At it's basic level, MV-DBMS works on four core principals:

1. Data should be expressible in what ever way makes sense. It should not be limited by size or other considerations.

2. Data should be expressible in its original form, and rather than being machine or human massaged or constrained into a set of standards.

3. Data is relatable even when its expression is inconsistent. For instance, '1/4" bolt,' and 'quarter-inch bolt' have identical "meaning" even though having a different expression.

4. Data should be expressible along multiple axes, rather than being expressed in a "flat" or "normalized" form, so that positional tools can be used to enhance meaning and simplify data management.

In short, MV-DBMS is a database engine explicitly designed to work with "raw text," which is the core information which businesses, organizations, and governments need to manage their existence. And when we speak of "raw text" we do not exclude numbers and numeric values. Numbers also are text, and by having "text" as well as numeric properties, they can be given meaning far in excess of some "binary expression."

The elegance of MV-DBMS comes from this openness and flexibility. Because virtually anything stored in a computer is "text," anything is ultimately manipulable and analyzable within MV-DBMS.

An example of the power of this openness is demonstrated through the "hot topic" of "objects." An object is merely a stand-alone program which is capable of transforming or validating certain classes or groups of data. Their advantage in being "stand-alone" is that they can be used almost anywhere where they are applicable. Few people know that the original concepts for "objects" were prototyped in an MV-DBMS environment: the only environment at the time flexible enough to allow virtually unlimited transformation and capable of supporting stand-alone operation rather than requiring an integrated job-stream.