42
M. E. U. POORE
which serve a larger number of purposes, are called natural, while those
serving a more limited number of purposes are termed artificial.
“(vi) It is clear from the above that there cannot be one ideal and
absolute scheme of classification for any particular set of objects, instead there must always be a number of classifications, differing in their
bases according to the. purpose for which they have been constructed.
One classification may, of course, be more natural than another in the
sense defined above, end if there is one factor influencing a group of
objects more powerfully than any other, then a classification based on
that factor will be more natural - that is to say useful for a greater
number of generalizations - than any other ; but the difference between this and other classifications is one of degree only and not of
kind. In the classifications of books for example, a classification based
on subject matter is probably more natural than any other, but one
based on authorship must run it fairly close bringing in, as it does,
factors of date and style, as well as to some extent, that of subject
matter.”
He concludes: “Clearly the primary aim of taxonomy, on this basis,
must be the construction of classes of living things about which
scientifically-useful generalizations can be made. Further there must be
a number of different classifications, each based on different characters
and each equally valid for its own purpose.” If this is true of taxonomy,
it cannot but be true for vegetation classification, where there is no
implication of phylogenetic descent.
In his consideration of the systematics of Darwin compared with the
great pre-Darwinian systematists, Cuvier, de Candolle, Adanson and
Lamarck, Cain (1959) discussed the dangers of weighting the characters
which are used in classification, because the exact significance of the
weighted characters is rarely understood and the assessments of the
importance of particular characters certainly differs from worker t o
worker. He maintains that taxonomy of organisms should be based on
all characters, and that “phylogenetic” weighting should be limited t o
those characters where there is positive evidence in its favour. For
example, fossils can and should be taken into account; and characters
which have been proved to be adaptive as a result of recent experimental work can be belittled.
These arguments can equally well be applied to the classification of
plant communities, and lead t o the conclusion that the most natural
classification should be one which is based on the widest possible range
of characters. In certain schemes of classification undue emphasis has
been given t o dominance, sera1 relationships, fidelity and li€e form, each of
which taken alone may give a misleading picture of vegetational relationships. (Poore, 195%; Braun-Blanquet, 1951; Gems, 1918 and 1941.)
M. E. U. POORE
which serve a larger number of purposes, are called natural, while those
serving a more limited number of purposes are termed artificial.
“(vi) It is clear from the above that there cannot be one ideal and
absolute scheme of classification for any particular set of objects, instead there must always be a number of classifications, differing in their
bases according to the. purpose for which they have been constructed.
One classification may, of course, be more natural than another in the
sense defined above, end if there is one factor influencing a group of
objects more powerfully than any other, then a classification based on
that factor will be more natural - that is to say useful for a greater
number of generalizations - than any other ; but the difference between this and other classifications is one of degree only and not of
kind. In the classifications of books for example, a classification based
on subject matter is probably more natural than any other, but one
based on authorship must run it fairly close bringing in, as it does,
factors of date and style, as well as to some extent, that of subject
matter.”
He concludes: “Clearly the primary aim of taxonomy, on this basis,
must be the construction of classes of living things about which
scientifically-useful generalizations can be made. Further there must be
a number of different classifications, each based on different characters
and each equally valid for its own purpose.” If this is true of taxonomy,
it cannot but be true for vegetation classification, where there is no
implication of phylogenetic descent.
In his consideration of the systematics of Darwin compared with the
great pre-Darwinian systematists, Cuvier, de Candolle, Adanson and
Lamarck, Cain (1959) discussed the dangers of weighting the characters
which are used in classification, because the exact significance of the
weighted characters is rarely understood and the assessments of the
importance of particular characters certainly differs from worker t o
worker. He maintains that taxonomy of organisms should be based on
all characters, and that “phylogenetic” weighting should be limited t o
those characters where there is positive evidence in its favour. For
example, fossils can and should be taken into account; and characters
which have been proved to be adaptive as a result of recent experimental work can be belittled.
These arguments can equally well be applied to the classification of
plant communities, and lead t o the conclusion that the most natural
classification should be one which is based on the widest possible range
of characters. In certain schemes of classification undue emphasis has
been given t o dominance, sera1 relationships, fidelity and li€e form, each of
which taken alone may give a misleading picture of vegetational relationships. (Poore, 195%; Braun-Blanquet, 1951; Gems, 1918 and 1941.)