Phytosociology is the study of the characteristics, classification, relationships, and distribution of plant communities. A phytosociological system is a system for classifying these communities. It is often suggested that it is not a Science in its formal sense. The aim of phytosociology is to achieve a sufficient empirical model of vegetation using plant taxa combinations that characterize univocally vegetation units. Vegetation units as understood by phytosociologists may express largely abstract vegetation concepts (e.g. the set of all hard-leaved evergreen forests of western Mediterranean area) or actual readily recognizable vegetation types (e.g. cork-oak oceanic forests on Pleistocene dunes with dense canopy in SW-Iberian Peninsula). Such conceptual units are called "syntaxa" (singular "syntaxon") and can be set in a hierarchy system called "synsystem" or syntaxonomical system. The act of creation, amelioration or adjusting the synsystem is called "syntaxonomy". Therefore, the syntaxonomical system is putatively a sufficient empirical representation of vegetation of a given territory. An International Code of Phytosociological Nomenclature, issuing the rules for naming ‘‘syntaxa’’ exists and its use has increased among vegetation scientists.
Sorry, that's a direct quote from Wikipedia. I came across the term while looking at Library of Congress Subject Headings. The LCSH term is "Numerical Syntaxonomy". It looked like it was related to language or linguistics or semantics, but it is apparently only related to plants and to the notion of plant "taxons" so the "taxonomy" isn't our usual notion of taxonomy.
This suggests that Wikipedia might be a good text source to consider using. Public Domain and all that.
---- And stop saying for Statistics: "The word is so rare that you have probably never seen it before, and may never see it again." is sounds like you're carping that I'm not telling you info about a word you know and *certainly* conradicts the "At Wordnik, we love new words!"