90
J. M. LAMBERT A N D M. B . DALE
geneous” vegetation for comparison : the selection of such stands can
only be made on the assumption that they exist, and the fact that
workers believe that they can recognize such stands in a given situation
is itself no proof of the general applicability of the method. Moreover,
if the concept of the stand as the basic unit is abandoned in favour of
more objective sampling, then the contentious concept of “minimal
area” can also disappear, together with a priori “tests of homogeneity”
for uniformity within a given stand. Apart from the saving in time, it is
surely better for ‘areas of relative homogeneity to emerge from an
analysis, rather than being imposed at the outset by the preconceptions
of the individual investigator.
If the foregoing argument is accepted, we are left with a deiinition
of vegetation which includes habitat relationships as well as the plants
themselves, but with no assumptions about the nature of the relationship; further, the only assumptions about the nature of the plant
component is that it consists of separately identifiable units, that
these can be recognized as species on accepted taxonomic criteria, and
that the use of species for categorizing the plant material is likely to
confer a basic heterogeneity on the system, On the assumption of such
heterogeneity, together with practical considerations (see pp. 76-77),
the decision is taken to classify rather than ordinate as the primary
approach. From this point onwards, therefore, we may now look at
other phytosociological concepts related to classification.
B. THE BASIC APPROACH
The place of phytosociology in vegetation studies is defined by
Poore (1962, p. 51) as follows: “The proper province of plant phytosociological studies should be to describe vegetation and to discover
and define problems for solution by more exact methods.” Elsewhere
(p. 36), he states: “Every description is an abstraction from the available data” and (p. 51) “current statistical methods are inappropriate
for [the description of stands for classification]”. Such statements,
which are representative of many others in phytosociological literature,
show a misunderstanding of the main function of a statistical approach.
The place of classification lies between description and abstraction, in
that it represents an ordering of the described material in a form suitable for abstraction. There are thus four distinct stages in the operation :
the irlitial selection of the phenomena to be described; the description
based on selection; the classijcation based on description; and the
ultimate abstraction of common material for further use, such as hypothesization. The criteria to be used for the selection, the method of
description, the form of classification, and the amount of abstraction,
are d subjective decisions; but, whereas “description” seeks to record
J. M. LAMBERT A N D M. B . DALE
geneous” vegetation for comparison : the selection of such stands can
only be made on the assumption that they exist, and the fact that
workers believe that they can recognize such stands in a given situation
is itself no proof of the general applicability of the method. Moreover,
if the concept of the stand as the basic unit is abandoned in favour of
more objective sampling, then the contentious concept of “minimal
area” can also disappear, together with a priori “tests of homogeneity”
for uniformity within a given stand. Apart from the saving in time, it is
surely better for ‘areas of relative homogeneity to emerge from an
analysis, rather than being imposed at the outset by the preconceptions
of the individual investigator.
If the foregoing argument is accepted, we are left with a deiinition
of vegetation which includes habitat relationships as well as the plants
themselves, but with no assumptions about the nature of the relationship; further, the only assumptions about the nature of the plant
component is that it consists of separately identifiable units, that
these can be recognized as species on accepted taxonomic criteria, and
that the use of species for categorizing the plant material is likely to
confer a basic heterogeneity on the system, On the assumption of such
heterogeneity, together with practical considerations (see pp. 76-77),
the decision is taken to classify rather than ordinate as the primary
approach. From this point onwards, therefore, we may now look at
other phytosociological concepts related to classification.
B. THE BASIC APPROACH
The place of phytosociology in vegetation studies is defined by
Poore (1962, p. 51) as follows: “The proper province of plant phytosociological studies should be to describe vegetation and to discover
and define problems for solution by more exact methods.” Elsewhere
(p. 36), he states: “Every description is an abstraction from the available data” and (p. 51) “current statistical methods are inappropriate
for [the description of stands for classification]”. Such statements,
which are representative of many others in phytosociological literature,
show a misunderstanding of the main function of a statistical approach.
The place of classification lies between description and abstraction, in
that it represents an ordering of the described material in a form suitable for abstraction. There are thus four distinct stages in the operation :
the irlitial selection of the phenomena to be described; the description
based on selection; the classijcation based on description; and the
ultimate abstraction of common material for further use, such as hypothesization. The criteria to be used for the selection, the method of
description, the form of classification, and the amount of abstraction,
are d subjective decisions; but, whereas “description” seeks to record