondary info, with no primary - what's the point of the distinction?
So - for these source-types, I take the GRO certificates and the hand-written SRO certificates (along with the facsimile SRO certificates) all to contain primary information.
Should she trust them outright? No way... I only said primary information is generally more reliable than secondary. Sources of error include:
On the last bullet, my granddad's birth certificate (which I regard as primary info) contains a date for his birth but my mother (secondary info) was convinced the date was 3d out because his parents had left it too late to register his birth so they massaged the date when they talked to the registrar. As his baptism (some years later) includes a birth date 3d before the registered birth date, I am believing this family story (on the basis of secondary info) and "rejecting" the primary info (still written down in notes).
I don't know whether it's an Irish phenomenon, but with regard to the civil registration of births, especially the earlier ones (from 1864 on), the date of birth on the civil registration should be compared with the church baptism record. If there is a discrepancy, the church record is usually the more accurate.
This is, in part, due to the fact that there was a fine for late registration and also that the christening often took place within hours of the birth. As a result, I will always attempt to check both the church and civil registration.
As said above, I would usually purchase the 'uncertified' copy of the civil registration as it is has the same information as the 'certified' copy. They both come from the same source within the GRO (i.e. they are copied from the same, now digital I believe, record) onto different paper.
With this in mind, I tend to vary between the record being primary or secondary depending on when & where the event took place.
ALWAYS view the original if you can - transcriber's have made mistakes since they started transcribing
I place more reliance on Certified Birth and Marriage records as primary sources because the persons involved were there to give the personal information at the time and place of the event.
Death records, I class as secondary sources, for everything except the attending physician's statement regarding the date, place, and cause of death.
This type of error is common, so I use this additional information only as a point of departure for more research.