Genealogical finding aids are currently a hot topic. Indexing existing records seems to be an obvious way to make information more available, particularly to beginning researchers. But in some cases, indexes hinder rather than assist finding a particular record.
To understand my concern, it is necessary to understand how and why indexes exist. Historically, an index was provided in conjunction with a written manuscript to aid the reader in finding particular information in a long document. For example, as a genealogical researcher, I commonly find indexes or lists of names in parish registers and New England Town Records. The priest or the town clerk spent the time to compile a list of the entries so that he could find the same entries in the future. However, parish registers are a good example of documents where indexes have limited utility.
Creating an index requires the indexer to select specific items within the text to include in the index list. An index differs from a catalog in that a catalog organizes information by subject and may be based on geography, chronologically or in some other fashion. An indexer reviews the entire document and makes a somewhat arbitrary selection of items to include in a usually alphanumeric list with page numbers showing where the terms can be found. Technology now provides an alternative; full text indexing. Documents can be digitized and then a complete "index" of every word becomes available through search programs (engines) and optical character recognition or OCR.
OCR has its limitations. The most significant one for genealogists is the lack of a reliable way to consistently and reliably read handwritten records. Hence the need for indexing. But as researchers we need to always be aware of the limitations of relying solely on indexes to find information. Unfortunately, there is always a background need to bulldoze the information, that is, to look at each entry.
Indexes may have a high level of reliability, but the rely heavily on the accuracy of the original record. For example, if a census enumerator wrote down your ancestor's name phonetically, the indexer will usually add the name the way it was spelled in the original record. Even if you, as a researcher, go back and examine the original record, you may not recognize the entry for your ancestor because it was so badly written in the first instance. Because indexes are derivitive, they add an additional layer of possible inaccuracy. It is always important to look at the original records, if they are available.
One serious mistake of unseasoned researchers is to assume that the index contains all of the information from the original record. This is usually not the case. Most indexes are selective and there may be much more information in the original.
The rule is that the absence of information about an ancestor in an index is not conclusive as to what information there may be in the original documents.
Of course, there are indexes that are the "original" document. For example, a telephone book or city directory is a form of index, but both can be considered as the "original." It is always a good practice to verify the information that is initially found in an index or index-like record with other sources.
None of these comments diminish in any way the importance of indexes as finding aids, but researchers should always be aware of and evaluate the reliability of any document used as a source for genealogical information.