Granularity: Determining the degree of detail in your term entries
To better understand granularity, think of coarse grains and fine grains, each one of them containing more or less matter depending on how much material they contain. The same applies to data granules. The Wikipedia provides a very easy example to understand granularity: recording your home address in one category would be a coarse granularity: recording it in several categories (street address, city, postal code, country) would be a finer granularity, and recording it under more categories (apartment number, state, postal code add-on) would be even finer.
So, if we transfer that to our term entries, we could decide to add more, or less, data fields to document our terminological information. Sue Ellen Wright gives the following example applied to term entries:
- /grammar/ m,n,s (masculine noun singular) has low (coarse) granularity,
but if we divide information units into finer categories such as
- /part of speech/ = noun
- /grammatical gender/ = masculine
- /number/ = singular
then, we have high (or fine) granularity.
The decision of having low or high granularity in your term entries depends on your needs, but also on how much time and effort you are willing to invest to make your information more retrievable and manageable. Do you want to include all grammatical information in a grammar data element or would you rather specify each one of them separately? It requires more time to make entries with finer granularity, so you have to make a decision on what is an acceptable level of effort to make it work in your favor.
For more information on termbase management principles, read my blog post “A termbase: What you should know.”
Sources and further reading:
Data modeling and data categories for terminology management, Sue Ellen Wright, 2009
Handbook of Terminology Management, Volume II. Sue Ellen Wright and Gerhard Budin. 2001