SDD proposal: Approximation modifiers

TDWG working group: Structure of Descriptive Data (SDD)

Introduction

A numerical or categorical value may be known to be true, but the exact value may only be approximate. This is different from an uncertainty about the truth of the statement itself, compare "Certainty modifiers".

Examples for approximation modifiers are "ca.", "approximately", "about", "roughly", "nearly", "almost", etc.

Approximation modifiers carry interesting information which would be useful to provide in a machine-readable way. No final decision has yet been made whether approximation modifiers should become a separate catetory of modifiers. They are related to certainty modifiers and could be handled there, or they could be subsumed under general or probability modifiers.

The question will have to be decided once the properties are better understood. For example, is it possible to specify the approximation quality for numerical measures exactly? If spores are measured in a microscope, should an approximation modifier be defined stating "plus/minus 0.5 µm" which has a numeric and measurement unit that allows a processor to understand the measurement uncertainty for a given value? Also, perhaps the latter would be more appropriately expressed in the character definition in the terminology, stating that most likely values entered in a character have a given margin of error, due to the measurement method.

Proposal

(None so far)

Request for discussion

Please comment on approximation modifiers!

Please send your criticism or suggestions to the SDD mailing list or to the author.

Gregor Hagedorn; Vers. 1; 14. March 2003