Natural Indexing Language 

One of the greatest hurdles in discussing natural indexing language is that it is not easy to identify or know as to what exactly constitutes a natural indexing language. We do not generally come across lists of natural indexing  languages. Obviously, a natural indexing language is the language  of the documents that are indexed or catalogued for a library. Hence, it could be static as long as the document collection remains static. As soon as a new batch of documents is added to the library, the terms of the indexing language are changed to accommodate the new terms contained in the new set of documents. Each system will have a different indexing language even if the documents cover the  same subject area.  Also,  since the indexing language is derived from the documents, added to the library or input into the system of different records, even if they represent the  same documents, they generate  a different indexing language. These variations affect the consistency associated with the library catalogue and so, present many problems. Most of the natural indexing languages are based upon the language of title, abstract and other text of documents. 

There is a debate going on as to the effective use of natural languages for indexing and subject cataloguing purposes. One school of thought believes that full exploitation  of the opportunities offered by computer systems could  only be done taking recourse to natural indexing language, whereas the other school holds the view that controlled indexing  language is the only proper way to  index documents. However, it may be mentioned that controlled indexing language is extremely time consuming, costly and uneconomic. One of  the applications of natural language indexing is the production of indexes based on words in titles of documents such as, Key Word In Context (KWIC) indexes.